23
SAALDBURG SONDERDRUCK AUS: SAALBURG JAHRBUCH 55-2005 HERAUSGEGEBEN VON EGON SCHALLMA YER VERL.AG PHILIPP VON ZABERN . MAINZ AM RHEIN

The adjacent forts

Embed Size (px)

Citation preview

SAALDBURG

SONDERDRUCK AUS

SAALBURG JAHRBUCH

55-2005

HERAUSGEGEBEN VON EGON SCHALLMAYER

VERLAG PHILIPP VON ZABERN MAINZ AM RHEIN

5 A A L [] 11 U R G I 133

THE ADJACENT FORTS OF GERMANIA SUPERIOR AND DACIA

Felix Marcu

The concern of this paper is to find an explanation for the existence of a fort or fortlet in the environs of anoshyther during the Principato Although the pattern in the frontier area is to maintain a distance of a days march between forts this is not always the case Of course the geographical and military particularities of each secshytion of the limes are prerequisite to a fix distance norm Consequently the distance between fortifications is of about 30-40 kilometer Interestingly enough the erecshytion of two forts in the close proximity of each other is sometimes a fact These examples are rare in the Roman Empire but noteworthy in Germania Superior (Fig 1) and Dacia (Fig 7) there is a significant degree of conshycentration

Although not an invention of the second century AD the prototype of adjacent forts was already established under the Republic Frontinus had clearly specified that Hasdrubal knew that the armies of Livius and Nero had united although by avoiding two separate camps they strove to conceal this fact 1 This leads obviously to the conclusion that the possibility of erecting two forts for different troops was a fact and more than that someshytimes this solution was preferred over the concept of joint garrisons1 Frontinus also indicates that Crassus during the Slave war fortified two forts beside the camp of the enemy3 Another evident proof is Caesar and his campaigns in Gaul especially the besiege of Alesia4 In the strategic point around the fortified hill Caesar aligshyned many forts for the garrisoning of his troops in a resshytricted circuit of C 20 km5 Nevertheless the mentioned examples reflect special needs in special times and there are no indications for the construction of more than one fortification except on campaign

This paper has been accomplished partially during the time I was awarded a scholarship by the Fritz Thyssen Foundation and Alexander von Humboldt-Foundation (Bonn) and part of it has been written at the Institute of Ancient History of the University of Cologne Germany Furthermore I wish to thank Prof loan PISO

and Dr C S Sommer for reading a draft of this paper and for some suggestion Naturally I alone am responsible for all flaws in the final version

Frontinus 12 2 For some examples see Redd6Schnurbein 2001 490-492

Frontinus 1 5 4 For the relation between literary sources and the archaeological

reality see ReddeJSchnurbein 2001 489-513

The association of troops was a principle of Roman strategy as well as under the Principato It is also well known how Masada was conquered but it should be remembered here that the Romans have built eight forts at the base of Masada rock and surrounded it with a high wall6 Probably the Romans did not even think there was a problem in associating military units when the need arose though at times they were aware of a certain danger This is clear when Tacitus remarks dushyring the Pannonian mutiny at the beginning of the first century AD that cum Pannonlcas legiones seditio incessit nullis nouzJis causis nisi quod mutatus princeps licentiam turbarum et ex ciuili bello spem praemioshyrum ostendebat Castris aestiuis tres simul legiones habebantur 7

But are they really conscious that the soldiers could be a menace It seems that at the beginning not quite so In the times of Germanicus the anger of the soldiers in leg I and XX would have been the result of Gods will tum (atalem increpans rabiem neque militum sed deum ira resurgere 8 Yet afterwards secuti exemplum ueterani haud multo post in Raetiam mittuntur specie de(endendae prouinciae ob imminentis Suebos ceterum ut auellerentur castris trttcibus adhuc non minus asperishytate remedii quam sceleris memoria9

Consequently it is obvious that there was no illusion regarding the power of more than one legion IO There are other illustrations of mutiny in the first century AD but a definite interdiction concerning the association of more than one legion was given only by Domitian after Saturninus revolt of AD 89 when geminari legionum castra prohibuitll Following there wont be more than one legion garrisoned in a single fort

6 Josephus Flavius Bellum Judaicum VII82

Tacitus Ann 116 8 Tacitus Ann 139 9 Tacitus Ann 144 10 Clearly Marius reforms of 107 BC nad been a beaKthrough for

the Roman army development and political danger of the soldiers will be reflected dUring the cvl wars of the first certGry BC

Caesar De bello Gailico 63-90 11 Suetonius Dom 3 5

134 I S I I l [J BUR G

GERMANIA SUPERIOR KM o shy 1r -0

Fig 1 1 Neckarburken 2 Miltenberg 3 Osterburken 4 Ohringen 5 Wezheim (based on Southern 1989 Fig 1

c

S 1 1 L [] BUR G I 135

During the first century until Domitians rule the literary sources are confirmed by archaeological evishydence and the examples of the comshybined legionary castra of Mainz or Vetera would suffice to make the point 12 Polybius proposed in refer-

to the Republican army that in any castra aestiva every legion has to be commanded by its tribunes and inside the fortification accomshymodation of the soldiers and horses have to be separate for each legion with a distance of fifty Roman feet between the tents of the tribunes or the tents of the cavalry of each legion 13 In the same time the conshysuls accommodation and the quaesshytorium 14 continue to be placed in the centre of the fortification 15 The reason was as expected purely technical The internal planning of the fortress or forts had to respond to the needs of the troops And what was the most important thing Obshyviously giving orders and avoiding confusion Consequently the orgashynization of the army inside the fort Concluding during the Republic or idea was to garrison the troops in one fortress in time of war or peace as there is no reason to do it separately Even the allies or auxiliaries were encamped theoretishycally jointly with legionaries Hyginus states that the combined garrison of a fortification should also include troops of nationes 16 Again we are dealing however with the army on campaign Evidently saving time and energy was the most important reason to build only one fort and just when the strategically needs justify it the construction of more than one fort would be the comshypromise

There is more than one eastra stativa garrisoned by a heterogeneous joint force of legionaries with auxiliaries

~

(~

Fig 2 Neekarburken (based on FiltzingerlPlaneklGimmerer 1976 Abb 215)

must be precise the Principat the

or auxiliaries with auxiliaries Additionally Rottweil17 Carnuntum 18 or Aquincum 19 are examples of forts placed in the neighbourhood of a fortress A combined garrison of auxiliaries is proposed where the area of an auxiliary fort extends to over 3-4 ha and no ala milliaria is attested Dacia displays three fortifications between 6-7 ha of Tibiscum Micia or Porolissum also with epigraphic evidence of a combined garrison2o For the other parts of the Empire there are other few examshyples at Vechten (c 45 hal Niederbieber (over 5 hal or Rottweil (c 6 hal where more than one unit is in garrison though there are uncertainties21 To my knowshyledge there are meagre evidences to the existence of two auxiliary neighbourhood forts garrisoned at one time by troops of the same status The exception is to be found

12 The fortress of Mainz was garrisoned by leg XVI Galica and liII Macedonica renamed after the Batavian revolt and dislocated to the east and at Burnum at first After the revolt the fortress was garrisoned by leg I Adiutrix and XIV Gemina At Vetera until the Batavlan revolt there were in garrison V Alaudae and XXI Rapax the last one replaced some time after by XV Primlgenia

13 This is necessary to give room for tribunes horses mules and baggage Polybius VI2728

14 For quaestorium see Hyginus sect 18 15 Polybius VI32 16 Hyginus sect 29 30

17 Conceivably Kastel I and Kasteli IVN are conternJca~ells

SornrnerKorurn 2005 18 Kardler 1997 19 Nemeth 990 677--678shy20 For a shor view over the forts and part of t~e )) ga~

reerences see Gudea 1997 32-34 37-39 46-49 -~ee 5

other fortifications in Dacia garrisoned only by -gt11 u~ ts t~eee with dimensions of c 3-4 ha as Booga 0 R~ 3 a a Porolissensis

BogaersROger 1974 62-65 ORL ~r la Pa~~~ L975 -13 k Rottweli there is still no evdence tlat he aested gar S5 ee contemporary Sommer 1992 272-282

136 I S 1 1 L [] BUR G

at Straubing where there were two contemporaneous forts at one time22 and possibly the forts of Ohringen

infra

GERMANIA SUPERIOR

Adjacent fons usually with different characteristics seem to be a pattern in Germania Superior There are not necessary too many details as the problem of this limes is relatively well documentedB yet there are not many excavations inside the fortifications and the chronology is still partially known24 Important here are those adshyjacent forts recognised at Neckarburken Walheim Milshy

Osterburken Ohringen and Welzheim

At Neckarburken (Fig 2) the fort of 1315 by 158 m (207 hal is probably garrisoned by coho III Aquintashynorum equitata transferred about the mid-second censhytury to Osterburken25 Eastward 200 m away from the fort there is a second one of only 80 by 80 m (064 hal of a more or less regular form with the eastern side a little shorter than the western one26 Three of the gates and the enclosure wall have been identified and inside has been uncovered the principia27 therefore one individual troop was in garrison This is numerus Brittonum Elantiensium28 present here for sure at least immediately after AD 14529 The unit is attested also in the nearby fortIet of Trienz30 Moreover it is known that the Brittones Elantienses restored the baths of the eastern fortlet of Neckarburken in AD 15831

A contemporaneous and similar situation seems to be encountered southwards at Walheim where two forts circa 300 m of each other were identified circa 150 m east of Neckar (Fig 31) One of the forts (Kastell I) of

Both forts are placed in the imediat vicinity and are contemposhyraneous starting with the Flavian emperors until Marcomannic wars when the western fort (Kastell IV) ceased to exist Czysz DietzFischerKellner 1995 518-521 Abb 220 The construction of two different forts is easy to understand because the erecting of the two forts was not perfectly simultaneous and the Romans probably considered unnecessary a joint garrison We should also not forget the early date of construction probably significant Also in Raetia it seems in an aerial photography that there are two forts at Theilenhofen but one of them is conSidered as a construction camp see Sommer 1988 544 Abb 27

23 For the problem of this sector of limes and also for the bblioshygraphy see Schonberger 1985 and Reuter 1999407-419

24 See for a short report KortOm 1998 34 25 FiltzingerPlanckCammerer 1976 425-427 26 FiltzingerPlanckCammerer 1976 Abb 215

Baatz 1973 124-127 FlltzingerPlanckCammerer 1976 427 28 For numeri Brittonum in general and their recruting history see

Southern 198994-98 Reuter 1999 21l 29 ell XIII 6490 30 Cil XIII 6498 31 Alfoldy 198367-69 Schallmayer 1984 451-455

134 by 156 m is probably occupied by coho I Asturum equitata because this is the garrison of the later fort of Mainhardt correspondent to Walheim on the outer limes of Antoninus Pius32 The turf and timber fort has been built at the end of 1st century or early in the 2nd

century The enclosure has been reconstructed of stone at one time and it was finally abandoned about AD 160)3 A numerus () fortlet (Kastell II) of 109 x 64 m was built this time only of turf and timber and has a rectangular plan34 The datable artefacts indicate a chronology starting under Trajan or early Hadrian reign until mid-second century35 The internal planning is not very traditional The interior is divided in two parts a relatively small praetentura and a retentura without any latera praetorii consequently without classical prinshycipia 36 This could be a reason to consider the garrishyson not as an independent unit but as a detachmentY Nevertheless the unit in garrison remains unknown but if there is a pattern regarding the existence of two neighshybourhood forts at least on this section of the limes than also in Walheim the second fortlet was garrisoned by a numeruss If this fort let was not intended for a numerus or for detachments it would be hard to explain the construction of two separate forts in the same time for two auxiliary units

At Miltenberg (Fig 31) there is a fort of 27 ha (Fig 32) where coh I Sequanorum et Rauracorum equitata is attested in mid-second century transferred here proshybably from Oberscheidenta139 At a distance of 2300 m from the first fort there has been identified another one of 06 ha (Fig 33) It is supposed that the fort was garrisoned by numerus exploratorum Seiopensium40

present here at least during the end of the second censhytury or in the beginning of the third jointly with exploshyratores Triputienses 41 There is no certain proof that the fortlet was built in the same time with the auxiliary fort

32 KortOmlauber 2004 34 33 KortumLauber 2004 34 34 Kortumlauber 2004 361 35 Until AD 150170 weredentlfied four phases of the interior

buldings KortGmlauber 2004 38 Abb 43 36 In the ntenor were excavated some barracks In the central part

of the fort probably a granary and other rectangular buildings with a clear plan of a bath in praetentura of the last phase KorshytOmlauber 2004 90-120

37 It is supposed on the basis of some fnds that at least part of regu ar units statlonedn the fortlet KortumLauber 2004 379

38 A pOSSible argument of the archaeOlogists in Walheim is that the special type of barrack Without arma has a good parallel in the numerus fort let of Hesselbach anc it is considered finally that this could be a special type of barrack intended for numeri KortOmlauber 2004 384f

Cl XIII 6604 Cil XIII 6597 AE 1977 593 Cil XIII 12460 3aatZi-errmann 1982 438

40 Cil XIII 6600 (7) 6605 41 Stell 1932 266f BaatzHerrmann 19824391

S II II L D BUR G I 137

0~ ~

~ Fort ~

~ ~~

m Hm +

IrJ I 1

0 SOm F9[j- ~ I 2 d

_I 3

Fig 3 1-3 Miltenberg based on BaatzlHerrmann 1982 Abb 402 403 405j

but it has been argued that it was occupied until the third century42 Nevertheless some new archaeological research shed new light on the history of the fortlet of Miltenberg It appears that sometime during the third century the garrison of the fortIet was drastically redushyced and probably the remaining soldiers occupied only the northern quarter of the fortlet 43 Anyway it is not certain if the soldiers belonged to the same unit

