A WORLD OF GOODS

din i ng in the sanctuary of deme ter and kore 1

Volume 8020 1 1

Copyright © The American School of Classical Studies at Athens, originally published in Hesperia 80 (2011), pp. 511–558. This offprint is supplied for personal, non-commercial use only. The definitive electronic version of the article can be found at <http://www.jstor.org/stable/10.2972/hesperia.80.4.0511>.

Hesperia The Journal of the Amer ic an School

of Cl assic al S tudies at Athens

hesperiaTracey Cullen, Editor

Editorial Advisory Board

Carla M. Antonaccio, Duke UniversityAngelos Chaniotis, Institute for Advanced Study, PrincetonJack L. Davis, American School of Classical Studies at Athens

A. A. Donohue, Bryn Mawr CollegeJan Driessen, Université Catholique de Louvain

Marian H. Feldman, University of California, BerkeleyGloria Ferrari Pinney, Harvard University

Sherry C. Fox, American School of Classical Studies at AthensThomas W. Gallant, University of California, San DiegoSharon E. J. Gerstel, University of California, Los Angeles

Guy M. Hedreen, Williams CollegeCarol C. Mattusch, George Mason University

Alexander Mazarakis Ainian, University of Thessaly at VolosLisa C. Nevett, University of Michigan

Josiah Ober, Stanford UniversityJohn K. Papadopoulos, University of California, Los Angeles

Jeremy B. Rutter, Dartmouth CollegeA. J. S. Spawforth, Newcastle University

Monika Trümper, University of North Carolina at Chapel Hill

Hesperia is published quarterly by the American School of Classical Studies at Athens. Founded in 1932 to publish the work of the American School, the jour-nal now welcomes submissions from all scholars working in the fields of Greek archaeology, art, epigraphy, history, materials science, ethnography, and literature, from earliest prehistoric times onward. Hesperia is a refereed journal, indexed in Abstracts in Anthropology, L’Année philologique, Art Index, Arts and Humanities Citation Index, Avery Index to Architectural Periodicals, Current Contents, IBZ: Internationale Bibliographie der geistes- und sozialwissenschaftlichen Zeitschriften- literatur, Numismatic Literature, Periodicals Contents Index, Russian Academy of Sciences Bibliographies, and TOCS-IN. The journal is also a member of CrossRef.

The American School of Classical Studies at Athens is a research and teaching institution dedicated to the advanced study of the archaeology, art, history, philosophy, language, and literature of Greece and the Greek world. Established in 1881 by a consortium of nine American universities, the School now serves graduate students and scholars from more than 180 affiliated colleges and uni-versities, acting as a base for research and study in Greece. As part of its mission, the School directs on going excavations in the Athenian Agora and at Corinth and sponsors all other American-led excavations and surveys on Greek soil. It is the official link between American archaeologists and classicists and the Ar-chaeological Service of the Greek Ministry of Culture and, as such, is dedicated to the wise management of cultural resources and to the dissemination of knowl-edge of the classical world. Inquiries about programs or membership in the School should be sent to the American School of Classical Studies at Athens, 6–8 Charlton Street, Princeton, New Jersey 08540-5232.

© The Amer i c an Sc hoo l o f C l a s s i c a l S tud i e s a t Athens

hesperia 80 (201 1)Pages 5 11–5 58

A WORLD OF GOODS

Transp ort Jars and Commodit y Exchange at the Late Bronz e Age Har bor of Kommos, Cre te

ABSTRACT

The harbor site of Kommos, Crete, has yielded rich evidence for long-distance exchange in the form of ceramic transport jars of types used not only for distribution within Crete and the Aegean, but also across the eastern Mediterranean. An integrated petrographic and chemical approach is here employed in order to determine the provenance of short-necked amphoras, transport stirrup jars, Egyptian jars, and Canaanite jars. The results reveal a detailed picture of local jar production within southern Crete, as well as jars that have their origins in the Nile Delta and at several specific locations along the Levantine coast.

TRANSPORT JARS AT KOMMOS

Situated on the coast of the large, fertile plain of the Mesara, the Minoan harbor town of Kommos (Fig. 1) has yielded more evidence for intercultural exchange in the form of imported ceramics than any other Bronze Age site in the Aegean.1 The unusual range and quantity of foreign pottery recovered at Kommos were initially recognized and documented by Vance Watrous.2 These studies have been cited repeatedly in subsequent synthetic overviews of Bronze Age trade in the eastern Mediterranean,3 and have been extended and fine-tuned by Jeremy Rutter in the light of the most recent excavations at the site during the 1990s.4

A substantial portion of the ceramic imports in question are contain-ers designed to ship commodities in bulk, often referred to as transport

1. Kommos has been excavated over 15 summers (1976–1985, 1991–1995) by a team directed by J. W. and M. C. Shaw of the University of Toronto, under a permit issued to the American School of Classical Studies at Athens. For permission to sample transport jars from the site, we are grateful to the

Directorate of Conservation and the 23rd Ephorate of Classical and Pre- historic Antiquities in Herakleion. Funding for this study was generously provided by the Institute of Aegean Prehistory. The results of the study have been previously presented in two conference papers (Day, Kilikoglou,

and Rutter 2004; Day et al. 2006). All dates are b.c.

2. Watrous 1985, 1989; Kommos III; Watrous, Day, and Jones 1998.

3. E.g., Knapp and Cherry 1994; Cline 1994.

4. Rutter 1999, 2006b, 2006c.

pe ter m . day e t al .512

vessels; there are also analogous Minoan vessels that seem to represent containers for outgoing exchange. When they occur in the archaeological record of the Graeco-Roman era, such transport vessels in the central and eastern Mediterranean are almost invariably described as “amphoras,” closed shapes with two vertical handles on the shoulder or linking the shoulder with the neck or rim. In the Middle and Late Bronze Age, however, the forms taken by such transport vessels tend to be characteristic of particular regions within a larger zone that witnessed increasingly intense intercul-tural exchanges between ca. 1750 and 1200.5 Within the southern Aegean, the most common form in the Late Bronze Age is a large version of the false-necked amphora or stirrup jar, conventionally termed the transport stirrup jar (Fig. 2).6 The form preferred along the Syro-Palestinian coast and widely exported to, as well as imitated in, both Cyprus and Lower Egypt is a shoulder-handled vessel often referred to as the Canaanite jar or amphora (Fig. 2).7 Differences in fabric and surface treatment make possible the recognition of amphoras of several subvarieties produced in Egypt, some closely comparable in shape to the Canaanite jar (Fig. 2).8 Finally, at Kommos itself, around the end of the first quarter of the 14th century, a distinctive variant of the traditional Minoan oval-mouthed amphora made its appearance, featuring a round mouth, a comparatively low neck, and two almost cylindrical handles attached at the shoulder and rim. Christened the short-necked amphora (Fig. 2),9 this form was soon being mass-produced, to judge from the thousands of fragments recovered from 14th- and 13th-century contexts at the site.10

Figure 1. Map of Crete showing the locations of Kommos and other archaeological sites. P. S. Quinn

5. For recent discussions of the absolute chronology of the Aegean Late Bronze Age, see Shelmerdine 2008, pp. 3–7; Manning and Bruce 2009, pp. 275–332 (including contributions by M. H. Wiener, W. L. Friedrich et al., and S. W. Manning et al.); Wild et al. 2010. For recent discussions of

later Minoan Palatial and Postpalatial terminology in terms of ceramic chro-nology, see Hatzaki 2007a, 2007b; Lan-gohr 2009, pp. 11–14, 181–233.

6. For transport stirrup jars, see Haskell et al. 2011; also Haskell 1981a, 1981b, 1984, 2005; Day and Haskell 1995; Day 1999.

7. Smith, Bourriau, and Serpico 2000; Bourriau, Smith, and Serpico 2001; Cohen-Weinberger and Goren 2004.

8. Aston 1998.9. Kommos III, pp. 135, 144.10. Rutter 2000.

tran sp ort jars at l ate b r onz e ag e kommo s 513

Clearly, these ceramic vessels have much to tell us about maritime trade and exchange, as well as about the production and distribution of value-added commodities from Crete. Lying as it does in a region of prime agricultural land and at a strategic position for the movement of goods through the Mediterranean, Kommos offers the possibility of investigating the relations between Crete and the wider Bronze Age world in a number of distinct but interrelated ways. It was against this background that the authors of this article undertook a program of thin-section petrography and instrumental neutron activation analysis (INAA) focused on the four groups of transport vessels described above. A total of 88 samples were se-lected, including 18 transport stirrup jars (TSJs), 13 short-necked amphoras (SNAs), 34 Canaanite jars (CJs), and 19 Egyptian jars (EJs) (Table 1).11

Figure 2. Late Minoan transport jars from Kommos

11. Rutter had not always correctly identified the sherds and larger vessel fragments selected for sampling in 1998 on the basis of their morphologi-cal features and macroscopic examina-tion of their fabrics. Thus three of the supposed SNAs were actually examples of earlier amphora types, in all cases probably locally produced (samples

98/18, 23, and 26 = C7989, C11093, and C11195, respectively). One of the CJs (sample 98/66 = C9100) was iden-tified by visiting authorities on Egyp-tian ceramics as an Egyptian import. Of four mistakenly identified EJs, three were reclassified as CJs (samples 98/38, 46, and 51 = C10723, C8726, and C7070, respectively) and one as a

nonlocal but probably Cretan import (sample 98/53 = C6949). The revised totals are those given in the text. For the identifications Rutter is extremely grateful to D. Aston and B. Bader of the SCIEM Project, Austrian Acad- emy of Sciences; P. Rose of the Tell el-Amarna excavations; and M. Serpico of the Cambridge Amphora Project.


TABLE 1. LATE MiNOAN TRANSPORT JAR SAMPLES FROM KOMMOS Sample No. Inv. No. Type1 Date

Kommos 98/1 C2849 TSJ LM IIIBKommos 98/2 C2893 TSJ LM IIIBKommos 98/3 C3318 TSJ LM IIKommos 98/4 C2470 TSJ LM IIIBKommos 98/5 C6355 TSJ LM IIIBKommos 98/6 C11077 TSJ LM IA FinalKommos 98/7 C7874 TSJ LM IIIBKommos 98/8 C4940 TSJ LM IIIA1Kommos 98/9 C9273 TSJ LM IIIBKommos 98/10 C11241 TSJ LM IIIBKommos 98/11 C11179 TSJ LM IIKommos 98/12 C10976 TSJ LM IIIA2Kommos 98/13 C10258 TSJ LM IIKommos 98/14 C9063 SNA LM IIIA2Kommos 98/15 C285 SNA LM IIIBKommos 98/16 C9276 SNA LM IIIBKommos 98/17 C9386 SNA LM IIIBKommos 98/18 C7989 SNA2 LM IIIA2 EarlyKommos 98/19 C9836 SNA LM IIIBKommos 98/20 C9662 SNA LM IIIBKommos 98/21 C10219 SNA LM IIIBKommos 98/22 C10348 SNA LM IIIBKommos 98/23 C11093 SNA2 LM IIIA2Kommos 98/24 C11152 SNA LM IIIA2Kommos 98/25 C11194 SNA LM IIIA2Kommos 98/26 C11195 SNA2 LM IIIA2Kommos 98/27 C11231 SNA LM IIIA2Kommos 98/28 C5582 TSJ LM IIKommos 98/29 C7986 TSJ LM IIIA2 EarlyKommos 98/30 C7819 TSJ LM IIIBKommos 98/31 C7987 TSJ LM IIIA2 EarlyKommos 98/32 C7981 TSJ LM IIIA2 EarlyKommos 98/33 C10765 EJ LM IIKommos 98/34 C10655 EJ LM IIIBKommos 98/35 C10218 EJ LM IIIA2Kommos 98/36 C7476 EJ LM IIIA2 EarlyKommos 98/37 C9837 EJ LM IIIAKommos 98/38 C10723 EJ3 LM IIKommos 98/39 C9504 EJ LM IIIA2Kommos 98/40 C9489 EJ LM IIIA2Kommos 98/41 C8837 EJ LM IBKommos 98/42 C8336 EJ LM IIIBKommos 98/43 C9625 EJ LM IIIA2 EarlyKommos 98/44 C8006 EJ LM IIIA1Kommos 98/45 C6392 EJ LM IIIBKommos 98/46 C8726 EJ3 LM IIIA2 EarlyKommos 98/47 C7448 EJ LM IIIA2Kommos 98/48 C4574 EJ LM IIIAKommos 98/49 C3559 EJ LM II


Kommos 98/50 C7072 EJ LM IIIA2 EarlyKommos 98/51 C7070 EJ3 LM IIIA2 EarlyKommos 98/52 C3350 EJ LM IIIA2Kommos 98/53 C6949 EJ4 LM IIIBKommos 98/54 C2763 EJ LM IB EarlyKommos 98/55 C8069 CJ LM IIIA2 EarlyKommos 98/56 C9167 CJ LM IIIA2Kommos 98/57 C6840 CJ historic levelsKommos 98/58 C11232 CJ LM IIIA2Kommos 98/59 C11141 CJ LM IIIBKommos 98/60 C10362 CJ LM IIIBKommos 98/61 C10360 CJ LM IIIBKommos 98/62 C9865 CJ LM IIIAKommos 98/63 C9624 CJ LM IIIA2 EarlyKommos 98/64 C10656 CJ LM IIIA2 EarlyKommos 98/65 C9398 CJ LM IIIBKommos 98/66 C9100 CJ5 LM IIIA2Kommos 98/67 C9941 CJ LM IIIBKommos 98/68 C8730 CJ LM IIIA2Kommos 98/69 C8058 CJ LM IIIA2 EarlyKommos 98/70 C8053 CJ LM IIIA2 EarlyKommos 98/71 C8728 CJ LM IIIA2Kommos 98/72 C8244 CJ LM IIIA2Kommos 98/73 C5140 CJ LM IIIBKommos 98/74 C6839 CJ LM IIIBKommos 98/75 C3351 CJ LM IIIA2Kommos 98/76 C7440 CJ LM IIIA2 EarlyKommos 98/77 C8729 CJ LM IIIA2 EarlyKommos 98/78 C8216 CJ LM IIIA2 EarlyKommos 98/79 C7115 CJ LM IIIA2 EarlyKommos 98/80 C8144 CJ LM IIIA2 EarlyKommos 98/81 C6990 CJ LM IIIA2 EarlyKommos 98/82 C7074 CJ LM IIIA2 EarlyKommos 98/83 C7638 CJ LM IIIA2 EarlyKommos 98/84 C7336 CJ LM IIIA2 EarlyKommos 98/85 C7428 CJ LM IIIA2 EarlyKommos 98/86 I47*A CJ LM IIIA2 EarlyKommos 98/87 C11086 SNA LM IIIBKommos 98/88 C11197 SNA LM IIIB

1 CJ = Canaanite jar; EJ = Egyptian jar; TSJ = transport stirrup jar; SNA = short-necked amphora.2 Subsequently reclassified as an earlier amphora type (see n. 11, above).3 Subsequently reclassified as a CJ (see n. 11, above).4 Subsequently reclassified as a probable Cretan import (see n. 11, above).5 Subsequently reclassified as an Egyptian import (see n. 11, above).

TABLE 1—Continued

Sample No. Inv. No. Type 1 Date


RESEARCH Q UEST iONS AND AiMS

The Kommos Transport Jar Project was focused on a specific functional class, the transport vessel, and included not only specimens strongly sus-pected of having been locally manufactured (the SNAs and some TSJs), but also those that were related either functionally (CJs and EJs) or formally (some TSJs), but were clearly not local products. The specific questions to which we hoped to find answers varied according to the formal category of the vessels. In characterizing the ceramics by both chemical and petro-graphic means, we hoped to be able to group the jars compositionally and, where possible, to suggest their provenance. By establishing the origins of the vessels we hoped to identify patterns in the production of the Cretan vessel types and in the movement of ceramic vessels and their contents around the eastern Mediterranean. Since the project as a whole was directly inspired by the discovery at Kommos of SNAs in such great quantities as to suggest that the form played an important role in the local economy, we begin with this class of amphora.

Short-Nec ked Amphora

The discovery of thousands of fragments of these vessels in the ruins of the monumental Late Minoan (LM) IIIA2–B Building P at Kommos, which was plausibly identified on architectural grounds as a facility for the storage and maintenance of ships during seasons considered unsuitable for sailing,12 suggested that the structure might also have served as a warehouse in which to stockpile SNAs. The vessels would then have been used as containers to ship one or more local products in bulk to consumers outside of Crete, if not indeed outside of the Aegean. Analysis of the distinctive form and chronology of the SNA had suggested that it should be viewed as a variant of the long-lived oval-mouthed amphora of the Mesara, expressly designed to advertise the newly recovered independence of the region following centuries of Knossian domination.13 The SNA initially drew some inspira-tion from its chief competitors for recognition as the optimally designed transport vessel within the eastern Mediterranean: thus the earliest SNAs imitated the angular shoulder of the Syro-Palestinian CJ and the decora-tion of the body with broad horizontal wavy bands typical of the Central Cretan TSJ.14 By the 13th century, the SNA had developed into a round-shouldered, plain shape (Fig. 2) produced in enormous quantities within narrowly prescribed limits of size and shape.

This reconstruction of the SNA’s history and its potential political and economic significance, however, is based entirely on the evidence from the single site of Kommos. No examples of this shape have been found in contemporary strata at the nearby sites of Ayia Triada and Phaistos, even though Kommos is universally believed to have functioned as the chief harbor for one or the other of these Minoan political centers throughout most of the Middle and Late Bronze Age, rather than as an independent political entity. In fact, aside from a single example published by Evans from Knossos,15 no example of a SNA has been identified to date at any

12. M. Shaw 1985; J. Shaw 2006a, pp. 70–85; 2006b, pp. 850–853.

13. Rutter 2000.14. Haskell 2005, pp. 208–209.15. PM II, pp. 627–629, fig. 392:3.


other Minoan site. Moreover, although Kommos may well have shipped hundreds or thousands of SNAs with their contents to one or more sites overseas, the fact remains that, as of 2008, no example of this distinctive ceramic form had yet been published from any Aegean or eastern Mediter-ranean site known to the authors.

Against this backdrop, there was a good deal to be learned about the SNA from a program of petrographic and neutron activation analyses. In how many distinct fabrics was the form produced? Were all these fabrics “local”—that is, was the vessel manufactured somewhere in the western Mesara in the immediate neighborhood of Kommos itself? Was there any evidence of a shift in the preferred clay recipe for SNAs over time, as there demonstrably was in their decoration and aspects of their shape? Of the 13 SNAs eventually sampled, four date from the LM IIIA2 period and nine probably from LM IIIB (Table 1). The number of samples is not large, in view of the thousands of examples that could have been selected for analysis, but we agreed that such a sample should at least allow several of the most basic questions to be addressed. Moreover, the results of the analyses of these samples should provide some indication of how future programs of analysis involving this particular shape might most usefully be structured.

Transp ort S tirrup Jar

In contrast to SNAs, TSJs have been recovered in substantial numbers at both coastal and inland sites on the Greek mainland and on Crete, as well as throughout the Aegean islands and along the western Anatolian coast, especially in contexts spanning the 15th to 13th centuries. They constitute an impressive percentage of the transport vessels recovered from three well-known shipwrecks of this period located within the Aegean (Cape Iria) or just outside it along the southern Turkish coast (Cape Gelidonya and Uluburun), and they have also been found in smaller quantities at coastal locales in Cyprus and the Levant, as well as at Tell el-Amarna, a good distance up the Nile in Egypt.16 Although this shape was clearly manufactured at a number of different sites across the southern Aegean, a majority of the examples that have so far been subjected to petrographic and trace-element analyses have proved to be Cretan products.17 Previous analyses (discussed below) have distinguished between TSJs made in the neighbor-hood of Chania in West Crete, those produced in Central Crete, sometimes specifically in the Mesara Plain, and those manufactured at one or more centers in East Crete.

The TSJs chosen for analysis in the present project come from well-dated contexts ranging from Neopalatial times through the Monopalatial era of LM II–IIIA1 to the 13th century (LM IIIB, variously termed Post- palatial, Final Palatial, or developed Third Palace period) (Table 1). The aim was to learn how homogeneous or heterogeneous the sources of the TSJs used at Kommos might have been over time, especially during the period when SNAs appear to have been the preferred local transport vessel (i.e., ca. 1375/1350–1200). The number of samples was modest, but this particular category of transport vessel, as noted above, had already been

16. Iria: Lolos 1999; Uluburun: Rutter 2005. For the Levant and other Mediterranean sites outside the Ae- gean, see Leonard 1994, pp. 45–47; Ben-Shlomo, Nodarou, and Rutter 2011, pp. 335–337; Haskell et al. 2011, pp. 125–131.

17. Haskell 2005; Haskell et al. 2011.


subjected to a large number of petrographic and chemical analyses, so a far larger comparative database existed for this group than for the SNAs.18 Equally, it was suspected that, while the SNA was a specifically local, south-central Cretan phenomenon, many of the TSJs would prove to have been manufactured in the northern part of Central Crete.

Canaanite Jar

During the developed Late Bronze Age, the shoulder-handled transport vessels known as CJs were the favored transport vessels throughout the Le- vant, including Cyprus and the Nile Delta, playing much the same role in the easternmost Mediterranean as that played by the TSJ within the Aegean. Just as the TSJ was produced at a number of centers throughout its zone of principal use, so too was the CJ manufactured throughout the Levant, from the Nile Delta to Cilicia, as well as imitated on Cyprus. Visual inspection of the more than 60 highly fragmentary to fully restorable examples of this form recovered at Kommos revealed a number of distinct fabrics, but how many different centers or regions of production these represented and how they were distributed over the 300 years of their use was unknown.

It is worth noting that CJs also occur at Chania during the LM III period, and are found in earlier contexts at a number of sites, such as Poros-Katsambas, Kato Zakros, and Pseira.19 Previous studies have examined the production, exchange, and use of the CJ in the Near East and Egypt.20 Its use as a container for resins and oil seems clear, but some vessels may also have contained wine.21

With these issues in mind, 32 CJs from Kommos were sampled for analysis in 1998 (Table 1). Subsequently, in 2002 and 2003, examination of the ceramic imports to Kommos by archaeologists with particular expertise in the study of Egyptian and Syro-Palestinian pottery resulted in a modest expansion of this sample from 32 to 34.22

Eg y p t ian Jar

Well over a decade after the initial publication by Watrous of the remarkable range of foreign ceramic imports to Kommos, this south Cretan harbor remains the only site in the Aegean where Late Bronze Age pottery from Egypt has been found. The Egyptian pottery is, almost without exception, undecorated.23 Aside from two recently recognized sherds of red-slipped carinated bowls produced in Nile silt fabrics, all the fragments belong to closed shapes that appear to have been manufactured from marl clays; most

18. See Day and Jones 1991 (Malia); Day and Haskell 1995 (Thebes); Day 1995a (Mycenae); Day 1999 (Iria); Rutter 2005 (Uluburun); Day and Joyner 2006 (Cannatello, Italy). See also Haskell et al. 2011, which includes petrographic analysis of 195 thin sec-tions of TSJs from a variety of sites in the Aegean and eastern Mediterranean, along with chemical analysis of 350.

19. For Chania, see, e.g., Stam- polidis, Karetsou, and Kanta 1998, pp. 57–58, nos. 4, 5. For other refer-ences, see Rutter 1999, p. 153, n. 34, and forthcoming.

20. Recent studies include Suger-man 2000; Smith, Bourriau, and Serpico 2000; Bourriau, Smith, and Serpico 2001; Smith et al. 2004; Cohen-Weinberger and Goren 2004.

21. Resins and oils: Serpico 1996, 2000; Bourriau, Smith, and Serpico 2001; Serpico et al. 2003. Wine: Leon-ard 1996.

22. See n. 11, above; Rutter 2006b, p. 712, nn. 215–216, 218.

23. A single flask fragment (C7550, from House X) is decorated with painted concentric circles.


are amphoras, but there are also several flasks and a pair of necked jars.24 Bourriau and her coworkers have published typologically similar vessels from Egypt.25 These jars have been linked primarily with the transportation of wine, mainly on the basis of their hieratic inscriptions.26 The association of the EJ with the elite-controlled production and exchange of an alcoholic drink makes their appearance on Crete very significant.

Essentially the same questions asked of the CJs were also asked of the Egyptian imports, although the latter exhibited far less variability of fabric at a macroscopic scale. Since the total number of Egyptian imports to Kommos was appreciably smaller, but at the same time represented a broader range of shapes within what appeared to be a single fabric, we decided to include samples from shapes other than amphoras, so that we could determine whether any significant petrographic or chemical differences could be detected among the distinct Egyptian closed shapes represented at the site (Table 1).27

Project Aims

Overall, then, the present analysis of transport jars from Kommos was concerned mainly with the determination of ceramic provenance, with the aim of contributing to a better understanding of the nature of contact and trade relations between the Minoan world and other Mediterranean centers in the Late Bronze Age. Specific goals included the separation of imported vessels from those produced on Crete, and the identification of compositional groups of common origin within both the local and imported groups of ceramics. By comparing the Cretan groups with our database of Minoan ceramics and Cretan geology, we hoped to locate their places of manufacture on the island. Where possible, we also identified comparative material for the nonlocal ceramics, drawing on studies conducted elsewhere in the Aegean, the Near East, and Egypt.

With these aims in mind, we decided that a combination of thin-section petrography and INAA was most appropriate for the project. The 88 coarse-ware transport jar samples are well suited to thin-section petrographic analysis, a technique that has proven very successful for the determination of provenance in coarse wares from Crete and the Aegean, as well as elsewhere in the Mediterranean.28 INAA has also been applied widely in provenance studies in the Aegean, the Near East, and Egypt.29 As a result, large databases of comparative INAA data exist for these areas, covering a wide range of elements, including many trace elements. Both analytical techniques have been used previously to characterize production of pottery at Kommos during the LM IA period.30

24. Flasks C10655 and C8006 were sampled as 98/34 and 98/44. The necked jars C9489 and C6392 (Rutter 2006a, p. 534, no. 57f/2, pl. 3.65, and p. 548, no. 61/7, pl. 3.73) were sampled as 98/40 and 98/45.

25. Bourriau, Smith, and Nicholson 2000; Bourriau 2004.

26. Bourriau 2004, p. 78; McGovern

1997; 2003, pp. 120–134, esp. fig. 6.4.27. As with the CJs, a number of

the 22 sampled Egyptian pieces turned out to have been inaccurately identified, so that the actual number of Egyptian samples was ultimately 19. See n. 11, above.

28. For Crete and the Aegean, see, e.g., Riley 1983; Wilson and Day 1994;

Whitelaw et al. 1997; Tomkins, Day, and Kilikoglou 2004.

29. Studies of Aegean material include Jones 1986; Kilikoglou et al. 2007; Mommsen and Sjöberg 2007; Newton et al. 2007.

30. Buxeda i Garrigós, Kilikoglou, and Day 2001; Shaw et al. 2001, pp. 111–133, 139–155.


PREViOUS ANALySES AND COMPARAT iVE MATERiAL

A number of previous petrographic and chemical analyses are relevant to our study of the Kommos transport jars. The analysis of prehistoric TSJs, especially those inscribed with Linear B, has been central to the field of Aegean ceramic analysis, and indeed has often been considered a test case for the effectiveness of analytical techniques. A series of papers published by the University of Oxford and the Fitch Laboratory of the British School at Athens from the 1960s to the 1980s created much interest and is summarized in an important review by Jones.31 This chemical work was supplemented by INAA conducted by the Manchester group in the early 1990s.32 Subsequent analyses, building on Riley’s pioneering petrographic study,33 have added detail to the picture of provenance by examining TSJs found on both Crete and the Greek mainland. A major program of analysis that includes large numbers of inscribed and uninscribed TSJs found over a wide area from Italy to Cyprus confirms that these jars had a number of different sources.34 Most of the “canonical” coarse-ware stirrup jars have an origin in West, Central, or East Crete. Although much attention had been directed toward north-central Crete and Knossos as a likely source, it appears that at least some of these Central Cretan ceramics were manu-factured in the south of the island.