42 Baatz Herrmann 1982440 43 See JaeScholz 2002 416f The authors idea is that this is not

the only fortlet with a reduced garrison and there is another clear example at Kapersburg and the scenario is that many troops were reduced because they have sent soldiers for the conflicts with the Persians during the third century JaeScholz 2002 418-420

Eastward of Neckarburken on the new limes built during mid-second century44 other adjoining forts have been constructed such as those of Osterburken (Fig 41 The first fort of 214 ha was built during the reign of Antoninus PiUS45 This one was occupied by cob III Aquintanorum equitata present here also in the third century as it is confirmed by the epithets Seterimr~ and Philippiana 47 Under Com modus another forrifiashy

44 For the bibliographical referecces see A1 983 S~a shymayer 1984

45 AlfOldy 1987406-408 46 Cil XIII 6568 47 Cil XIII 6566

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

5 A A L [] 11 U R G I 133

THE ADJACENT FORTS OF GERMANIA SUPERIOR AND DACIA

Felix Marcu

The concern of this paper is to find an explanation for the existence of a fort or fortlet in the environs of anoshyther during the Principato Although the pattern in the frontier area is to maintain a distance of a days march between forts this is not always the case Of course the geographical and military particularities of each secshytion of the limes are prerequisite to a fix distance norm Consequently the distance between fortifications is of about 30-40 kilometer Interestingly enough the erecshytion of two forts in the close proximity of each other is sometimes a fact These examples are rare in the Roman Empire but noteworthy in Germania Superior (Fig 1) and Dacia (Fig 7) there is a significant degree of conshycentration

Although not an invention of the second century AD the prototype of adjacent forts was already established under the Republic Frontinus had clearly specified that Hasdrubal knew that the armies of Livius and Nero had united although by avoiding two separate camps they strove to conceal this fact 1 This leads obviously to the conclusion that the possibility of erecting two forts for different troops was a fact and more than that someshytimes this solution was preferred over the concept of joint garrisons1 Frontinus also indicates that Crassus during the Slave war fortified two forts beside the camp of the enemy3 Another evident proof is Caesar and his campaigns in Gaul especially the besiege of Alesia4 In the strategic point around the fortified hill Caesar aligshyned many forts for the garrisoning of his troops in a resshytricted circuit of C 20 km5 Nevertheless the mentioned examples reflect special needs in special times and there are no indications for the construction of more than one fortification except on campaign

This paper has been accomplished partially during the time I was awarded a scholarship by the Fritz Thyssen Foundation and Alexander von Humboldt-Foundation (Bonn) and part of it has been written at the Institute of Ancient History of the University of Cologne Germany Furthermore I wish to thank Prof loan PISO

and Dr C S Sommer for reading a draft of this paper and for some suggestion Naturally I alone am responsible for all flaws in the final version

Frontinus 12 2 For some examples see Redd6Schnurbein 2001 490-492

Frontinus 1 5 4 For the relation between literary sources and the archaeological

reality see ReddeJSchnurbein 2001 489-513

The association of troops was a principle of Roman strategy as well as under the Principato It is also well known how Masada was conquered but it should be remembered here that the Romans have built eight forts at the base of Masada rock and surrounded it with a high wall6 Probably the Romans did not even think there was a problem in associating military units when the need arose though at times they were aware of a certain danger This is clear when Tacitus remarks dushyring the Pannonian mutiny at the beginning of the first century AD that cum Pannonlcas legiones seditio incessit nullis nouzJis causis nisi quod mutatus princeps licentiam turbarum et ex ciuili bello spem praemioshyrum ostendebat Castris aestiuis tres simul legiones habebantur 7

But are they really conscious that the soldiers could be a menace It seems that at the beginning not quite so In the times of Germanicus the anger of the soldiers in leg I and XX would have been the result of Gods will tum (atalem increpans rabiem neque militum sed deum ira resurgere 8 Yet afterwards secuti exemplum ueterani haud multo post in Raetiam mittuntur specie de(endendae prouinciae ob imminentis Suebos ceterum ut auellerentur castris trttcibus adhuc non minus asperishytate remedii quam sceleris memoria9

Consequently it is obvious that there was no illusion regarding the power of more than one legion IO There are other illustrations of mutiny in the first century AD but a definite interdiction concerning the association of more than one legion was given only by Domitian after Saturninus revolt of AD 89 when geminari legionum castra prohibuitll Following there wont be more than one legion garrisoned in a single fort

6 Josephus Flavius Bellum Judaicum VII82

Tacitus Ann 116 8 Tacitus Ann 139 9 Tacitus Ann 144 10 Clearly Marius reforms of 107 BC nad been a beaKthrough for

the Roman army development and political danger of the soldiers will be reflected dUring the cvl wars of the first certGry BC

Caesar De bello Gailico 63-90 11 Suetonius Dom 3 5

134 I S I I l [J BUR G

GERMANIA SUPERIOR KM o shy 1r -0

Fig 1 1 Neckarburken 2 Miltenberg 3 Osterburken 4 Ohringen 5 Wezheim (based on Southern 1989 Fig 1

c

S 1 1 L [] BUR G I 135

During the first century until Domitians rule the literary sources are confirmed by archaeological evishydence and the examples of the comshybined legionary castra of Mainz or Vetera would suffice to make the point 12 Polybius proposed in refer-

to the Republican army that in any castra aestiva every legion has to be commanded by its tribunes and inside the fortification accomshymodation of the soldiers and horses have to be separate for each legion with a distance of fifty Roman feet between the tents of the tribunes or the tents of the cavalry of each legion 13 In the same time the conshysuls accommodation and the quaesshytorium 14 continue to be placed in the centre of the fortification 15 The reason was as expected purely technical The internal planning of the fortress or forts had to respond to the needs of the troops And what was the most important thing Obshyviously giving orders and avoiding confusion Consequently the orgashynization of the army inside the fort Concluding during the Republic or idea was to garrison the troops in one fortress in time of war or peace as there is no reason to do it separately Even the allies or auxiliaries were encamped theoretishycally jointly with legionaries Hyginus states that the combined garrison of a fortification should also include troops of nationes 16 Again we are dealing however with the army on campaign Evidently saving time and energy was the most important reason to build only one fort and just when the strategically needs justify it the construction of more than one fort would be the comshypromise

There is more than one eastra stativa garrisoned by a heterogeneous joint force of legionaries with auxiliaries

~

(~

Fig 2 Neekarburken (based on FiltzingerlPlaneklGimmerer 1976 Abb 215)

must be precise the Principat the

or auxiliaries with auxiliaries Additionally Rottweil17 Carnuntum 18 or Aquincum 19 are examples of forts placed in the neighbourhood of a fortress A combined garrison of auxiliaries is proposed where the area of an auxiliary fort extends to over 3-4 ha and no ala milliaria is attested Dacia displays three fortifications between 6-7 ha of Tibiscum Micia or Porolissum also with epigraphic evidence of a combined garrison2o For the other parts of the Empire there are other few examshyples at Vechten (c 45 hal Niederbieber (over 5 hal or Rottweil (c 6 hal where more than one unit is in garrison though there are uncertainties21 To my knowshyledge there are meagre evidences to the existence of two auxiliary neighbourhood forts garrisoned at one time by troops of the same status The exception is to be found

12 The fortress of Mainz was garrisoned by leg XVI Galica and liII Macedonica renamed after the Batavian revolt and dislocated to the east and at Burnum at first After the revolt the fortress was garrisoned by leg I Adiutrix and XIV Gemina At Vetera until the Batavlan revolt there were in garrison V Alaudae and XXI Rapax the last one replaced some time after by XV Primlgenia

13 This is necessary to give room for tribunes horses mules and baggage Polybius VI2728

14 For quaestorium see Hyginus sect 18 15 Polybius VI32 16 Hyginus sect 29 30

17 Conceivably Kastel I and Kasteli IVN are conternJca~ells

SornrnerKorurn 2005 18 Kardler 1997 19 Nemeth 990 677--678shy20 For a shor view over the forts and part of t~e )) ga~

reerences see Gudea 1997 32-34 37-39 46-49 -~ee 5

other fortifications in Dacia garrisoned only by -gt11 u~ ts t~eee with dimensions of c 3-4 ha as Booga 0 R~ 3 a a Porolissensis

BogaersROger 1974 62-65 ORL ~r la Pa~~~ L975 -13 k Rottweli there is still no evdence tlat he aested gar S5 ee contemporary Sommer 1992 272-282

136 I S 1 1 L [] BUR G

at Straubing where there were two contemporaneous forts at one time22 and possibly the forts of Ohringen

infra

GERMANIA SUPERIOR

Adjacent fons usually with different characteristics seem to be a pattern in Germania Superior There are not necessary too many details as the problem of this limes is relatively well documentedB yet there are not many excavations inside the fortifications and the chronology is still partially known24 Important here are those adshyjacent forts recognised at Neckarburken Walheim Milshy

Osterburken Ohringen and Welzheim

At Neckarburken (Fig 2) the fort of 1315 by 158 m (207 hal is probably garrisoned by coho III Aquintashynorum equitata transferred about the mid-second censhytury to Osterburken25 Eastward 200 m away from the fort there is a second one of only 80 by 80 m (064 hal of a more or less regular form with the eastern side a little shorter than the western one26 Three of the gates and the enclosure wall have been identified and inside has been uncovered the principia27 therefore one individual troop was in garrison This is numerus Brittonum Elantiensium28 present here for sure at least immediately after AD 14529 The unit is attested also in the nearby fortIet of Trienz30 Moreover it is known that the Brittones Elantienses restored the baths of the eastern fortlet of Neckarburken in AD 15831

A contemporaneous and similar situation seems to be encountered southwards at Walheim where two forts circa 300 m of each other were identified circa 150 m east of Neckar (Fig 31) One of the forts (Kastell I) of

Both forts are placed in the imediat vicinity and are contemposhyraneous starting with the Flavian emperors until Marcomannic wars when the western fort (Kastell IV) ceased to exist Czysz DietzFischerKellner 1995 518-521 Abb 220 The construction of two different forts is easy to understand because the erecting of the two forts was not perfectly simultaneous and the Romans probably considered unnecessary a joint garrison We should also not forget the early date of construction probably significant Also in Raetia it seems in an aerial photography that there are two forts at Theilenhofen but one of them is conSidered as a construction camp see Sommer 1988 544 Abb 27

23 For the problem of this sector of limes and also for the bblioshygraphy see Schonberger 1985 and Reuter 1999407-419

24 See for a short report KortOm 1998 34 25 FiltzingerPlanckCammerer 1976 425-427 26 FiltzingerPlanckCammerer 1976 Abb 215

Baatz 1973 124-127 FlltzingerPlanckCammerer 1976 427 28 For numeri Brittonum in general and their recruting history see

Southern 198994-98 Reuter 1999 21l 29 ell XIII 6490 30 Cil XIII 6498 31 Alfoldy 198367-69 Schallmayer 1984 451-455

134 by 156 m is probably occupied by coho I Asturum equitata because this is the garrison of the later fort of Mainhardt correspondent to Walheim on the outer limes of Antoninus Pius32 The turf and timber fort has been built at the end of 1st century or early in the 2nd

century The enclosure has been reconstructed of stone at one time and it was finally abandoned about AD 160)3 A numerus () fortlet (Kastell II) of 109 x 64 m was built this time only of turf and timber and has a rectangular plan34 The datable artefacts indicate a chronology starting under Trajan or early Hadrian reign until mid-second century35 The internal planning is not very traditional The interior is divided in two parts a relatively small praetentura and a retentura without any latera praetorii consequently without classical prinshycipia 36 This could be a reason to consider the garrishyson not as an independent unit but as a detachmentY Nevertheless the unit in garrison remains unknown but if there is a pattern regarding the existence of two neighshybourhood forts at least on this section of the limes than also in Walheim the second fortlet was garrisoned by a numeruss If this fort let was not intended for a numerus or for detachments it would be hard to explain the construction of two separate forts in the same time for two auxiliary units

At Miltenberg (Fig 31) there is a fort of 27 ha (Fig 32) where coh I Sequanorum et Rauracorum equitata is attested in mid-second century transferred here proshybably from Oberscheidenta139 At a distance of 2300 m from the first fort there has been identified another one of 06 ha (Fig 33) It is supposed that the fort was garrisoned by numerus exploratorum Seiopensium40

present here at least during the end of the second censhytury or in the beginning of the third jointly with exploshyratores Triputienses 41 There is no certain proof that the fortlet was built in the same time with the auxiliary fort

32 KortOmlauber 2004 34 33 KortumLauber 2004 34 34 Kortumlauber 2004 361 35 Until AD 150170 weredentlfied four phases of the interior

buldings KortGmlauber 2004 38 Abb 43 36 In the ntenor were excavated some barracks In the central part

of the fort probably a granary and other rectangular buildings with a clear plan of a bath in praetentura of the last phase KorshytOmlauber 2004 90-120

37 It is supposed on the basis of some fnds that at least part of regu ar units statlonedn the fortlet KortumLauber 2004 379