Ceramics from the site of Kommos itself have been relatively well studied in such analyses. Myer and Betancourt characterized the products of the western Mesara at Kommos using thin-section petrography.35 Imported pottery has also received attention.36 The discovery of a Neopalatial kiln at Kommos led to a detailed set of analyses that aimed to establish a reference group for the site.37 INAA has also been conducted on a limited set of ceramics from Kommos by Hancock and Betancourt.38 The wider area of the Mesara and the foothills of the Asterousia Mountains to the south are also generally well covered by ceramic analyses of Bronze Age material.39

There are several petrographic studies of New Kingdom transport jars from Egypt, which are related typologically to the Late Minoan EJ samples found at Kommos.40 INAA has been the dominant chemical technique ap-plied to Egyptian pottery, and the results have been presented in a number of publications.41 McGovern has suggested that amphoras belonging to the Marl D group have an origin in the vicinity of Thebes, on the basis of comparisons with both natural clays and a control group of samples of

31. Jones 1986, pp. 477–493.32. Tomlinson 1995, 1996.33. Riley 1981.34. Haskell et al. 2011.35. Myer and Betancourt 1990.36. Watrous, Day, and Jones 1998.37. Shaw et al. 1997, 2001; Buxeda i

Garrigós, Kilikoglou, and Day 2001. For the products from another roughly contemporary kiln at a nearby site, see Belfiore et al. 2007.

38. Hancock and Betancourt 1987; Tomlinson, Rutter, and Hoffmann 2010.

39. We have an extensive collection of data for most major Bronze Age sites on the island. In Central and East Crete, coverage is most complete for the Early Bronze Age, but we also have substantial comparative datasets for the Late Bronze Age. Our datasets for the site of Knossos are especially detailed for both periods. In West Crete we

have nearly complete coverage for the Early Bronze Age, with some Late Bronze Age material from Chania.

40. Nicholson and Rose 1985; Bourriau and Nicholson 1992; Bour-riau, Smith, and Nicholson 2000.

41. For INAA of Egyptian jars, see Allen et al. 1982; Allen and Hamroush 1984; Hancock, Millet, and Mills 1986; Redmount and Morgenstein 1996; McGovern 1997, 2000.


supposed Theban origin.42 The validity of this group has been challenged, however, as it appears that some samples have been assumed rather than demonstrated to be local Theban products. Bourriau and her colleagues have analyzed samples of the Marl D group by INAA and petrography and suggested that they were made in the Memphis region, close to the Delta and Faiyum vineyards, from Nile clays, or a mix of Nile and marl clays.43

Several petrographic studies have addressed the origins of CJs with considerable success. Sugerman has studied the movements of these jars,44 while Smith’s thin-section analyses have established a petrographic typol-ogy, not only suggesting source areas for each type, but also correlating individual sources with the transport of specific goods.45 This work has benefited greatly from the petrographic analyses of Goren and his cowork-ers, who have carried out extensive work on the ceramics of the Levantine coast, including the characterization of the clays used to produce the Amarna tablets.46 A more recent study examined the transport jars found at the site of Tell el Dab‘a in the Nile Delta.47

With the availability of such a range of comparative chemical and geological information, and with the emphasis of the project focused firmly on provenance, an integrated petrographic and chemical approach was chosen in order to better characterize and discriminate groupings. Some of the data from the project has been used to develop and test the first true mixed-mode data procedures.48

P ETRO GRAP H iC ANALySiS

The 88 transport jar samples from Kommos were prepared as ceramic thin sections and studied with a polarizing microscope. Petrographic fabrics were classed according to a variety of criteria, including the mineralogy and texture of the dominant inclusions and the nature of the clay micromass, and described according to a modified version of the system proposed by Whitbread.49

We divided the thin sections into a total of 26 petrographic fabric groups (Table 2). When compared with the typology of the transport jars, these groups fall neatly into fabrics considered Cretan (TSJ and SNA = fabrics A–J) and those considered imports (CJ and EJ = fabrics 1–12). The broad provenances of these groups on petrographic grounds are in close agreement with those based on typology.50

Summaries of the main characteristics of each of the 26 petrographic fabric groups are given below and illustrated with thin-section micrographs of a sample of each fabric (Figs. 3–6). These summaries indicate the min-eralogical, petrographic, and textural features used to group and separate

42. McGovern 2000.43. Bourriau, Smith, and Nicholson

2000.44. Sugerman 2000.45. Smith, Bourriau, and Serpico

2000; Bourriau, Smith, and Serpico

2001; Smith et al. 2004.46. Goren et al. 2003; Goren, Fink-

elstein, and Na’aman 2003. 47. Cohen-Weinberger and Goren

2004.48. Baxter et al. 2008.

49. Whitbread 1989, 1995.50. One exception is fabric 9, which

consists of a sample originally classified as an EJ (98/53), but which appears to be Cretan on the basis of its petrogra-phy (see n. 11, above).


the thin sections, and give an impression of the range of variation in groups containing many samples. The probable origin of each group is discussed in the light of comparisons with our database of thin sections of Bronze Age ceramics and the previous studies summarized above. We have also commented on a range of technological aspects of the fabrics, where these can be interpreted from detailed analysis of the thin sections. Full descrip-tions of each petrographic fabric can be found in online Appendix 1.51

Cre tan Pe trographic Fabr ics A–J

Fabric A: Main South-Central Cretan

Kommos 98/1, 2, 4, 7, 9, 11, 14, 16, 17, 22, 24, 26–30, 32

This fabric is characterized by the occurrence of a wide range of coarse-grained, rounded, aplastic inclusions, which do not occur within every sample (Fig. 3:a–c). The aplastics include volcanic rock fragments, meta-morphic rock fragments, and siltstones. In spite of the differences between individual samples, they form a coherent, if heterogeneous, fabric.

51. See http://dx.doi.org/10.2972/hesperia.80.4.0511.app1.

TABLE 2. P ETRO GRAP H iC FABRiC CLASSiFiCAT iON OF LATE MiNOAN TRANSPORT JAR SAMPLES FROM KOMMOS

Fabric1 Description Samples

A Main south-central Cretan Kommos 98/1, 2, 4, 7, 9, 11, 14, 16, 17, 22, 24, 26–30, 32B Medium/coarse igneous Kommos 98/15, 19, 20, 87, 88C Serpentine Kommos 98/21, 25D Fine calcareous phyllite Kommos 98/3, 13E Metasedimentary Kommos 98/18, 23F Quartzite, phyllite, and schist Kommos 98/6, 31G Phyllite, siltstone, and chert Kommos 98/5H Phyllite Kommos 98/8I Quartz, polycrystalline quartz, schist, and microfossil Kommos 98/12J Schist Kommos 98/101 Quartz and calcite Kommos 98/33–37, 39–45, 47–49, 52, 66, 761a Quartz and calcite with chert Kommos 98/51, 64, 73, 842 Macrofossiliferous clay pellet Kommos 98/56, 59, 68, 71, 72, 74, 75, 78, 80, 81, 863 Quartz and clay pellet Kommos 98/60, 63, 67, 853r Quartz and clay pellet related sample Kommos 98/824 Quartz, chert, and macrofossil Kommos 98/70, 79, 832/4r Fabrics 2 and 4 related sample Kommos 98/615 Chert Kommos 98/58, 655r Chert related sample Kommos 98/626 Fusilinid Kommos 98/50, 547 Serpentine and micrite Kommos 98/55, 698 Polycrystalline quartz and calcite Kommos 98/779 Quartz and metamorphic Kommos 98/5310 Micaceous quartz and feldspar Kommos 98/3811 Quartz and schist Kommos 98/4612 Alkali feldspar Kommos 98/57

1 A–J = fabrics represented by TSJ and SNA samples (Cretan); 1–12 = fabrics that include CJ or EJ samples (imported). Abbreviations: a = subfabric; r = related sample.


Comparative material for fabric A comes from a variety of Early Mi- noan (EM) sites in the Mesara Plain of Crete, from kilns at Kommos and Ayia Triada,52 and from Neopalatial storage jars found at Kannia, Phais-tos, Kommos, and Ayia Triada. Petrographically, it is very similar to the fabric encountered in previous analyses of coarse-ware stirrup jars, includ-ing many of those from Kommos.53 The petrographic variability in the fabrics of the Mesara and the difficulty of discriminating between them has been noted elsewhere.54 For this reason, it is impossible to determine the precise provenance of fabric A. The presence of fine-grained volcanic rock fragments is well documented in ceramics at Kommos, however, and has been taken to be diagnostic of an origin in the western Mesara.55 It is notable that fabric A includes both TSJ and SNA samples, some of which are indistinguishable in thin section. This is a clear indication that at least some vessels of these two types were manufactured at the same location in the western Mesara.

Fabric B: Medium/Coarse Igneous

Kommos 98/15, 19, 20, 87, 88

Fabric B is characterized by the occurrence of very coarse to very fine, me-dium-grained, acid or intermediate igneous rock fragments (cf. granodiorite) and their constituent minerals, set in a clay-rich micromass (Fig. 3:d). Most samples seem to have been moderately well fired in an oxidizing atmosphere, although the low optical activity of sample 98/87 suggests that it was fired at a much higher temperature. Granodiorite inclusions have been found in EM fabrics from Moni Odigitrias and Ayia Kyriaki in southern Crete. Rocks of this composition occur in the foothills of the nearby Asterousia Mountains, which border the southern part of the Mesara Plain.56 Fabric B differs from the well-documented granodiorite fabric found in the Mirabello Bay area of East Crete.57

Fabric C: Serpentine

Kommos 98/21, 25

The two samples in this fabric are characterized by the presence of very coarse to fine inclusions of altered serpentine, igneous rock fragments, siltstone, phyllite, and schists, set in a noncalcareous clay with sparse, very fine quartz and biotite (Fig. 3:e). Fabric C is strongly associated with TSJs, which contain chert and serpentine.58 It has its origin in the ophiolite

52. Shaw et al. 2001; Belfiore et al. 2007.

53. Day 1995a; Day and Haskell 1995; Haskell et al. 2011, pp. 58–60, fabric 11.

54. Wilson and Day 1994, pp. 57– 77; Shaw et al. 2001, pp. 116–119, 141–150.

55. Myer and Betancourt 1990, pp. 9–10, pl. A; Shaw et al. 2001, p. 118.

56. Wilson and Day 1994, pp. 54– 57; Bonneau, Jonkers, and Meulenkamp 1984.

57. Betancourt 1984; Day 1995b; Tomkins and Day 2001.

58. For the chert and serpentine fabric, see Day and Haskell 1995 (Thebes); Haskell et al. 2011, pp. 52– 54, fabric 8.


Figure 3. Thin-section micrographs of LM transport jars from Kommos: (a) fabric A (Kommos 98/1); (b) fabric A (Kommos 98/27); (c) fabric A (Kommos 98/9); (d) fabric B

(Kommos 98/15); (e) fabric C (Kommos 98/25); (f ) fabric D (Kommos 98/3); (g) fabric E (Kommos 98/23); (h) fab-ric F (Kommos 98/6). Field of view = 2.0 mm.


Figure 4. Thin-section micrographs of LM transport jars from Kommos: (a) fabric G (Kommos 98/5); (b) fabric H (Kommos 98/8); (c) fabric i (Kommos 98/12); (d) fabric J

(Kommos 98/10); (e) fabric 1 (Kommos 98/43); (f ) sub-fabric 1a (Kommos 98/84); (g) fabric 2 (Kommos 98/56); (h) fabric 3 (Kommos 98/60). Field of view = 2.0 mm.



Figure 5 (opposite). Thin-section micrographs of LM transport jars from Kommos: (a) fabric 3 related sample (Kommos 98/82); (b) fabric 4 (Kommos 98/70); (c) fabrics 2 and 4 related sample (Kommos 98/61); (d) fabric 5 (Kommos 98/58); (e) fab-ric 5 related sample (Kommos 98/62); (f ) fabric 6 (Kommos 98/54); (g) fab-ric 7 (Kommos 98/69); (h) fabric 8 (Kommos 98/77). Field of view = 2.0 mm.

Figure 6 (above). Thin-section micrographs of LM transport jars from Kommos: (a) fabric 9 (Kom- mos 98/53); (b) fabric 10 (Kommos 98/38); (c) fabric 11 (Kommos 98/46); (d) fabric 12 (Kommos 98/57). Field of view = 2.0 mm.

complex of Crete, which is present in a variety of locations in the central and eastern parts of the island. Comparable samples come from Knossos and a number of other sites, including Thebes, Enkomi, and the Uluburun shipwreck, as well as from Malia,59 although the geology of the area around this last site does not seem compatible with their petrography.

Fabric D: Fine Calcareous Phyllite

Kommos 98/3, 13

This fabric is characterized by the presence of rare, very coarse phyllite grains set in fine calcareous clay containing monocrystalline quartz, mica, micrite, and microfossils (Fig. 3:f ). The two samples are texturally very different, although this can be explained by the wide variation in grain size and relative quartz content of calcareous clays from Crete,60 and by the deliberate addition of phyllite temper. Both samples show strong optical activity, indicating a relatively low firing temperature. This fabric stands out from that of all other TSJs in the study, a fact that most likely reflects the very early date (LM II) of the vessels, and perhaps their provenance as well.

Fabric E: Metasedimentary

Kommos 98/18, 23

This fabric is characterized by inclusions of micrite and a range of meta-morphic, sedimentary, and metasedimentary rock fragments (Fig. 3:g). The paste appears to have been produced by the mixing of a light-colored

59. Day and Jones 1991.60. Hein et al. 2004.


calcareous clay with a darker, noncalcareous clay. Both samples were fired at relatively high temperatures in an oxidizing atmosphere. Fabric E has no matches in the comparative material that we have examined. In geological terms, however, it is not incompatible with an origin in Central Crete.

Fabric F: Quartzite, Phyllite, and Schist

Kommos 98/6, 31

The two samples in this fabric are characterized by the presence of very coarse to very fine mono- and polycrystalline quartz, schist, and phyllite inclusions, plus textural features (TFs), set in a fine, noncalcareous clay with quartz, micas, and metamorphic rock fragments (Fig. 3:h). The two samples, which are of different dates, contain different proportions of these inclusions. The fabric is related to fabric I (discussed below), but differs in texture. Fabric I also contains micrite and microfossils, which are absent from fabric F. Sample 98/31 is very similar to Day’s West Cretan transport jar group 1,61 as well as Riley’s West Cretan group of transport jars from Mycenae.62 The sample contains frequent quartzite, which is characteristic of fabrics found in West Crete, specifically in EM pottery and modern geological samples from the Chania Plain.63

Fabric G: Phyllite, Siltstone, and Chert

Kommos 98/5

Fabric G is represented by one sample containing coarse, usually well-rounded phyllite, chert, and siltstone inclusions in a noncalcareous, mica-ceous clay (Fig. 4:a). It has been highly fired in an oxidizing atmosphere. The sample is very similar to Day’s siltstone/igneous/chert transport jar group 9 and fabric 2B found in the Kommos kiln.64 It may also be related to our fabric C, which is dominated by serpentine. Fabric G is likely to have an origin in Central Crete.

Fabric H: Phyllite

Kommos 98/8

This sample is characterized by the presence of large phyllite rock fragments and equant quartz inclusions, contained in a fine, noncalcareous clay with quartz, phyllite, and biotite (Fig. 4:b). Phyllite fabrics of this type can be related to the phyllite-quartzite series of rocks that occur in various loca-tions on Crete. The closest comparative archaeological material is from East Crete, although exact parallels are difficult to find.65 The petrography of the sample does not seem to support a typological link with Kos. While Koan amphoras sometimes have phyllite inclusions, they are usually found in association with volcanic glass, which is not present in this sample.66

61. Day 1995a, pp. 311–312; Day and Haskell 1995, pp. 90–91; Haskell et al. 2011, pp. 42–46.

62. Riley 1981.63. Nodarou 2011, pp. 20–26,

42–46, pls. 14, 15.64. For group 9, see Day 1999,

pp. 65–67; Haskell et al. 2011, pp. 54–56. For fabric 2B from the Kommos kiln, see Shaw et al. 2001,

pp. 116, 145.65. Day 1995b, 1997; Poursat and

Knappett 2005, pp. 18–19.66. Whitbread 1995, pp. 83–106.


Fabric I : Quartz, Polycrystalline Quartz, Schist, and Microfossils

Kommos 98/12

Fabric I is characterized by the presence of very coarse- to medium-grained quartz, polycrystalline quartz, micrite, quartzite, quartz/mica schist, and phyl-lite inclusions and dark red-brown TFs in an orange-brown noncalcareous micromass containing abundant fine quartz, mica, and microfossils (Fig. 4:c). It appears to be related to fabric F, but differs in texture. Although the mixing of calcareous and noncalcareous material is known in West Cre-tan pottery, the large proportion of micrite inclusions in sample 98/12 is uncharacteristic of ceramics from this area of Crete. The provenance of fabric I remains uncertain.

Fabric J : Schist

Kommos 98/10

This sample contains a distinctive range of medium- to coarse-grained quartz, biotite, and hornblende schist fragments (Fig. 4:d). These occur along with mineral grains of metamorphic origin in a fine-grained, clay-rich micromass containing quartz, biotite, and calcite. The suite of metamorphic rock fragments in the sample appears to have been added as temper to a fine, noncalcareous base clay. Comparable samples come from EM pottery found at sites in the Asterousia Mountains to the south of the Mesara, deriving from hornblende schists in the foothills surrounding the plain.

Imp orted Pe trographic Fabr ics 1–12

Fabric 1 : Quartz and Calcite

Kommos 98/33–37, 39–45, 47–49, 52, 66, 76

This fabric is characterized by the presence of calcite and well-rounded quartz inclusions, set in red-firing clay containing fine calcite, quartz, microfossils, and hornblende (Fig. 4:e). There is clear evidence of clay mixing in all samples except 98/45, which is more homogeneous. It ap-pears that a fine, marly clay was mixed with a clay firing red-brown, which contained the rounded quartz grains. All samples are high-fired. Most of the calcareous inclusions are micritic. Their internal features have been lost during firing and the subsequent formation of secondary calcite. The size and abundance of quartz, calcite, TFs, and voids vary. Sample 98/76 contains a well-rounded basaltic inclusion; sample 98/45 has less frequent calcite inclusions.

This fabric is dominant among EJ samples from Kommos. The inclusions are compatible with those found in Egyptian ceramics, which commonly contain rounded sand-grade quartz (aeolian quartz), together with feldspar and mica.67 The calcareous nature of the fabric could link it with the marl groups in Egypt, although its relationship with the mixed marl and Nile silt fabrics is not clear.68 Differences between our petrographic analysis and the approach taken in comparative studies of Egyptian pottery make

67. Bourriau and Nicholson 1992, pp. 37–41.

68. For such mixed fabrics, see Bourriau, Smith, and Nicholson 2000.


it difficult to place this fabric with confidence. Nevertheless, it is more or less compatible with the marl groups described from Memphis, Saqqara, and Amarna, especially those with higher proportions of fine biotite in the matrix. Sample 98/76, containing a volcanic rock inclusion, is the only sample of this fabric classified typologically as a CJ.

Subfabric 1a: Quartz and Calcite with Chert

Kommos 98/51, 64, 73, 84

The samples in this fabric, while containing inclusions broadly similar to those found in fabric 1, also show some important differences. They contain chert, volcanic rock fragments, and ferruginous concentrations, as well as more foraminifera, less plagioclase feldspar, and markedly less hornblende (Fig. 4:f ). There are textural differences as well: subfabric 1a has a finer coarse fraction that is better sorted and more densely packed. The groundmass is optically more active than that of fabric 1, reflecting a difference in firing temperature between the two groups of samples.

While the samples are petrographically similar, it appears that those of subfabric 1a may have a very different provenance from that of the main group. The presence of rounded quartz and calcite is characteristic of Egyptian fabrics, but subfabric 1a is a better match for Smith’s group 1 of Canaanite jars from Israel.69 This group is likewise characterized by rounded quartz inclusions, which are thought to have come from beach sand on the Israeli coast. It also contains fossiliferous remains and may be linked to the Jezreel Valley, in the environs of Tell Abu Hawam, specifically between Haifa and Akko.70

Fabric 2 : Macrofossiliferous Clay Pellet

Kommos 98/56, 59, 68, 71, 72, 74, 75, 78, 80, 81, 86 (related: 98/61)

This homogeneous, well-defined fabric is characterized by the presence of large macrofossils and clay pellets, with medium-sized calcite, chert, and microfossil inclusions, set in calcareous clay with fine calcite and quartz (Fig. 4:g). The microfossils (foraminifera) occur as part of the marly base clay, whereas the macrofossils (coralline algae) may have come from a bio-clastic limestone added as temper. The presence of red-brown clay pellets attests to the mixing of terra rossa with a marly clay body (e.g., sample 98/71). In this respect, fabric 2 is related technologically to fabric 3. In terms of provenance it is also clearly related to fabric 4, only differing in the relative proportion of macrofossil fragments, monocrystalline quartz grains, clay pellets, and chert.

Fabric 2 is identical to Smith’s petrographic group 5 of Canaanite jars from the New Kingdom.71 The distinctive fossils and other inclusions in this group permitted a relatively precise ascription to a source on the Lebanese coastal plain.72 Similar fossils have been found in thin sections of tablets sent to Amarna from Amurru, which also suggests a Lebanese provenance for fabric 2.73 Our samples correspond to the low-quartz variety of Smith’s group 5, which would place them on the coastal Akkar Plain in the Lebanon/Syria region.74

69. Bourriau, Smith, and Serpico 2001, pp. 116–121; Smith et al. 2004, pp. 57–58.

70. Bourriau, Smith, and Serpico 2001, pp. 116–121, 140, pl. 7:22–26; Smith et al. 2004, pp. 56–58.

71. Bourriau, Smith, and Serpico 2001, pp. 132–136, 143, pl. 7:37–40; Smith et al. 2004.

72. Smith et al. 2004, pp. 71–73.73. Goren, Finkelstein, and

Na’aman 2003.74. Smith et al. 2004, pp. 62–63,

71–73.


Fabric 3 : Quartz and Clay Pellet

Kommos 98/60, 63, 67, 85 (related: 98/82)

This fabric is characterized by very coarse micrite inclusions, rounded TFs, monocrystalline quartz, and microfossils in a calcareous clay containing fine quartz and micrite (Fig. 4:h). Like fabric 1, it has abundant micrite and quartz. In comparison with fabric 2, it has fewer macrofossils and a greater abundance of monocrystalline quartz inclusions. Fabric 3 corresponds to Smith’s group 2 from Aphek, Jaffa, Tel Hefer, and other sites on the Israeli coastal plain.75 The red staining present in some of the samples derives from an iron-stained soil known as hamra, which occurs on the coast.76 This group has an origin along the part of the coastal plain that borders the Carmel ridge, from Haifa to Yavneh Yam.77

Sample 98/82 (Fig. 5:a), which is related to fabric 3, features a large calcareous inclusion containing a coralline macrofossil body of the type seen in fabrics 2 and 4. It is possible that the bioclastic limestone inclusion is the origin of the macrofossil fragments in these two fabrics. Smith’s group 5 (related to fabrics 2 and 4) and group 2 (related to fabric 3) have adjacent source areas on the Levantine coast.78

Fabric 4 : Quartz, Chert, and Macrofossils

Kommos 98/70, 79, 83 (related: 98/61)

This fabric is characterized by the presence of coarse, rounded calcite, macrofossils, quartz, and chert inclusions set in a calcareous micromass con-taining microfossils, fine quartz, calcite, and opaques (Fig. 5:b). It appears to have been produced by the addition of weathered fossiliferous mate-rial and quartz temper to a fine calcareous clay. The base clay may have been prepared by the mixture of two or more materials, one of which was high in organic matter, as indicated by streaking in samples 98/70 and 98/79, and by the organic-rich TFs that occur in many sections. Fabrics 2 and 4 are closely related by their calcareous composition and the occurrence of both rounded quartz and macrofossils of coralline algae. Like fabric 2, fabric 4 is related to Smith’s group 5 and is likely to have an origin in the Lebanese coastal plain.79 Fabric 4 corresponds to the higher-quartz subgroup of Smith’s group 5, which is believed to be from the coastal area between Akko and Sidon.

Sample 98/61 is related to fabrics 2 and 4 (Fig. 5:c). It contains many of their characteristic inclusions in its coarse fraction, including large cor-alline macrofossils, chert, several different types of clay pellets, planktonic and benthic foraminifera, and monocrystalline quartz grains. This sample has a less abundant coarse fraction than fabrics 2 and 4.

Fabric 5 : Chert

Kommos 98/58, 65 (related: 98/62)

The two samples that constitute fabric 5 are composed of sand-tempered calcareous clays containing rounded chert inclusions, radiolaria, foramin-ifera, spinel, opaques, and rare serpentine (Fig. 5:d). Both samples have

75. Bourriau, Smith, and Serpico 2001, pp. 121–125, 140, pl. 7:27–30; Smith et al. 2004, pp. 63–64.

76. Cohen-Weinberger and Goren 2004, pp. 77–78.

77. Singer-Avitz and Levy 1992; Smith et al. 2004, pp. 58–60, 63–64.

78. Smith et al. 2004.79. Bourriau, Smith, and Serpico

2001, pp. 132–136, 143, pl. 7:37–40; Smith et al. 2004, pp. 71–73.


a bimodal grain-size distribution, but differ in other textural and micro-structural characteristics. Sample 98/58 has a greater frequency of voids, a more abundant coarse fraction, and a stronger alignment of inclusions than sample 98/65. This fabric appears to have been prepared from a mixture of calcareous and noncalcareous clay, as indicated by the ptygmatic texture in sample 98/58 and the TFs in both sections. The presence of calcite, microfossils, radiolarian chert, and serpentine inclusions links this fabric with subgroup 1.1 of Smith’s group 4, which has a “high percentage of ophiolite.”80 It is thus related to fabric 7, which is linked to the ophiolite complex of the northwestern Syrian coast.81

Sample 98/62 is related to fabric 5, but differs in its elongate phyllite inclusions and igneous rocks, as well as the nature of its chert inclusions (Fig. 5:e). It is similar to a distinctive but problematic group of material in Smith’s classification. That group was originally considered to be related to the ophiolite series and classified as subgroup 4.2.2.82 In the new scheme, however, it is placed in a separate group (Smith’s group 6) and assigned an origin in southern Cyprus.83 It is therefore possible that sample 98/62 may also have an origin in Cyprus.

Fabric 6 : Fusilinid

Kommos 98/50, 54

This highly distinctive fabric consists of two almost identical sections characterized by the occurrence of large fusilinid foraminifera (microfossils) and micrite inclusions, together with medium-sized rounded quartz, set in a dark, high-fired clay (Fig. 5:f ). Both samples contain a high percentage of large vughs and elongate voids, which are aligned to the margins of the sections. They were high-fired in a reducing atmosphere. Fabric 6 is unlike that of any other transport jar sample analyzed in this project and cannot be correlated with any of the comparative material. It may be compatible with a marl mixed with large fragments of bioclastic limestone. Further comparative material should be sought in Egypt.

Fabric 7 : Serpentine and Micrite

Kommos 98/55, 69

This fabric is characterized by the presence of rounded, medium- to coarse-grained serpentine and micrite inclusions (Fig. 5:g). It differs from fabric C by the presence of micritic inclusions and microfossils and matches Smith’s group 4.1.1, which is characteristic of the ophiolites of the Baër-Bassit complex of northwestern Syria.84 Ceramics with this composition have been found at Ras Shamra (Ugarit),85 while thin-section analysis of two letters sent from Ugarit to Amarna have revealed a similar composition.86

80. Smith et al. 2004, pp. 65–68.81. Cohen-Weinberger and Goren

2004, pp. 71–73.82. Bourriau, Smith, and Serpico

2001, pp. 127–132.83. Smith et al. 2004, pp. 68–70.

84. Bourriau, Smith, and Serpico 2001, pp. 127–132, 142–143, pl. 7:33–36; Cohen-Weinberger and Goren 2004, pp. 71–73.

85. Smith et al. 2004, pp. 60–61.86. Goren et al. 2003.


Fabric 7 is therefore suspected to have originated on the Syrian coast north of Latheqieh, possibly around Ras Shamra.

Fabric 8 : Polycrystalline Quartz and Calcite

Kommos 98/77

This fabric is characterized by medium-grained, rounded polycrystalline quartz and micrite inclusions (Fig. 5:h). It was produced by a mixture of rounded, medium-sized temper with a fine, highly calcareous, mi-crofossiliferous clay. It is related to several other fabrics in our study; in particular, its calcareous, fossiliferous character and the metamorphic polycrystalline quartz suggest a link with sample 98/62, which is related to fabric 5 and may have originated on Cyprus. It does not, however, match any one group in Smith’s classification.

Fabric 9 : Quartz and Metamorphic

Kommos 98/53

This fabric is characterized by coarse- to fine-grained, rounded monocrystalline quartz, polycrystalline quartz, metamorphic rock fragments, and elongate phyllite and schist inclusions in a red, noncalcareous, fine-grained micromass with extensive clay-mixing features (Fig. 6:a). While it is not closely related to any other samples in our study, its mineralogy is compatible with the low-grade metamorphic geology of south-central Crete. We take this vessel to be Cretan and to have been misclassified in macroscopic sorting.

Fabric 10 : Micaceous Quartz and Feldspar

Kommos 98/38

This fabric contains moderately well-sorted, coarse- to fine-grained, rounded quartz and feldspar inclusions in a dense, noncalcareous micro-mass containing fine quartz, plagioclase feldspar, and abundant mica laths (Fig. 6:b). The feldspar and quartz coarse fraction is likely to have been added as temper to a finer base clay. The presence of TFs with abundant fine micas, quartz, and plagioclase suggest that the base clay may have been produced by blending two or more different types of material. The source of the temper is likely to have been an intermediate igneous rock, and the fabric is suspected of having been produced outside of Crete.

Fabric 11 : Quartz and Schist

Kommos 98/46

Fabric 11 is characterized by coarse- to very fine-grained monocrystalline quartz, polycrystalline quartz, micrite, quartzite/cataclasite, and phyllite inclusions set in a fine, noncalcareous micromass with quartz and micas (Fig. 6:c). In addition to the range of metamorphic rock fragments, the sample contains rare igneous rocks. It was clearly imported to Crete, but its origin is uncertain.


Fabric 12 : Alkali Feldspar

Kommos 98/57

This fabric is characterized by the presence of very coarse- to fine-grained alkali feldspar inclusions, intermediate and acid volcanic igneous rock, schist and phyllite fragments, TFs, and opaques set in a noncalcareous clay with fine quartz, feldspar, micas, and opaques (Fig. 6:d). Many of the rock and mineral inclusions in this fabric may have originated from a dark red-brown clay, which forms the distinctive TFs. This clay may have been mixed with another, lighter clay. Although it is clearly an import to Crete, as yet we have no comparative data for this fabric.

iNSTRUMENTAL NEU TRON ACT iVAT iON ANALySiS

All 88 transport jar samples were analyzed chemically by INAA. The ex-ternal surface of each sample was cleaned with a tungsten carbide drill bit. The sherds were then ground to a fine powder in an agate mortar and dried at 110°C overnight. Approximately 130 mg of each dry, powdered sample was aliquoted into a polyethylene vial, which was then heat-sealed. The vials were irradiated in batches of 10, along with two reference samples, at a thermal neutron flux of 5 x 1013 n.cm–2.s–1, in a swimming pool reactor at the National Center for Scientific Research “Demokritos” in Athens. Eight days after irradiation, the γ-spectra of the samples and reference materials were recorded in order to determine the concentrations of the elements Sm, Lu, Yb, U, As, Sb, Ca, Na, K, and La. The same samples were also measured two weeks later in order to determine the concentra-tions of the following isotopes with longer half-lives in spectra with lower backgrounds: Ce, Th, Cr, Hf, Cs, Tb, Sc, Rb, Fe, Ta, Co, Eu, Ba, Ni, Zr, Zn, and Nd. The full analytical data for the 27 elements analyzed in each of the 88 samples are presented in online Appendix 2.87

Sixteen of the 27 elements analyzed (Sm, Lu, Yb, Ca, Na, La, Ce, Th, Cr, Hf, Cs, Sc, Rb, Fe, Co, and Eu) were chosen for statistical analysis. The elements As and Sb were excluded because of their high natural fluctuation, which could introduce unrealistic variability to the data. Seven elements (U, Tb, Ta, Ba, Ni, Zr, and Nd) were not included because of their low count in the transport jar samples. Finally, K and Zn were omitted because they are not present in the comparative data sets that we have used.

A comprehensive picture of the chemical variability within the INAA dataset can be obtained by the calculation of a variation matrix using the 16 elements listed above, along with U and Ta (Table 3).88 In a variation matrix, the variance of each element (columns) is expressed as a logarith-mic ratio over that of all other elements (rows). These values are used to calculate the sum of the variances, or total variance, for each individual element when it is used as a divisor. The results provide important infor-mation about the covariance structure of the data and the effect of one elemental concentration on the others. We have also calculated the total variation (Var.T.) of the dataset from the sum of all variances divided by twice the number of elements included in the matrix. This calculation helps

87. See http://dx.doi.org/10.2972/hesperia.80.4.0511.app2.

88. On the variation matrix and the logarithmic ratio transformation of data, see Buxeda i Garrigós and Kiliko-glou 2003.


TABLE 3. CHEMiCAL VARiATiON MATRix FOR 18 OF THE ELEMENTS ANALyzED wi TH iNAA iN LATE MiNOAN TRANSPORT JAR SAMPLES FROM KOMMOS

Sm Lu U Yb Ca Na La Ce Th

Sm 0.0000 0.0201 0.0678 0.0183 0.5474 0.3619 0.0103 0.008 0.1094Lu 0.0201 0.0000 0.0667 0.0020 0.5900 0.2868 0.0217 0.0165 0.0641U 0.0678 0.0667 0.0000 0.0767 0.5853 0.3571 0.0690 0.0826 0.1355Yb 0.0183 0.0020 0.0767 0.0000 0.5963 0.2897 0.0189 0.0131 0.0623Ca 0.5474 0.5900 0.5853 0.5963 0.0000 1.1546 0.6343 0.6294 0.9052Na 0.3619 0.2868 0.3571 0.2897 1.1546 0.0000 0.3411 0.3291 0.2490La 0.0103 0.0217 0.0690 0.0189 0.6343 0.3411 0.0000 0.0064 0.0710Ce 0.0085 0.0165 0.0826 0.0131 0.6294 0.3291 0.0064 0.0000 0.0687Th 0.1094 0.0641 0.1355 0.0623 0.9052 0.2490 0.0710 0.0687 0.0000Cr 0.8705 0.6995 0.8585 0.7107 1.3845 0.7090 0.8429 0.8360 0.6601Hf 0.0750 0.0830 0.1839 0.0771 0.7077 0.3880 0.0816 0.0643 0.1638Cs 0.6863 0.5770 0.5783 0.5836 1.6921 0.5137 0.6018 0.6270 0.3850Sc 0.1209 0.0601 0.1468 0.0666 0.7064 0.1998 0.1222 0.1075 0.0918Rb 0.4017 0.3209 0.3280 0.3281 1.3249 0.3160 0.3465 0.3549 0.1900Fe 0.1026 0.0523 0.1211 0.0608 0.6480 0.2376 0.1136 0.0954 0.1064Ta 0.0345 0.0516 0.1126 0.0500 0.6042 0.3482 0.0485 0.0338 0.1463Co 0.2836 0.1854 0.3183 0.1950 0.8146 0.2754 0.2994 0.2686 0.2432Eu 0.0080 0.0262 0.0804 0.0263 0.4899 0.3731 0.0293 0.0231 0.1405

Total 3.7269 3.1240 4.1687 3.1757 14.014 6.7300 3.6586 3.5651 3.7927vt/ 0.7716 0.9205 0.6898 0.9056 0.2052 0.4273 0.7860 0.8066 0.7582rTotal 0.9379 0.9826 0.9585 0.9808 0.8878 0.8638 0.9627 0.9656 0.9484

Cr Hf Cs Sc Rb Fe Ta Co Eu

Sm 0.8705 0.0750 0.6863 0.1209 0.4017 0.1026 0.0345 0.2836 0.0080Lu 0.6995 0.0830 0.5770 0.0601 0.3209 0.0523 0.0516 0.1854 0.0262U 0.8585 0.1839 0.5783 0.1468 0.3280 0.1211 0.1126 0.3183 0.0804Yb 0.7107 0.0771 0.5836 0.0666 0.3281 0.0608 0.0500 0.1950 0.0263Ca 1.3845 0.7077 1.6921 0.7064 1.3249 0.6480 0.6042 0.8146 0.4899Na 0.7090 0.3880 0.5137 0.1998 0.3160 0.2376 0.3482 0.2754 0.3731La 0.8429 0.0816 0.6018 0.1222 0.3465 0.1136 0.0485 0.2994 0.0293Ce 0.8360 0.0643 0.6270 0.1075 0.3549 0.0954 0.0338 0.2686 0.0231Th 0.6601 0.1638 0.3850 0.0918 0.1900 0.1064 0.1463 0.2432 0.1405Cr 0.0000 1.0722 0.8451 0.5227 0.7649 0.5778 0.9614 0.3871 0.8552Hf 1.0722 0.0000 0.8404 0.1904 0.5187 0.1665 0.0473 0.3615 0.0822Cs 0.8451 0.8404 0.0000 0.5168 0.0991 0.5780 0.8056 0.6484 0.7378Sc 0.5227 0.1904 0.5168 0.0000 0.2785 0.0106 0.1295 0.0730 0.1121Rb 0.7649 0.5187 0.0991 0.2785 0.0000 0.3123 0.4776 0.4097 0.4397Fe 0.5778 0.1665 0.5780 0.0106 0.3123 0.0000 0.1034 0.0772 0.0904Ta 0.9614 0.0473 0.8056 0.1295 0.4776 0.1034 0.0000 0.2864 0.0373Co 0.3871 0.3615 0.6484 0.0730 0.4097 0.0772 0.2864 0.0000 0.2578Eu 0.8552 0.0822 0.7378 0.1121 0.4397 0.0904 0.0373 0.2578 0.0000

Total 13.558 5.1035 11.316 3.4557 7.2113 3.4539 4.2780 5.3846 3.8094vt/ 0.2122 0.5635 0.2541 0.8322 0.3988 0.8326 0.6722 0.5341 0.7549rTotal 0.5547 0.9483 0.6019 0.9695 0.6384 0.9807 0.9372 0.8267 0.9187

S.T.V 103.526824Var.T. 2.875745


to determine whether one or more groups are present in the dataset. The total variation of 2.9 in our dataset indicates that we can expect to find chemically distinct groups in the data.

From the variation matrix it is clear that the elements Ca, Cr, and Cs introduce the highest variation when used as a divisor. In comparison, Lu, Yb, Sc, and Fe produce the lowest individual variance values and the lowest total variance. This means that the variance in these elements is probably natural or due to mixing or tempering, rather than to alteration or con-tamination. Scandium (Sc) has the second lowest variability in the dataset and is a lithophilic, immobile element. We therefore chose to express the concentration of each of the 18 elements in our dataset as a logarithmic ratio over the concentration of Sc.

The logarithmic ratios of all elemental concentrations over Sc were then submitted to cluster analysis based on Euclidean distance and aver-age linkage. The resulting dendrogram is shown in Figures 7–9. From the general structure of the dendrogram we can expect to distinguish several well-defined chemical groups, a result in accordance with high total varia-tion in the matrix. The nine groups circled and labeled with roman numerals on the dendrogram are discussed in detail below. The mean elemental con-centrations as well as the percentage standard deviations of all nine groups are presented in Table 4. The dashed line that separates the dendrogram into two slightly unequal halves corresponds to the split between samples of Cretan origin (to the left of the line) and those from the Near East and Egypt (to the right). The sample numbers of the members of each group are shown in Figure 7, the correspondence between the chemical groups and the typological classification of the vessels is shown in Figure 8, and the correspondence between the chemical groups and the petrographic fabrics is shown in Figure 9.

Cre tan Chemic al Groups

Chemical Group I

This is the largest and most compact group in the dendrogram, contain-ing 22 samples, of which 13 are TSJ and 7 SNA (Fig. 8). The remaining two samples (98/46 and 98/61) are classed as CJs on stylistic grounds. A close examination of sample 98/46 (see online Appendix 2) reveals that it has a lower concentration of the rare earth elements Th, Fe, and Sc than the rest of chemical group I. Since Sc was used as a divisor, its low value elevated the concentration of other elements, with the result that this sample clustered in group I.

Chemical group I contains most of the transport jar samples from petrographic fabric A, as well as individuals from Central Cretan fabrics D and E and fabrics G and J (Fig. 9). The standard deviations for most elements in chemical group I are less than 10%, except for Ca, Na, Cs, and Rb, which have standard deviations greater than 20%. This situation is indicative of marine sediments in which alkalis are selectively leached, or Na is enriched, in the case of high-fired ceramics, through the fixa-tion of analcime. All samples with Na values close to or greater than 1% behave in this way. This phenomenon has been documented in pottery from the Minoan kiln at Kommos and may be a trademark of the site.89

89. Buxeda i Garrigós, Kilikoglou, and Day 2001.

TA

BL

E 4

. M

EA

N A

ND

PE

RC

EN

TA

GE

ST

AN

DA

RD

DE

ViA

TiO

N O

F C

HE

MiC

AL

GR

OU

PS

OF

LA

TE

MiN

OA

N

TR

AN

SP

OR

T J

AR

SA

MP

LE

S F

RO

M K

OM

MO

S A

NA

Ly

zE

D B

y i

NA

A

Gro

upSm

Lu

YbC

aN

aL

aC

eT

hC

rH

fC

sSc

Rb

FeC

oE

u

Im

ean

4.93

0.34

2.68

6.04

0.94

27.2

059

.58

9.76

362.

104.

235.

7820

.25

89.5

55.

7133

.17

1.13

%9.

687.

867.

7926

.62

26.1

110

.59

10.3

811

.95

17.6

59.

4931

.33

7.90

27.4

67.

6512

.51

9.38

IIm

ean

4.48

0.34

2.62

6.37

0.71

24.1

557

.15

9.97

607.

703.

532.

4923

.60

51.6

76.

8145

.64

1.07

%6.

1111

.45

9.78

26.8

419

.40

11.0

08.

3111

.87

51.3

128

.95

58.9

08.

6941

.36

5.40

9.13

7.33

III

mea

n5.

180.

352.

753.

391.

2828

.30

63.4

310

.19

347.

004.

843.

3717

.32

73.3

35.

0429

.00

1.17

%6.

6810

.92

9.41

9.59

8.59

5.84

7.98

10.0

911

.44

2.56

52.4

74.

6231

.49

4.24

5.27

1.97

IVm

ean

7.03

0.41

3.46

8.95

0.39

35.3

386

.43

9.08

116.

809.

451.

5313

.26

32.5

4.35

22.4

21.

60

%14

.14

11.0

910

.81

27.1

028

.63

15.0

912

.30

7.67

13.2

93.

9726

.87

10.3

115

.38

7.56

18.5

717

.08

Vm

ean

5.97

0.34

2.53

17.6

10.

3628

.11

60.1

65.

6113

3.70

4.09

1.55

13.1

135

.00

4.82

21.3

21.

44

%21

.60

22.1

021

.96

17.1

116

.23

23.7

525

.85

25.9

618

.28

22.2

729

.43

22.1

424

.28

20.1

323

.20

22.6

8

VI

mea

n7.

160.

383.

0311

.47

0.53

35.4

981

.86

7.90

107.

507.

261.

3117

.05

35.0

05.

5023

.65

1.71

%7.

327.

676.

2616

.52

38.0

67.

337.

8610

.05

8.03

9.26

32.9

38.

2926

.87

8.80

12.6

77.

50

VII

mea

n6.

240.

322.

5918

.21

0.23

34.5

770

.57

9.45

85.5

02.

923.

5710

.88

66.6

73.

1010

.41

1.33

%8.

379.

657.

602.

694.

959.

887.

829.

6213

.90

14.1

220

.43

8.92

17.3

26.

617.

9811

.93

VII

Im

ean

7.13

0.39

2.99

10.2

40.

6139

.50

79.9

310

.73

123.

704.

753.

9212

.22

66.6

73.

489.

091.

49

%2.

082.

562.

419.

6229

.72

3.73

2.90

1.08

5.38

1.59

9.57

1.87

8.66

0.00

5.49

4.04

IXm

ean

5.16

0.34

2.76

12.5

10.

2828

.55

56.6

06.

0231

07.0

02.

861.

2112

.24

20.0

03.

1827

.49

1.19

%16

.87

14.7

817

.93

8.82

15.1

518

.08

26.2

421

.28

5.28

22.5

412

.86

22.9

970

.71

22.4

920

.09

23.7

0

Fig

ure

7. iN

AA

den

drog

ram

of L

M tr

ansp

ort j

ar sa

mpl

es w

ith sa

mpl

e nu

mbe

rs a

nd in

terp

rete

d ch

emic

al g

roup

s. w

ell-

defin

ed c

hem

ical

gro

ups a

re c

ircl

ed.

The

ver

tical

das

hed

line

divi

des s

ampl

es o

f Cre

tan

orig

in (l

eft)

from

thos

e of

Nea

r Eas

tern

and

Egy

ptia

n or

igin

(rig

ht).

Con

cent

ratio

ns a

re e

xpre

ssed

as

log-

ratio

s ove

r Sc.


The elemental profile of chemical group I is consistent with that of the Neogene marine sediments of Central Crete, but distinct from the refer-ence group of pottery associated with the Kommos kiln.90

Chemical Group II

This group, which is related to chemical group I, consists of four SNA samples belonging to fabric A, as well as one sample (98/25) belonging to fabric C, which was later reclassified as an earlier amphora type. Chemical group II has a higher average concentration of Cr than chemical group I (see online Appendix 2). In Central Crete ultrabasic rocks, which contain Cr in the form of chromite inclusions, are abundant, and the distribution of chromite and Cr in clays can be uneven. It can therefore be safely suggested that chemical groups I and II are subgroups of a larger cluster.

Chemical Group III

Three SNA samples form the small, tight chemical group III (Fig. 8). All three belong to petrographic fabric B and are connected with igneous rocks from the foothills of the western Asterousia Mountains.

Other TSJ and SNA samples

The remaining TSJ and SNA samples do not belong to any specific cluster, but rather appear as outliers in the dendrogram (Fig. 8). Samples 98/6 and 98/31 are located on the opposite side of the dendrogram from chemical groups I, II, and III (Fig. 7). They have chemical compositions very similar to comparative material from Chania with a high concentration of light rare earth elements (see online Appendix 2). This composition is also similar to that of a group of TSJs found at Mycenae and attributed to West Crete.91 These two samples form petrographic fabric F (Fig. 9), which is considered to be West Cretan. Sample 98/8 is also located on the right side of the dendrogram, but is more closely related chemically to East Crete than to West Crete. The two ends of the island are quite similar to each other in chemical terms, however. Samples 98/12 and 98/81 are outliers that do not match any chemical reference groups.

Imp orted Chemic al Groups

The right side of the dendrogram consists of CJ and EJ samples (Fig. 8). The clusters here are looser than those on the left side. In general there are three larger groups (IV, V, VI), as well as at least three more groups containing two or three members each (VII, VIII, IX).

Chemical Group IV

Chemical group IV is a cluster of four samples belonging to petrographic fabric 3 (Fig. 9), which is linked to the Carmel coast of northern Israel, together with a fifth, closely related sample. This is a very compact group and one that is chemically well differentiated from the rest (Table 4).

90. Shaw et al. 2001.91. Tomlinson 1996.

Fig

ure

8. iN

AA

den

drog

ram

of L

M tr

ansp

ort j

ar sa

mpl

es w

ith o

rigi

nal t

ypol

ogic

al c

lass

ifica

tion

of v

esse

ls a

nd in

terp

rete

d ch

emic

al g

roup

s. w

ell-

defin

ed

chem

ical

gro

ups a

re c

ircl

ed. T

he v

ertic

al d

ashe

d lin

e di

vide

s sam

ples

of C

reta

n or

igin

(lef

t) fr

om th

ose

of N

ear E

aste

rn a

nd E

gypt

ian

orig

in (r

ight

).

Con

cent

ratio

ns a

re e

xpre

ssed

as l

og-r

atio

s ove

r Sc.

T =

tran

spor

t stir

rup

jar;

S =

shor

t-ne

cked

am

phor

a; E

= E

gypt

ian

jar;

C =

Can

aani

te ja

r.


Chemical Group V

Chemical group V is a relatively loose group, but one composed entirely of individuals from petrographic fabric 2 (Fig. 9). It is chemically distinct from the other samples (Table 4).

Chemical Group VI

The largest chemical group on the right side of the dendrogram is group VI, which consists of 12 of the 19 EJ samples (Fig. 8). It is homogeneous and exhibits only minor variation in most elements, except for the alkalines, which show a standard deviation of over 20%, possibly due to alterations while buried (Table 4). The chemical signature of this group matches that of the amphoras from Malkata analyzed by McGovern92 and the marl D fabric in the Vienna system, comprising amphoras from Thebes, Memphis, Amarna, and Qantir analyzed by Bourriau and her coworkers.93 It is safe to suggest that these EJ samples from Kommos were made from the same raw materi-als as the Egyptian vessels in those studies. McGovern suggests that they were produced in the area of Thebes and transported to the Delta region for filling, whereas Bourriau argues for a likely source in the Memphis region.94

Chemical Groups VII and VIII

Two small groups of three samples each appear on the far right side of the dendrogram. Chemical group VII is composed of members of petrographic fabric 4 (Fig. 9), which may come from the Lebanese coast. Despite the small number of samples, the group is relatively tight and the average values have low standard deviations (Table 4). Next to this cluster is chemical group VIII, composed of samples from petrographic fabric 1a (Fig. 9). This is also a tight group with very low standard deviations.

Chemical Group IX

On the far left side of the dendrogram are two very distinctive CJ samples (98/58 and 98/65), which we have labeled chemical group IX (Fig. 7). This group corresponds to petrographic fabric 5 (Fig. 9), and its most distinctive chemical characteristic is its high Cr concentration (Table 4). Its likely origin lies in coastal Syria among the ophiolite series, which would account for the elevated Cr levels.

Other CJ and EJ Samples

Four EJ samples of petrographic fabric 1 (98/33, 36, 48, 49) form a loose cluster, along with samples 98/50 and 98/53, between chemical groups IV and V in the dendrogram (Fig. 7). The chemical profile of this group, when compared to that of chemical group VI, exhibits minor differences in most elements except Na, which is twice as abundant. Another loose cluster of two samples (98/55 and 98/69) is located just to the right of the line divid-ing Cretan from imported ceramics (Fig. 7). Both samples are members of petrographic fabric 7 (Fig. 9) and are chemically distinct from the rest of the samples, although the chemical affinity between the two samples is low compared to that of the more tightly clustered groups discussed above.

92. McGovern 1997.93. Bourriau 2004.94. McGovern 1997; Bourriau 2004.

See also Cohen-Weinberger and Goren 2004, pp. 84–85, with references, for a comparison of the results of petro-graphic analysis with INAA data and interpretation.

Fig

ure

9. iN

AA

den

drog

ram

of L

M tr

ansp

ort j

ar sa

mpl

es w

ith p

etro

grap

hic

fabr

ic c

lass

ifica

tion

and

inte

rpre

ted

chem

ical

gro

ups.

wel

l-de

fined

che

mic

al

grou

ps a

re c

ircl

ed. T

he v

ertic

al d

ashe

d lin

e di

vide

s the

sam

ples

of C

reta

n or

igin

(lef

t) fr

om th

ose

of N

ear E

aste

rn a

nd E

gypt

ian

orig

in (r

ight

). C

once

ntra

-tio

ns a

re e

xpre

ssed

as l

og-r

atio

s ove

r Sc.


THE PROVENANCES OF THE TRANSPORT JARS FROM KOMMOS

Petrographic and chemical analyses of the Late Minoan transport jars from Kommos have provided us with two different lines of evidence with which to address the question of provenance. Although the two analytical techniques are concerned with different compositional characteristics, their results correspond closely with one another, as well as with the typological classification of the transport jars (Figs. 8, 9). For example, both chemistry and petrography confirm the split between the Cretan TSJ and SNA samples and the imported CJ and EJ samples. This agreement is one of the clearest to have appeared in any analytical study of Minoan ceram-ics, and it demonstrates the potential benefits of a combined application of the two techniques in a “mixed-mode” approach.95 The analysis of the petrographic and chemical data within the context of the comparative material described above has allowed us to distinguish local and imported transport jars at Kommos and to suggest a geographical provenance for each group.

Cre tan versus Non-Cre tan Ceramics

The 88 samples can be clearly separated into Cretan and non-Cretan ves-sels on the basis of the petrographic classification and the INAA cluster analysis. The resulting groups range from well-defined clusters of samples that are petrographically and chemically homogeneous and can be closely linked to geology and/or comparative material, to unique samples that lack clear comparative data but can nevertheless be ascribed to a general area.

The typological distinction between the TSJ and SNA samples, which are considered to be products of Crete,96 and the non-Cretan CJ and EJ samples is evident in both datasets. This split is particularly visible in the petrographic classification, where fabrics A–J consist of TSJ and SNA samples, while fabrics 1–12 consist of EJ and CJ samples. Chemically, the Cretan TSJ and SNA samples form tighter clusters in the INAA dendro-gram than the imported CJ and EJ samples. Both sets of samples, however, contain several distinctive chemical groups, a fact that has implications for the provenance of the vessels. Some 27 of the 31 TSJ and SNA samples and 47 of the 53 CJ and EJ samples appeared on opposite sides of the INAA dendrogram (Fig. 8), confirming the broad chemical distinction between Cretan and non-Cretan material.

With the exception of sample 98/53, which is taken to be Cretan, the petrography of the 88 transport jars also supports the broad division between Cretan and non-Cretan samples, as suggested by their typological classification. We have found it useful to structure the following commen-

95. Mixed-mode statistical analysis was conducted on some of the chemical and petrographic data in collaboration with the GEOPRO Training and Mo- bility Network funded by the European

Commission. See Beardah et al. 2003; Moustaki and Papageorgiou 2004; Bax-ter et al. 2008. The last of these dem-onstrates that when both chemical and petrographic aspects of the composition

are included in the same statistical pro-cedure, the data may show new struc-ture and clearer groupings than when either dataset is considered separately.

96. Rutter 1999.


tary by petrographic fabric, adding observations on the typological and chemical properties of each group. Other comments will be found in the general discussion that follows.

Cre tan Transp ort Jars

Fabric A: Main South-Central Cretan

Of the total of 17 pieces assigned to this petrographic fabric, 10 are TSJs and 6 are SNAs. The date range for the former is LM II–IIIB, that for the latter LM IIIA2–IIIB. Both shapes were evidently produced contempora-neously within the western Mesara during the 14th and 13th centuries, but the local production of TSJs clearly anticipated that of SNAs by as much as a century or more. The dearth of LM IIIA1 TSJs is a result of the fact that the samples included only one piece of that date in any fabric. Nine of the ten TSJs in fabric A were placed in chemical group I by INAA (the single exception, sample 98/7, is a chemical outlier), while the six SNAs were divided between chemical groups I and II.

Fabric B: Medium/Coarse Igneous

A total of five samples were assigned to this fabric, all of them from SNAs of LM IIIB date. Two of the five were assigned to chemical group I, but the other three separated out into the tightly clustered chemical group III. This petrographically (and, to a lesser extent, chemically) distinct set of samples provides interesting evidence for what may be a second workshop engaged in the production of SNAs in the western Mesara during the 13th century.