38 A pOSSible argument of the archaeOlogists in Walheim is that the special type of barrack Without arma has a good parallel in the numerus fort let of Hesselbach anc it is considered finally that this could be a special type of barrack intended for numeri KortOmlauber 2004 384f

Cl XIII 6604 Cil XIII 6597 AE 1977 593 Cil XIII 12460 3aatZi-errmann 1982 438

40 Cil XIII 6600 (7) 6605 41 Stell 1932 266f BaatzHerrmann 19824391

S II II L D BUR G I 137

0~ ~

~ Fort ~

~ ~~

m Hm +

IrJ I 1

0 SOm F9[j- ~ I 2 d

_I 3

Fig 3 1-3 Miltenberg based on BaatzlHerrmann 1982 Abb 402 403 405j

but it has been argued that it was occupied until the third century42 Nevertheless some new archaeological research shed new light on the history of the fortlet of Miltenberg It appears that sometime during the third century the garrison of the fortIet was drastically redushyced and probably the remaining soldiers occupied only the northern quarter of the fortlet 43 Anyway it is not certain if the soldiers belonged to the same unit

42 Baatz Herrmann 1982440 43 See JaeScholz 2002 416f The authors idea is that this is not

the only fortlet with a reduced garrison and there is another clear example at Kapersburg and the scenario is that many troops were reduced because they have sent soldiers for the conflicts with the Persians during the third century JaeScholz 2002 418-420

Eastward of Neckarburken on the new limes built during mid-second century44 other adjoining forts have been constructed such as those of Osterburken (Fig 41 The first fort of 214 ha was built during the reign of Antoninus PiUS45 This one was occupied by cob III Aquintanorum equitata present here also in the third century as it is confirmed by the epithets Seterimr~ and Philippiana 47 Under Com modus another forrifiashy

44 For the bibliographical referecces see A1 983 S~a shymayer 1984

45 AlfOldy 1987406-408 46 Cil XIII 6568 47 Cil XIII 6566

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

134 I S I I l [J BUR G

GERMANIA SUPERIOR KM o shy 1r -0

Fig 1 1 Neckarburken 2 Miltenberg 3 Osterburken 4 Ohringen 5 Wezheim (based on Southern 1989 Fig 1

c

S 1 1 L [] BUR G I 135

During the first century until Domitians rule the literary sources are confirmed by archaeological evishydence and the examples of the comshybined legionary castra of Mainz or Vetera would suffice to make the point 12 Polybius proposed in refer-

to the Republican army that in any castra aestiva every legion has to be commanded by its tribunes and inside the fortification accomshymodation of the soldiers and horses have to be separate for each legion with a distance of fifty Roman feet between the tents of the tribunes or the tents of the cavalry of each legion 13 In the same time the conshysuls accommodation and the quaesshytorium 14 continue to be placed in the centre of the fortification 15 The reason was as expected purely technical The internal planning of the fortress or forts had to respond to the needs of the troops And what was the most important thing Obshyviously giving orders and avoiding confusion Consequently the orgashynization of the army inside the fort Concluding during the Republic or idea was to garrison the troops in one fortress in time of war or peace as there is no reason to do it separately Even the allies or auxiliaries were encamped theoretishycally jointly with legionaries Hyginus states that the combined garrison of a fortification should also include troops of nationes 16 Again we are dealing however with the army on campaign Evidently saving time and energy was the most important reason to build only one fort and just when the strategically needs justify it the construction of more than one fort would be the comshypromise

There is more than one eastra stativa garrisoned by a heterogeneous joint force of legionaries with auxiliaries

~

(~

Fig 2 Neekarburken (based on FiltzingerlPlaneklGimmerer 1976 Abb 215)

must be precise the Principat the

or auxiliaries with auxiliaries Additionally Rottweil17 Carnuntum 18 or Aquincum 19 are examples of forts placed in the neighbourhood of a fortress A combined garrison of auxiliaries is proposed where the area of an auxiliary fort extends to over 3-4 ha and no ala milliaria is attested Dacia displays three fortifications between 6-7 ha of Tibiscum Micia or Porolissum also with epigraphic evidence of a combined garrison2o For the other parts of the Empire there are other few examshyples at Vechten (c 45 hal Niederbieber (over 5 hal or Rottweil (c 6 hal where more than one unit is in garrison though there are uncertainties21 To my knowshyledge there are meagre evidences to the existence of two auxiliary neighbourhood forts garrisoned at one time by troops of the same status The exception is to be found

12 The fortress of Mainz was garrisoned by leg XVI Galica and liII Macedonica renamed after the Batavian revolt and dislocated to the east and at Burnum at first After the revolt the fortress was garrisoned by leg I Adiutrix and XIV Gemina At Vetera until the Batavlan revolt there were in garrison V Alaudae and XXI Rapax the last one replaced some time after by XV Primlgenia

13 This is necessary to give room for tribunes horses mules and baggage Polybius VI2728

14 For quaestorium see Hyginus sect 18 15 Polybius VI32 16 Hyginus sect 29 30

17 Conceivably Kastel I and Kasteli IVN are conternJca~ells

SornrnerKorurn 2005 18 Kardler 1997 19 Nemeth 990 677--678shy20 For a shor view over the forts and part of t~e )) ga~

reerences see Gudea 1997 32-34 37-39 46-49 -~ee 5

other fortifications in Dacia garrisoned only by -gt11 u~ ts t~eee with dimensions of c 3-4 ha as Booga 0 R~ 3 a a Porolissensis

BogaersROger 1974 62-65 ORL ~r la Pa~~~ L975 -13 k Rottweli there is still no evdence tlat he aested gar S5 ee contemporary Sommer 1992 272-282

136 I S 1 1 L [] BUR G

at Straubing where there were two contemporaneous forts at one time22 and possibly the forts of Ohringen

infra

GERMANIA SUPERIOR

Adjacent fons usually with different characteristics seem to be a pattern in Germania Superior There are not necessary too many details as the problem of this limes is relatively well documentedB yet there are not many excavations inside the fortifications and the chronology is still partially known24 Important here are those adshyjacent forts recognised at Neckarburken Walheim Milshy

Osterburken Ohringen and Welzheim

At Neckarburken (Fig 2) the fort of 1315 by 158 m (207 hal is probably garrisoned by coho III Aquintashynorum equitata transferred about the mid-second censhytury to Osterburken25 Eastward 200 m away from the fort there is a second one of only 80 by 80 m (064 hal of a more or less regular form with the eastern side a little shorter than the western one26 Three of the gates and the enclosure wall have been identified and inside has been uncovered the principia27 therefore one individual troop was in garrison This is numerus Brittonum Elantiensium28 present here for sure at least immediately after AD 14529 The unit is attested also in the nearby fortIet of Trienz30 Moreover it is known that the Brittones Elantienses restored the baths of the eastern fortlet of Neckarburken in AD 15831

A contemporaneous and similar situation seems to be encountered southwards at Walheim where two forts circa 300 m of each other were identified circa 150 m east of Neckar (Fig 31) One of the forts (Kastell I) of

Both forts are placed in the imediat vicinity and are contemposhyraneous starting with the Flavian emperors until Marcomannic wars when the western fort (Kastell IV) ceased to exist Czysz DietzFischerKellner 1995 518-521 Abb 220 The construction of two different forts is easy to understand because the erecting of the two forts was not perfectly simultaneous and the Romans probably considered unnecessary a joint garrison We should also not forget the early date of construction probably significant Also in Raetia it seems in an aerial photography that there are two forts at Theilenhofen but one of them is conSidered as a construction camp see Sommer 1988 544 Abb 27

23 For the problem of this sector of limes and also for the bblioshygraphy see Schonberger 1985 and Reuter 1999407-419

24 See for a short report KortOm 1998 34 25 FiltzingerPlanckCammerer 1976 425-427 26 FiltzingerPlanckCammerer 1976 Abb 215

Baatz 1973 124-127 FlltzingerPlanckCammerer 1976 427 28 For numeri Brittonum in general and their recruting history see

Southern 198994-98 Reuter 1999 21l 29 ell XIII 6490 30 Cil XIII 6498 31 Alfoldy 198367-69 Schallmayer 1984 451-455

134 by 156 m is probably occupied by coho I Asturum equitata because this is the garrison of the later fort of Mainhardt correspondent to Walheim on the outer limes of Antoninus Pius32 The turf and timber fort has been built at the end of 1st century or early in the 2nd

century The enclosure has been reconstructed of stone at one time and it was finally abandoned about AD 160)3 A numerus () fortlet (Kastell II) of 109 x 64 m was built this time only of turf and timber and has a rectangular plan34 The datable artefacts indicate a chronology starting under Trajan or early Hadrian reign until mid-second century35 The internal planning is not very traditional The interior is divided in two parts a relatively small praetentura and a retentura without any latera praetorii consequently without classical prinshycipia 36 This could be a reason to consider the garrishyson not as an independent unit but as a detachmentY Nevertheless the unit in garrison remains unknown but if there is a pattern regarding the existence of two neighshybourhood forts at least on this section of the limes than also in Walheim the second fortlet was garrisoned by a numeruss If this fort let was not intended for a numerus or for detachments it would be hard to explain the construction of two separate forts in the same time for two auxiliary units

At Miltenberg (Fig 31) there is a fort of 27 ha (Fig 32) where coh I Sequanorum et Rauracorum equitata is attested in mid-second century transferred here proshybably from Oberscheidenta139 At a distance of 2300 m from the first fort there has been identified another one of 06 ha (Fig 33) It is supposed that the fort was garrisoned by numerus exploratorum Seiopensium40

present here at least during the end of the second censhytury or in the beginning of the third jointly with exploshyratores Triputienses 41 There is no certain proof that the fortlet was built in the same time with the auxiliary fort

32 KortOmlauber 2004 34 33 KortumLauber 2004 34 34 Kortumlauber 2004 361 35 Until AD 150170 weredentlfied four phases of the interior

buldings KortGmlauber 2004 38 Abb 43 36 In the ntenor were excavated some barracks In the central part

of the fort probably a granary and other rectangular buildings with a clear plan of a bath in praetentura of the last phase KorshytOmlauber 2004 90-120

37 It is supposed on the basis of some fnds that at least part of regu ar units statlonedn the fortlet KortumLauber 2004 379

38 A pOSSible argument of the archaeOlogists in Walheim is that the special type of barrack Without arma has a good parallel in the numerus fort let of Hesselbach anc it is considered finally that this could be a special type of barrack intended for numeri KortOmlauber 2004 384f

Cl XIII 6604 Cil XIII 6597 AE 1977 593 Cil XIII 12460 3aatZi-errmann 1982 438

40 Cil XIII 6600 (7) 6605 41 Stell 1932 266f BaatzHerrmann 19824391

S II II L D BUR G I 137

0~ ~

~ Fort ~

~ ~~

m Hm +

IrJ I 1

0 SOm F9[j- ~ I 2 d

_I 3

Fig 3 1-3 Miltenberg based on BaatzlHerrmann 1982 Abb 402 403 405j

but it has been argued that it was occupied until the third century42 Nevertheless some new archaeological research shed new light on the history of the fortlet of Miltenberg It appears that sometime during the third century the garrison of the fortIet was drastically redushyced and probably the remaining soldiers occupied only the northern quarter of the fortlet 43 Anyway it is not certain if the soldiers belonged to the same unit

42 Baatz Herrmann 1982440 43 See JaeScholz 2002 416f The authors idea is that this is not

the only fortlet with a reduced garrison and there is another clear example at Kapersburg and the scenario is that many troops were reduced because they have sent soldiers for the conflicts with the Persians during the third century JaeScholz 2002 418-420

Eastward of Neckarburken on the new limes built during mid-second century44 other adjoining forts have been constructed such as those of Osterburken (Fig 41 The first fort of 214 ha was built during the reign of Antoninus PiUS45 This one was occupied by cob III Aquintanorum equitata present here also in the third century as it is confirmed by the epithets Seterimr~ and Philippiana 47 Under Com modus another forrifiashy

44 For the bibliographical referecces see A1 983 S~a shymayer 1984

45 AlfOldy 1987406-408 46 Cil XIII 6568 47 Cil XIII 6566

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

c

S 1 1 L [] BUR G I 135

During the first century until Domitians rule the literary sources are confirmed by archaeological evishydence and the examples of the comshybined legionary castra of Mainz or Vetera would suffice to make the point 12 Polybius proposed in refer-

to the Republican army that in any castra aestiva every legion has to be commanded by its tribunes and inside the fortification accomshymodation of the soldiers and horses have to be separate for each legion with a distance of fifty Roman feet between the tents of the tribunes or the tents of the cavalry of each legion 13 In the same time the conshysuls accommodation and the quaesshytorium 14 continue to be placed in the centre of the fortification 15 The reason was as expected purely technical The internal planning of the fortress or forts had to respond to the needs of the troops And what was the most important thing Obshyviously giving orders and avoiding confusion Consequently the orgashynization of the army inside the fort Concluding during the Republic or idea was to garrison the troops in one fortress in time of war or peace as there is no reason to do it separately Even the allies or auxiliaries were encamped theoretishycally jointly with legionaries Hyginus states that the combined garrison of a fortification should also include troops of nationes 16 Again we are dealing however with the army on campaign Evidently saving time and energy was the most important reason to build only one fort and just when the strategically needs justify it the construction of more than one fort would be the comshypromise

There is more than one eastra stativa garrisoned by a heterogeneous joint force of legionaries with auxiliaries