Fabric C: Serpentine

The two members of this fabric, both originally identified as SNAs, were sampled from body sherds. The earlier, LM IIIA2 vessel (sample 98/25) has comparatively thin walls and may in fact belong to an oval-mouthed amphora, or perhaps even to an undecorated TSJ. The later, LM IIIB vessel (sample 98/21), however, has too slim and thick-walled a lower body profile to belong to a stirrup jar or earlier oval-mouthed amphora; although it is atypical for a SNA in bearing the impressions of what appear to be three diagonally transverse string marks on its exterior,97 it is difficult to imagine to what other shape such a fragment could belong, especially in view of its very close resemblance in profile to the more fully preserved SNA from which sample 98/19 was taken.98 In chemical terms, both jars would be considered Central Cretan products.

Fabric D: Fine Calcareous Phyllite

These two TSJ samples are early (LM II). Their mineralogy is very distinctive, with no clear parallels in the comparative material. They are based on common Neogene marls, however, so their chemical composition unsurprisingly falls within chemical group I. After the conclusion of the analysis they were identified macroscopically as products of the island of Gavdos, which is compatible with the analytical profile.99

97. Rutter 2006a, p. 555, no. 67a/22, pls. 3.78, 3.93:e.

98. Rutter 2006a, p. 569, no. 75/6, pls. 3.83, 3.94:c.

99. Our thanks to K. Kopaka and C. Papadaki for this identification. The two TSJs are both unusual in their painted decoration, though only 98/13 is substantially preserved. The latter was published as a Gavdiot import in Rutter 2006b, pp. 672–674. Sample 98/3 was not identified as Gavdiot in Rutter 2006b.


Fabric E: Metasedimentary

The two samples of this fabric were originally identified as SNAs. Although they have a distinct petrographic composition, they fall into chemical group I, and are therefore thought to be Central Cretan products. The shape of both vessels has been reconsidered since the sampling: sample 98/18 is in fact a necked amphora that is arguably the immediate typological ancestor of the SNA, dating from LM IIIA2 Early, while sample 98/23 comes from a large closed vase, probably another amphora of pre-SNA type. This, then, may represent an early SNA fabric.

Fabric F: Quartzite, Phyllite, and Schist

Petrographically, these two vessels are clearly products of West Crete, matching both West Cretan control groups and the TSJs from Mycenae deemed to be West Cretan. Typologically, sample 98/31, as a LM IIIA2 Early light-on-dark TSJ, is clearly West Cretan. Sample 98/6, however, is cream-slipped over a red-firing clay and was for this reason originally thought likely to be an East Cretan import.100 The comparatively early date of this piece—LM IA Final, a period before and during which the number of documented TSJs is still comparatively small—suggests that it may be an early West Cretan TSJ.

Fabric G: Phyllite, Siltstone, and Chert

Sample 98/5, from a TSJ fragment found in a LM IIIB context, is worthy of comment as an example of a common Central Cretan TSJ fabric that seems not to be local to the western Mesara (judging from its rarity at Kommos relative to fabric A).

Fabric H: Phyllite

Watrous suggested that the source of TSJ sample 98/8, dated to LM IIIA1, was East Cretan.101 The East Cretan origin suggested by both the petro-graphic and chemical analysis is welcome confirmation of this.

Fabric I : Quartz, Polycrystalline Quartz, Schist, and Microfossils

TSJ sample 98/12, dated to LM IIIA2, is a false neck and handle fragment in a fabric that is clearly not local to Kommos. Although possibly related to petrographic fabric F, this sample is not closely identifiable with any particular region of Crete by petrography, chemistry, or typology.

Fabric J : Schist

The distinctive fabric of TSJ sample 98/10 is not matched by anything noteworthy in its chemistry (group I) or decoration (linear body and handle fragment). This piece comes from one of the later LM IIIB contexts in Building P and is chiefly of interest for further expanding the range of discrete, probably South Cretan petrographic fabrics attested in TSJs at Kommos during this period.

100. The vessel is published in Rutter 2006a, p. 429, no. 29/5, as an “import from unknown Minoan pro-duction center.” See Haskell 2005, p. 208, for the significance of the pale slip and red clay body.

101. Kommos III, pp. 31, 153, no. 520, pl. 18.


The Provenances of Cre tan Transp ort Jars

The samples of demonstrable Cretan origin have been classified into six fabrics and four unique samples, which can in most cases be assigned a specific provenance on the island. Several fabrics correspond to known petrographic groups of stirrup jars and other vessels.102 The main south-central Cretan fabric A is known from several analyses in the western Mesara. The fact that both TSJ and SNA samples belong to this fabric clearly indicates that transport jars of these two types share an area of production, if not an exact location. Chemical analysis confirms the compositional similarity between these samples (chemical groups I and II), as well as their relationship to other ceramics from the Mesara. However, it also demonstrates that they differ chemically from the reference group estab-lished for the LM IA kiln at Kommos.103 The INAA results highlight the possible separation of some SNA samples from fabric A (chemical group II). This grouping may represent variation in the origin of the sand inclusions in these samples, which was not picked up in petrographic analysis, or may simply reflect the uneven distribution of Cr within the south-central Cretan source area.

It is not possible to pinpoint the exact origin of the SNA and TSJ samples in fabric A because of the wide variability in this common fabric. The range of sand-sized inclusions in the samples, however, is compatible with an origin at one or more locations in south-central Crete. In particular, the presence of fine volcanic rock fragments suggests that this fabric may have been produced in the western Mesara (Fig. 10). The elemental profile of chemical groups I and II, to which most of the samples of fabric A be-long, is closely related to that of the Neogene marine sediments on Crete.

The medium/coarse igneous fabric B is a variant of the main south-central composition represented by fabric A. The samples belonging to this smaller fabric all come from SNAs. They are also chemically distinctive (chemical group III). The mineralogy of the samples and their similarity to Minoan ceramics found at Moni Odigitrias and Ayia Kyriaki suggest that their raw materials originated somewhere in the foothills of the Asterousia Mountains, south of the Mesara Plain, not far from Kommos (Fig. 10).104

The two other fabrics represented among the SNAs, the serpentine fabric C and the metasedimentary fabric E, may also be related to fabric A. The samples belonging to these fabrics cluster with or near chemical groups I and II. While both fabrics could, on the basis of their petrography, come from a number of places in central and eastern Crete, they are not incom-patible with the Mesara (Fig. 10).

Thus, the petrographic and chemical analyses of the SNA samples in fabrics A, B, C, and E and in chemical groups I, II, and III indicate that they originate from one or more locations in the Mesara Plain. We have studied 13 different samples of this type of jar from several levels at Kommos and can conclude with confidence that they were a local product.

The majority of the TSJ samples from Kommos, which belong to the main south-central Cretan fabric A, are also likely to have been produced in the Mesara. The compositional similarity between TSJ and SNA samples in fabric A and chemical group I (Fig. 9) strongly suggests that they were

102. For stirrup jars, see the groups in Haskell et al. 2011.

103. For the kiln, see n. 37, above. These differences, however, in no way rule out the possibility of production within the Kommos area, since there is a substantial chronological gap between the sampled jars and the kiln, which went out of use before the end of LM IA.

104. For red-firing schist-related fabrics at Kommos, see Shaw et al. 2001, pp. 117–119, 152–155.


produced from the same raw materials, perhaps even at the same location. The INAA data do indicate some differences between the two vessel types within fabric A, as shown, for example, by the SNA samples that form chemical group II. Most of the other TSJ and SNA samples in fabric A, however, are indistinguishable from one another in composition. The TSJ samples belonging to the fine calcareous phyllite fabric D, the phyllite, siltstone, and chert fabric G, and the schist fabric J are also compatible with an origin in the geological series of Central Crete, clustering as they do within chemical group I. Fabric D may have been produced from metamorphic rock and calcareous Neogene sediments, both of which oc-cur in the Mesara, as well as in other areas of Crete. Fabrics G and J are compatible with a source in the Mesara and have parallels among other Minoan ceramics from this area.

While the majority of the TSJs from Kommos analyzed in this study are likely to have been produced in the Mesara, four samples may have originated elsewhere on the island. These are the samples that constitute the quartzite, phyllite, and schist fabric F, the phyllite fabric H, and the quartz, polycrystalline quartz, schist, and microfossil fabric I. Petrographi-cally and chemically, fabric F matches samples from West Crete, specifically in the Chania Plain (Fig. 10). The related fabric I, which cannot be placed with certainty, may also have originated in the west of the island (Fig. 10). Fabric H is compatible with several locations on Crete where outcrops of phyllite-quartzite rock occur, but the best matches are in East Crete. It is worth noting that, with the exception of fabric D, the four possible non-Mesaran TSJ samples are all located at the opposite end of the dendrogram from the rest of the TSJ samples from Kommos.

Figure 10. Map showing proposed provenances of Cretan LM transport jars from Kommos. 1 = Kommos, 2 = Ayia Triada, 3 = Phaistos, 4 = Moni Odigitrias, 5 = Ayia Kyriaki, 6 = Knossos. P. S. Quinn


Imp orted Transp ort Jars

Fabric 1 : Quartz and Calcite

The chronological range of fabric 1 at Kommos extends from LM IB to LM IIIB, with over half of the sampled pieces coming from LM IIIA contexts of the 14th century. Most of these vessels are represented by only one or two small fragments found in construction fills. It is thus quite likely that many of the complete vessels actually arrived at Kommos in the course of the 15th century during the LM II–IIIA1 phases of the Monopalatial era, when Knossos, on present evidence, was the only functioning palatial center on Crete. Nothing about the two flasks included in this fabric (samples 98/34 and 98/44) distinguishes them in any way from the larger jars that represent the remainder of the sampled Egyptian imports. The INAA results correlate particularly well with the petrographic assignments. Of the 18 members of fabric 1, 14 were assigned to chemical group VI, and the remaining four were considered to be related to that group (Fig. 9).

Subfabric 1a: Quartz and Calcite with Chert

Subfabric 1a is a particularly tight grouping from all perspectives and may originate in the Jezreel Valley of Israel. The four samples assigned to this subfabric come from CJs recovered from LM IIIA2 Early (samples 98/51, 64, 84) and LM IIIB (sample 98/73) contexts, the last a tiny fragment from what could well be a much earlier vessel. Three of the four samples are assigned to chemical group VIII (the exception is the chemical outlier 98/64), and that group in turn includes only the three samples assigned to this petrographic subfabric.105

Fabric 2 : Macrofossiliferous Clay Pellet

This fabric contains 11 samples, all of them CJs and eight of them also assigned to and accounting for all members of chemical group V (Fig. 9). Within this petrographic fabric, samples 98/72 and 98/81 are chemical outliers, which indicates a probable difference in provenance. The date range of the fabric is LM IIIA2 Early–IIIB.

Fabric 3 : Quartz and Clay Pellet

The four members of this fabric are all assigned to chemical group IV, along with a single related sample (98/82) (Fig. 9). The chronological range is once again LM IIIA2 Early–IIIB.

Fabric 4 : Quartz, Chert, and Macrofossils

The three samples in this fabric are also linked chemically as the three members of chemical group VII. All these pieces come from LM IIIA2 Early contexts. They have links with Smith’s group 5, which has its origins in the Lebanese coastal plain between Akko and Sidon (Fig. 11).

Fabrics 2 and 4 Related Sample

Sample 98/61 is chemically distinct from both fabrics 2 and 4. It may origi-nate on the Lebanese coastal plain (Fig. 11). Found in a LM IIIB context,

105. After the conclusion of the sampling program, the visiting Syro-Palestinian specialist M. Serpico identi-fied eight additional examples of this fabric among inventoried CJ fragments from Kommos.


the lower body fragment that was the source of this sample is extremely heavily worn, suggesting that the original CJ to which the sherd belonged may have been considerably earlier in date.

Fabric 5 : Chert

The two samples of fabric 5 are chemically similar, and are the only two samples in chemical group IX (Fig. 9). Their date is LM IIIA2–IIIB. Sample 98/65 was taken from a base fragment of a Syrian flask or spindle bottle, a significantly smaller one-handled shape quite different from the shoulder-handled CJs that constitute the bulk of Syro-Palestinian ceramic imports to Kommos.

Fabric 5 Related Sample

The chemical composition of this sample (98/62) does not appear to be linked with that of any other sampled CJ from Kommos. It could have originated in Cyprus if the similarities noted above are valid. The fragment from which the sample was taken was recovered from a LM IIIA context.

Fabric 6 : Fusilinid

These two EJ samples are virtually identical petrographically but not par-ticularly close chemically. One of the samples (98/54) comes from a LM IB Early context, the earliest at the site to have furnished an Egyptian import. The second vessel (98/50), preserved as a large number of sherds, comes from a LM IIIA2 Early building fill.

Fabric 7 : Serpentine and Micrite

These samples, from vases found in the same LM IIIA2 Early floor deposit in House X, are only loosely related both petrographically and chemically, but may nevertheless be assigned an origin along the northern Syrian coast in the neighborhood of Ras Shamra (Fig. 11). Sample 98/55 is from a fully

Figure 11. Map showing proposed provenances of non-Cretan LM transport jars from Kommos. P. S. Quinn


restorable vessel and was presumably imported to Kommos not long before being abandoned on the floor ca. 1370.

Fabric 8 : Polycrystalline Quartz and Calcite

This sample is a chemical outlier of unknown provenance.

Fabric 9 : Quartz and Metamorphic

The pale-slipped and burnished, medium-coarse closed shape from which sample 98/53 was taken is certainly an import to Kommos, but analysis suggests it comes from another Minoan center of production.106

Fabric 10 : Micaceous Quartz and Feldspar

The single example of this distinctive fabric was considered to be of Egyp-tian origin when the pale-slipped jar was sampled in 1998. It is as unique chemically as it is petrographically, and is among the comparatively few CJs to have been recovered at Kommos from a LM II context.107

Fabric 11 : Quartz and Schist

Petrographically, this sample is clearly an import to Crete, and it is also a chemical outlier. The rim fragment preserves traces of a handle attach-ment on the upper neck, and thus belongs to a rather different large closed shape than the standard shoulder-handled CJ. It was found in a LM IIIA2 Early context.108

Fabric 12 : Alkali Feldspar

Like the single examples of the two previous petrographic fabrics, the sole representative of fabric 12 is also a chemical outlier. It is clearly an import to Crete.

The Provenances of Imp orted Transp ort Jars

Among the samples of non-Cretan CJs and EJs, several petrographic and chemical groups are evident. Comparison with previously published ceramic analyses from contemporaneous archaeological sites and/or the geology of potential source areas make it possible to draw some conclusions about the provenances of these groups.

The majority of the EJ samples from Kommos fall within the quartz and calcite fabric 1, which is equivalent to the homogeneous chemical group VI (Fig. 9). Mineralogically and chemically these samples match comparative material from Egypt, particularly marl fabrics described from the Memphis region of the Nile Delta (Fig. 11:a). The highly distinctive fusilinid fabric 6, formed by two EJ samples containing large fusilinid foraminifera, may also come from the Nile area of Egypt (Fig. 11:a). Petrographically, these samples are unlike those from any other analyzed transport jars, either from Kommos or from other sites represented in the comparative data. They seem to have been produced in Egypt by the

106. Visiting Near Eastern ceramic specialists have been unanimous in rejecting this piece as either Egyptian or Syro-Palestinian, although when sampled it was considered to be an Egyptian import.

107. Both the Egyptian and Syro-Palestinian ceramic specialists who examined this sample recently agreed that it belonged to a CJ and was not Egyptian.

108. While the jar was considered to be Egyptian during the sampling in 1998, it has been subsequently reclassi-fied as Syro-Palestinian (Rutter 2006a, p. 580, no. MI/UP/1).


addition of bioclastic limestone to a marly paste such as that used for the manufacture of fabric 1.

Samples belonging to subfabric 1a, while petrographically similar to the Egyptian ceramics in fabric 1, are chemically distinct and form a sepa-rate cluster in the dendrogram (chemical group VIII). These samples bear more similarities to fabrics described from other Canaanite jars in Israel than to the marl fabrics of Egypt. In particular, they provide a petrographic match for fabrics thought to originate in the Jezreel Valley near modern Haifa (Fig. 11:c).

The homogeneous compositional group of CJ samples represented by fabric 3 and chemical group IV can also be ascribed a provenance in northern Israel. This group corresponds petrographically to fabrics de-scribed from the New Kingdom sites of Aphek, Jaffa, and Tel Hefer. The origin of such fabrics is thought to be a stretch of the northern Israeli coast south of Haifa (Fig. 11:c).

Fabrics 2 and 4 comprise well-defined groups of CJs that are further confirmed by their chemistry. Related by the presence in both of rounded quartz and distinctive coralline macrofossil inclusions, these fabrics have also been recognized in CJs from the New Kingdom and attributed to a source on the Lebanese coastal plain. The striking similarity between this comparative material and fabrics 2 and 4 suggests that the Lebanese coast may be the origin of many of the imported CJs found at Kommos. More precisely, it is possible that the samples of fabric 4 may have originated in the southern part of this large coastal area, between Sidon and Akko (Fig. 11:c).

The two CJ samples that form fabric 5 are chemically very distinct from all other samples in this study. Petrographically, they have parallels in a Canaanite fabric thought to have originated in the ophiolite complex of the northern Syrian coast. Fabric 7 likewise contains inclusions that may have come from this ophiolite complex and can be linked to fabrics found in the same area, notably at the coastal site of Ras Shamra (Fig. 11:b).

Sample 98/62, which is related petrographically to fabric 5, is chemi-cally very different from these Syrian samples. It can be distinguished from fabric 5 by the nature of its chert inclusions and the presence of additional fragments of igneous and metamorphic rocks. In this respect it resembles the fabric of a previously published Canaanite sample that has been ascribed to southern Cyprus (Fig. 11:b).109 Another CJ sample, 98/77 of fabric 8, is related petrographically to sample 98/62 and may also have originated in Cyprus, although its chemistry is very different.

Fabrics 10, 11, and 12 are single CJ samples that bear no petrographic resemblance to any other material known from Kommos, Crete, or else-where. Fabrics 10 and 12 are also chemically very different from other samples in our analysis. All three samples are clearly non-Cretan, but at present we cannot place them with any confidence. Sample 98/53 in fabric 9 is also problematic. It was originally sampled as an Egyptian import, but its petrography suggests that it may have originated in south-central Crete. Other discrepancies in our petrographic and chemical assessment of the provenance of the imported transport jars at Kommos are the existence of a CJ sample within the otherwise uniformly Egyptian fabric 1 and chemical group VI.109. Smith et al. 2004, pp. 68–70.


110. See n. 11, above.111. Haskell et al. 2011.112. See n. 99, above.113. Hallager and Hallager 2000,

pp. 144–146; 2003, pp. 214–217.

DiSCUSSiON

The analysis of Late Bronze Age transport jars from Kommos has shown excellent agreement between chemical and petrographic data, each set com-plementing the other. Along with a remarkable correspondence between the petrographic and chemical groups in this study and those published previously, this agreement has enabled us to make confident statements of provenance, and even to identify possible misclassifications of individual samples.110 General assumptions based on archaeological evidence alone, such as a north-central Cretan origin for TSJs as opposed to a presumed source in the Mesara for SNAs, have been challenged, and a new picture has emerged with surprising detail.

The 16 pieces originally sampled as SNAs have been reduced to no more than 13, and possibly as few as 12 or 11, by the elimination of pieces that were certainly (samples 98/18, 23, 26) or possibly (samples 98/25 and, much less likely, 98/21) misidentified. The unambiguously identified SNAs are distributed within just two petrographic fabrics (A and B) and three chemical groups (I–III).

One of the petrographic fabrics (B) and one of the chemical groups (III) are restricted to the LM IIIB period; neither these nor chemical group II are represented among the 18 TSJs sampled and analyzed as part of this project. The analysis of the SNA samples demonstrates that there was more than one production center in south-central Crete, and, in the case of the fabric most frequently encountered (fabric A), shows that the same raw materials were used in the production of TSJs and SNAs.

The picture of TSJ sources that emerges from this program of pe-trographic analysis and INAA is one of great interest. Rather than being linked to the north-central part of the island, the majority of the TSJ samples have been shown to be broadly local to the western Mesara. These include examples presumably from the same workshops as the main SNA group. The vessels extend chronologically from LM II to the time of the effective abandonment of Kommos in later LM IIIB, a span broader than that documented in most previous analyses.111

The remainder of the sampled TSJs represent six different petrographic fabrics and four different chemical groups, with no single combination of the two sets of analytical groupings represented by more than two samples; these include a characteristic LM II fabric identified as a product of Gavdos112 and vessels of LM IIIA2 Early date imported from West Crete. The only Cretan site from which TSJs have been published in numbers roughly equivalent to those now published from Kommos is the coastal center of Chania.113

The 18 samples of non-Cretan fabric 1 belong to imported Egyptian transport vessels (two flasks, two necked jars, and 14 probable amphoras). These, together with the two amphora samples of fabric 6, likewise con-sidered Egyptian, represent roughly half of the total number of Egyptian ceramic imports so far identified in LM IB–IIIB contexts at Kommos. Apart from the fusilinid foraminifera, which set the two representatives of fabric 6 apart petrographically, and the atypical chemical composition of one of these samples (98/54), there is no evidence for subdivisions within


this corpus of almost 20 analyzed samples. The two flasks, the two necked jars, and the abundant amphoras occur in the same fabric, which is familiar from petrographic and chemical comparative material found in Egypt.

The 34 samples now considered to belong to Syro-Palestinian trans-port vessels (one spindle bottle and 33 CJs) represent exactly half of the 68 Late Bronze Age Syro-Palestinian imports thus far inventoried from LM IB and later contexts at Kommos.114 Several fully restorable CJs have been found in LM IIIA2 Early floor deposits in House X, as well as in the court in front of Building P in a somewhat later LM IIIA2 context. These discoveries of imported CJs in what are, in effect, contexts of primary use at Kommos confirm the broader picture of their chronological distribution provided by all of the recovered CJ fragments at the site, namely that the peak period of their importation was the 14th century. Yet it is also clear that the earliest CJs to appear at Kommos had arrived before the end of the Neopalatial era (two from LM IB contexts, neither sampled for analysis), and that the bulk of those inventoried (a total of 37 pieces) come from contexts of the Monopalatial era (LM II–IIIA2 Early), when a Knossian administration dominated Central Crete and perhaps utilized Kommos as its principal southern outlet for intercultural exchange.115

Petrographic and chemical analyses have been able to identify groups of CJs that were produced in several regions of the Levantine coast, as well as distinctive fabrics from coastal Syria and probably Cyprus. The links between the CJ groups at Kommos and those published by Smith’s team are remarkably close and allow us to posit well-defined geographical areas of origin. Almost half (30) of the 65 imported CJs inventoried from Kom-mos belong to just three fabric categories, which have been assigned by the Cambridge Amphora Project to production zones in the northwestern Jezreel Valley (Smith’s group 1 = our subfabric 1a and chemical group VIII), along the coast of northern Israel south of Haifa (Smith’s group 2 = our fabric 3 and chemical group IV), and along the Lebanese coast further to the north (Smith’s group 5 = our fabrics 2 and 4 and chemical groups V and VII). The production zone of a fourth group of CJs, only about half as popular at Kommos as any of the preceding three, has been identified as the Syrian coast in the neighborhood of Ras Shamra (ancient Ugarit) (Smith’s group 4 = our fabrics 5 and 7 and chemical group IX). Notwith-standing the popularity of these four Syro-Palestinian fabrics at Kommos, one of the more striking features of the corpus of CJs from the site is its overall heterogeneity. Among the 33 sampled CJs, a total of 12 distinct, petrographically defined fabric types are represented, most of them also exhibiting discrete patterns of chemical composition as well.

Overall, then, this analysis of the principal types of transport jar present at Kommos in the Late Bronze Age has revealed a wealth of information. The area of the Mesara has been shown to have hosted the production not only of SNAs, but also of TSJs, the latter over a longer period of time. Petrographic fabrics representing production from the same raw materi-als have demonstrated that at least some of the SNAs and TSJs shared a production location.

Turning to the imported Egyptian amphoras and other shapes, we note a remarkable uniformity of composition, perhaps reflecting their origin at

114. Rutter 2006b, pp. 649–653. About 10% of this total (n = 7) occur as kick-ups in historical levels, a clear indication that the dates of the contexts in which these and other similar pieces have been found are simply termini post quos and do not necessarily consti-tute reliable dates for the arrival of such imports at Kommos, much less for their actual dates of production at various centers in the Levant. This caveat is particularly applicable to the examples of such containers recovered in the form of small fragments within chrono-logically mixed building fills.

115. Rutter 2006b, pp. 684–687; 2006c.


a single location over a long period of time. Although both petrography and chemistry suggest a provenance in the Memphis region, further work in Egypt is required to be more confident about the origins of this specific fabric. In contrast to the EJs, the CJs from Kommos represent a variety of sources that have been isolated with some precision, and their origins, from a number of locations along the Syro-Palestinian coast, reveal maritime trade connections between Kommos and specific areas of the Levant.

Comparative evidence also enables us to speculate about the prob-able contents of these jars. While there is reason to believe that TSJs may have been used for the transportation of perfumed oils,116 we have as yet little indication of what the SNAs may have held. Hieratic inscriptions on Egyptian amphoras from the Nile Delta indicate that they held wine, but what of the Canaanite jars, with their variety of origins? The work of the Cambridge Amphora Project holds the key to such questions, and their data correspond well to the fabrics identified in this study. Through an array of techniques, Serpico and others have shown that those CJs with origins on the Lebanese and Syrian coasts carried resin, probably terebinth resin, while those from the coastal region of Israel and the Jezreel Valley usually contained oil.117 The integrated physical and chemical analysis of coarse-ware pottery fabrics has thus provided a window on the movement of various goods and commodities around the eastern Mediterranean, as well as the place of the harbor of Kommos in that activity.

116. Shelmerdine 1984.117. Bourriau, Smith, and Serpico

2001, pp. 140–144, with references.

REFERENCES

Allen, R. O., and H. Hamroush. 1984. “The Application of Geochemical Techniques to the Investigation of Two Predynastic Sites in Egypt,” in Archaeological Chemistry III (Advances in Chemistry 205), ed. J. B. Lambert, Washington, D.C., pp. 51–65.

Allen, R. O., M. S. Rogers, R. S. Mitchell, and M. A. Hoffman. 1982. “A Geochemical Approach to the Understanding of Ceramic Technology in Predynastic Egypt,” Archaeometry 24, pp. 199–212.

Aston, D. A. 1998. Die Keramik des Grabungsplatzes Q1 1: Corpus of Fabrics, Wares, and Shapes (For- schungen in der Ramses-Stadt: Die Grabungen des Pelizaeus-Museums Hildesheim in Qantir–Pi-Ramesse 1), Mainz.

Baxter, M. J., C. C. Beardah, I. Papa-georgiou, M. A. Cau, P. M. Day, and V. Kilikoglou. 2008. “On Statis-tical Approaches to the Study of Ceramic Artefacts Using Geochem-ical and Mineralogical Data,” Archaeometry 50, pp. 142–157.

Beardah, C. C., M. J. Baxter, I. Papa-georgiou, and M. A. Cau. 2003. “‘Mixed-Mode’ Approaches to the Grouping of Ceramic Artefacts Using S-Plus,” in The Digital Heri-tage of Archaeology: Computer Appli-cations and Quantitative Methods in Archaeology. Proceedings of the 30th Conference, Heraklion, Crete, April 2002, ed. M. Doerr and A. Sarris, Athens, pp. 261–265.

Belfiore, C., P. M. Day, A. Hein, V. Kilikoglou, V. La Rosa, P. Maz-zoleni, and A. Pezzino. 2007. “Petrographic and Chemical Char-acterisation of Pottery Production of the Late Minoan I Kiln at Haghia Triada, Crete,” Archaeometry 49, pp. 621–653.

Ben-Shlomo, D., E. Nodarou, and J. B. Rutter. 2011. “Transport Stirrup Jars from the Southern Levant: New Light on Commodity Exchange in the Eastern Mediterranean,” AJA 115, pp. 329–353.

Betancourt, P. P. 1984. East Cretan White-on-Dark Ware: Studies on a Handmade Pottery of the Early to


Middle Minoan Periods (University Museum Monograph 51), Philadel-phia.

Bonneau, M., H. A. Jonkers, and J. E. Meulenkamp. 1984. Geological Map of Greece, 1:50,000: Timbakion Sheet, Athens.

Bourriau, J. 2004. “The Beginnings of Amphora Production in Egypt,” in Invention and Innovation: The Social Context of Technological Change 2: Egypt, the Aegean, and the Near East, 1650–1150 B.C., ed. J. Bourriau and J. Phillips, Oxford, pp. 78–95.

Bourriau, J. D., and P. T. Nicholson. 1992. “Marl Clay Pottery Fabrics of the New Kingdom from Mem-phis, Saqqara, and Amarna,” JEA 78, pp. 29–91.

Bourriau, J. D., L. M. V. Smith, and P. T. Nicholson. 2000. New Kingdom Pottery Fabrics: Nile Clay and Mixed Nile/Marl Clay Fabrics from Memphis and Amarna (EES Occasional Pub-lication 14), London.

Bourriau, J., L. Smith, and M. Serpico. 2001. “The Provenance of Canaan-ite Amphorae Found at Memphis and Amarna in the New Kingdom,” in The Social Context of Technological Change: Egypt and the Near East, 1650–1550 B.C. Proceedings of a Conference Held at St. Edmund Hall, Oxford, 12–14 September 2000, ed. A. J. Shortland, Oxford, pp. 113–146.

Buxeda i Garrigós, J., and V. Kiliko-glou. 2003. “Total Variation as a Measure of Variability in Chemical Data-Sets,” in Patterns and Process: A Festschrift in Honor of Dr. Edward V. Sayre, ed. L. van Zelst, Suitland, Md., pp. 185–198.

Buxeda i Garrigós, J., V. Kilikoglou, and P. M. Day. 2001. “Chemical and Mineralogical Alteration of Ceram-ics from a Late Bronze Age Kiln at Kommos, Crete: The Effect on the Formation of a Reference Group,” Archaeometry 43, pp. 349–371.

Cline, E. H. 1994. Sailing the Wine-Dark Sea: International Trade and the Late Bronze Age Aegean (BAR-IS 591), Oxford.

Cohen-Weinberger, A., and Y. Goren. 2004. “Levantine-Egyptian Interac-tions during the 12th to the 15th Dynasties Based on the Petrography

of the Canaanite Pottery from Tell el-Dab‘a,” Ägypten und Levante 14, pp. 69–100.

Day, P. M. 1995a. “Appendix C: Petro-graphic Analysis of Transport Stir-rup Jars from the House of the Oil Merchant, Mycenae,” in I. Tour-navitou, The “Ivory Houses” at My- cenae (BSA Suppl. 24), London, pp. 309–316.

———. 1995b. “Pottery Production and Consumption in the Siteia Bay Area during the New Palace Pe- riod,” in Achladia: Scavi e ricerche della Missione greco-italiana in Creta orientale (1991–1993) (Incunabula graeca 97), ed. M. Tsipopoulou and L. Vagnetti, Rome, pp. 148–175.

———. 1997. “Ceramic Exchange between Towns and Outlying Set-tlements in Neopalatial East Crete,” in The Function of the “Minoan Villa.” Proceedings of the Eighth International Symposium at the Swedish Institute at Athens, 6–8 June 1992 (SkrAth 4º, 46), ed. R. Hägg, Stockholm, pp. 219–228.

———. 1999. “Petrographic Analysis of Ceramics from the Shipwreck at Point Iria,” in Phelps, Lolos, and Vichos 1999, pp. 59–75.

Day, P. M., and H. Haskell. 1995. “Transport Stirrup Jars as Indicators of Trade and Production in the Late Bronze III Aegean,” in Trade and Production in Premonetary Greece: Aspects of Trade. Proceedings of the Third International Workshop, Athens 1993 (SIMA-PB 134), ed. C. Gillis, C. Risberg, and B. Sjöberg, Jon-sered, pp. 87–109.

Day, P. M., and R. E. Jones. 1991. “Petrographic and Chemical Anal- ysis of Inscribed and Other Stirrup Jars from Malia,” BCH 115, pp. 94– 97.

Day, P. M., and L. Joyner. 2006. “Coarseware Stirrup Jars from Can-natello, Sicily: New Evidence from Petrographic Analysis,” SMEA 47, pp. 309–314.

Day, P. M., V. Kilikoglou, and J. B. Rut- ter. 2004. “A World of Goods: Trans- port Jar Sources and Commodity Trade in the Late Bronze Age East-ern Mediterranean” (paper, 34th International Symposium on Archaeometry, Zaragoza 2004).

Day, P. M., P. S. Quinn, V. Kilikoglou, and J. Rutter. 2006. “‘Του κόσμου τα καλά’: Το λιμάνι του Kομμού και το εμπόριο αγαθών στην ανατο-λική Μεσόγειο κατά την Ύστερη Εποχή του Χαλκού” (paper, 10th International Cretological Congress, Chania 2006).

Goren, Y., S. Bunimovitz, I. Finkel-stein, and N. Na’aman. 2003. “The Location of Alashiya: New Evi-dence from Petrographic Investi- gation of Alashiyan Tablets from El-Amarna and Ugarit,” AJA 107, pp. 233–255.

Goren, Y., I. Finkelstein, and N. Na’a- man. 2003. “The Expansion of the Kingdom of Amurru according to the Petrographic Investigation of the Amarna Tablets,” BASOR 329, pp. 1–11.

Hallager, E., and B. P. Hallager, eds. 2000. The Greek-Swedish Excava-tions at the Agia Aikaterini Square, Kastelli, Khania, 1970–1987 II: The Late Minoan IIIC Settlement (SkrAth 4º, 47.2), Stockholm.

———. 2003. The Greek-Swedish Exca-vations at the Agia Aikaterini Square, Kastelli, Khania, 1970–1987 III: The Late Minoan IIIB:2 Settlement (SkrAth 4º, 47.3), Stockholm.

Hancock, R., and P. P. Betancourt. 1987. “INAA of Minoan Ceramics from Kommos, Crete,” Journal of Radioanalytical and Nuclear Chemis-try 114, pp. 393–401.

Hancock, R. G. V., N. B. Millet, and A. J. Mills. 1986. “A Rapid INAA Method to Characterize Egyp- tian Ceramics,” JAS 13, pp. 107–117.

Haskell, H. W. 1981a. “The Coarse Ware Stirrup Jars of Crete and the Cyclades” (diss. Univ. of North Carolina, Chapel Hill).

———. 1981b. “Coarse-Ware Stirrup-Jars at Mycenae,” BSA 76, pp. 225–238.

———. 1984. “Pylos: Stirrup Jars and the International Oil Trade,” in Pylos Comes Alive: Industry and Administration in a Mycenaean Pal-ace. Papers of a Symposium, ed. C. W. Shelmerdine and T. G. Palaima, New York, pp. 97–107.

———. 2005. “Region to Region Ex- port of Transport Stirrup Jars from


LM IIIA2/B Crete,” in Ariadne’s Threads: Connections between Crete and the Greek Mainland in Late Minoan III (LM IIIA2 to LM IIIC). Proceedings of the International Work-shop Held at Athens, Scuola archeolo- gica italiana, 5–6 April 2003 (Tripo-des 3), ed. A. L. D’Agata, J. Moody, and E. Williams, Athens, pp. 205–221 (with response by A. Kanta [pp. 223–234] and discussion [pp. 235–241]).

Haskell, H. W., R. E. Jones, P. M. Day, and J. T. Killen. 2011. Transport Stirrup Jars of the Bronze Age Aegean and East Mediterranean (Prehistory Monographs 33), Philadelphia.

Hatzaki, E. M. 2007a. “Neopalatial (MM IIIB–LM IB): KS 178, Gyp-sades Well (Upper Deposit), and SEX North House Groups” in Knossos Pottery Handbook: Neolithic and Bronze Age (Minoan) (BSA Studies 14), ed. N. Momigliano, London, pp. 151–196.

———. 2007b. “Final Palatial (LM II– IIIA2) and Postpalatial (LM IIIB–LM IIIC Early): MUM South Sec-tor, Long Corridor Cists, MUM Pits (8, 10–11), Makritikhos ‘Kitchen,’ MUM North Platform Pits, and SEX Southern Half Groups,” in Knossos Pottery Hand-book: Neolithic and Bronze Age (Mi- noan) (BSA Studies 14), ed. N. Mo- migliano, London, pp. 197–251.

Hein, A., P. M. Day, P. S. Quinn, and V. Kilikoglou. 2004. “Geochemical Diversity of Neogene Clay Deposits in Crete and Its Implications for Provenance Studies of Minoan Pot-tery,” Archaeometry 46, pp. 357–384.

Jones, R. E. 1986. Greek and Cypriot Pottery: A Review of the Scientific Studies (Fitch Laboratory Occa-sional Paper 1), London.

Kilikoglou, V., A. P. Grimanis, A. Tso-lakidou, A. Hein, D. Malamidou, and Z. Tsirtsoni. 2007. “Neutron Activation Patterning of Archaeo-logical Materials at the National Center for Scientific Research ‘Demokritos’: The Case of Black-on-Red Neolithic Pottery from Macedonia, Greece,” Archaeometry 49, pp. 301–319.

Knapp, A. B., and J. F. Cherry. 1994. Provenience Studies and Bronze Age

Cyprus: Production, Exchange, and Politico-Economic Change (Mono-graphs in World Archaeology 21), Madison.

Kommos = Kommos: An Excavation on the South Coast of Crete, Princeton

II = P. P. Betancourt, The Final Neolithic through Middle Minoan III Pottery, 1990.

III = L. V. Watrous, The Late Bronze Age Pottery, 1992.

V = J. W. Shaw and M. C. Shaw, eds., The Monumental Minoan Buildings at Kommos, 2006.

Langohr, C. 2009. Périphéreia: Étude régional de la Crète aux Minoen récent II–IIIB (1450–1200 av. J.-C.) 1: La Crète centrale et occidentale (Aegis 3), Louvain-la-Neuve.

Leonard, A. Jr. 1994. An Index to the Late Bronze Age Aegean Pottery from Syria-Palestine (SIMA 114), Jonsered.

———. 1996. “ ‘Canaanite Jars’ and the Late Bronze Age Aegeo-Levantine Wine Trade,” in The Origins and Ancient History of Wine (Food and Nutrition in History and Anthro-pology 1), ed. P. E. McGovern, S. J. Fleming, and S. H. Katz, pp. 233–254.

Lolos, Y. G. 1999. “The Cargo of Pot-tery from the Point Iria Wreck: Character and Implications,” in Phelps, Lolos, and Vichos 1999, pp. 43–58.

Manning, S. W., and M. J. Bruce, eds. 2009. Tree-Rings, Kings, and Old World Archaeology and Environment: Papers Presented in Honor of Peter Ian Kuniholm, Oxford.

McGovern, P. E. 1997. “Wine of Egypt’s Golden Age,” JEA 83, pp. 69–108.

———. 2000. The Foreign Relations of the “Hyksos”: A Neutron Activation Study of Middle Bronze Age Pottery from the Eastern Mediterranean (BAR-IS 888), Oxford.

———. 2003. Ancient Wine: The Search for the Origins of Viniculture, Prince- ton.

Mommsen, H., and B. L. Sjöberg. 2007. “The Importance of the ‘Best Relative Fit Factor’ When Evaluat-ing Elemental Concentration Data of Pottery Demonstrated with Mycenaean Sherds from Sinda,

Cyprus,” Archaeometry 49, pp. 359–371.

Moustaki, I., and I. Papageorgiou. 2004. “Latent Class Models for Mixed Variables with Applications in Archaeometry,” Computational Statistics and Data Analysis 48, pp. 659–675.

Myer, G. H., and P. P. Betancourt. 1990. “The Fabrics at Kommos,” in Kommos II, pp. 1–13.

Newton, G. W. A., J. Bourriau, E. B. French, and A. J. N. W. Prag. 2007. “INAA of Archaeological Samples at the University of Man-chester,” Archaeometry 49, pp. 289–319.

Nicholson, P. T., and P. Rose. 1985. “Pottery Fabrics and Ware Groups at el-Amarna,” in Amarna Reports II (EES Occasional Publication 2), ed. B. J. Kemp, London, pp. 133–174.

Nodarou, E. 2011. Pottery Production, Distribution, and Consumption in Early Minoan West Crete: An Ana-lytical Perspective (BAR-IS 2210), Oxford.

Phelps, W., Y. Lolos, and Y. Vichos, eds. 1999. The Point Iria Wreck: Interconnections in the Mediterranean, ca. 1200 B.C. Proceedings of the Inter-national Conference, Island of Spetses, 19 September 1998, Athens.

PM = A. J. Evans, The Palace of Minos, 4 vols., London 1921–1935.

Poursat, J.-C., and C. Knappett. 2005. Fouilles exécutées à Malia: Le Quar- tier Mu IV: La poterie du Minoen moyen 2: Production et utilisation (ÉtCrét 33), Paris.

Redmount, C. A., and M. E. Morgen-stein. 1996. “Major and Trace Ele-ment Analysis of Modern Egyptian Pottery,” JAS 23, pp. 741–762.

Riley, J. A. 1981. “Petrological Exami-nation of Coarse-Ware Stirrup-Jars from Mycenae,” BSA 76, pp. 335–340.

———. 1983. “The Contribution of Ceramic Petrology to Our Under-standing of Minoan Society,” in Minoan Society. Proceedings of the Cambridge Colloquium 1981, ed. O. Krzyszkowska and L. Nixon, Bristol, pp. 283–292.

Rutter, J. B. 1999. “Cretan External Relations during Late Minoan


IIIA2–B (ca. 1370–1200 b.c.): A View from the Mesara,” in Phelps, Lolos, and Vichos 1999, pp. 139–186.

———. 2000. “The Short-Necked Amphora of the Post-Palatial Me- sara,” in Πεπραγμένα του Η΄ Διεθνούς Κρητολογικού Συνεδρίου, Ηράκλειο, 9–14 Σεπτεμβρίου, 1996 A3, ed. A. Karetsou, Herakleion, pp. 177–188.

———. 2005. “Assessing the Shipboard Profile of a Regional Ceramic Assemblage: The Aegean Pottery from the Ulu Burun Shipwreck” (paper, 106th Annual Meeting of the AIA, Boston 2005).

———. 2006a. “Neopalatial and Later Minoan Pottery,” in Kommos V, pp. 377–630, 694–710, 1115–1187.

———. 2006b. “Ceramic Imports of the Neopalatial and Later Bronze Age Eras,” in Kommos V, pp. 646–688, 712–715.

———. 2006c. “Ceramic Evidence for External Contact: Neopalatial and Postpalatial,” in Kommos V, pp. 859–863.

———. Forthcoming. “The Canaanite Transport Amphora within the Late Bronze Age Aegean: A 2010 Per-spective on a Frequently Changing Picture,” in KE-RA-ME-JA: Fest-schrift in Honor of Cynthia W. Shel- merdine, Philadelphia.

Serpico, M. 1996. “Mediterranean Resins in New Kingdom Egypt: A Multidisciplinary Approach to Trade and Usage” (diss. University College London).

———. 2000. “Resins, Amber, and Bitumen,” in Ancient Egyptian Materials and Technology, ed. P. T. Nicholson and I. Shaw, Cambridge, pp. 430–474.

Serpico, M., J. Bourriau, L. Smith, Y. Goren, B. Stern, and C. Heron. 2003. “Commodities and Contain-ers: A Project to Study Canaanite Amphorae Imported into Egypt during the New Kingdom,” in The Synchronisation of Civilisations in the Eastern Mediterranean in the Second Millennium B.C. 2. Proceedings of the SCIEM 2000 EuroConference, Hain-dorf, 2–7 May 2001 (DenkschrWien 29), ed. M. Bietak, Vienna, pp. 365–375.

Shaw, J. W. 2006a. “The Architecture and Stratigraphy of the Civic Build-ings,” in Kommos V, pp. 1–116.

———. 2006b. “Architectural Forms and Their Uses,” in Kommos V, pp. 846–854, 875–877.

Shaw, J. W., A. van der Moortel, P. M. Day, and V. Kilikoglou. 1997. “A LM IA Pottery Kiln at Kommos, Crete,” in TEXNH: Craftsmen, Craftswomen, and Craftsmanship in the Aegean Bronze Age. Proceedings of the 6th International Aegean Con-ference, Philadelphia, Temple Univer-sity, 18–21 April 1996 (Aegaeum 16), ed. R. Laffineur and P. P. Betan-court, Liège, pp. 317–331.

———. 2001. A LM IA Ceramic Kiln in South-Central Crete: Function and Pottery Production (Hesperia Suppl. 30), Princeton.

Shaw, M. C. 1985. “Late Minoan I Building J/T, and Late Minoan III Buildings N and P at Kommos: Their Nature and Possible Uses as Residences, Palaces, and/or Empo-ria,” in A Great Minoan Triangle in South-Central Crete: Kommos, Hagia Triadha, Phaistos (Scripta Mediterranea 6), ed. J. W. Shaw and M. C. Shaw, Toronto, pp. 19– 30.

Shelmerdine, C. W. 1984. “The Per-fumed Oil Industry at Pylos,” in Pylos Comes Alive: Industry and Administration in a Mycenaean Pal-ace. Papers of a Symposium, ed. C. W. Shelmerdine and T. G. Palaima, New York, pp. 81–96.

———. 2008. “Background, Sources, and Methods,” in The Cambridge Companion to the Aegean Bronze Age, ed. C. W. Shelmerdine, Cambridge, pp. 1–18.

Singer-Avitz, L., and Y. Levy. 1992. “An MB IIa Kiln at the Nahal Soreq Site,” ‘Atiqot 21, pp. 9–14.

Smith, L. M. V., J. D. Bourriau, Y. Go- ren, M. J. Hughes, and M. Serpico. 2004. “The Provenance of Canaan-ite Amphorae Found at Memphis and Amarna in the New Kingdom: Results 2000–2002,” in Invention and Innovation: The Social Context of Technological Change 2: Egypt, the Aegean, and the Near East, 1650–1150 B.C., ed. J. Bourriau and J. Phillips, Oxford, pp. 55–77.

Smith, L. M. V., J. D. Bourriau, and M. Serpico. 2000. “The Provenance of Late Bronze Age Transport Am- phorae Found in Egypt,” Internet Archaeology 9 (http://intarch.ac.uk).

Stampolidis, N., A. Karetsou, and A. Kanta, eds. 1998. The Eastern Mediterranean: Cyprus-Dodecanese-Crete, 16th–6th Century B.C. (Exhibi-tion catalogue, Archaeological Museum of Herakleion 1998), Herakleion.

Sugerman, M. O. 2000. “Webs of Commerce: The Archaeology of Ordinary Things in Late Bronze Age Israel and Palestine” (diss. Har-vard Univ.).

Tomkins, P., and P. M. Day. 2001. “Pro-duction and Exchange of the Earli-est Ceramic Vessels in the Aegean: A View from Early Neolithic Knos-sos, Crete,” Antiquity 75, pp. 259–260.

Tomkins, P. D., P. M. Day, and V. Kili- koglou. 2004. “Knossos and the Earlier Neolithic Landscape of the Herakleion Basin,” in Knossos: Pal-ace, City, State. Proceedings of the Conference in Herakleion Organised by the British School at Athens and the 23rd Ephoreia of Prehistoric and Clas-sical Antiquities, in November 2000, for the Centenary of Sir Arthur Evans’s Excavations at Knossos (BSA Studies 12), ed. G. Cadogan, E. Hatzaki, and A. Vasilakis, London, pp. 51– 59.

Tomlinson, J. E. 1995. “Multivariate Analysis of Neutron Activation Data on Heavy Ware Stirrup Jars from Mycenae,” in I. Tournavi- tou, The “Ivory Houses” at Mycenae (BSA Suppl. 24), London, pp. 305–308.

———. 1996. “Chemical Evidence for a Cretan Origin of Heavy Ware Stirrup Jars Found at Mycenae,” in Archaeometrical and Archaeological Research in Macedonia and Thrace. Proceedings of the 2nd Symposium of the Hellenic Archaeometrical Society, Thessaloniki (26–28 March 1993), ed. I. Stratis, M. Vavelidis, K. Ko- tsakis, G. Tsokas, and E. Tsoukala, Thessaloniki, pp. 371–378.

Tomlinson, J. E., J. B. Rutter, and S. M. A. Hoffmann, 2010. “Myce-naean and Cypriot Late Bronze Age


Ceramic Imports to Kommos: An Investigation by Neutron Activation Analysis,” Hesperia 79, pp. 191–231.

Watrous, L. V. 1985. “Late Bronze Age Kommos: Imported Pottery as Evidence for Foreign Contact,” in A Great Minoan Triangle in South-Central Crete: Kommos, Hagia Triadha, Phaistos (Scripta Mediterranea 6), ed. J. W. Shaw and M. C. Shaw, Toronto, pp. 7–11.

———. 1989. “A Preliminary Report on Imported ‘Italian’ Wares from the Late Bronze Age Site of Kom-mos on Crete,” SMEA 27, pp. 69– 79.

Watrous, L. V., P. M. Day, and R. E. Jones. 1998. “The Sardinian Pottery from the Late Bronze Age Site of Kommos in Crete: Description, Chemical and Petrographic Analy-ses, and Historical Context,” in Sardinian and Aegean Chronology: Towards the Resolution of Relative and Absolute Dating in the Mediter-

ranean. Proceedings of the Interna-tional Colloquium “Sardinian Stratig-raphy and Mediterranean Chronol-ogy,” Tufts University, Medford, Massachusetts, March 17–19, 1995 (Studies in Sardinian Archaeology 5), ed. M. S. Balmuth and R. H. Tykot, Oxford, pp. 337–340.

Whitbread, I. K. 1989. “A Proposal for the Systematic Description of Thin Sections towards the Study of Ancient Ceramic Technology,” in Archaeometry. Proceedings of the 25th International Symposium, ed. Y. Maniatis, Amsterdam, pp. 127–138.

———. 1995. Greek Transport Ampho-rae: A Petrological and Archaeological Study (Fitch Laboratory Occasional Paper 4), Athens.

Whitelaw, T., P. M. Day, E. Kiriatzi, V. Kilikoglou, and D. E. Wilson. 1997. “Ceramic Traditions at EM IIB Myrtos, Fournou Korifi,” in TEXNH: Craftsmen, Craftswomen,

and Craftsmanship in the Aegean Bronze Age. Proceedings of the 6th International Aegean Conference, Philadelphia, Temple University, 18–21 April 1996 (Aegaeum 16), ed. R. Laffineur and P. P. Betan-court, Liège, pp. 265–275.

Wild, E. M., W. Gauss, G. Forsten-pointner, M. Lindblom, R. Smetana, P. Steier, U. Thanheiser, and F. We- ninger. 2010. “14C Dating of the Early to Late Bronze Age Strati-graphic Sequence of Aegina Ko- lonna, Greece,” in Beam Interactions with Materials and Atoms (Nuclear Instruments and Methods in Phys-ics Research B 268), ed. M. B. H. Breese, L. E. Rehn, C. Trautman, and I. Vickridge, Amsterdam, pp. 1013–1021.

Wilson, D. E., and P. M. Day. 1994. “Ceramic Regionalism in Prepalatial Central Crete: The Mesara Imports at EM I to EM IIA Knossos,” BSA 89, pp. 1–87.

Peter M. Day

Universit y of Sheffielddepartment of archaeolog ynorthgate housewest streetsheffield s1 4etunited kingdom

p.m.day@sheff ie ld.ac .uk

Patrick S. Quinn

Universit y Col lege Londoninstitute of archaeolog y31–34 gordon squarelondon wc1h 0p yunited kingdom

patr ick.quinn@uc l .ac .uk

Jeremy B. Rutter

Dartmouth Col legedepartment of classics310 reed hallhanover, new hampshire 03755

jeremy.b.rutter@dar tmouth.edu

Vassilis Kilikoglou

NCSR Demokr itosinstitute of material scienceayia paraskevi153 10 athensgreece

ki l [email protected] itos .gr

© The Amer i c an Sc hoo l o f C l a s s i c a l S tud i e s a t Athens h t t p : / /dx .do i . o rg /10 .2972/he spe r i a . 80 .4 .0511 . app1

APPENDIX 1

PETROGRAPHIC FABRIC DESCRIPTIONS OF LATE MINOAN TRANSPORT JARS FROM KOMMOS

An appendix to “A World of Goods: Transport Jars and Commodity Ex-change at the Late Bronze Age Harbor of Kommos, Crete,” by P. M. Day, P. S. Quinn, J. B. Rutter, and V. Kilikoglou: Hesperia 80 (2011), pp. 511– 558; http://dx.doi.org/10.2972/hesperia.80.4.0511

Summaries of the main characteristics of the petrographic fabric groups appear in the main text of the article, together with a series of thin-section micrographs (Figs. 3–6). More detailed petrographic descriptions of each fabric are presented below, according to a modified version of the system proposed by Whitbread.1

FABRIC A: MAIN SOU TH-CENTRAL CRETAN

Kommos 98/1, 2, 4, 7, 9, 11, 14, 16, 17, 22, 24, 26–30, 32

INCLUSIONS10%–20%.2 Eq and el, wr–sr, randomly oriented. Bimodal grain-size distribution; moderately (e.g., samples 98/14, 22) to poorly (e.g., samples 98/9, 17) sorted. Consisting of coarse inclusions set in a very fine-grained matrix (e.g., samples 98/9, 17) or just in the clay matrix (e.g., samples 98/14, 22).

Coarse Fraction3.25–0.1 mm.3

Quartz-biotite schist: variable grain size and texture. Most inclusions have a general alignment of the biotite mica laths, some have biotite and quartz-rich layers (e.g., sample 98/32). Some inclusions contain calcite in addition to biotite and quartz, others exhibit a crenulation cleavage.

1. Whitbread 1989, 1995. Abbrevia-tions used in this appendix: a = angular; el = elongate; eq = equant; PPL = plane polarized light; r = rounded; sa = sub- angular; sr = subrounded; TF = textural feature; wr = well rounded; XP = crossed polars.

2. Boundary between inclusions and matrix is 10 μm.

3. This fabric is characterized by a wide range of coarse nonplastic inclu-sions, which is not always represented within individual samples. However, the range and frequency of rock and

mineral inclusions overlap between the samples of this fabric, forming an over-all continuum. It is therefore not pos-sible to state the semiquantitative fre-quency of individual types of inclusions, as these may be abundant in one sample but absent in another.

hesperia 80 (201 1)Online Supplement

patr ic k s . q u inn2

Phyllite: fine-grained, composed of quartz and biotite crystals with strong alignment.

Slate: fine-grained with aligned chloritic material.Fine-grained volcanic rock fragments, two types: 1. Wr grains with a por-

phyritic texture, composed of feldspar laths set in a red-brown, devitrified matrix. The feldspar is often partially altered. Probably of basic composition (e.g., samples 98/1, 2, 29). 2. Medium-grained igneous rock fragments composed of pyroxene, plagioclase feldspar in a red-brown devitrified groundmass (cf. dolerite).

Polycrystalline quartz: usually straight extinction, but can be undulose. Some grains exhibit a stretched metamorphic texture and others grade into cataclasites (e.g., sample 98/9).

Sandstone: generally composed of moderately well-sorted quartz grains set in either a clay-rich or a fine-grained quartz-rich matrix. Quartz arenite composi-tion, but can grade into quartzwackes (e.g., samples 98/11, 28, 32) with increasing matrix. Some are partially metamorphosed, the quartz grains breaking down to form polycrystalline quartz and the clay matrix recrystallized to form biotite mica.