~

(~

Fig 2 Neekarburken (based on FiltzingerlPlaneklGimmerer 1976 Abb 215)

must be precise the Principat the

or auxiliaries with auxiliaries Additionally Rottweil17 Carnuntum 18 or Aquincum 19 are examples of forts placed in the neighbourhood of a fortress A combined garrison of auxiliaries is proposed where the area of an auxiliary fort extends to over 3-4 ha and no ala milliaria is attested Dacia displays three fortifications between 6-7 ha of Tibiscum Micia or Porolissum also with epigraphic evidence of a combined garrison2o For the other parts of the Empire there are other few examshyples at Vechten (c 45 hal Niederbieber (over 5 hal or Rottweil (c 6 hal where more than one unit is in garrison though there are uncertainties21 To my knowshyledge there are meagre evidences to the existence of two auxiliary neighbourhood forts garrisoned at one time by troops of the same status The exception is to be found

12 The fortress of Mainz was garrisoned by leg XVI Galica and liII Macedonica renamed after the Batavian revolt and dislocated to the east and at Burnum at first After the revolt the fortress was garrisoned by leg I Adiutrix and XIV Gemina At Vetera until the Batavlan revolt there were in garrison V Alaudae and XXI Rapax the last one replaced some time after by XV Primlgenia

13 This is necessary to give room for tribunes horses mules and baggage Polybius VI2728

14 For quaestorium see Hyginus sect 18 15 Polybius VI32 16 Hyginus sect 29 30

17 Conceivably Kastel I and Kasteli IVN are conternJca~ells

SornrnerKorurn 2005 18 Kardler 1997 19 Nemeth 990 677--678shy20 For a shor view over the forts and part of t~e )) ga~

reerences see Gudea 1997 32-34 37-39 46-49 -~ee 5

other fortifications in Dacia garrisoned only by -gt11 u~ ts t~eee with dimensions of c 3-4 ha as Booga 0 R~ 3 a a Porolissensis

BogaersROger 1974 62-65 ORL ~r la Pa~~~ L975 -13 k Rottweli there is still no evdence tlat he aested gar S5 ee contemporary Sommer 1992 272-282

136 I S 1 1 L [] BUR G

at Straubing where there were two contemporaneous forts at one time22 and possibly the forts of Ohringen

infra

GERMANIA SUPERIOR

Adjacent fons usually with different characteristics seem to be a pattern in Germania Superior There are not necessary too many details as the problem of this limes is relatively well documentedB yet there are not many excavations inside the fortifications and the chronology is still partially known24 Important here are those adshyjacent forts recognised at Neckarburken Walheim Milshy

Osterburken Ohringen and Welzheim

At Neckarburken (Fig 2) the fort of 1315 by 158 m (207 hal is probably garrisoned by coho III Aquintashynorum equitata transferred about the mid-second censhytury to Osterburken25 Eastward 200 m away from the fort there is a second one of only 80 by 80 m (064 hal of a more or less regular form with the eastern side a little shorter than the western one26 Three of the gates and the enclosure wall have been identified and inside has been uncovered the principia27 therefore one individual troop was in garrison This is numerus Brittonum Elantiensium28 present here for sure at least immediately after AD 14529 The unit is attested also in the nearby fortIet of Trienz30 Moreover it is known that the Brittones Elantienses restored the baths of the eastern fortlet of Neckarburken in AD 15831

A contemporaneous and similar situation seems to be encountered southwards at Walheim where two forts circa 300 m of each other were identified circa 150 m east of Neckar (Fig 31) One of the forts (Kastell I) of

Both forts are placed in the imediat vicinity and are contemposhyraneous starting with the Flavian emperors until Marcomannic wars when the western fort (Kastell IV) ceased to exist Czysz DietzFischerKellner 1995 518-521 Abb 220 The construction of two different forts is easy to understand because the erecting of the two forts was not perfectly simultaneous and the Romans probably considered unnecessary a joint garrison We should also not forget the early date of construction probably significant Also in Raetia it seems in an aerial photography that there are two forts at Theilenhofen but one of them is conSidered as a construction camp see Sommer 1988 544 Abb 27

23 For the problem of this sector of limes and also for the bblioshygraphy see Schonberger 1985 and Reuter 1999407-419

24 See for a short report KortOm 1998 34 25 FiltzingerPlanckCammerer 1976 425-427 26 FiltzingerPlanckCammerer 1976 Abb 215

Baatz 1973 124-127 FlltzingerPlanckCammerer 1976 427 28 For numeri Brittonum in general and their recruting history see

Southern 198994-98 Reuter 1999 21l 29 ell XIII 6490 30 Cil XIII 6498 31 Alfoldy 198367-69 Schallmayer 1984 451-455

134 by 156 m is probably occupied by coho I Asturum equitata because this is the garrison of the later fort of Mainhardt correspondent to Walheim on the outer limes of Antoninus Pius32 The turf and timber fort has been built at the end of 1st century or early in the 2nd

century The enclosure has been reconstructed of stone at one time and it was finally abandoned about AD 160)3 A numerus () fortlet (Kastell II) of 109 x 64 m was built this time only of turf and timber and has a rectangular plan34 The datable artefacts indicate a chronology starting under Trajan or early Hadrian reign until mid-second century35 The internal planning is not very traditional The interior is divided in two parts a relatively small praetentura and a retentura without any latera praetorii consequently without classical prinshycipia 36 This could be a reason to consider the garrishyson not as an independent unit but as a detachmentY Nevertheless the unit in garrison remains unknown but if there is a pattern regarding the existence of two neighshybourhood forts at least on this section of the limes than also in Walheim the second fortlet was garrisoned by a numeruss If this fort let was not intended for a numerus or for detachments it would be hard to explain the construction of two separate forts in the same time for two auxiliary units

At Miltenberg (Fig 31) there is a fort of 27 ha (Fig 32) where coh I Sequanorum et Rauracorum equitata is attested in mid-second century transferred here proshybably from Oberscheidenta139 At a distance of 2300 m from the first fort there has been identified another one of 06 ha (Fig 33) It is supposed that the fort was garrisoned by numerus exploratorum Seiopensium40

present here at least during the end of the second censhytury or in the beginning of the third jointly with exploshyratores Triputienses 41 There is no certain proof that the fortlet was built in the same time with the auxiliary fort

32 KortOmlauber 2004 34 33 KortumLauber 2004 34 34 Kortumlauber 2004 361 35 Until AD 150170 weredentlfied four phases of the interior

buldings KortGmlauber 2004 38 Abb 43 36 In the ntenor were excavated some barracks In the central part

of the fort probably a granary and other rectangular buildings with a clear plan of a bath in praetentura of the last phase KorshytOmlauber 2004 90-120

37 It is supposed on the basis of some fnds that at least part of regu ar units statlonedn the fortlet KortumLauber 2004 379

38 A pOSSible argument of the archaeOlogists in Walheim is that the special type of barrack Without arma has a good parallel in the numerus fort let of Hesselbach anc it is considered finally that this could be a special type of barrack intended for numeri KortOmlauber 2004 384f

Cl XIII 6604 Cil XIII 6597 AE 1977 593 Cil XIII 12460 3aatZi-errmann 1982 438

40 Cil XIII 6600 (7) 6605 41 Stell 1932 266f BaatzHerrmann 19824391

S II II L D BUR G I 137

0~ ~

~ Fort ~

~ ~~

m Hm +

IrJ I 1

0 SOm F9[j- ~ I 2 d

_I 3

Fig 3 1-3 Miltenberg based on BaatzlHerrmann 1982 Abb 402 403 405j

but it has been argued that it was occupied until the third century42 Nevertheless some new archaeological research shed new light on the history of the fortlet of Miltenberg It appears that sometime during the third century the garrison of the fortIet was drastically redushyced and probably the remaining soldiers occupied only the northern quarter of the fortlet 43 Anyway it is not certain if the soldiers belonged to the same unit

42 Baatz Herrmann 1982440 43 See JaeScholz 2002 416f The authors idea is that this is not

the only fortlet with a reduced garrison and there is another clear example at Kapersburg and the scenario is that many troops were reduced because they have sent soldiers for the conflicts with the Persians during the third century JaeScholz 2002 418-420

Eastward of Neckarburken on the new limes built during mid-second century44 other adjoining forts have been constructed such as those of Osterburken (Fig 41 The first fort of 214 ha was built during the reign of Antoninus PiUS45 This one was occupied by cob III Aquintanorum equitata present here also in the third century as it is confirmed by the epithets Seterimr~ and Philippiana 47 Under Com modus another forrifiashy

44 For the bibliographical referecces see A1 983 S~a shymayer 1984

45 AlfOldy 1987406-408 46 Cil XIII 6568 47 Cil XIII 6566

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

136 I S 1 1 L [] BUR G

at Straubing where there were two contemporaneous forts at one time22 and possibly the forts of Ohringen

infra

GERMANIA SUPERIOR

Adjacent fons usually with different characteristics seem to be a pattern in Germania Superior There are not necessary too many details as the problem of this limes is relatively well documentedB yet there are not many excavations inside the fortifications and the chronology is still partially known24 Important here are those adshyjacent forts recognised at Neckarburken Walheim Milshy

Osterburken Ohringen and Welzheim

At Neckarburken (Fig 2) the fort of 1315 by 158 m (207 hal is probably garrisoned by coho III Aquintashynorum equitata transferred about the mid-second censhytury to Osterburken25 Eastward 200 m away from the fort there is a second one of only 80 by 80 m (064 hal of a more or less regular form with the eastern side a little shorter than the western one26 Three of the gates and the enclosure wall have been identified and inside has been uncovered the principia27 therefore one individual troop was in garrison This is numerus Brittonum Elantiensium28 present here for sure at least immediately after AD 14529 The unit is attested also in the nearby fortIet of Trienz30 Moreover it is known that the Brittones Elantienses restored the baths of the eastern fortlet of Neckarburken in AD 15831

A contemporaneous and similar situation seems to be encountered southwards at Walheim where two forts circa 300 m of each other were identified circa 150 m east of Neckar (Fig 31) One of the forts (Kastell I) of

Both forts are placed in the imediat vicinity and are contemposhyraneous starting with the Flavian emperors until Marcomannic wars when the western fort (Kastell IV) ceased to exist Czysz DietzFischerKellner 1995 518-521 Abb 220 The construction of two different forts is easy to understand because the erecting of the two forts was not perfectly simultaneous and the Romans probably considered unnecessary a joint garrison We should also not forget the early date of construction probably significant Also in Raetia it seems in an aerial photography that there are two forts at Theilenhofen but one of them is conSidered as a construction camp see Sommer 1988 544 Abb 27

23 For the problem of this sector of limes and also for the bblioshygraphy see Schonberger 1985 and Reuter 1999407-419

24 See for a short report KortOm 1998 34 25 FiltzingerPlanckCammerer 1976 425-427 26 FiltzingerPlanckCammerer 1976 Abb 215

Baatz 1973 124-127 FlltzingerPlanckCammerer 1976 427 28 For numeri Brittonum in general and their recruting history see

Southern 198994-98 Reuter 1999 21l 29 ell XIII 6490 30 Cil XIII 6498 31 Alfoldy 198367-69 Schallmayer 1984 451-455

134 by 156 m is probably occupied by coho I Asturum equitata because this is the garrison of the later fort of Mainhardt correspondent to Walheim on the outer limes of Antoninus Pius32 The turf and timber fort has been built at the end of 1st century or early in the 2nd

century The enclosure has been reconstructed of stone at one time and it was finally abandoned about AD 160)3 A numerus () fortlet (Kastell II) of 109 x 64 m was built this time only of turf and timber and has a rectangular plan34 The datable artefacts indicate a chronology starting under Trajan or early Hadrian reign until mid-second century35 The internal planning is not very traditional The interior is divided in two parts a relatively small praetentura and a retentura without any latera praetorii consequently without classical prinshycipia 36 This could be a reason to consider the garrishyson not as an independent unit but as a detachmentY Nevertheless the unit in garrison remains unknown but if there is a pattern regarding the existence of two neighshybourhood forts at least on this section of the limes than also in Walheim the second fortlet was garrisoned by a numeruss If this fort let was not intended for a numerus or for detachments it would be hard to explain the construction of two separate forts in the same time for two auxiliary units

At Miltenberg (Fig 31) there is a fort of 27 ha (Fig 32) where coh I Sequanorum et Rauracorum equitata is attested in mid-second century transferred here proshybably from Oberscheidenta139 At a distance of 2300 m from the first fort there has been identified another one of 06 ha (Fig 33) It is supposed that the fort was garrisoned by numerus exploratorum Seiopensium40

present here at least during the end of the second censhytury or in the beginning of the third jointly with exploshyratores Triputienses 41 There is no certain proof that the fortlet was built in the same time with the auxiliary fort

32 KortOmlauber 2004 34 33 KortumLauber 2004 34 34 Kortumlauber 2004 361 35 Until AD 150170 weredentlfied four phases of the interior

buldings KortGmlauber 2004 38 Abb 43 36 In the ntenor were excavated some barracks In the central part

of the fort probably a granary and other rectangular buildings with a clear plan of a bath in praetentura of the last phase KorshytOmlauber 2004 90-120

37 It is supposed on the basis of some fnds that at least part of regu ar units statlonedn the fortlet KortumLauber 2004 379

38 A pOSSible argument of the archaeOlogists in Walheim is that the special type of barrack Without arma has a good parallel in the numerus fort let of Hesselbach anc it is considered finally that this could be a special type of barrack intended for numeri KortOmlauber 2004 384f

Cl XIII 6604 Cil XIII 6597 AE 1977 593 Cil XIII 12460 3aatZi-errmann 1982 438

40 Cil XIII 6600 (7) 6605 41 Stell 1932 266f BaatzHerrmann 19824391

S II II L D BUR G I 137

0~ ~

~ Fort ~

~ ~~

m Hm +

IrJ I 1

0 SOm F9[j- ~ I 2 d

_I 3

Fig 3 1-3 Miltenberg based on BaatzlHerrmann 1982 Abb 402 403 405j

but it has been argued that it was occupied until the third century42 Nevertheless some new archaeological research shed new light on the history of the fortlet of Miltenberg It appears that sometime during the third century the garrison of the fortIet was drastically redushyced and probably the remaining soldiers occupied only the northern quarter of the fortlet 43 Anyway it is not certain if the soldiers belonged to the same unit

42 Baatz Herrmann 1982440 43 See JaeScholz 2002 416f The authors idea is that this is not

the only fortlet with a reduced garrison and there is another clear example at Kapersburg and the scenario is that many troops were reduced because they have sent soldiers for the conflicts with the Persians during the third century JaeScholz 2002 418-420

Eastward of Neckarburken on the new limes built during mid-second century44 other adjoining forts have been constructed such as those of Osterburken (Fig 41 The first fort of 214 ha was built during the reign of Antoninus PiUS45 This one was occupied by cob III Aquintanorum equitata present here also in the third century as it is confirmed by the epithets Seterimr~ and Philippiana 47 Under Com modus another forrifiashy

44 For the bibliographical referecces see A1 983 S~a shymayer 1984

45 AlfOldy 1987406-408 46 Cil XIII 6568 47 Cil XIII 6566

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

S II II L D BUR G I 137

0~ ~

~ Fort ~

~ ~~

m Hm +

IrJ I 1

0 SOm F9[j- ~ I 2 d

_I 3

Fig 3 1-3 Miltenberg based on BaatzlHerrmann 1982 Abb 402 403 405j

but it has been argued that it was occupied until the third century42 Nevertheless some new archaeological research shed new light on the history of the fortlet of Miltenberg It appears that sometime during the third century the garrison of the fortIet was drastically redushyced and probably the remaining soldiers occupied only the northern quarter of the fortlet 43 Anyway it is not certain if the soldiers belonged to the same unit

42 Baatz Herrmann 1982440 43 See JaeScholz 2002 416f The authors idea is that this is not

the only fortlet with a reduced garrison and there is another clear example at Kapersburg and the scenario is that many troops were reduced because they have sent soldiers for the conflicts with the Persians during the third century JaeScholz 2002 418-420

Eastward of Neckarburken on the new limes built during mid-second century44 other adjoining forts have been constructed such as those of Osterburken (Fig 41 The first fort of 214 ha was built during the reign of Antoninus PiUS45 This one was occupied by cob III Aquintanorum equitata present here also in the third century as it is confirmed by the epithets Seterimr~ and Philippiana 47 Under Com modus another forrifiashy

44 For the bibliographical referecces see A1 983 S~a shymayer 1984

45 AlfOldy 1987406-408 46 Cil XIII 6568 47 Cil XIII 6566

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

138 I 5 1 1 L [] BUR G

r-nT

i

-~~ ~

J IfIt

a O~~middot t~~lcJ~~ bull lt1 --~

oOm

Fig 4 Osterburken

tion of irregular plan with an enclosure having gates on three sides following more or less the edge of the slope has been attached to the south-eastern side of the first fort

Brittones Elantienses are transferred also to Oster burshyken probably during the reign of Commodus48 and it is supposed that they garrisoned the newly attached fort Nevertheless the area occupied by the new fort is of 135 ha more than double than the fortlet of NeckarburkenshyOst The reason for this is the steep slope where the fort has been built hypothetically a large part of the interior should have remained unoccupied49 Another possibility is that other detachments to be in garrison which is of leg VIII Augusta attested here on inscriptions where it is confirmed that the legionaries have built something here probably the fort itself50 However this is not sufshyficient to believe the presence of legionary detachments but the existence of three baths close to each otherSl

suggests also troops in garrison Two neighbouring forts were built at Ohringen (Fig

5) The first one Ohringen-Biirgkastell of c 1545 by 141 m (218 ha)52 has been garrisoned at first by coho I Helvetiorum 53 The second fort Ohringen-Rendelkastell

48 Alfoldy 1987408 49 FiltzingerlPlanckiCammerer 1976 446 50 FiltzlngerlPlanckCammerer 1976 446 51 Kortum 2004 52 Information by C S Sommer See also Schbnberger 1973 53 AE 1987 148 AE 1987 784 a b Between AD 222-235 coh

Septima Belgarum (Cll XIII 11758--9) has been probably budt the fort in its third phase FiltzingerPlanckiCammerer 1976437 his is confirmed by the inscriptions discovered at Ohnngen where it has the epithets Alexandriana (Cll XIII 11758) or Gordiana (Cll XIII 11759)

of c 130 by 150 (1 95 hal was built later than the first 54 To mv knowledge there is no building archaeoloshygically investigated inside the fort therefore it would be useless to report anything precise concerning the forts chronologys eyertheless two tile-stamps indicate the presence of nlumerus) Brittonum) Cal )56 and an altar of the Ilumlerus) B(rittonum) A1urrensium S7bull The name of another unit Britt(onum) Aure(lianienses) is to

be found on two dedication slabs of AD 178 along with coho I Helvetiorum 58 Also the name of numerus Aurelianensium is abbreviated on other tile-stamps and on another inscription from Falerii 59 Regardless it was proposed in the two forts a garrison of auxiliaries6o It seems that this judgment is in direct relation only with the area occupied by the forts yet the presence of more than one unit of mlfflert is obvious 61 Consequently E Stein has proposed that the Ohringen-Biirgkastell was garrisoned by numerus Brittonum Aurelianensium also because here it is mentioned a centurion of leg VIII Augusta62 On the other hand H Schonberger belieshyves that the eastern fort (Rendelkastell) was garrisoned from the beginning by coho 1 Helvetiorum and the wesshytern fort (Biirgkastell) was garrisoned during the third century by coho 1 Septima Belgarum63 However after Septimius Severus there is almost no information neither for the deployment of the numeri of Ohringen nor for

54 FltzingerPlanckiCammerer 1976 438 Some of information are from C S Sommer

~ow there are some nformatlofls about area of the headshyquarter building of the West~astel I am indebted to C S Somshy

for the results 0 some unpublished aralyses inside some fores mentioned Jere

Cl Xli 12498

Cil XI 6471 58 Cil XII i 6542f E Stein proposed that there was only one numerus

here fomed from an amalgamation of n(umerus Brit(tonum) Cal() and num(erus) Blnttonum) M(urrensium) Stein 1932 247 Because the abbreViated form of coho I Helveliorum appear along witn he name of the n(umerus) Brit(tonum) Call ) also on tiles Cl Schbrberger proposed rightly that both troops used a common brickyard Schonberger 1973 293

Cil XIII 12497 1 Cl XI 3104 60 One Of tle concluslols is tnat coh I Helvetiorum will be the

garrson of during the third century FiltzngerPancklCamerer 1976 437f In any case there is ro archaeological proof that It was so

61 Probably 1ere It is also present a part of exploratio Seiopensis from Mdtenberg-Ost Stein 1932 248f

CII_ XII 11757 Stein 1932 250 act that the unit could lave been comnrandec by a certuriof1 is demonstrated by the Inscnptol from FaerH (Cl XI 3140) where its mentioned a praeposlto numen Aurelianensis ald Brittonum as it is the one of tile sae egior mentionec on the Inscnptlon of Welzheim-Ost (Cll XIII 6526)

63 a~thor qLestIOf1ng aso the possibility that coho I Helveshy[lorum m be replaced by Bnttones Aurelianenses Schbnberger 1973 294f

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

S 1 1 L n BUR G I 139

the precise chronology of Ohringen-Rendelkastell yet there are signs of occupation of both forts even during the third century64

At Welzheim at the southern end of a 81 km long and straight stretch of frontier were identified other two

11 Westernfort (Biirgkastell)

LJ 100 200 300 400m

Fig 5 Ohringen (based on Schonberger 1973 Abb 2)

~

~Om

rCJ----~ Westemfort

IL~I bd

Fig 6 Wezheim (based on FiltzingeriPlancklGimmerer 1976 Abb 228 332)

64 SchOnberger 1973 296

forts located in the vicinity of each other at a distance of c 530 m (Fig 6) The western fort has been garrishysoned by ala I Scubulorum and it has occupied an area of 236 by 181 m (43 ha)65 Concerning the interior we have information only about principia parts of the

enclosure and two wells close to the intervallum street66 A straight road made the connection between the two forts The eastern one has an area of 123 bv 130(N)1136(S) (163 hal Inshyside the fort has been uncovered a small bath and perhaps a horreumP A votive inscription dedicated to Jushypiter was discovered in the bath and it names a centurion of leg VIII Aushygusta as praepositus Brit(tonum) et expl(oratorwn)68 Plus there were discovered several tile-stamps of nushymerus Brittonum L69 and three of numerus Brittonmn Cr () or GrJo Accordingly even if the area of the eastern fort is theoretically too large for a numerus fort the garrison here it is unquestionably formed by units of Brittones and exploratores

DACIA

In Dacia Inferior on the so-called limes TransAlutanus four out of the thirteen fortsfort lets are double that is to say there are two fortifications in close proximity of each other at Gimpulung-Jidava Sapata de Jos Urshyluieni and Baneasa In the same province west of the limes TransAlutanus on the Olt line adjacent forts at Racovita-Copikeni Romula and westward at Bumbe~ti are documented Other strongholds of the same type are known from archaeological studies on the westernmost route in Dacia from Lederata to Sarmizegetusa at Greshybenac and Tibiscum Finally the lasts adjacent forts are indicated on the northern frontier of Dacia Porolissensis at Porolissum and Ca~eiu Here are not mentioned the adjacent fortifications from the Ora~tie mountains beshycause they are castra aestiva -[ or the forts of Variidia

65 FltzirgerPanckiCamrnerer 1976 559 66 Planck 1979

FiltzlngerlPlanckiCamrnerer 1976 560 It is 0 5~e f~ara -e bath was bUilt when the fort was sI1 Il Cr- aa e abandolrnent of the fort It is wobable a -e lt ri N2

dscoverea In re south-west corner of tre - 8 ae ar

penoo (AD 190 c 230) when there ~ rae ee- - garrison inside the fort Oriel-Murray Ha-~ilrr 99 6

68 CI XIII 6526 69 CIL XIII l2500

70 Cil XIII 12499 71 OaicovlcluFerenczlGlodaru 1989217-223

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

o

~ -l t1 lI

~

10 j) YJ o~pilyen1lt1I figto ~

Vl

gt gt

J CD

C

7J

C

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

because one of them existed only during the wars of the early second century before the foundation of a real province 72

The Roman forts at Campulung-Jidava (Fig 8) lay on the limes TransAlutanus close to Bran pass on the road linking Transylvania to Wallachia During the arshychaeological excavations initiated in the nineteenth censhytury there were identified two forts of 13235 by 9865 m (13 hal and 50 by 60 m (03 hal built 200-300 m apart The first fort was erected in stone combined with brick Within latera praetorii excavators identified principia a building with an apse and a horreum In retentura were detected portions of a barrack Regarshyding the chronology of this fort scholars argued for a possible Hadrianic or late second century foundation date but there is no definite archaeological proof73

Nevertheless the stamped tiles of leg Xl Claudia could indicate the existence of the fort in Trajans reign as 1 Bogdan-Cataniciu pointed outJ4 Other brick stamps and triangular cross-section arrowheads deposits attesshyting coho I Flavia Commagenorum sagittariorum were found precisely in the third century contexts5

The twentieth century archaeological excavations from the second fort at Campulung-Jidava situated roughly 250 m southwards yielded many finds but no interior building was identified6 However the gates of the forts also built of brick and in the same consshytruction method like the enclosure of the northern fort suggest the same construction date for both forts There is no conclusive evidence for one troop or another in garrison but considering the size of the fortlet (50 by 60 m) one can suppose that it was occupied by a nushymerus7

Southward at Sapata de Jos on the same frontier there are two other forts positioned 35 m apart (Fig 9) The largest fort with an enclosure of brick sill-wall and a timber superstructure is C 125 by 90 m (112 hal and the other one built only of earth and timber measured 35 by 45 m (015 ha)J8 Some argued that these forts functioned simultaneously because there appears to be a joint surrounding ditch79 Additionally we know that

72 The two forts of Vanidia are located on the hill Chilii and in the valley of Cara~ at Pusta The first fort has been probably a temporary camp as Its position and plan suggest Information by E Nemeth

73 Tudor 1936 115 Bogdan-Cataniciu 1997 44 Gudea 1997 80

74 Bogdan-Cataniciu 1997 44f

75 Petolescu 2002 96

76 Unfortunately there is still missing a report concerning the finds and the archaeological excavation

77 See Gudea 1997 80

78 Christescu 1936435-447

79 Bogdan-Cataniciu 1997 95

S A A LOB U R G I 141

the building material for both fortifications is brick80

The dimensions of the forts are very small and for that matter it is most unlikely that complete units occupied these forts The only arguments to establish a chronoshylogy are two monetary hoards discovered inside and in the precincts of the largest fort The coins of the first hoard range from AD 205-248 and of the second one

p==w===a~ I ~

middot rk r-IT] ~

II

J I l gt- tilLt JId =-=~1- bull

iIIJ C bullN ___L--ccA

~ il II I ~) ~~~~

50 ------=shy

Fig 8 Gimpulung-]idava (based on Bogdan-Cataniciu 1997)