Siltstone, two types: 1. wr, orange-brown in PPL and XP (x100). Consisting of quartz and mica in a red clay matrix. Fairly well sorted, can have graded bedding (e.g., sample 98/16). Can grade into fine-grained schists/phyllites, with the alignment of biotite mica laths. 2. r, pale brown in PPL and dark gray in XP (x100). Consisting of sparse quartz and biotite mica laths in a very fine-grained/glassy matrix. Has a crude alignment/lamination. Grades into slate described above (e.g., sample 98/1).

Chert: can have an irregular muddy appearance. Possibly grading into cherty radiolarian mudstone. Sometimes contains chalcedony radiolaria (e.g., sample 98/22).

Micrite: r–sr grains, rarely a (e.g., sample 98/24).Mudstones: similar in color to the matrix. Can exhibit faint polygonal cracked

structure (e.g., sample 98/14).Chlorite: fibrous texture, wrQuartz and feldspar rock fragment: possibly an acid igneous rock (granite/

microgranite)Monocrystalline quartzBiotite micaTFsOpaquesAlkali feldsparPyroxene

Fine Fraction0.1–0.01 mm.4

Dominant to common

Monocrystalline quartz

Dominant to absent

Micrite

Few to rare

Polycrystalline quartz

Rare to absent

Biotite mica

4. The fine fraction varies from almost absent to quite abundant. Sam-ples without a significant fine-grained fraction may be the result of levigation prior to the addition of coarse-grained temper.

fabr ic de scr ip t ions of transp ort jars fr om kommos 3

OpaquesMudstone

Very rare to absent

Alkali feldsparEpidoteWhite micaMonocrystalline calciteChertQuartz biotite schist

MATRIX60%–85%. Red-brown, orange-brown to yellow-brown to green-brown to gray-brown in PPL (x40) and red-brown to dark brown to green-gray in XP (x40). Generally homogeneous; optically inactive to slightly active in most of the samples, but can be moderately active (e.g., samples 98/4, 11, 14, 26); noncalcareous.

TFs few to absent, eq and el, sr–r. Size = <3.25 mm; mode = 0.45 mm. Dark brown in PPL (x40) and dark red-brown in XP (x40). Clear to diffuse boundaries; high optical density; discordant with the groundmass (e.g., samples 98/4, 14, 17, 27, 30). Composed of sparse or, less commonly, fairly well-packed fine-grained inclusions of monocrystalline quartz and rare polycrystalline quartz, chert, and altered igneous rock fragments (e.g., sample 98/17).

VOIDS5%–10%. Common to rare voids, consisting of common to very few mesovughs, rare to few macrovughs, rare to very few megavughs (e.g., sample 98/14), com-mon to absent macro- and megaplanar voids, and very rare to few mesoplanar voids. Double- to open-spaced; moderate to crude orientation of the long axes of el voids parallel with the margins of the sections. Some samples exhibit secondary micritic lining of voids.

FABRIC B: MEDIUM/COARSE IGNEOUS

Kommos 98/15, 19, 20, 87, 88

INCLUSIONS15%–20%. Eq and el, sa–sr, open- to single-spaced, crudely (sample 98/88) to moderately (sample 98/20) aligned to the margins of the samples. Weak bimodal grain-size distribution, except for sample 98/19, which is moderately bimodal.

Coarse Fraction3.0–0.1 mm.

Common

Monocrystalline quartz: eq, sa–sr. Size = <0.1 mm; mode = 0.25 mm. Straight to slightly undulose extinction.

Few

Plagioclase feldspar: el, a–sr. Size = <1 mm; mode = 0.5 mm. Polysynthetic twinning.


Few to rare

Alkali feldspar: el, sa–sr. Size = <0.6 mm; mode = 0.35 mm.

Few to absent

TFs: eq and el, wr. Size = <1.1 mm; mode = 0.3 mm.

Very few

Microperthite: el, sr. Size = <2.9 mm; mode = 0.25 mm (e.g., sample 98/87).Polycrystalline quartz: el, sa–r. Size = <3.0 mm; mode = 0.45 mm. Inequi-

granular, grades into cataclasite in samples 98/15, 19, 20, 87.Quartz, alkali feldspar, and biotite mica rock fragments: eq and el, sa–sr.

Size = <1.25 mm; mode = 0.6 mm. Possibly from a fine/medium-grained acid/intermediate igneous rock, perhaps granodiorite. Not every grain contains all of the above minerals.

Rare

Chert: eq and el, sr–r. Size = <0.6 mm; mode = 0.3 mm.Micrite: el, sr–r. Size = <1.5 mm; mode = 0.45 mm.Opaques: eq and el, sa–wr. Size = <0.35 mm; mode = 0.15 mm.

Very rare to absent

Microfossils (foraminifera): eq, wr. Size = 0.35–0.3 mm (e.g., sample 98/88).Altered igneous rock fragments: eq and el, r–wr. Size = <2.0 mm; mode =

0.28 mm (e.g., samples 98/88, 20). Composed of feldspars in a red-brown altered matrix (cf. diorite), or completely altered with remnant mineral structure (e.g., sample 98/15).

Biotite mica: eq and el, sa. Size = <0.35 mm; mode = 0.15 mm.Altered mineral (cf. olivine): eq, r. Size = 0.3 mm. Very altered (sample 98/88).Kyanite: el, a. Size = 0.5 mm (sample 98/20).

Fine Fraction0.1–0.01 mm.

Common

Monocrystalline quartzMicriteBiotite mica

Few

Opaques

Rare

ChertPolycrystalline quartzAlkali feldspar

Very rare

Plagioclase feldsparWhite mica


MATRIX75%–83%. Orange-brown in PPL (x40) and red-brown in XP (x40). Homoge-neous; optically inactive (e.g., sample 98/87) to moderately active (e.g., sample 98/20); noncalcareous.

TFs: few to absent, eq and el, wr. Size = <1.1 mm; mode = 0.3 mm. Two types:1. Rare to absent, orange-brown in PPL (x100) and red-brown in XP (x100).

Clear to diffuse boundaries; high optical density; concordant with the groundmass. Composed of noncalcareous clay containing monocrystalline quartz, feldspar, opaques, and white mica (e.g., sample 98/15).

2. Few to absent, gray-brown in PPL, dark gray in XP (x100). Usually clear but can exhibit merged boundaries and tails (e.g., sample 98/87), dark margins. Com-posed of poorly sorted polycrystalline quartz, chert, and rare plagioclase feldspar.

VOIDS2%. Few voids, consisting of very rare to absent megavughs, very rare macrovughs, few mesovughs, and very rare to abundant macrochannels (e.g., samples 98/15, 20). Open-spaced, with very crude orientation parallel to the margins of the samples. There is some secondary calcite deposition in some samples (e.g., sample 98/87).

FABRIC C: SERP ENT INE

Kommos 98/21, 25

INCLUSIONS10%–15%. Eq and el, sa–r, single- to open-spaced, moderately aligned to the margins in sample 98/25. Strongly bimodal grain-size distribution; coarse fraction moderately well sorted.


Frequent to common

Serpentine: eq and el, sa–r. Size = <2.3 mm; mode = 0.1 mm. Yellow-orange to red-orange in PPL (x100) and yellow-gray (sample 98/21) to red-orange and dark red-brown in XP (x100). Most grains exhibit a prominent mesh structure in XP; some have a remnant crystal structure in PPL (e.g., sample 98/21: possible altered olivine crystals). Some grains have dark margins in XP (e.g., sample 98/21); many grains are cracked and fragmented with associated voids, and may contain opaques (e.g., sample 98/21).

Common to few

Altered igneous rock fragments: eq, sr–r. Size = <1.2 mm; mode = 0.8 mm. Orange-brown to opaque in PPL (x100) and red-brown to black in XP (x100). Heavily (sample 98/25) to extremely (sample 98/21) altered. Contains randomly oriented (decussate) plagioclase laths (<0.3 mm), and less commonly epidote, in a totally devitrified matrix. Can be cracked and fragmented and have associated voids (sample 98/21). Composed of fine-grained intermediate or basic igneous rock.


Common to very rare

Siltstone: eq and el, sa–sr. Size = <2.8 mm; mode = 0.9 mm. Gray-brown in PPL (x100) and gray and green-brown in XP (x100). Composed of moderately well-sorted sa quartz grains with biotite, clay minerals, plagioclase, and opaques. Some bedding with quartz-rich and clay/mica-rich layers; grains can be enriched with secondary calcite deposition. Some grains appear to be partially metamorphosed.

Very few to very rare

Opaques: eq and el, sr–r. Size = <0.25 mm; mode = 0.15 mm. May be the remains of totally altered rock fragments (e.g., sample 98/21).

Rare to very rare

Chert: el, sa–sr. Size = <1.2 mm; mode = 0.3 mm. Fine-grained with rare larger grains. One large inclusion has coarser quartz in a vein as well as a possible replaced foraminifer. Can be weathered and contain opaques.

Micrite: el, sr–r. Size = 1.8–0.45 mm. Has a muddy cracked appearance. One large grain has been partly dissolved and the clay matrix surrounding it is enriched with calcite (sample 98/25).

Monocrystalline quartz: eq, sa. Size = <0.25 mm; mode = 0.2 mm. Straight extinction.

Polycrystalline quartz: el, sa. Size = <0.45 mm; mode = 0.3 mm. Composed of rare large crystals and several small eq crystals with very small biotite laths, or large and small crystals with a cataclastic texture.

TFs: eq, sr–r. Size = <0.55 mm; mode = 0.3 mm.

Rare to absent

Quartz and biotite schist: el, sa–sr. Size = <0.7 mm; mode = 0.5 mm. Re-crystallized quartz grains with biotite and rare white micas. Moderate to strong alignment; some quartz/mica banding in a few grains.

Plagioclase feldspar: eq and el, sa–sr. Size = <0.3 mm; mode = 0.2 mm.Quartz and feldspar medium-fine igneous rock fragments: eq and el, sa.

Size = <0.6 mm; mode = 0.3 mm. Composed of two or more (<0.5 mm) interlock-ing grains of two or more of the minerals quartz, plagioclase, and alkali feldspar. Some intergrowth of quartz and feldspar in places. These inclusions are likely to be the source of the isolated monocrystalline quartz and feldspar in the samples.

Very rare

Alkali feldspar: el, sa–r. Size = <0.25 mm; mode = 0.1 mm. Weathered.


Frequent to common


Common

Biotite mica

Common to few

Micrite


Few to rare

Opaques

Rare

SerpentineWhite mica

Rare to very rare


Very rare

Epidote

MATRIX75%–83%. Red and green-brown (sample 98/21) and gray and green-brown to yellow and orange-brown (sample 98/25) in PPL (x40) and orange-brown (sample 98/25) to red-brown (sample 98/21) in XP (x40). Homogeneous (sample 98/21) to slightly heterogeneous (sample 98/25); optically active (sample 98/25) to optically inactive (sample 98/21); noncalcareous.

VOIDS7%–10%. Few voids, consisting of few to absent megavughs, common mesovughs, few macrovughs, very few macroplanar voids, and rare to absent megaplanar voids. Open- to single-spaced; moderately aligned in sample 98/25. Some voids are as-sociated with large inclusions, especially in sample 98/21. Some calcareous lining of voids in sample 98/21.

FABRIC D: FINE CALCAREOUS P HYLLI TE

Kommos 98/3, 13

INCLUSIONS10%–20%. Eq and el, a–r, open- (sample 98/3) or double- to open-spaced (sample 98/13), moderately aligned to the margins of the samples. Bimodal grain-size distribution; coarse fraction moderately well sorted.


Predominant

Phyllite: eq and el, a–sa. Size = <4 mm; mode = 1 mm. Orange and red-brown in PPL (x40) and dark orange and red-brown in XP (x40). Consisting of sa–sr (<0.2 mm) quartz crystals and strongly aligned biotite and white mica. The long axes of the quartz crystals can be aligned with the metamorphic fabric that folds around the grains. The fabric exhibits a crenulation cleavage in places. Some inclusions exhibit a stronger metamorphic fabric than others. One grain in sample 98/3 is a partially metamorphosed siltstone and a few appear to have the beginnings of schistose banding.


Few

TFs: eq and el, r. Size = <0.45 mm; mode = 0.25 mm (e.g., sample 98/13).

Fine Fraction (Kommos 98/3)0.4–0.01 mm.5

Frequent


Common

MicriteBiotite micaOpaques

Few

Polycrystalline quartzWhite micaMonocrystalline calcite

Very few

ChertMicrofossils (foraminifera)TFs

Rare

Phyllite

Fine Fraction (Kommos 98/13)0.4–0.01 mm.

FrequentMonocrystalline quartzBiotite mica

Common

Micrite

Few

White micaOpaques

Rare

PhylliteTFs

Very rare

ChertMetamorphic rock fragmentsMicrofossils (foraminifera)

5. The fine fractions of the two sec-tions that constitute this fabric are described separately, since they differ in

composition and in the size range of their inclusions.


MATRIX70%–85%. Red-brown (sample 98/3) to brown (sample 98/13) in PPL (x40) and yellow-brown (sample 98/3) to orange-brown (sample 98/13) in XP (x40). Homogeneous throughout the sections; moderately optically active; calcareous.

TFs rare, more abundant in sample 98/13 than in sample 98/3, eq and el, r. Size = <0.45 mm; mode = 0.25 mm. Brown-black and opaque in PPL (x40) and dark red-brown in XP (x40). Clear to diffuse boundaries; high optical density; roughly concordant to discordant with groundmass. Composed of opaque, dark brown matrix with fine quartz, micas, and rare calcite. Can be hard to distinguish from areas enriched with opaque matter.

VOIDS5%–10%. Few (sample 98/3) to rare (sample 98/13) voids. Sample 98/3 has rare macrovughs, frequent mesovughs, common vesicles, and very rare megachannels, which are single- to open-spaced. Sample 98/13 has few macrovughs, common mesovughs, and few mesovesicles, which are double- to open-spaced and crudely oriented to the margins of the samples. Some calcareous lining of voids (sample 98/13).

FABRIC E: METASEDIMENTARY

Kommos 98/18, 23

INCLUSIONS12%. Eq and el, a–sr, single-spaced, randomly oriented (sample 98/23) to very crudely aligned with the margins of the section (sample 98/18). Strongly bimodal grain-size distribution; coarse fraction moderately well sorted.

Coarse Fraction3.8–0.24 mm.6

Frequent to common

Phyllite and siltstones: el, sr. Size = <3.1 mm; mode = 0.75 mm. Color varies from gray and gray-brown to orange-brown in PPL (x100) and dark gray-brown in XP (x100). Composed of dark, very fine clay matrix containing eq and el, sa–r, fine (<0.06 mm) monocrystalline quartz, biotite and white mica, and opaques. The micas and quartz are usually strongly aligned in a straight to often crenulate meta-morphic fabric; some grains, however, have no such alignment, but are brecciated siltstones that form part of the poorly bedded, disturbed, partly recrystallized fine sandstone inclusions in sample 98/23. Phyllite also occurs as part of a disturbed polycrystalline quartz/quartzite inclusion in sample 98/23. A few phyllite grains contain large el quartz crystals and are almost schistose.

Common to few

Polycrystalline quartz/quartzite/cataclasite: el and eq, a–sr. Size = <2.5 mm; mode = 0.75 mm. Consisting of variably sized (<0.2 mm) interlocking quartz crystals with undulose extinction, and very rare plagioclase, often in a finer quartz-rich

6. Within the coarse fraction is a range of very fine-grained (<0.03 mm) to fine-grained (<0.2 mm) metamor-phic (phyllite and quartzite) and dis-turbed/brecciated sedimentary rocks

(siltstone, sandstone, and cataclasite). These appear to grade into one another and represent the differential meta- morphism of a suite of sedimentary rocks with variable grain size, ratio of

inclusions to matrix, and composition. The above categories represent an approximate subdivision of this variable intergrading suite of rocks and their relative abundance.


matrix. Can have small secondary accessory minerals and some clay matrix and opaques (revealing the often folded texture) and areas of associated phyllite. These inclusions vary in grain size and grade into the sandstones and brecciated siltstones; they are related to the phyllite (see sample 98/23).

Few

Micrite: eq and el, sr. Size = <1.3 mm; mode = 0.55 mm. Mostly discrete inclusions with sharp boundaries, but some are less well defined and have clear to merging irregular boundaries. Can contain rare polycrystalline quartz inclusions.

Very few

Partially metamorphosed very fine sandstone: eq and el, sr. Size = <1.75 mm; mode = 1.25 mm. Composed of fine (<0.15 mm) sa–sr mono- and polycrystalline quartz crystals with undulose extinction, rare plagioclase, amphibole, and biotite mica, in a clay-rich, folded, and disturbed fabric. Grades into the quartzite and disturbed siltstone inclusions described above, and can be associated with fine siltstone and phyllite.

Rare to very rare

Igneous rock fragments: eq and el, sr. Size = <0.7 mm; mode = 0.5 mm. Fine- to medium-grained, equigranular, with randomly oriented mica laths. Contains two or more of quartz, plagioclase, biotite, and alkali feldspar with granophyric texture.

Very rare

Monocrystalline quartz: el, sa. Size = <0.45 mm; mode = 0.25 mm. Even extinction.

Biotite: el, sa. Size = 0.35 mm. Weathered inclusion consisting of two or three biotite crystals with core and margin color differentiation.

Opaques: eq and el, sr–r. Size = <0.6 mm; mode = 0.2 mm.


Frequent to common

Monocrystalline quartzMicrite and monocrystalline calcite

Common to few

Metamorphic, sedimentary, and metasedimentary rock fragments

Few

Polycrystalline quartzBiotite mica

Few to rare

Opaques

Rare

Plagioclase feldsparWhite micaChert


MATRIX83%. Light orange-brown to green and green-gray red-brown in PPL (x40), and yellow-brown to dark orange- and red-brown and dark brown in XP (x40). Relatively homogeneous (sample 98/23) to heterogeneous (sample 98/18); optically inactive to optically slightly active; composed of calcareous and noncalcare-ous clay.

Both samples exhibit signs of clay mixing in the form of elongated swirls and streaks (<4 mm) of fine, noncalcareous clay, red-brown in PPL and XP (x100), with clear to merging boundaries. This fine clay contains rare monocrystalline quartz inclusions, can be cracked in appearance (sample 98/23), and has been mixed with a coarser, more calcareous clay. In sample 98/18 there is a large micritic inclusion (el, 0.35 mm), which appears to be mixed and is associated with the fine, dark brown clay.

VOIDS5%. Rare voids, consisting of few to very few macrovughs, common mesovughs, very rare meso- and macrochannels, and rare mesovesicles. Double- to open-spaced, but can be single-spaced (sample 98/18); moderately well aligned to the margins of the section. Some voids contain a secondary calcareous lining (e.g., sample 98/18).

FABRIC F: QUARTZITE, PHYLLITE, AND SCHIST

Kommos 98/6, 31

INCLUSIONS17%–20%. Eq and el, sa–r, single-spaced or less, roughly aligned with the mar-gins of samples. Bimodal grain-size distribution; coarse fraction is moderately to poorly sorted.


Common to few

Polycrystalline quartz/quartzite: eq and el, sa–r. Size = <3.4 mm; mode = 0.3 mm. Generally inequigranular, but can contain layers of equigranular interlocking quartz. Undulose extinction, sometimes with zigzag sutures. Some grains have a distinct alignment and banding. Can contain biotite and white mica laths, usually aligned with the quartz fabric; rare grains have an almost cataclastic texture (sample 98/6). These inclusions grade into the quartz biotite schist.

Common

Monocrystalline quartz: eq, sa–r. Size = <0.35 mm; mode = 0.15 mm. Usually undulose extinction. Can have associated biotite and white mica.

Few to very rare

Phyllite: el, sa–sr. Size = <1.26 mm; mode = 0.7 mm. Two main types: 1. Biotite and white mica-rich, with very sporadic el quartz crystals. Orange-

brown in PPL (x100) and yellow-orange in XP (x100). Strong homogeneous micaceous fabric with undulose extinction.


2. More quartz-rich, representing a finer version of the quartz and biotite schist. Has a strongly inequigranular, aligned fabric, often rich in el quartz, with biotite and less commonly white mica.

Very few

Quartz mica schist: el, sa–sr. Size = <3.1 mm; mode = 0.6 mm. Distinct banding of moderately to strongly aligned, interlocking, usually inequigranular polycrystalline quartz with biotite and white mica and opaques. Clearly related to the polycrystalline quartz/quartzite described above, which contains rare micas and also grades into the phyllite.

Rare to very rare

Poorly metamorphosed siltstone: eq, r. Size = <1.75 mm. Composed of eq and el quartz, biotite mica, and opaques, randomly arranged or roughly aligned.

TFs: eq–el, r–sr. Size = <0.75 mm; mode = 0.27 mm.

Very rare



Frequent


Common

White mica

Few


Few to very few

Biotite micaOpaques

Rare

MicritePhyllite

Very rare

AndaluciteTourmaline

MATRIX70%–76%. Orange-brown in PPL (x40) and orange to red-brown in XP (x40). Homogeneous; optically active; noncalcareous.

TFs rare to very rare, eq and el, r–sr. Size = <0.75 mm; mode = 0.27 mm. Red-brown to opaque black in PPL (x100) and dark red-brown in XP (x100). Consisting of dark homogeneous clay containing poorly sorted, randomly oriented, usually sa quartz, polycrystalline quartz, and rarely biotite and white mica and spherical opaques. Many contain voids and are cracked. May have an incomplete


ring void around margin. Boundaries are usually clear but can be merging; high optical density; discordant to concordant.

VOIDS7%–10%. Few to rare voids, consisting of few to rare macrovughs, few mesovughs, common microvughs, rare to very rare mesochannels, and very rare to absent macrochannels. Voids are single- to open-spaced. In both samples, the long axes of voids are roughly aligned with the margins of sections. Voids very rarely contain secondary calcite precipitation.

FABRIC G: P HYLLI TE, SILTSTONE, AND CHERT

Kommos 98/5

INCLUSIONS20%. Eq and el, sa–r, open- to single-spaced, moderately aligned parallel to the margins of the sample. Bimodal grain-size distribution; coarse fraction moderately well sorted.


Dominant

Phyllite: el, sa–sr. Size = <2.5 mm; mode = 1.3 mm. Gray in PPL (x100) and gray to yellow-gray in XP (x100). Consisting of sparse (<0.12 mm) quartz grains with moderately to strongly aligned biotite and white mica. All grains have margins of varying thickness, dark gray to black in PPL (x100) and orange-red in XP (x100). Some grains have areas with recrystallized fine polycrystalline quartz, some have opaques, many have voids where large quartz grains have been removed.

Few

TFs: eq, r. Size = <0.3 mm; mode = 0.1 mm.

Very few

Chert: eq, r–sr. Size = <1.5 mm; mode = 0.6 mm. Irregular muddy appear-ance; possibly grading into cherty mudstone. Can be partially metamorphosed. Can contain replaced siliceous planktonic and benthic foraminifera, and possible chalcedony radiolaria. Some rare unaltered chert grains without microfossils.

Siltstone: eq and el, sr. Size = 1.7–0.75 mm. Orange-red in PPL and XP (x100). Consisting of often recrystallized quartz and biotite mica in a red clay matrix. Some alignment of micas. May be partially metamorphosed.

Rare

Polycrystalline quartz: eq and el, sa. Size = <0.5 mm; mode = 0.4 mm. Straight or undulose extinction, sometimes with a consertal texture. May have small altered biotite crystals (metamorphic origin).

Fine-grained volcanic rock fragment: eq, r. Size = 0.8 mm. Remnant crystal structure set in a red-brown devitrified glassy matrix.

Mudstone: eq, r. Size = 0.7 mm. Contains faint polygonal cracking, has a dark margin.



Frequent

Monocrystalline quartzBiotite mica

Very few

TFs

Rare

MicriteCalciteOpaques

Very rare


Few to rare

Opaques

MATRIX73%. Red-brown and green-brown to green in PPL (x40), and red green-brown, gray-brown, and orange-brown in XP (x40). Generally homogeneous; optically inactive to slightly active; noncalcareous.

TFs few, eq, r. Size = <0.3 mm; mode = 0.1 mm. Red-brown to opaque black in both PPL and XP (x100), clay-rich. Boundaries usually diffuse; high optical density; discordant.

VOIDS7%. Few voids, consisting of rare megavughs, rare macrovughs, and very few macrochannels and meso- and microvughs. Double- to open-spaced; generally oriented perpendicular to the margins of the sample.

FABRIC H: P HYLLI TE

Kommos 98/8

INCLUSIONS30%. El, sr–sa, single-spaced or less, moderately to crudely aligned to the long axis of the sample. Weakly bimodal grain-size distribution; coarse fraction poorly sorted.


Frequent

Phyllite: el, sr–sa. Size = <3.8 mm; mode = 1.0 mm. Green-gray and brown in PPL (x100) and gray to gray-brown in XP (x100). Composed of sparse, aligned fine quartz grains (<0.06 mm) in a fabric of strongly aligned fine white mica with


rare amphibole. All grains show a strong metamorphic fabric and several exhibit folding. Many have gray-brown clay and layers rich and poor in opaques. Micas go into extinction together. Some grains have dark margins. There are rare very fine-grained metamorphic inclusions (cf. slate) and one coarser inclusion (cf. schist) with r strained quartz grains (<0.16 mm), but these grade into the phyllites.

Few

Monocrystalline quartz: eq, sa–r. Size = <1.3 mm; mode = 0.15 mm. Usually undulose extinction.

Rare

Mudstone: eq, sa. Size = 1.75 mm. Dark red-brown. Contains fine quartz and micas, with very sporadic silt-sized quartz crystals in a clay matrix.

TFs: eq, wr. Size = <0.6 mm; mode = 0.15 mm.

Very rare

Micrite: eq, sr. Size = 0.12 mm.Polycrystalline quartz: eq, r. Size = <0.75 mm; mode = 0.13 mm. Can grade

into cataclasite. One large grain contains an opaque alteration product. Chert: el, sr. Size = 0.13 mm.


Frequent


Common

Phyllite

Few

Biotite micaPolycrystalline quartz

Very few

Opaques

Rare

Micrite

MATRIX63%. Green gray-brown (core) to red-brown (margin) in PPL (x40) and dark gray-brown (core) to dark red and orange-brown (margin) in XP (x40). Generally homogeneous; moderately optically active, but can be more active in some areas; noncalcareous.

TFs rare, eq, wr. Size = <0.6 mm; mode = 0.15 mm. Clear boundaries, often with a thin, ring-shaped void around the periphery, which can be partly infilled with calcite. High optical density; discordant to moderately concordant with groundmass. Composed of brown-black (PPL x100) and dark brown-black (XP x100), opaque-rich clay, containing unsorted, unoriented (<0.24 mm) quartz and rare chert and biotite crystals. TFs appear to have a different thermal expansion from that of the main clay body and have therefore broken away.


VOIDS7%. Few voids, consisting of few macrovughs, few macrochannels, rare mesochan-nels, common mesovughs. Single- to open-spaced; moderately to crudely aligned to the margins of the sample. Some secondary calcareous lining of voids.

FABRIC I: Q UARTZ, POLYCRYSTALLINE Q UARTZ, SCH IST, AND MICROFOSSILS

Kommos 98/12

INCLUSIONS5%. Eq and el, sa–r, single-spaced, no preferred orientation. Bimodal grain-size distribution; coarse fraction moderately sorted.


Common

Polycrystalline quartz and mica schist fragments: eq and el, sa–r. Size = <2.15 mm; mode = 1.25 mm. Suite of fragments originating from mica schist rock, ranging from large grains of two or more large (<1.5 mm) interlocking quartz crystals with undulose extinction and grains of inequigranular cataclastic quartz (all with or without very sparse isolated fine white mica and less commonly biotite, which may be aligned) to distinctly banded grains with aligned quartz and mica or more chaotic inequigranular, almost cataclastic quartz/mica schist fragments with complex interlocking sutures. These fragments are all likely to have the same origin, coming from a highly deformed low-grade metamorphic rock. The finer-grained quartzite fragments below are likely to be related to these inclusions, as are the monocrystalline quartz grains in this sample.