80 Stone and probably timoer were scarce In thiS part of Dacia Inferior but good quality clay to make bricks is at hand

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

142 I 5 1 1 L [1 BUR G

from Trajan to Valentinians reign Other scant findings from the forts are some coins issued starting with Comshymodus reign Except for the numismatic evidence there is a brooch with the inscription ROMA dated in the late second or early third century 8 1

On the same frontier the 30 m apart forts of Urluieni located 30 km southward from Sapata de Jos where is a junction of two main roads north-south and eastshywest measured 123 by 104 m (120 hal and 112 by 85 m (095 hal respectively (Fig 10)82 The rampart of the first mentioned fort is built with of a cobblestone layer and a brick sill-wall above in the same manner as the largest fort of Sapata de Jos The only known building within the fort are the principia 1 Bogdan-Cataniciu

~~~~~ ~ ~~ ~=~ ~= ~---- -- -- ------- 1 A 7

I I I I I I I I

_-----shy

tD--=======~~

I II I

IIII Ir

1I

I II I I I

I II I II i I II I I II I 1 I

---shy

_~--I

J

--------------~~

-==J___ _o 100m

Fig 9 Siipata de Jos (based on Gudea 1997)

81 Christescu 1936 445f Fig 132 82 Bogdan-Cataniciu 199780

who carried out the excavations on the site argued that this fort vas constructed at the beginning of the second centuryP The support materials for the garrisoning of this fort with archers from eastern Mediterranean area are three arrowheads with triangular section and a bow ear lath Unfortunately it was not possible to establish the chronological relationship between the two forts because scant attention was paid to the southeastern fortler The only datable findings found in the intershyuallmn area of this fortlet are a coin from the reign of Elagabalus and a vheel-made lamp from the third censhytury as well hence the dating in the second century is by no means certain

On the limes TransAlutanus the most southern neighshybouring placed forts of Baneasa (Fig 11) lie 50 m beshyhind the frontier The position is a strategic one again and the forts kept under surveillance the Calmatui valshyley the access road to Oltenia The nineteenth century archaeological excavations involve a few trial trenshyches 84 The construction of the 126 by 130 m (163 hal fort is argued to have taken place on numismatic basis sometimes during mid-second century This fort is one of those few forts on limes TransAlutanus which could have been after its dimensions garrisoned by a complete cohors quingenaria Interestingly the largest fort was divided by an earthen rampart into two almost equal parts 150 m towards north-east there is a 45 by 63 m (028 hal fortlet but there is no evidence about it

It is difficult to assess anything certain on the chronoshylogy of forts on the limes TransAlutanus Only regarshyding the neighbouring forts of Sapata de Jos and Jidava we know with probability that were at one moment contemporaneolls yet there is meagre evidencess The attribution of the small camps from the limes TransAlutanus to numeri units was made only on dimenshysional basis although it is well known that this is not always an applicable criteria Other arguments for the identification of a numerus emerged only at one of the forts of Urluieni where some samples of east Mediterrashynean weaponry were identified On one hand these weashypons could have been used by detachments of coho I Flauia Commagenorum sagittariorum garrisoned proshybably nearby at Gimpulung-Jidava and on the other

83 199798 84 Cantacullno 1944

I Bogoan-Catanciu conSiders that the existence of tre adjacent fois confirms that the limes TransAlutanus was used in more than perod 1997 95 In the case of the adjaCent forts ~e dmenslors one of the forts bel1g sometimes smaller or 1uch s1aller than he other one and the strategical positon could rdlcate the act that ~le situator In Dacia Irerior IS al1051 cenlica Wltl tle one II Germanla Supenor that is to say the forts couio have beer contemporareous at one 11omenl

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

S 1 1 LOB U R G I 143

I

t N

I

middot I

D D

~ ~~ ~ --_ ~

ffn 191~~11 II II I IiI I I iI 1II

W 50

---~-~-

Fig 10 Urluieni (based on Bogdan- Cataniciu 1997)

hand by Suri sagittarii attested in the province Dacia Inferior as well 86

The road along the river Olt was guarded around the Cozia massif by two forts at Copaceni and Racovita (Fig 12) The western half of Copaceni forder was

86 The only numeri units attested in Dacia Inferior are Suri sagitam which became probably numerus Surorum sagittariorum and numerus equitum IIlyricorum or numerus burgariorum et vere~

Fig 11 Baneasa (based on Gudea 1997)

flooded and destroyed by the air therefore it has only one completely preserved side measuring c 64 m This fortification was erected under Hadrian as it is attested by the inscription of AD 138 where numerus Burgashyriorum et Veredariorum is mentioned when they rebuilt

dariorum Anyway the last two units are not probaoy of nationes

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

144 I 5 L rJ BUR G

t

50m ~

Fig 12 Copaceni-RacoviJa (based on Gudea 1997)

Fig 13 Romula (based on Gudea 1997)

the forr after two years from the construction date s7

The evidences showing that this fortIet functioned in the third century too are the coins founded within the fort which range from Antoninus Pius to Gordian and a milestone dated under Maximinus Thrax discovered in one tower of the only gate preserved88

The fort measuring 11240 by 101 m (113 hal at Racovita lays 500 m north of the Copikeni fortlet The only buildings we know about are those from lashytera praetorii namely the headquarters and a granary without having any clue about the building sequence of the fort Because the gate towers are rectangular and have a slight exterior projection like other forts on the Olr line some scholars considered the construction date of the forr as the first half of the second century89 This could not be true if we take a look only in the same area at the gate towers from the forts of Bumbe~ti for instance also rectangular and even without any exterior projection although the date of erection is for sure AD 201 (Fig 14)90 Other scholars assumed that the fort has been garrisoned by the same numerus Burgariorum et Veredariorum because the Copaceni fortlet became obsolete 91 This is difficult to believe as we have already seen that the Copaceni fort was still functioning in late second century Anyway the Racovita fort is almost double in size than the Copaceni fortlet whence imposshysible to have been occupied by exactly the same unit

Elsewhere on the same limes Alutanus some scholars emphasize the existence in the hinterland of the urban settlement at Romula-Reljca of other two forts (Fig 13) The existence of these forts was inferred on the basis of P Polonics sketch from the early XXth century But it was not confirmed by any archaeological survey92 Since so little is known about these forts I Bogdan-Cihiiniciu doubts whether the presence of some military troops and subsequently the existence of the forrs is real Moshyreover the same author argues that the ramparts of both forts are actually the citys enclosure rebuilt under Philip the Arab93 The units attested on stamped tiles or bricks are leg XI Claudia V Macedonica coho I Flavia

87 Cil ill 13796 IlS 9180 88 Tudor 1982 76 89 See with the bibliography Gudea 199793 90 ell III 14485A IDR II 174 9 Tudor 1981 81 85 92 Tudor 1978 194 Gt-dea 1997 85

93 Bogdar-Catariciu 1997 64f There arent any archaeological excavatons 11 the area where the forts are attested therefore is flO eason to doubt tre topographists P Poionlc draft whose piars ard s~etches from the late nineteenth and early twentieth centuries are usually very pecise

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

S A A L D BUR G I 145

~j

Fig 14 Bumbe~ti (based on Gudea 1997)

Commagenorum and numerus Syrorum 94 At Numerus Syrorum (Mauretania Tingitana) on some inscriptions of the third century is recorded formula domus ROnlula a very important argument for the troops dislocation from Dacia to Mauretania under Septimus Severus95

The transfer would have been possible in the same time with the movement of Sex Iulius Iuljanus tribun of n(umerus) Syrorum M(a)lvensium96 at the command of 1000 Thracian recruits from Tracia to Mauretania97

94 The unit appear as numerus on an inscription from RomJla (ell III 8032) and on a dedication to Sol Invictus (AE 12605) On another inscription it is named n(umerus) Syrorum M(a)lvensium (ell VIII 9381 20945 ILS 2763) the last being the name of the Roman Romula (Malva)

95 Speidel 1973 171 n 24 with the bibliography This possiblity is questIOned by H Wolff and C C Petolescu who considers that the unit from Dacia should be different from the one of Mauretania see Wolff 1975 139ff Petolescu 198344 H Wolff suspected that the formula domus Romula used only at Numerus Syrorum is not a proff for the presence of that numerus In Mauretania and the term sagittariorum IS an annex only of the troops name of Dacia thus the units are different Nevertheless on one of the funerary stones at Numerus Syrorum It is recorded Quadrati Suri Sagittari Speidel 1977 n 1 C C Petolescu conshysiders that on the inSCription at Sevilla IS mentioned the name Malvensis for distinguishmg betweer the homonym units of numeri see Petolescu 2002 144 This could be quite true in the case of two units with the same name garrisoned in the same province but it would be hard to believe such an argume1t for units of two different provinces situated miles away Anyhow the use of the term Malvensis attached to the troops name it would not be necessary when the units has moved to another place and it is explained by M P Speidel as a proof of an earlier date of inscription and later the term being probably dropped Spedel 1973 172

9euro Cil VIII 9381 20945 IlS 2763

97 Rowell 19362554

or more probable the numerus has been already preshysent there when the Thracians arrived98 The presence of military units at Romula it is clearly a fact also if we have a look at the career of Sex Iulius Possessor at one moment praepositus of numerus Syrorum Sagittashyriorum and in the same time of ala prima Hispanorum and curator in Romula (elL II 1180) consequently it is reasonable to accept the existence at Romula of some fortifications for garrisoning the troops for the units garrisoned inside a civil settlement it is a reality only in the Eastern provinces

Guarding the entrance from south to the Jiu gorge on the road leading from Oltenia to Transylvania at Bumbesectti (Fig 14) there were identified two other forts 800 m apart 99 The western part of one fort is deshystroyed by river Jiu and the only complete side measures 167 m The excavators assumed on tile-stamp basis that the fort was garrisoned by coho IIII Cypria and deshytachments of leg V Macedonica or VII Claudia lOll yet at least for the presence of legionaries the argument is not sufficient Additionally it is known fot sure that the rampart was rebuilt in stone at the very beginning of the third century by coh I Aurelia Brittonum miliaria lUI

Inside the fort there are some excavated buildings but

98 Spedel 1973 172 Spece 1977172f

99 Tudor 1978 269 309

100 Marlnolu 2003 57 6Of

101 ell III 14485A lOR II 174

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

146 I S 1 L [] BUR G

I

50 rrgt ~~t-=-

Fig 15 Grebetae (based on Gudea 1997)

we dont know their function The second fort in the area is at Viirtop and measures 115 by 126 m (145 hal being also very disturbed The dating evidence within the fort some coins which range from Trajan to Commodus suggests a construction date during the second century102 This would agree with the discovery of the only tile-stamp with the symbol of leg llll Flavia Felix 103

The chronology of the forts is not fully comprehenshysive and it is also possible that the forts are not built sishymultaneously Yet regarding the tile-stamps of the three legions it is probable an early date for both of them but without knowing the building sequence

On the most important route of Dacia in the wesshyternmost limit of the province at Grebenac (Fig 15) there are other adjoining forts almost identical in plan and dimensions with the one of Urluieni (Fig 10) menshytioned above The situation at Grebenac is not properly understood as there was no archaeological excavation thus we know only the forts dimensions of 110 by 130 m (143 hal and 60 by 110 (066 hal respectivelylo4

102 Marinoiu Hortopan 2003 36-40 It is not quite sure tnat there IS

not a civilian settlement in that area developed on the place oT an earlier fort

103 MarinoiuHortopan 2003 36

104 Gudea 1997 25f

Very provocative is the situation at Jupa-Tibiscum (Fig 16) vhere a total number of five forts was idenshytified although not all contemporaneous Initially proshybably between the two wars of the early second century there had been a fortlet 60 by 60 m (036 hal then in the same area it was another fort 110 by 101 m (111 hal of earth and timber at the beginning then rebuilt in stone The last general report of the excavation carried out by D Benea and P Bona taking into consideration the dating elements concludes that the last fort was built under Trajan and rebuilt under Hadrianlos In the 1990s at a distance of 15 m south from the stone forts southern side E ~emeth identified a part of another enclosure of a new fort which lay in the south-western part of the already known fortification lOb The archaeshyologists argued that this rampart is identical to another one discovered in 80 under the western part of the largest enclosure built in mid-second centuryIll7 The military units garrisoning these two fortifications were supposed to be coho I sagittariorum attested here by a tile-stamp detachments of leg III Flavia Felix and XIII Gemina and starting with Hadrians reign Palmyshyrenii sagittarii Vhatsoever the presence of legionaries detachments as garrison is improbable and maybe they only have sent building material or took part in a joint building construction team The stone fort has been atshytributed to coho I sagittarioruln and the new discovered one to the P1lm)renii sagittarii I08 However considering the size of the known fort of just 11 hectare it would have been difficult to garrison here a troop which is milshyliaria at least in the mid-second century In any event even if the cohort was quingenaria the fort is too small whence it is probable that this stone fort was occupied by the numerus and the other one by that cohort The two forts were replaced by a larger fort 195 by 310 m (604 hal built perhaps after mid-second century

Even more peculiar evidence is to be found in the complex archaeological situation of the Tibiscum site namely the discovery of a new fortification nearby on the other bank of the river the dimension from the main fort being about 600 m towards the east The fort had a short life since sometime in the second century it was sealed by a civil settlement The excavations are very scarce on this site therefore little is to prove or disprove the identity of the garrisoning troop for instance109 The fort of 60 by 90 m (054 hal if there is one could have been garrisoned by ~1oors brought at Tibiscum probably under Antoninus Pius 110 When the largest

BelteaBora 199432 36

106 I arD grateful to Dr E Nernetr wfjo explalrs to me the archaeoshyogcal condtlons of the dscovery

BeneaBona 199437

BenealBora 1994 37

Arde~iArde~ 2004 38--55

ihe Moers are atested at Toisclm as numerus Maurorum TioiscenSlurn (CL Iii 1343 Cll III 1295) and in Dacia in the dcora 0 AD 158 (CIL XVi 108 lOR I 16)

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

fort 600 m away was built maybe late in the second century Mauri equites could have been accommodated here together with cohors I sagittariorum or cohors 1 Vindelicorum 111 and with Palmyrene archers

I

o 20 40 6C 80 10CO)M

======jij) Q========------------------shyI I

------------------shy

S L [1 8 U R G I 147

Late in the 1990s in the close proximity of Ca~ei (Fig 17) fort on the northern frontier of Dacia were identified under the civilian settlement parts of two ditches Being the archaeologist of the site D Isac conshysiders that this second enclosure belongs to a fort with a short existence but contemporary at one moment with the southern fort occupied at the beginning by coho n Britannorum milliaria ll2 The area where this short-lived fort was built is liable to flooding and it was in antiquity as well therefore today the conditions of excavation became more difficult The garrison of the newly discovered fortification is thought to be coho I Brittonum milliaria which is known later as the garshyrison of the southern fort l13

The short existence of the newly discovered fort at Ca~ei its topographical position which is prone to flooshyding indicate that this fort could have been either a construction camp or a proper fort initially misplaced and abandoned in short time probably after it was flooded as the archaeological excavations indicate 1l4

It is likely therefore that the fort from the immediate vicinity towards south was an afterthought replacing at some time the former fort and being erected in a suishytable place where the ground is naturally higher than the rest

~rn

o 10 20 30 40 SOm -=-----=~- -

Fig 16 Tibiscum (based partially on Arder 2003 and Fig 17 Cii~eiu (based on Isac 2003 Fig 2) BenealBona 1994)

112 Isac 2003 40

113 Isac 2003 40

111 The unit will replace here coho I sagittariorum sometime during 114 Even with t~e best agrimensores tre Romals could rrake

the second half of the 2nc century when saglttaril are present at mistakes as was ~he case for exal1ple wher In 49 BC Caesar at

Drobeta see Piso 2001230 Petolescu 2002 120f Ierda misplaced a for between two Ivers Caes Bell CIV 148

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

148 I S A A LOB U R G

At Porolissum in one of the most important strateshygical point on the northern frontier of Dacia (Fig 18) c 500 m eastward from the well-known fort at Pomet hill on Citera hill a fortlet was built 10110 by 6665 m

omel

D FortifIcation ~ Double Turf Rampart o Arnphitheate bull Watchtower Stone Wall Ditch

Turf Rampart Roman Road

Ii I

I I II

1

1

I

~~~aJ1__ =1-I

U~ bull LJI il

~

HJ20 3140 50

Fig 18 Porolissium (Pomet)

(067 ha)15 Scholars argued that the fortlet was consshytructed in the first half of the second century and proshybably it was garrisoned by Palmyrene archersll6 Their arguments are on one hand the smaller size of the fort and on the other hand a triangular in section arrowhead identified in the southern tower of north-western gate Due to the discovery of a tile-stamp of cohors III ll7

probably Campestris unit existing in Porolissum accorshyding to I Piso under Septimius Severus or Caracalla ll8

it is probable that the fort was still functioning in that period In any event it is hard to believe that cohors III Campestris a military unit occupied such a fortshylet Probably at first Palmyrene archers were a smaller unit11 than later in mid 3rd century when it is clear an equivalent of an ala as it is mentioned on one inscripshytion of Salonic 120

During the 3rd century the presence of the Palmyrene archers in the fort on Pomet hill is confirmed by the discovery inside the fort of tile-stamps many triangushylar arrowheads and the existence close to this fort of a temple dedicated to Bel restored in the third centuryl21

CONCLUSION

In theory it seems that these adjacent forts could fulshyfilled a series of functions have had a set of functions fOfts housing auxiliary units forts garrisoned by an auxiliary troop and a numerus respectively one of the forts could have been castra aestiva and the other one castra statiua one of the two forts could have been a construction camp or used as training ground or as a fort-annexe In order to distinguish between a variety of possibilities it is essential to understand the chronoshylogical relationship between two neighbouring forts at least to establish if the forts existed simultaneously Unshyfortunately dearth of the basic knowledge concerning most of the forts in Germania Superior and Dacia due to the lack of archaeological excavations impedes us to establish their chronology very precisely

Recently has oeen proposed the existence In the immediate vlcnlty of the orts on PO1let of another fortification located on an pfenor platea~ south to tle fort out we dont rave yet sufficient eVidences Matei 2003

16 elL 111803837 Gudea 199746-50

Gudea 1989 93

118 PISO 2001 231

9 For rstance early in the 30 tre commander of tf]e troop was a centurion of leg V Macedomca (AE 1980 755)

G X2 1 n 146 (I~S 9472 iDRE 1356) The ipscriptlon records hat G Meslrlus Servllanus has been praefectus of this troop see also Petolescu 2002 141-143

12 AE 1979501 g 1980755 AE 1977 666 For the temple see Pse 1980 279-282 Gudea 1989 1741 For arrowheads see Gucea 1989 286 545 PI exxxi Probably dUring the third certury the unit has a colort staus If this is he unit abbreviated CH I P P r tre inscrptlon of Potalssa (Cll ill 908)

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

The garrisoning of adjacent forts by two auxiliary units would theoretically not be impossible but in the case of regular troops the norm was to accommodate where it was necessary more than one unit in a single stronghold Until Domitian in the case of legionary fortresses the situation was similar and there are no adshyjacent strongholds for two separate legions Therefore the model for garrisoning two auxiliary military units in one fort was used whenever necessary It is certain that Domitians compulsory rule concerning the legions would not apply in the case of auxiliary and if two units of similar rank had to be garrisoned at one place it looks as they were combined in one fortification 122

In other places fortifications located in the vicinity of one another but one garrisoned by legionaries and the other by auxiliaries were discovered This is the case of Rottweil Carnuntum and Aquincum It would be difshyficult to say if this is a reflection of a different status between troops but surely it reflects a different funcshytion

In Germania Superior there is a concentration of adshyjacent-forts on one sector of the limes and its successor yet in Dacia the adjacent forts are to be found in all of the three provinces Dacia Porolissensis Superior and Inferior with a concentration in the latter Vhen units are attested they usually are of different rank Namely the numeri of Brittons Palymrenes or Moors of Gershymania Superior Dacia Porolissensis and Dacia Superior clearly testified additionally to the auxiliary units at the same place It would be difficult to establish what troops of nationes garrisoned the forts on the limes TransAlushytanus The existence of another numeri not attested in Dacia Inferior would not be impossible 123

The size of one of the neighbouring forts or fortlets presented above indicates that these forts were occupied either by an auxiliary detachment or by a numerus It was assumed that only fort-sizes as a guide to garrisons may not always be very precise but as a general rule those of roughly 06 ha are usually labelled numerus

122 It is not quite certain if the Domitians solution to the iegionaries mutinies wasnt only temporary Under Trajan at Apulum on some tile-stamps there are abbreviated jointly the names of the legions I Adiutrix and XIII Gemina Baluta 1997 1671 ThiS IS the most important argument for C Oprealu (1998 42) to postUlate here the existence of two forts for each legion Nevertheless the situashytion it seems to be quite different and there are only detachments of I Adiutrix in Dacia see for the entire discussion and biblioshygraphy Piso 2000 2051 If there are some legionaries of I Adiutrix at ApLilum they should have been garrisoned in the fort of leg XIII Gemina but this is not surely as the tile-stamps strictly prove that the legionaries had a joint brickyard that is the legioraries of I Adiutrix worked in the brickyard of the Xllfth legion

It could be representative that numerus burganorum et vereshydariarum it is attested in this province OnlY once (Cll III 13796 = ILS 9180) The scant attestation of numen either on diploma or on inscriptions it is well known

S A 1 L [1 BUR G I 149

forts Out of the seventeen fortifications nine belong to

this category lvloreover related to these forts we have elements attesting the presence of a national numerus the proof being some elements such as inscription and tile-stamps In Germania Superior the presence of nushymeri in the majority of those adjacent fortifications is a fact Three of the six forts in Germania Superior are of over 06-07 ha about 15 ha and one of c 2 ha but here are always attested more troops of numeri Conseshyquently it is useless to take here the area of the forts as a guide The plan of the forts irregular in most cases is probably the result of the participation of these l1Ushy

meri also in the construction124 not quite specialized yet the only building inscription is of leg VIII Augusta at Osterburken Consequently considering the irregular shape of the eastern fort at Ohringen it is probable that this is the one garrisoned by the numeri and not the western one The distance between fortifications range from degto 2300 m but in most cases the distance is some hundreds metres Curiously the largest distance is between the forts at Miltenberg Also here is the only place among the adjacent forts where we have clear inshydication of garrisoning even in the third century Moreshyover it is not certain whether the unit of exploratores is of nationes or not

In the other six adjacent forts of Germania Superior there are indications concerning the presence of the nushymeri only during the second century125 until Septimius Severus yet the existence of the numeri during the third century it is not excluded

Along with epigraphical evidence which has been disshycovered in some forts of Dacia archaeological material typical for eastern archers is known as the so-called dreifiigelige arrowheads This is the case at Urluieni Copaceni Romula Tibiscum and Porolissum 126 Since so little is known with certainty about the forts of Greshybenac it was labelled numerus fort only on dimensional basis The interest for archaeological excavations in the other three fordets of 03 hectares on the limes transashy[utanus are sadly almost lacking but it is difficult to assume that these fordets had a principia and therefore probably did not house tactically independent units

124 As the numerus Burganorum et Veredariorum for exampe JJit the fort at Copaceni il AD 138 (Cll III 13796 = ILS 9180

Brittones were In garnson also in some forts on the other pamiddot~s Of

the limes of Gerrrania Superior probably at SaalbJg Lgmantel Niederbieber or Obernburg see for a short aceen and part of the bibliography Southern 1989 12Of Latey ~5 supposed that generaly the f011ets on the Taunus or WetteraJilmeS are garrisoned by aUXiliary detachments Reuter 1999 418t

126 The numeri from lIblscum and PorolssLJrr are mentioned on some inscrptions as we Fa these forts see for instance Benea Bona 1994 and Gudea 1989

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

150 I S 1 1 LOB U R G

Concluding these remarks I would say that the exisshytence of a common pattern in Germania Superior and Dacia is clear That is in the second century on the Odenwald-Neckar limes and on the outer limes between Miltenberg and Welzheim and in Dacia in different areas of the province in the most vulnerable locations or in the strategical points on the frontier were erected adjacent fortifications As so few adjacent forts have been excavated and none of them on any appreciable scale it is not possible to attribute a definitive function to them but it seems that a usually smaller numerus fortlet often co-exists with an auxiliary base 12- Maybe this was at the beginning the response to a military problem that is the garrisoning of the national troops future numeri different in organization from auxiliary units therefore necessary to be accommodated in a seshyparate fort or fortlet either individual or in relation to an auxiliary fort

When exactly this process begins it would be diffishycult to establish but it seems it was under Hadrian and Antoninus Pius Anyway it has to be in relation with the appearance of the national numeri The theories concerning the history of these troops indicate a deveshylopment completed under Trajan Hadrian or Marcus Aurelius 128 The archaeological and epigraphic evidences suggest also the existence of the numeri of nationes at the latest under Hadrian The first evidences come from Germania Superior l29 and Dacia 130 Curiously in other provinces adjacent forts are lacking with some exceptions in Raetia and not accidentally also in these provinces the evidences regarding numeri are missing or they are a characteristic of the third centuryUl Thus it

In the most vulnerable sections of the western limes there were garrisoned in Dacia more than one aUXiliary unit out always in one stronghold This IS the case at Porolissum Micia or Tibiscum

128 Th Mommsen stated that numeri developed fram the native irregulars mentioned by Hyginus (19 29 30 43) the author thought under Trajan until their final form under Marcus AureliJs Mommsen 1884 219-234 later A v Domaszewski concluded probably correct that the process of numeri development was fnished under Hadrian Domazsewski 1908 59-61

129 The fortle of 06 ha at Hesselbach seems to be garrisoned by an independent unit as it has a principia on its own and it was built between AD 95-100 and abandoned in the mid-second century Baatz 1973 661 Whatever the earliest date for the preserce of Brittones in the Upper Germany is AD 145-6 (ell XIII 6511 6514 6517 6518 6490) For the discussion and bibiography regarding the probematis 0 Bnttones from Germania Supenor and their arrival here see Southern 1989 95~98