Few

Monocrystalline quartz: eq and el, sa–r. Size = <1.0 mm; mode = 0.65 mm. Undulose extinction.

Quartzite: eq and el, sr–r. Size = <1.75 mm; mode = 0.6 mm. Composed of fine (<0.14 mm, but usually <0.03 mm), eq, interlocking quartz crystals, usually equigranular, but can be slightly inequigranular in places. Can rarely be weathered yellow-brown in PPL (x100) and may contain rare opaques. One very fine-grained inclusion is chertlike, but is part of the range of inclusions related to the coarser-grained inequigranular polycrystalline quartz inclusions in this section.

Micrite: eq, sr–r. Size = <1.75 mm; mode = 0.6 mm. Composed of very fine yellow-gray micritic calcite containing very sparse fine (<0.06 mm) quartz and feld-spar, opaques, microfossils (foraminifera), and in some inclusions, non-birefringent siliceous microfossil remains (diatoms).

Phyllite: el, sr–r. Size = <1.05 mm; mode = 0.9 mm. Color varies from yellow orange-brown to red-brown in PPL (x100) and orange-brown to dark red-brown in XP (x100). Fine-grained, aligned, mica-rich, with sparse (<0.2 mm) quartz grains occurring as small lenses concordant with the mica fabric or, less commonly, more distinct bands. Most inclusions contain opaques.

TFs: eq and el, sr–wr. Size = <1.35 mm; mode = 0.2 mm. Including large opaque bodies (see below).


Very rare

Perovskite: el, r. Size = 0.45 mm. Brown fractured mineral with strong absorption color, high birefringence, complex twinning, anomalous brown-gold interference colors. Occurs with a small crystal of quartz.

Macrofossil: el, sa. Size = 0.75 mm. A longitudinal section of a tubular organic structure. El body composed of four layers of fine micrite separated by two thin, dis-continuous gaps and a large central channel, which contain oxidized organic matter.


Dominant


Few

Micrite

Very few

Microfossils (foraminifera)Biotite micaOpaquesPolycrystalline and mica schist fragmentsQuartziteTFs

Rare

PhylliteWhite mica

Very rare

HornblendeAugiteChert

MATRIX65%. Light gray orange-brown to reddish orange-brown in PPL (x40) and red gray-brown to dark red-brown in XP (x40). Homogeneous; optically inactive; noncalcareous.

TFs few, eq and el, sr–wr. Size = <1.35 mm; mode = 0.2 mm. Color varies from orange and red-brown to opaque black-brown in PPL (x100) and dark orange-brown to opaque red-black in XP (x100). Consisting of fine, amorphous, dark orange and red clay with opaques, containing a variable quantity of fine to very fine (<0.12 mm), eq and el, sr–r monocrystalline quartz grains and rare micas with little or no preferred orientation. Some bodies have large (<0.5 mm) polycrystalline quartz or quartz/mica schist fragments within or associated with them. Boundaries are clear to sharp; high to very high optical density; usually concordant to slightly discordant with the groundmass. Included here are opaque red-black inclusions that are clearly related to the less opaque red and orange-brown TFs, both of which contain quartz, polycrystalline quartz, and schist inclusions. One related TF, dark red-brown in PPL and XP (x100), contains well-packed el, r, medium-grained (<0.45 mm) phyllite fragments, randomly oriented in a dark red amorphous clay.


VOIDS30%. Very few voids, consisting of very rare macrovughs, rare el mesovoids, and common mesovughs and mesovesicles. Single- to double-spaced, with no preferred orientation.

FABRIC J : SCH IST

Kommos 98/10

INCLUSIONS18%. Eq and el, sr, single- to double-spaced, crude orientation to the long axis of the sample. Strongly bimodal grain-size distribution; coarse fraction moderately well sorted.


Common

Quartz, biotite, hornblende schist: eq and el, sr–r. Size = <3.2 mm; mode = 0.55 mm. Composed of quartz, biotite, hornblende, and opaques in varying amounts. Eq quartz grains are <0.1 mm but usually finer and have an interlocking arrange-ment. Grains can contain distinct quartz and biotite or hornblende banding or have these minerals more or less interdispersed. Many grains exhibit a strong alignment of the biotite and hornblende crystals in a metamorphic fabric, but in other grains these minerals are more or less randomly arranged; in one very large (3.2 mm) example, the biotite crystals are ordered in several radiating acicular fronds. Some grains are composed almost entirely of interlocking fine eq quartz crystals with rare biotite or hornblende. Despite the large variation in the proportions of the main components of these metamorphic rock fragments, they are likely to be related to one another, as well as to the finer-grained phyllite inclusions that also occur in this section (see below).

Few

Phyllite: el, sr. Size = 1.5–0.35 mm. Consisting of strongly aligned fine biotite mica laths with less common thin areas of fine-grained, eq and el, interlocking quartz crystals. In one grain the micas are aligned in a crenulation cleavage. Related to the coarser-grained quartz, biotite, and hornblende schist fragments described above, and the two grade into each other.

Monocrystalline quartz: eq, sa–sr. Size = <0.45 mm; mode = 0.15 mm. Un-dulose extinction.

Very few

Polycrystalline quartz: eq and el, sa–sr. Size = <0.40 mm; mode = 0.2 mm. Variable grain size, undulose extinction.

Rare

Micrite: eq and el, r. Size = <0.85 mm; mode = 0.25 mm. Composed of loosely packed fine micrite with opaques and very rare fine quartz, hornblende, and biotite inclusions (<0.08 mm).

Cataclasite: eq, sa–sr. Size = <0.5 mm; mode = 0.55 mm. Polycrystalline quartz rock fragment with large variation in the size of the individual interlocking quartz


crystals. The undulose extinction and arrangement of bands of strained quartz indicate that it was formed under high stress. Related to the metamorphic rock frag-ments, polycrystalline quartz, and strained monocrystalline quartz described above.

Very rare

Biotite: el, sr. Size = 0.22 mm.Opaques: el, sr. Size = 0.40–0.25 mm.

Very rare to absent

Mudstone: el, r. Size = 0.28 mm. Consists of dark red-brown clay with rare fine-grained (<0.025 mm) quartz and biotite.


Frequent


Few

MicritePolycrystalline quartz

Very few

HornblendeTFsOpaques

Very rare

Microfossils (foraminifera)

MATRIX76%. Yellow gray-brown to orange green-brown in PPL (x40) and dark yellow-brown to dark orange-brown in XP (x40). Homogeneous; moderately optically active to highly active; noncalcareous.

TFs very few. Two main types:1. Very few, eq and el, r. Size = <0.44 mm; mode = 0.08 mm. Red and orange-

brown in PPL (x100) and dark red-brown in XP (x100). Boundaries are clear; high optical density; usually concordant with groundmass. Composed of red clay with rare very fine biotite, quartz, calcite grains, and opaques.

2. Very rare, el, sa. Size = 0.6–0.55 mm. Color red-brown in PPL (x100) to dark red-brown in XP (x100). Boundaries are clear to merging; high optical density; concordant with groundmass. Composed of red clay with fine biotite laths and rare fine eq quartz grains. Contains streaks of clay with el tails.

VOIDS6%. Very few voids, consisting of very rare megaplanar voids, rare macrovughs, rare macrochannel voids, few mesochannel voids and el mesovughs, very few mesovesicles, and few microvesicles. Open-spaced; crudely oriented to the long axis of the sample. Thin micritic lining in some of the larger voids.


FABRIC 1: Q UARTZ AND CALCI TE

Kommos 98/33–37, 39–45, 47–49, 52, 66, 76

INCLUSIONS25%. Eq and el, sa–r, single-spaced or less, moderately to strongly (e.g., sample 98/52) aligned with the margins of the sections. Moderately to strongly bimodal grain-size distribution; coarse fraction moderately to poorly sorted.


Dominant to frequent

Calcite: eq and el, sa–r. Size = <3.4 mm; mode = 0.3–0.4 mm. Usually fine homogeneous micrite, but can be fine and less commonly coarser sparite associated with the micrite (e.g., sample 98/76). Often a weathered gray with color differentia-tion between core and margin in PPL (opaque-rich margin in sample 98/47). Can be fragmentary and easily confused with secondary calcite-lined voids (e.g., samples 98/36, 42). Some grains exhibit a remnant fossiliferous structure (e.g., sample 98/76). Can contain opaque bodies and rare quartz grains. Some variation in the size of the calcite grains, from finer (e.g., sample 98/44) to coarser (e.g., sample 98/48).

Common to few

Monocrystalline quartz: eq and el, sr–r. Size = <0.68 mm; mode = 0.24– 0.40 mm. Predominantly undulose extinction.

Rare

Polycrystalline quartz: eq and el, sr–r. Size = <0.96 mm; mode = 0.32 mm. Consisting of two or more eq interlocking quartz crystals with undulose extinction.

Few to very rare

TFs: see below.

Rare to absent

Opaques: fragmentary with associated voids. Remains of vascular plant matter in sample 98/44.

Very rare to absent

Microcline: el, sr. Size = <0.5 mm; mode = 0.3 mm. Can be altered.Biotite: eq and el, sr. Size = <0.35 mm; mode = 0.3 mm.



Calcite

Frequent to common


Rare to absent

Plagioclase feldspar


HornblendeMicrofossils (foraminifera) (sample 98/66)Microcline


Biotite mica (e.g., sample 98/44)Opaques

MATRIX55%–68%. Dark gray and slightly green-brown to red orange-brown in PPL (x40) and green, red, and gray-brown to red orange-brown at the margins of the section in XP (x40) (e.g., samples 98/33, 76), but can be yellow gray-brown where enriched with micrite (e.g., sample 98/49). Relatively homogeneous, but some sections exhibit heterogeneity in the abundance of fine calcite (e.g., sample 98/49), which indicates that there may have been an admixture of a fine, dark brown, noncalcareous clay (as in sample 98/48) with a quantity of marl. Optically inactive in all sections; calcareous.

TFs few to very rare. At least three types:1. Most common: red-brown to gray green-brown in PPL (x100) and dark

gray green-brown to almost black in XP (x100). Usually eq, but can be el, with tails (e.g., samples 98/39, 49), r. Size = <2.25 mm; mode = 0.25 mm (e.g., sample 98/76). Boundaries sharp (can have voids around the edges) to very diffuse (e.g., sample 98/43); high to very high optical density; concordant to discordant with groundmass. Consisting of sparse very fine quartz and biotite mica and opaques in a dark clay matrix, usually noncalcareous but can be slightly calcareous.

2. Less common: gray-green in PPL (x100) and brown-green in XP (x100). Clear to merging boundaries; neutral optical density; concordant with groundmass. Consisting of moderately aligned quartz, biotite, and opaques in a dark, slightly calcareous clay matrix with fine opaques (e.g., sample 98/49).

3. Less common: red-brown in PPL (x100) and dark red-brown in XP (x100). Eq, r. Size = <1.25 mm. Consisting of very dark fine clay with randomly oriented fine and medium r and sr quartz crystals, plus white mica and calcite (e.g., sample 98/66).

VOIDS7%–20%. Rare (e.g., sample 98/40) to common (e.g., sample 98/48) voids, but usually few, consisting mainly of common or few el meso- and macrovughs and channels; megavughs may also be present. Usually single-spaced or less, strongly aligned with the margins of the samples. Voids in some sections contain secondary calcite deposition (e.g., sample 98/39), although this may be difficult to distinguish from poorly preserved, decayed micrite inclusions (e.g., sample 98/36).

SUBFABRIC 1A: Q UARTZ AND CALCI TE WI TH CHERT

Kommos 98/51, 64, 73, 84

INCLUSIONS20%. Eq and el, wr–sr, single-spaced, not aligned to margins of samples. Unimodal grain-size distribution (1.92–0.01 mm); well sorted (sample 98/73) to moderately well sorted (sample 98/64).


Dominant

Quartz: eq, wr–sr. Size = <0.4 mm; mode = 0.2 mm. Undulose extinction.


Calcite: eq–el, wr–r. Size <1.04 mm; mode = 0.24 mm. Fine micritic calcite; less commonly fine sparry calcite or monocrystalline calcite. Can have relic fossil structure (cf. coralline algae). May contain ferruginous matter (sample 98/64). In samples 98/51 and 98/84 inclusions are degraded.

Very few to absent

TFs: eq–el, wr. Size = <1.92 mm; mode = 0.4 mm. Two types:1. Gray- and black-brown to red- and orange-brown. Sharp to merging

boundaries; high optical density; concordant to discordant. Ferruginous clay-rich inclusions with rare very fine quartz grains. Can contain fine calcite, perhaps of secondary origin (sample 98/84). Related to fine ferruginous inclusions. Clay pellets.

2. Green orange- and yellow-brown. Clear to diffuse boundaries; neutral opti-cal density; discordant. Fine, clay-rich, composed of several discrete parts, including highly optically active aligned mudstone particle with rare chert and calcite inclu-sions. Surrounded by less structured clay-rich material containing calcite, quartz, and ferruginous inclusions. Clay pellet containing argillaceous rock fragment.

Few to rare

Microfossils: eq and el, wr. Size <0.4 mm; mode = 0.2 mm. Planktonic and benthic foraminifera. Can be infilled with calcite (sample 98/84).

Few to absent

Chert, eq and el, wr–sr. Size = <0.32 mm; mode = 0.16 mm. Composed of microcrystalline quartz.

Very few to rare

Ferruginous inclusions, eq and el, wr–r. Size = <0.24 mm; mode = 0.05 mm. Opaque inclusions related to TFs.

Very rare to absent

Hornblende: eq, wr. Size = 0.12 mm.Epidote: eq, wr. Size = 0.08 mm.Plagioclase feldspar: eq and el, r. Size = 0.24–0.16 mm.

MATRIX65%–78%. Orange-brown to gray- and green-brown in PPL (x40). Orange- and yellow-brown to dark gray-brown in XP (x40). Generally homogeneous, except for color differences between core and margin in samples 98/51 and 98/84 and uneven ferruginous staining in sample 98/73. Moderately optically active (sample 98/73) to optically active (sample 98/84); slightly (sample 98/84) to highly (sample 98/64) calcareous.

VOIDS2%–15%. Mainly el meso- and macrovoids and macrovughs. Strongly aligned with margins of samples 98/51 and 98/84. Voids in samples 98/51 and 98/84 contain secondary calcite and can be associated with degraded calcite inclusions.


FABRIC 2: MACROFOSSILIFEROUS CLAY PELLET

Kommos 98/56, 59, 68, 71, 72, 74, 75, 78, 80, 81, 86 (related: 98/61)

INCLUSIONS10%–25%. Eq and el, r–wr, usually single-spaced or less, but can be double-spaced in places (e.g., sample 98/81), moderately (e.g., sample 98/81) to poorly (e.g., sample 98/71) aligned to the margins of the samples. Moderately bimodal grain-size distribution; coarse fraction is moderately well sorted.


Frequent to common

Macrofossils: eq and el, r–sr. Size = <3.3 mm; mode = 0.5 mm. Composed of micrite and rarely a single crystal of calcite (e.g., sample 98/78). Dominated by eq and el, wr coral bodies and fragments (e.g., sample 98/75). Less common are mollusk shell fragments (e.g., sample 98/78) and unidentifiable inclusions, which are almost certainly of biological origin.

Very common to rare

TF: eq and el, r. Size = <1.75 mm; mode = 0.8 mm.

Common to few

Microfossils: eq and el, wr. Size = <1.1 mm; mode = 0.2 mm. Small globular planktonic foraminifera, small biserial benthic foraminifera and larger complex benthic foraminifera. Composed of micrite or a single crystal of calcite. Can have associated opaques (oxidized organic matter), and the chambers of specimens can be infilled by sparry calcite. The specimens appear to be in situ in the clay of this fabric.

Rare to absent

Calcite: eq and el, a–r. Size = <0.9 mm; mode = 0.2 mm. Monocrystalline, but can be micritic with quartz inclusions (e.g., sample 98/81). In sample 98/59 (reduction-fired) there appears to be a large number of micritic inclusions; these are macrofossil inclusions that have lost most of their internal structure. Have very rare sparry limestone fragments in sample 98/86, consisting of eq, well-formed, interlocking monocrystalline calcite crystals.

Very few to rare

Monocrystalline quartz: eq and el, sr. Size = <1.05 mm; mode = 0.2 mm. Undulose extinction.

Chert: eq and el, sr–r. Size = <0.75 mm; mode = 0.3 mm. Dark gray in PPL (x100) and yellow-brown in XP (x100). Very fine, but can be coarser in places (e.g., sample 98/68). Can contain opaques.

Rare to very rare

Opaques: eq and el, sa–r. Size = <0.65 mm; mode = 0.125 mm. Some may be altered inclusions (e.g., chert).




Calcite

Common to few


Very few to few

Opaques

Very few to rare


Few to rare

TFs

Rare

Macrofossil debrisChert

Very rare

Plagioclase feldsparBiotite micaAltered fibrous mineral (cf. serpentine)

MATRIX60%–85%. Light gray-brown, yellow-brown, green orange-brown, and green-brown in PPL (x40) and red- and orange-brown, dark green-brown, yellow-brown, and yellow green-brown in XP (x40). Some color differentiation between core and margin in some sections (e.g., sample 98/78). Relatively homogeneous, although sample 98/71 exhibits some heterogeneity in the form of isolated areas of merg-ing TFs (clay mixing). Usually moderately optically active, but can be more (e.g., sample 98/81) or less active (e.g., sample 98/59).

TFs common to rare, eq and el, r. Size = <1.75 mm; mode = 0.8 mm. Orange red-brown in PPL (x100) and dark red-brown in XP (x100), but can be green gray-brown in PPL (x100) to dark green-brown in XP (x100) where reduction fired (e.g., sample 98/59), and dark red-brown to almost black where rich in opaque matter (e.g., sample 98/81). Clear boundaries, but can be diffuse in places; high optical density; usually concordant with the groundmass. Can be cracked and have associated ring-shaped voids. Composed of a clay-rich, noncalcareous matrix containing sparse, fine (size = <0.3 mm; mode = 0.04 mm) monocrystalline quartz, with rare hornblende, white and biotite mica, and chert. Sample 98/59 contains an interesting TF that consists of dark, bedded marl (clay and fine calcite), and could therefore be an argillaceous rock fragment (Arf ), but has slightly diffuse boundaries and appears to be rather plastic. These TFs appear to represent unmixed red-firing clay pellets and are well paralleled by mixes of red clays with marls elsewhere in the eastern Mediterranean.

Also present are dark red-brown TFs with microfossils and sparry calcite in-clusions (sample 98/72). Eq, r. Size = <1.1 mm; mode = 0.3 mm. Red gray-brown and gray green-brown to gray-brown in PPL (x100) and dark red-brown to black-brown in XP (x100). Boundaries vary from sharp to diffuse; high optical density; most are discordant with the groundmass. Composed of dark, noncalcareous clay with very fine quartz inclusions and some fine opaques.


VOIDS5%–15%. Few to very few voids, consisting of very rare to absent megavughs, few to rare macro- and mesovughs, and frequent to few macro- and mesochannels (often very complex). Single-spaced or less; crudely (sample 98/71) to moderately well aligned to the margins of the sections. Voids in sample 98/59 exhibit calcareous lining, although this may be a product of the decomposition of micritic (macrofos-sil) inclusions rather than actual secondary calcite precipitation.

FABRIC 3: Q UARTZ AND CLAY P ELLET

Kommos 98/60, 63, 67, 85 (related: 98/82)

INCLUSIONS25%. Eq and el, r–sr, single-spaced, less commonly closed or double-spaced, with poor to moderate alignment of inclusions to the margins of the samples (strongest in sample 98/60). Trimodal grain-size distribution.

Very Coarse Fraction2.69–0.4 mm.

Predominant

TFs: eq, r. Size = <1.9 mm; mode = 0.64 mm.

Dominant to common

Micrite: eq and el, sr–r. Size = <2.69 mm. Homogeneous very fine calcite, often containing sparse very fine sa–sr monocrystalline quartz grains and planktonic foraminifera, and less commonly containing benthic foraminifera and ostracods (e.g., sample 98/67). Can contain some sparry calcite and voids. One inclusion is ring-shaped with a central void and may represent the micritic infilling of a macrofossil structure.

Very rare to absent

Macrofossil shell fragments: el, sa. Size = <0.7 mm; mode = 0.6 mm. Straight fragments of a macrofossil shell with a multilayered structure, preserved as a single crystal of calcite (e.g., sample 98/63).

Very rare

Opaques/organic remains: eq, r. Size = <0.8 mm; mode = 0.45 mm. Very dark red and orange-brown to opaque black in PPL (x100) and very dark brown to opaque black in XP (x100). Roughly spherical bodies composed of several (up to five) concentric layers separated by narrow ring voids. Usually contain sparse fine silt-sized quartz grains incorporated in the structure. May be the remains of some sort of plant (e.g., sample 98/67).

Very rare to absent

Microfossils (foraminifera): eq, r. Size = 0.45 mm.




Monocrystalline quartz: eq and el, sr–r. Size = <0.32 mm; mode = 0.25– 0.2 mm. Usually undulose extinction.

Common to few

Foraminifera: eq, r–wr. Size = <0.4 mm; mode = 0.2 mm. Planktonic and benthic foraminifera. Chambers commonly infilled with micrite, a single crystal of calcite, or rarely chert (sample 98/67). Can sometimes be difficult to distinguish from other calcite inclusions of similar size (e.g., samples 98/63, 67). Chambers not infilled in sample 98/60.

Few

TFs: eq and el, r. Size = <0.4 mm.

Few to very few

Micrite: eq and el, sr–r. Size = <0.4 mm; mode = 0.2 mm. Inclusions of micrite that cannot be identified as infilled foraminifera. Composed of fine microcrystalline calcite, but can contain sparite and have sparse fine quartz grains, foraminiferal remains, chert, and areas of opaque staining.


Chert: eq and el, sr–r. Size = <0.35 mm; mode = 0.225 mm. Can be spheri-cal bodies that resemble radiolaria (e.g., sample 98/67), amorphous chert bodies (e.g., sample 98/60), or amorphous bodies with sparse micrite and remains of foraminifera, formed by the replacement of micrite inclusions (e.g., sample 98/67).

Rare to very rare

Polycrystalline quartz: eq and el, sr. Size = <0.3 mm; mode = 0.25 mm. Rough- ly equigranular (0.05–0.15 mm crystals). Can be interlocking; usually undulose extinction.


Very rare

Microfossils (ostracods): el, sr; mode = 0.35 mm. Ostracod valve with associated micrite, suggesting that it originated from the large micrite inclusions described above (sample 98/60).


Frequent


Common to few

MicriteMicrofossils (foraminifera)TFs

Few to very few

Opaques


Rare to very rare

ChertCalcite

Very rare

Plagioclase feldsparBiotite micaMicroclineHornblende

MATRIX72%. Gray green-brown, gray-brown, orange green-brown, and gray yellow-brown in PPL (x40) and dark red-brown, dark gray orange-brown, yellow, and orange-brown in XP (x40). Homogeneous; optically inactive to optically very slightly active; calcareous.

TFs frequent, eq and el, wr. Size = <1.9 mm; mode = 0.25 mm. Gray orange green-brown to red orange-brown in PPL (x100) and dark orange green-brown to dark red-brown and almost black in XP (x100). Boundaries sharp (often separated by ring-shaped voids) to slightly merging; high optical density; usually discordant to less commonly concordant with the groundmass. Composed of dark noncalcareous clay containing opaques with usually abundant eq and el, sr–sa monocrystalline quartz, phyllite (e.g., sample 98/63), chert-infilled radiolaria (e.g., sample 98/60), amphibole, and biotite mica laths.

VOIDS3%. Few to very few voids, consisting of common to few macrovughs, few mesovughs, very few to absent macrochannels, very few to rare mesochannels, and very few macro- and microvesicles. Double- to open-spaced, but can be single-spaced in places; slight alignment of the long axes of voids to the edge of the section in samples 98/63 and 98/67 and moderate alignment in sample 98/60. Some calcare-ous lining of voids in sample 98/67.

FABRIC 4: QUARTZ, CHERT, AND MACROFOSSILS

Kommos 98/70, 79, 83 (related 98/61)

INCLUSIONS15%–30%. Eq and el, r, single- to double-spaced, moderately well aligned to the margins of the samples. Bimodal grain-size distribution; coarse fraction poorly sorted.



Macrofossils: eq and el, r–wr. Size = <1.2 mm; mode = 0.32 mm. Mostly coralline bodies composed of micritic calcite with remnant structure visible in PPL (and also in XP in the case of the oxidized samples 98/70 and 98/79). Can have dark margins. Less abundant are el mollusk shell fragments composed of sparry calcite and less commonly monocrystalline calcite, with remnant layered structure visible in PPL.


Common to few

Monocrystalline quartz: eq and el, sr–wr. Size = <0.5 mm; mode = 0.16– 0.32 mm. Undulose extinction.

Few to very few

Microfossils (foraminifera): eq and el, wr. Size = <0.7 mm; mode = 0.2 mm. Dominantly globular planktonic foraminifera. Small planktonics are often as-sociated with red and orange-brown opaques infilling chambers and surrounding tests. Some large opaque TFs contain several planktonic foraminifera (e.g., samples 98/70, 79).

Very few to rare

TFs: eq and el, r–wr. Size = <4.0 mm; mode = 0.3 mm.

Rare to very rare

Chert: eq and el, r. Size = <0.75 mm; mode = 0.35 mm. Fine-grained, but can more rarely contain coarser quartz crystals (<0.025 mm) (e.g., sample 98/70) and calcite inclusions. Usually unaltered, but can rarely have associated opaques. Usually have polygonal fracture in PPL.

Very rare

Monocrystalline calcite: eq and el, sr. Size = <0.36 mm; mode = 0.16 mm.Polycrystalline quartz: eq and el, sa–sr. Size = <0.46 mm; mode = 0.3 mm.

Undulose extinction.Opaques: eq and el, r–sr. Size = <0.85 mm; mode = 0.25 mm. Red-brown to

black in PPL (x100). May have remnant structure of decayed associated calcare-ous matter. May be carbonized organic-rich inclusions or altered rock inclusions. Some similar features have less distinct boundaries and are described below as TFs.

Very rare to absent

Micrite: el, r. Size = 2.88 mm. There are many micritic inclusions in these sections, but most of them exhibit at least a faint remnant structure in PPL or XP, which indicates that they have a biological origin; these are described above. Only very few micritic inclusions have no such structures. Most of these may be of the same origin as the coralline bodies, to judge from their similar shape, composition, and modal size. One large micritic inclusion in sample 98/70 may have a different origin. It is composed of fine micrite grains (<0.01 mm) and contains very rare foraminifera, possible shell fragments, fine eq quartz grains, and opaques. Sample 98/79 contains a large, compositionally similar micritic feature, but this has less distinct boundaries and is described as a TF below.

Microcline: el, sr. Size = 0.4 mm.


Common


Common to few



Few

OpaquesMonocrystalline calcite

Few to rare

Micrite

Very rare

ChertMacrofossils

MATRIX73%–80%. Orange and red orange-brown to red gray-brown to gray and gray green-brown in PPL (x40) and yellow-brown, light yellow green-brown, orange-brown, red-brown, and red gray-brown to dark green-brown and dark red-brown in XP (x40). Relatively homogeneous, except that samples 98/70 and 98/79 have prominent dark streaks in the otherwise light-colored groundmass; optically very active to slightly active; calcareous.

TFs of two main types:1. Calcareous TFs (sample 98/79), eq and el, r. Size = 4.0–0.8 mm. The larger

example is composed of fine microcrystalline calcite with rare fine quartz and monocrystalline calcite inclusions (<0.08 mm), planktonic and benthic foramin-ifera, and opaques. Color is orange and yellow-brown in PPL (x100) and yellow green-brown in XP (x100). Diffuse to merging boundaries; low optical density; discordant with the groundmass. The smaller example is composed of fine granu-lar calcite with some larger calcareous crystals, opaques, and possible microfossil remains. Color is gray orange-brown in PPL (x100) and yellow orange-brown in XP (x100). Clear to merging boundaries; low optical density; concordant with the groundmass and has tails.