30 The first datable evidences are the diplomas issued to the Palmyrene soldiers of AD 120 (Cll XVI 68 =0 lOR I 5 RMD 17 lOR I 6) and AD 126 (RMD 27 lOR I 8 RMD 28 lOR I 9) =0 =0

see Mann 1985

131 Except Germania Superior and Dacia only in Britannia and Africa have been present numeri Here the numeri are attested only from the early thrd century onward Southern 1989 116 126shy131 Also in Germania Inferior the numeri present on the Rline are clearly attestec again only n the third century AlfOldy 1968 79f

would not be fanciful to believe that most of the adjashycent forts were the result of the necessity of garrisoning numeri where the need demand This is also the reason why the dating of limes transalutanus under Hadrian or Antoninus Pius is confirmed132

It would be hard to believe that a separate garrison for natiol1es is only a matter of trust as MP Speidel concludes133 but mere a problem of different status The legal status of the Ilationes has been also the subject of many studies 134 Exhaustively the scholars agree that the evolution of the numeri has been from irregular barbashyrian troops or unromanized during the second century to the more or less regular units very similar to auxilia during the third century when some of the numeri have been elevated to cohors or ala status 135 Probably this change of status is also the reason for the units of nashytiones to be at first garrisoned separately from auxiliary troops but sooner or later probably at the end of the second or early in the third century nationes will be accommodated jointly with auxilia

It is difficult to establish a pattern concerning for example the distance benveen these forts sometimes being placed in the immediate vicinity and at times the extent is of c 500 m Hence in the cases discussed the distance is a matter of topography and strategy rather than a problem of establishing territoria boundaries

As anticipated it may be right that it is about a new type of frontier and a new type of garrisoll to strengthen it thus relieving the auxilia of the more mundane tasks 136

132 Tre date of the limes TransAutanus it is the subject of conshysiderable debate Te construction of the Imes was attributed to Hadrian Antnlus Pius or Septimus SeverLS Has been conshysidered a creatlor f1adrian or Antonrus Pius Llecause the earl est evidences are some cOins of Trajan and Hadrian and the analogies With other fronters of t~e empire Zangemeister 1895 8H Kornemarl~ 1907 Fabncius 1926 COl 645 Their opInion seems to be confrrrec by serre of the arelaeologlea evidences see 1997 86-91 Contra trere are an entire group 0 scilolars The consensus of opilion thac the limes TransAutanus has been constructec late In the second century andlor early in the tlrd century Tocilescu 1900 123f Christeseu 934 73 Tldor 1978 253 Petolescu 2000 207 It is trJe that most of the eVldelces Inecate a ate date of the limes TransAutanus but some celns and the existence of this adjacent forts impy an oCCJpatloll tlIS part Of province also in the second certury probaby n the first half and or a shor period beng reoccupied during thrd century

133 Tile author referring to t1e description of Hyg rus fortlflcaton Spelde 1975 228

The ost Irrpoiant are Stein 1932 233f Rowell 1936 Vttnghof 1950 Mann 1954 Calles 1964 Baatz 1973 72-76 Spece 1975 Mann 1985 Southerl 1989 10amp110 Reuter 1999

135 1932 237 Rowell 1936 Calles 1964 189-198 Baatz 197373 187

136 Southerr 1989 82

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

5 1 1 L [] BUR G I 151

BIBLIOGRAPHY

Alfoldy 1983

Alfoldy 1987

ArdetArdet 2004

Baatz 1973

BaatzHerrmann 1982

Baluta 1997

BeneaBona 1994

BogaersRiiger 1974

Bogdan-Cataniciu 1997

Callies 1964

Cantacuzino 1944

Christescu 1934

Christescu 1936

CzyszDietzlFischeriKellner 1995

DaicoviciuFerencziGlodariu 1989

Domaszewski 1908

Driel-MurrayHartmann 1999

Fabricius 1926

FiltzingeriPlanckiCammerer 1976

Alf6ldy G Caius Popilius Carus Pedo und die Vorverlegung des Obergermashynischen Limes FBW 8 1983 55-67

Aif6ldy G Romische Heeresgeschichte Beitrage 1962-1985 Mavors vol 3 (Amsterdam 1987)

Ardet AlArdet L c Tibiscum A~ezarile romane (Cluj-Napoca 2004)

Baatz D Kastell Hesselbach und andere Forschungen am Odenwaldlimes Limeforschungen 12 (Berlin 1973)

Baatz DlHerrmann F-R Die Romer in Hessen (Stuttgart 1982)

Balura c L Tipuri de ~tampile tegulare militare inedite descoperite la Apulum II ~tampile fara antroponim Apulum 34 1997 133-168

Benea DlBona P Tibiscum (Bucure~ti 1994)

Bogaers J ElRiiger C B Der Niedergermanische Limes (Bonn 1974)

Bogdan-Cataniciu L Wallachia in the Defensive System of the Roman Empire 1st - 3th (sic) centuries AD - Muntenia in sistemul defensiv al imperiului roman sec I-III pChr (Alexandria 1997)

Callies H Die fremden Truppen im Romischen Heer des Prinzipats und die sogenannten Nationalen Numeri BRGK 45 1964 130-227

Cantacuzino Gh Le grand camp romain situe pres de la commune de Baneasa (Departement de Teleorman) Premier rapport Dacia 9-10 1944 441-472

Christescu v Le tresor de monnaies de Silpata-de-Jos et la date du limes romain de la Valachie Istros 1 1934 73-80

Christescu v Le Castellum romain de Sapara de ]05 Dacia 5-6 1935shy1936 435-447

Czysz XTDietz K-HlFischer ThlKellner H-J Die Romer in Bayern (Stuttgart 1995)

Daicoviciu HlFerenczi ~tlGlodariu t Cetati ~i a~ezilri dacice in sud-vestul Transilvaniei I (Bucure~ti 1989)

Domaszewski A V Die Rangordnung des Romischen Heeres B] 117 1908 1-278

Driel-Murray C vlHartmann H-H Das Ostkastell von Welzheim RemsshyMurr-Kreis Die romischen Lederfunde Die Terra Sigillata Forsch u Ber z Vor- u Friihgesch in Baden-Wiirttemberg 42 (Stuttgart 1999)

Fabricius Limes RE XXV 1926 col 641-645

Filtzinger PhlPlanck DlCammerer B Die Romer in Baden-Wiirttemberg (StuttgartAalen 1976)

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

152 I S A A L [] BUR G

Gudea 1989

Gudea 1997

Isac 2003

JaeScholz 2002

Kandler 1997

Kornemann 1907

Kortum 2004

KortumlLauber 2004

Mann 1954

Mann 1985

Marinoiu 2003

MarinoiuIHortopan 2003

Matei 2003

Mommsen 1884

Nemeth 1990

Opreanu 1998

Petolescu 1983

Petolescu 2000

Petolescu 2002

Piso 1980

Gudea N Porolissum Un complex arheologi dao-rona b margri1ea de nord a Imperiului roman 1 Acta ~lus Porohssensis 13 1989 Iff

Gudea N Der dakische Limes ~laterialien zu seiner Geschichte JRGZ 44 1997 497ff (1-113)

Isac D The Roman fort SAMV-l-Casecteiu (Cluj-)apoca 2003)

Jae MlScholz M Reduktion von numerus- und Kleinkastellen des obershygermanischen Limes im 3 Jahrhundert in Limes XVIII Proceedings of the XVIII Int Congress of Roman Frontier Studies held in Amman Jordan (Sept 2000) BAR IS 10841 (Oxford 2002) 415-423

Kandler M (Hrsg) Das Auxiliarkastell Carnuntum 2 Forschungen seit 1989 Sonderschr OAI 30 (Wien 1997)

Kornemann E Die neueste Limesforschung (1900-1906) im Lichte der roshymisch-kaiserlichen Grenzpolitik Klio 7 1907 73-121

Kortiim K Neue Untersuchungen zum romischen Kastellbad von Osterburshyken Neckar-Odenwald-Kreis Arch Ausgr Bad-Wurttemb 2004 144-148

Kortiim KlLauber ] Walheim 1 Forsch u Ber z Vor- und Fruhgesch in Baden-Wurttemberg 95 (Stuttgart 2004)

Mann ] C A note on the numeri Hermes LXXXII 1954 501-506

Mann J C The Palmyrene Diplomas in Roxan M Roman Military Diplomas II 1978-1984 (London 1985)

Marinoiu v Unitatile milirare care au stationat in castrele romane de la Bumbe~ti Jiu Litua 9 2003 57-65

Marinoiu VlHortopan D Cercetarile arheologice efectuate la BumbesecttishyJiu - Vartop ~i Ciocadia - Codri~oare judetul Gorj (campaniile 2000shy2001) Litua 9200335-55

Matei AI v Moigrad-Porolissum com Mirsectid judo Salaj [Porolissum] in Cronica Cercetarilor Arheologice (Campania 2003)

Mommsen Th Die Conscriptionsordnung der romischen Kaiserzeit Hermes XIX 1884219-234

Nemeth M Forschungen im Alenkastell von Aquincum in Wetters HI Kandler M (Hrsgg) Akten des 14 Internationalen Limeskongresses 1986 in Carnuntum RLiO 362 (Wien 1990) 675-681

Opreanu C Dacia romana secti Barbaricum (Timisectoara 1998)

Petolescu C C Sex Iulius Possessor SCIVA 341 198342-56

Petolescu C c Dacia ~i Imperiul roman (Bucuresectti 2000)

Petolescu C C Auxilia Daciae (Bucure~ti 20021

Piso 1 Beitrage zu den Fasten Dakiens im 3 Jahrhundert ZPE 80 1980 273-282

--~--------------- shy

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

Pis a 2000

Piso 2001

Planck 1975

Planck 1979

ReddeSchnurbein 2001

Reuter 1999

Rowell 1936

Schall mayer 1984

Schonberger 1973

Schonberger 1985

Sommer 1988

Sommer 1992

SommerKortum 2005

Southern 1989

Speidel 1973

Speidel 1975

Speidel 1977

Stein 1932

Tocilescu 1900

S 1 1 L [] BUR G I 153

Piso I Les legions dans la province Dacie in Le Bohec Y (ed) Les legions de Rome sous la Haut-Empire Acres du Congres de Lyon 17-19 septembre 1998 (Lyon 2000)

Piso I Studia porolissensia (I) Le temple Dolichenien Acta Mus Naposhycensis 381 2001 221-237

Planck D Arae Flaviae I Forsch u Ber z Vor- und Friihgesch in BadenshyWiirttemb 6 (Stuttgart 1975)

Planck D Ausgrabungen im Ostkastell von Welzheim Rems-Murr-Kreis (Baden-Wiirttemberg) AKB 9 1979411-417

Redde MSchnurbein S v (Hrsgg) Alesia Fouilles et recherches francoshyallemandes sur les travaux militaires romains autour du Mont-Auxois (1991shy1997) 1 Les fouilles Memoires de IAcademie des Inscriptions et Belles Lettshyres 22 (Paris 2001)

Reuter M Die numeri des romischen Heeres in der Mittleren Kaiserzeit BRGK 80 1999356-569

Rowell H T Numerus RE 17 (1936) 1327-1341 2537-2554

Schall mayer E Der Odenwaldlimes (Stuttgart 1984)

Schonberger H Das Romerkastell Ohringen-West (Biirgkastell) BRGK 53 1972 233-296

Schonberger H Die romischen Truppenlager der friihen und mittleren Kaishyserzeit zwischen Nordsee und Inn BRGK 66 1985321-497

Sommer C S Kastellvicus und Kastel FBW 13 1988 457-707

Sommer C S MVNICIPIVM ARAE FLAVIAE - Militarisches und ziviles Zentrum im rechtsrheinischen Obergermanien BRGK 73 1992 269-313

Sommer C SlKortum K Rottweil RW Kastelle und Stadt Municipium Arae Flaviae in Planck D Die Romer in Baden-Wiirtemberg (Stuttgart 2005) 292-301

Southern P The numeri of the Roman Imperial Army Britannia 20 1989 81-140

Speidel M P Numerus Syrorum Malvensium The transfer of a Dacian army unit to Mauretania and its implications Dacia NS 17 19-3 169shy177

Speidel M P The rise of ethnic units in the Roman imperial army A-Rx III3 (BerlinINew York 1975) 202-231

Speidel M P A tribune of Cohors III Campestris Apulum 15 19-- 631shy633

Stein E Die Kaiserlichen Beamten und Truppenkorper im romischen Deutschland unter dem Prinzipat (Wien 1932)

Tocilescu G G Fouilles et recherches archeologiques en Roumanie (Bucarest 1900)

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq

154 I S 1 1 L [] BUR G

Tudor 1936

Tudor 1978

Tudor 1981

Tudor 1982

Vittinghoff 1950

Wolff 1975

Zangemeister 1895

Tudor D I Castri romani di Jidava (Romania) Castrele romane de la Jidava langa Campulung in Muscel Bucure~ti 2 (1936) 89-117

Tudor D Oltenia romana (Bucure~ti4 1978)

Tudor D Comandamentele militare de la Praetorium in Dacia SCIVA 3211 1981 76-88

Tudor D Materiale arheologice din castrul Praetorium I (Copaceni judo Valshycea) descoperite de Tocilescu G Drobeta 5 198249-78

Vittinghoff E Zur angeblichen Barbarisierung des romischen Heeres durch die Verbande der Numeri Historia 1 1950389-407

Wolff H Miscellanea Dacica Acta Mus Napocensis 12 1975 139-158

Zangemeister K Neues Heidelberger Jahrbuch 5 1895 81 sqq