2. Red-brown opaque features with microfossil and calcite remains (samples 98/70, 79, 83), eq and el, r–wr. Size = <1.15 mm; mode = 0.3 mm. Yellow orange-brown through red-brown to black-brown in PPL (x100) and orange-brown to dark black-brown in XP (x100). Boundaries vary from sharp to merging between and within individual TFs; high optical density; mostly concordant, but can be slightly discordant with groundmass. Composed of fine, red-brown-stained clay that varies from translucent to opaque and contains fine quartz and calcite (<0.05 mm) and foraminifera. These TFs range from very indistinct bodies to clay pellets; some very dark opaque bodies have been described as opaque inclusions above. The dark red-brown and opaque color is likely to be a product of the oxidation of organic matter; this appears to be associated with foraminiferal tests, which may form the nucleus of some bodies. One TF in sample 98/79 contains possible woody vascular material.

VOIDS5%–7%. Few to very few voids, consisting of very few to absent megavughs, very few to rare macrovughs, rare mesovughs, frequent to common macrochannels, and few to common mesochannels. Generally single-, less commonly double-spaced (sample 98/79); strongly aligned to the margins of the section.


FABRICS 2 AND 4 RELATED SAMPLE

Sample 98/61 illustrates the overlap between fabrics 2 and 4, displaying most of their characteristic inclusions in its coarse fraction. It has a single large eq, r micrite inclusion; a single large el, r coralline macrofossil; several types of eq, r clay pellets; numerous medium-sized eq and el, r–sr, fine-grained chert inclusions; medium to fine planktonic and benthic foraminifera with associated organic matter; and medium-sized eq and el, r–sr monocrystalline quartz grains with undulose extinc-tion. It has a less abundant coarse fraction in its bimodal grain-size distribution than those of fabrics 2 and 4, described above. It is clearly related to both of these fabrics and cannot be assigned to the variation within either.

FABRIC 5: CHERT

Kommos 98/58, 65 (related: 98/62)

INCLUSIONS20%–27%. Eq and el, sr–wr, single- to double-spaced, moderately (sample 98/58) to poorly (sample 98/65) aligned to the margins of the section. Moderately to strongly bimodal grain-size distribution; coarse fraction moderately well sorted (sample 98/58) to less well sorted (sample 98/65).


Frequent

Chert: eq and el, r–wr. Size = <0.75 mm; mode = 0.25 mm. Gray to red-brown in PPL (x100). Fine, roughly equigranular with some quartz grains (>0.4 mm). Can contain circular equigranular and chalcedonic radiolaria, isolated examples of which are also present as separate inclusions (see below). Can be moderately to heavily altered to red-brown opaques as small spherical bodies, in fractures or amorphous and ubiquitous. In altered grains radiolaria are usually less altered than the chert matrix and therefore stand out in PPL. Some very badly altered grains are almost unrecognizable as chert. Most grains are polygonally fractured.

Common

Monocrystalline quartz: eq and el, sr–r. Size = <0.35 mm; mode = 0.2 mm. Straight or slightly undulose extinction.

Few to very few

Radiolaria: eq, wr. Size = <0.3 mm; mode = 0.125 mm. Infilled with chalcedony quartz and less commonly homogeneous very fine quartz. Can contain opaques as bodies or amorphous; these may appear as the outer lattice shell of the radiolaria, which may also be visible as a ghost in PPL. These inclusions are related to the chert in both samples.

Very few

Micrite: eq and el, a–r. Size = <1.5 mm; mode = 0.25 mm. Generally fine-grained, but can contain some coarser sparry regions with crystals <0.02 mm. In some areas may be lightly stained orange in PPL (x100). One grain in sample 98/58 has a remnant coralline structure. May be related to the larger, less distinct areas of micritic material described below.


Very few to rare

TFs: eq and el, r. Size = <2.4 mm; mode = 1.2 mm.

Rare to very rare

Monocrystalline calcite: eq, sa–r. Size = <0.36 mm; mode = 0.20 mm. Grains composed of eq and el (<0.2 mm) monocrystalline calcite crystals. Some are infilled spherical fossil structures (possibly foraminifera), one of which has small isotropic lath-shaped minerals (sample 98/65).

Foraminifera: eq, r. Size = <0.34 mm; mode = 0.15 mm. Mostly globular plank-tonic foraminifera, but some complex benthic foraminifera in sample 98/58. Cham-bers may be infilled. Specimens can be associated with opaque matter. Can occur in calcareous TFs (see below); this may be the origin of the isolated tests in the matrix.

Augite: eq and el, sa–sr. Size = <0.3 mm; mode = 0.1 mm.

Very rare to absent

Spinel: eq, r. Size = <0.16 mm. Dark red-brown in PPL and very dark red-brown in XP (x100). Roughly spherical body composed of several concentric layers of opaque material and a central body.

Sandstone: el, r. Size = 0.85 mm. Coarse-grained (<0.14 mm) quartzwacke. Grains are composed of eq and el, sa–sr monocrystalline quartz with undulose extinction set in very fine-grained chert cement with red-brown opaques. Probably related to the chert grains described above.

Hornblende: el, sr. Size = 0.75 mm.Clino- and orthopyroxene intergrowth: eq, a. Size = 0.18 mm. Lamellar

structure (sample 98/58).


Frequent

Monocrystalline calcite

Common to frequent


Common to few

Chert

Rare

Microfossils (radiolaria)

Rare to very rare

Microfossils (foraminifera)OpaquesAugite

MATRIX67%–78%. Yellow and green gray-brown to red green-brown and dark gray green-brown in PPL (x40) and yellow and orange green-brown to dark orange and red green-brown in XP (x40). Moderately homogeneous (sample 98/65) to heteroge-neous (sample 98/58), with ptygmatic texture in sample 98/58 from the incomplete mixing of more and less calcareous material; moderately optically active (sample 98/65) to optically active (sample 98/58).


TFs very few to rare, eq and el, r. Size = <2.4 mm; mode = 1.2 mm. Yellow-brown to gray and gray green-brown in PPL (x100), mainly yellow gray-brown in XP (x100). Composed of very fine micritic clay with rare fine monocrystalline calcite grains. Contain common foraminifera (mainly planktonics and some with opaques) and some very fine opaques. These are areas of unmixed calcareous mate-rial that forms the ptygmatic texture together with darker, less calcareous material in sample 98/58. It appears to be the source of the foraminifera and many of the calcareous inclusions in the matrix. Boundaries diffuse to merging; low optical density; concordant to moderately discordant.

VOIDS2%–6%. Rare to very few voids, consisting of very rare to absent megavughs, few macrovughs, few mesovughs, common to few macrovughs and mesochannels, and few to very few microvesicles. Single- to open-spaced; strongly (sample 98/58) to poorly (sample 98/65) aligned to the margins of the section. Some calcareous deposition in some voids.

FABRIC 6: F USILINID

Kommos 98/50, 54

INCLUSIONS17%. Eq and el, a–sr, single- to double-spaced, moderately (sample 98/50) to strongly (sample 98/54) aligned with margins of the samples. Bimodal grain-size distribution; coarse fraction poorly sorted.


Predominant

Microfossils (foraminifera), eq and el, a–sr. Size = <2.0 mm; mode = 0.7 mm. Fragments of large fusilinid foraminifera with multilayered curved (0.1–0.15 mm thick), concentric shells, consisting of a single crystal of calcite with a faint laminated structure perpendicular to the circumference of the layers. The spaces between successive layers are often infilled with sparry calcite and some fragments have associated micrite.

Very few to rare

Micrite: eq and el, sr. Size = <1.5 mm; mode = 0.5 mm. Cloudy in PPL; can be cracked and exhibit mosaic extinction, indicating that the micrite may have replaced a sparry calcite structure. Commonly have yellow, orange, or brown weathered margins. The rare association of calcareous fusilinid shell fragments and micrite in both samples indicates that the latter may represent the matrix in which these microfossils were enclosed.

Rare to very rare

Monocrystalline quartz: eq and el, r. Size = <0.4 mm; mode = 0.25 mm. Usu-ally undulose extinction. In sample 98/54 one grain is associated with several fine crystals of biotite.

TFs: eq and el, r–wr. Size = <1.2 mm; mode = 0.75 mm.


Very rare to absent

Biotite: el, sa. Size = 0.35 mm (sample 98/50). Hornblende: eq and el, r. Size = 0.2 mm.Opaques: eq and el, sr–r. Size = <0.25 mm; mode = 0.2 mm.


Dominant

Calcite

Frequent


Common

Biotite

Few

Opaques

Rare to absent

Plagioclase feldspar

Rare to very rare

Hornblende

Very rare to absent


MATRIX73%. Gray green-brown, orange-brown, red-brown, and dark red-brown in PPL (x40) and very dark red-brown and dark red green-brown in XP (x40). Homoge-neous; optically inactive; noncalcareous.

TFs rare, eq and el, r–wr. Size = <1.2 mm; mode = 0.75 mm. Red orange-brown, red-brown, and dark red gray-brown in PPL (x100) and red-brown and dark red orange-brown in XP (x100). Clay pellets, streaks, and swirls composed of a fine, very dark, usually red, noncalcareous clay with sparse (<0.04 mm) bio-tite mica laths and (<0.14 mm) monocrystalline quartz and opaques. Often have cracked appearance. Boundaries are clear, very diffuse, and often merging; high to neutral optical density; usually concordant but can be slightly discordant with the groundmass.

Sample 98/50 may contain a textural depletion feature. It is a distinct, eq region (0.25 mm) of fine clay, gray-yellow in PPL (x100) and gray-brown in XP (x100), with clear boundaries and some fine opaques and biotite grains, but otherwise very few inclusions. It appears to have low (negative) optical density and is highly translucent.

VOIDS10%. Few voids, consisting of rare to absent megavughs, common macrovughs, few mesovughs, common macrochannel to macroplanar voids, very few mesochannel


voids, and rare mesovesicles. Usually single- or less than single-spaced; strongly aligned with the margins of the sections (especially in sample 98/54). Some sec-ondary calcareous lining of the voids.

FABRIC 7: SERP ENT INE AND MICRI TE

Kommos 98/55, 69

INCLUSIONS15%–22%. Eq and el, sr–r, single- to double-spaced, strongly (sample 98/55) to moderately (sample 98/69) aligned to the margins of the samples. Moderately (sample 98/69) to strongly (sample 98/55) bimodal grain-size distribution; coarse fraction moderately (sample 98/69) to poorly (sample 98/55) sorted.


Frequent to common

Serpentine: eq and el, sr–r. Size = <1.2 mm; mode = 0.5 mm. Light gray-yellow through amber to red-orange in PPL (x100) and dark yellow-orange, red-orange, and red-brown to black-brown and dark gray-green in XP (x100). Most grains have some sort of mesh structure in XP, most prominent in the lighter-colored inclusions. This structure is less visible in PPL, although some inclusions have a remnant crystal structure. Can have small opaque inclusions. Most have a cracked structure visible in PPL. Many inclusions have associated voids around their mar-gins, which may contain a thin secondary deposit of micrite; this is best seen in XP, where it contrasts with the inclusions themselves.

Common to few

Micrite: eq and el, sa–r. Size = <1.01 mm; mode = 0.45 mm. Composed of very fine micritic calcite, but can be coarser (<0.0075 mm). Many inclusions contain opaques. Often light gray-brown in PPL (x100). Can be fragmented and very faint in places, perhaps due to weathering.

Few

Mudstone: eq and el, r–wr. Size = <1.35 mm; mode = 0.35 mm. Two main types: Most common is a dark, slightly silty mudstone. Light gray and slightly yellow gray-brown to dark gray-brown and dark gray in PPL (x100) and very dark gray-brown, very dark gray and almost black in XP (x100). Has polygonally cracked structure in PPL. Composed of very fine clay, often with rare, sparsely distributed fine (<0.04 mm) quartz inclusions. Less common is a reddish-brown mudstone, which is more abundant in sample 98/69. It differs from those described above in that it is orange and red-brown in PPL and XP (x100) and has no polygonal cracking. Fine quartz grains (<0.05 mm) occur sparsely within these clay-rich inclusions and they may be related to the red siltstone that occurs in sample 98/69 (see below).

Few to very rare

Microfossils (foraminifera): eq and el, wr. Size = <0.4 mm; mode = 0.25 mm. Planktonic foraminifera. Chambers are commonly infilled with sparry calcite or a single crystal of calcite in sample 98/55. In sample 98/69 the chambers may or may not be infilled by opaques or contain small spherical opaque bodies.


Very few to rare

Chert: eq and el, sa–r. Size = <0.65 mm; mode = 0.45 mm. Fine- or medium-grained (<0.04 mm), with occasional chalcedony radiolaria. Some grains are slightly weathered and appear light in PPL (x100).

Rare

Macrofossils: el, sa–r. Size = <0.9 mm; mode = 0.55 mm. Composed of fine micrite or rarely sparry calcite (sample 98/55). Inclusions have remnant layered shell structure preserved to a greater or lesser degree.

Biotite and pyroxene: eq and el, sa–r. Size = <1.1 mm; mode = 0.65 mm. Biotite crystals occurring alongside and intergrown with colorless, low first-order birefringent pyroxene. There are grains composed almost entirely of pyroxene, with very small areas of pleochroic biotite, and vice versa, as well as isolated biotites and pyroxenes, some with twinning and zoning (cf. augite).

Rare to absent

Siltstone: el, r. Size = 0.5–0.8 mm. Red and orange-brown in PPL (x100) and dark red-brown in XP (x100). Composed of eq and el, sa–sr monocrystalline quartz grains (<0.04 mm; coarse silt and below) and small opaque bodies in a dark, clay-rich matrix. The grains appear to be well sorted and exhibit some bedding in grain size and color of the matrix, which can be nearly opaque in places (sample 98/69).

Marl: eq and el, r–wr. Size = <0.85 mm; mode = 0.45 mm. Composed of a clay-rich, fine micritic matrix that varies in color from light yellow-gray to green gray-brown in PPL (x100) and light to dark green-brown and yellow green-brown in XP (x100). This matrix contains foraminifera and other fos-siliferous remains, together with rare fine quartz and some fine opaque bodies. There can be differences in the color of the matrix within individual inclusions (e.g., sample 98/55).

Very rare

Hornblende: el, sa. Size = 0.65–0.45 mm.Altered igneous rock fragments: eq and el, r. Size = 0.45–0.4 mm. Composed

of el plagioclase laths (<0.12 mm) set in an orange red-brown matrix of altered biotite and opaques.

Opaques: el, sa–sr. Size = 0.35 mm. Dark black or slightly red black. Can be associated with voids. There are many dark features in these two sections, including altered serpentine and opaque stained clay matrix; these are described elsewhere. These inclusions represent distinctive opaque bodies, which cannot be identified as altered rock or mineral fragments.

TFs: eq and el, r.

Very rare to absent

Phyllite: el, r. Size = 0.3–0.45 mm. Light gray-brown in PPL (x100) and yellow gray-brown in XP (x100). Composed of very fine aligned white mica with rare fine opaques. Micas go extinct together parallel to the direction of alignment. These phyllite inclusions have a more or less cracked structure parallel to the direction of the mica alignment.

Monocrystalline calcite: eq, r. Size = 0.3 mm (sample 98/55).Sanidine: el, sr. Size = 0.65 mm. Simple twinning. Opaque alteration products

in large irregular fractures. Cataclasite: el, sr. Size = 0.55 mm. Composed of large (<0.24 mm) quartz

crystals with undulose extinction set in a fine quartz matrix.



Frequent to common

Monocrystalline quartzMicrite

Common to rare

Biotite mica

Few to rare


Few to absent


Very few to rare

Opaques

Rare to very rare

Serpentine

Very rare

Microfossils (radiolaria)ChertAugitePlagioclase feldsparAltered igneous rock fragments

Very rare to absent

SiltstoneQuartz and biotite rock fragmentPhylliteEpidote

MATRIX74%–79%. Light and dark gray-brown to light gray-yellow, orange-green, and gray-brown in PPL (x40) and yellow and red dark green-brown to dark red and orange-brown in XP (x40). Homogeneous; optically inactive.

TFs very rare, eq and el, r. Gray-green to orange green-brown in PPL (x100) and dark red green-brown in XP (x100). Consisting of poorly packed clay with evenly distributed micrite, sparse eq, r (<0.06 mm) monocrystalline quartz, and fine opaques. Diffuse boundaries; high optical density; concordant to discordant with the groundmass.

VOIDS15%–22%. Very few to few voids, consisting of few to absent megavughs, few to very few macrovughs, very few to few mesovughs, very rare to rare macrochannels, and few mesochannels. Usually single-spaced (more in places in sample 98/69); strongly (sample 98/55) to moderately (sample 98/69) aligned to the margins of the samples.


FABRIC 8: POLYCRYSTALLINE Q UARTZ AND CALCI TE

Kommos 98/77

INCLUSIONS7%. Eq and el, r, open-spaced, randomly arranged. Bimodal grain-size distribution; coarse fraction moderately well sorted.


Common

Polycrystalline quartz: eq and el, sr–r. Size = <0.6 mm; mode = 0.45 mm. Vari-able grain size, interlocking crystals with undulose extinction. May have cataclastic texture or small associated opaques and less commonly fine micas, indicating a metamorphic origin.

Calcite: eq and el, r. Size = <0.65 mm; mode = 0.25 mm. Composed entirely of micrite, micrite with rare, variably sized monocrystalline calcite crystals, in-terlocking eq grains of monocrystalline calcite, or a single large monocrystalline calcite crystal. The micritic calcite grains may be “muddy” or have a remnant fossil structure. One grain, composed of fine monocrystalline calcite crystals, contains rare fine monocrystalline quartz crystals, while another has large monocrystalline calcite crystals intergrown with fine eq monocrystalline quartz crystals.

Few

Monocrystalline quartz: eq and el, r. Size = <0.55 mm; mode = 0.2 mm. Undulose extinction.

Very few

Serpentine and opaques: eq and el, r. Size = <0.45 mm; mode = 0.175 mm. Color varies from light yellow-orange through dark orange and red-brown to opaque black in PPL (x100). These inclusions are altered serpentine and range from partially altered yellow-gray grains with distinctive mesh structure to totally altered opaque bodies with no discernable structure.

Chert: eq and el, r. Size = <0.45 mm; mode = 0.35 mm. Fine- to very fine-grained; can be yellow in PPL (x100).

Rare

Altered igneous rock fragments: eq and el, r. Size = 0.55–0.25 mm. Orange-brown in PPL (x100). Composed of el randomly oriented feldspar laths (0.3– 0.1 mm) in a devitrified groundmass. Altered fine-grained basic igneous rock. One grain has aggregates of yellow acicular crystals in the altered matrix.

TFs: eq–el, r. Size = <0.3 mm; mode = 0.15 mm.

Very rare

Metasedimentary rock fragment: el, r. Size = 0.65 mm. Composed of fine-grained, eq, monocrystalline quartz grains set in a matrix of fine biotite mica. Has a bedded structure and two vertical bands of polycrystalline quartz. Partially metamorphosed siltstone.

Phyllite: el, r. Size = 0.2 mm.Zoisite: eq, r. Size = 0.2 mm.


Augite: eq, r. Size = 0.4 mm.Hornblende: eq, r. Size = 0.25 mm.Microfossils (foraminifera): eq, r. Size = <0.2 mm; mode = 0.15 mm. Often

have associated oxidized organic matter.Acid igneous rock fragments: el, sr. Size = 0.25 mm. Composed of lath-shaped

alkali feldspars surrounded by interlocking quartz crystals and brown, heavily altered biotite. Is a fine-grained equigranular acid igneous rock (cf. microgranite).

Quartz and epidote rock fragment: el, r. Size = 0.55 mm. Highly altered igne-ous or metamorphic rock fragment.


Frequent


Common


Few

Microfossils (foraminifera)Opaques

Very few

White micaSmall, highly birefringent mineral grains (including hornblende, augite, and

epidote)

Rare

Polycrystalline quartzBiotiteChert

Very rare

Alkali feldsparQuartz and biotite rock fragmentSerpentinePhylliteTFs

MATRIX78%. Light yellow and orange gray-green brown to dark green-brown in PPL (x40) and yellow and orange-brown to red-green brown in XP (x40). Moderately homogeneous, except for a group of large, complex voids in the center of the sec-tion (possibly created during the production of the thin section). Optically slightly active to optically moderately active.

TFs rare. Two types:1. Rare, eq and el, r. Size = <0.3 mm; mode = 0.15 mm. Color varies from

orange and red-brown in PPL (x100) and dark orange and red-brown in XP (x100). Boundaries are sharp to diffuse; high optical density; concordant with groundmass. Composed of fine dark clay with opaques, micrite, and rare biotite and white mica.


2. Very rare, eq, r. Size = 0.25 mm. Gray yellow-brown in PPL (x100) and orange yellow-gray brown in XP (x100). Diffuse boundaries; neutral optical den-sity; concordant with groundmass. Composed of micritic clay with some opaques.

This sample contains many small red and orange-brown to opaque bodies that are not included in the TFs.

VOIDS15%. Few voids (including large voids produced during the making of the thin section), consisting of rare megavughs, rare megachannel voids, few macrovughs, rare macrochannel voids, few mesovughs, few mesovesicles, and rare microvesicles. Single- to open-spaced; generally randomly arranged. One mesovugh has a slightly darkened rim, perhaps produced by the destruction of an organic inclusion.

FABRIC 9: Q UARTZ AND METAMORP H IC

Kommos 98/53

INCLUSIONS23%. Eq and el, sr–r, single- to double-spaced, moderately aligned to the margins of the samples. Slightly bimodal grain-size distribution; coarse fraction poorly sorted.


Dominant

Monocrystalline quartz: eq and el, sr–r. Size = <0.30 mm; mode = 0.175 mm. Usually undulose extinction.

Few

Polycrystalline quartz metamorphic rock fragments: eq and el, sr–r. Size = <2.4 mm; mode = 0.15 mm. Range of polycrystalline quartz fragments, includ-ing grains composed of two or three large interlocking quartz crystals to more complex grains composed of many finer crystals of quartz with rare fine biotite and/or white mica. Less commonly can have cataclastic texture in places, align-ment of crystals and opaques. There is one very large grain of inequigranular polycrystalline quartz with almost cataclastic texture and yellow (in PPL x100) isotropic minerals replacing micas. These inclusions are polycrystalline quartz fragments of metamorphic origin and are likely to be related to the schists and phyllite inclusions described below.

Opaque altered inclusions: eq and el, sr–r. Size = <0.375 mm; mode = 0.15 mm. Conspicuous black opaque bodies, often cracked and fragmented and associated with voids. They are altered inclusions. The color varies from dark gray and red-brown to opaque black in PPL and XP (x100). While many inclusions are totally opaque, some contain rare fine eq monocrystalline quartz grains and very rare micas, and have a remnant structure, which suggests that they may have had a sedimentary origin.

Very few

Schist and phyllite: el, sr–r. Size = <0.9 mm; mode = 0.75 mm. Group of phyllite and fine-grained schist fragments that grade into each other. Ranging from very


fine-grained phyllite composed of strongly aligned white mica with opaques and rare, sparse, aligned, very fine-grained (<0.01 mm) eq and el monocrystalline quartz, through biotite and white mica-rich phyllites with more abundant coarser-grained (<0.05 mm) el quartz crystals, to aligned, quartz-rich, fine-grained (<0.1 mm) schist with less fine micaceous component. Color may be light gray-brown, orange-brown, or gray to black-brown in PPL (x100) and dark gray-brown, dark red-brown, and black opaque in XP (x100).

Rare

TFs: eq and el, sr–r. Size = <0.9 mm; mode = 0.32 mm.

Very rare

Alkali feldspar: el, sa. Size = 0.28 mm. Within a streak of dark clay.Chert: eq and el, sr–r. Size = 0.2–0.22 mm. Fine, equigranular.


Dominant


Common

Biotite mica

Few

Polycrystalline quartz metamorphic rock fragmentsWhite micaOpaques

Very rare

Phyllite

MATRIX73%. Green-gray orange and red-brown in PPL (x100) and orange-brown to red and red-gray brown in XP (x100). Heterogeneous, with conspicuous streaks and bodies of a clay groundmass that is gray-brown to gray-green brown in PPL (x100) and dark red-gray and black brown to black-brown in XP (x100). Optically inactive; noncalcareous.

TFs rare, eq and el, sr–r. Size = <0.9 mm; mode = 0.32 mm. Three main types:1. Rare micritic TFs. Composed of micrite-rich clay (marl), sometimes with

rare fine eq quartz grains and pure fine micrite with a varying abundance of opaques. Occur as discrete bodies with sharp boundaries or less discrete features with diffuse and merging boundaries, which are highly concordant with the groundmass. Many have associated voids and can be represented by a thin micritic rim enclosing a large void where the main body of the feature has been removed. Low optical density.

2. Very rare brown, clay-rich TFs. Color varies from orange and red-brown in PPL (x100) and dark orange and red-brown in XP (x100). Composed of amor-phous fine red clay, some with rare fine quartz and opaques. Boundaries are clear to diffuse; high optical density; concordant with groundmass.

3. Very rare gray quartz and feldspar TF. Light gray-brown in PPL (x100) and dark gray-brown in XP (x100). Composed of very fine (<0.06 mm) eq and


el quartz and alkali feldspar grains, randomly arranged in a dark clay matrix with opaques. Sharp boundaries; high optical density; concordant with groundmass. Is el and has a tail. This sample is characterized by the presence of many long, com-plex streaks of a dark clay with high optical density. Color varies from gray-brown to gray-green brown in PPL (x100) and dark red-gray and black-brown in XP (x100). This clay contains many of the inclusions that occur in the red, clay-rich groundmass that constitutes the main body of this section, including monocrystal-line quartz, polycrystalline quartz, metamorphic rock fragments, phyllite, feldspar, opaques, chert, and very fine micas. The inclusions are aligned with the long axes of these extensive TFs.

VOIDS4%. Very few voids, consisting of rare el macrovughs, common mesovughs, rare mesovesicles, and rare microvughs and microvesicles. Double- to open-spaced; mod-erately well aligned to margins of sample. Some larger voids are associated with micritic material, although this is a product of the decay of calcareous TFs and not of secondary calcite deposition. Some voids associated with opaque altered inclusions.

FABRIC 10: MICACEOUS QUARTZ AND FELDSPAR

Kommos 98/38

INCLUSIONS27%. Eq and el, sr, single-spaced, no preferred orientation. Bimodal grain-size distribution; coarse fraction moderately well sorted.


Common

Monocrystalline quartz: eq, r. Size = <0.6 mm; mode = 0.3 mm. Extinction usually slightly undulose.

Plagioclase feldspar: eq and el, sa–sr. Size = <1.84 mm; mode = 0.42 mm. Eq or lathlike, unaltered plagioclase with multiple twinning and often zoning.

Very few

Alkali feldspar: eq and el, sa–r. Size = <1.15 mm; mode = 0.32 mm. Unaltered, eq or lathlike, twinned or untwinned.

Opaque altered inclusions: eq and el, r. Size = <0.36 mm; mode = 0.2 mm.

Rare

Polycrystalline quartz: eq, sr. Size = 0.95 mm to 0.55 mm. Composed of large and small interlocking crystals with undulose extinction and very rare fine biotite and/or white mica. Metamorphic origin.

Chert: eq and el, sr–r. Size = <0.65 mm; mode = 0.25 mm. Generally fine-grained but can be coarser (<0.04 mm), equigranular to inequigranular. Can be colorless to yellow and orange-gray brown in PPL (weathered).

TFs: eq and el, r–wr. Size = <1.6 mm; mode = 0.45 mm.Altered igneous rock fragments: el, r. Size = <0.5 mm; mode = 0.45 mm.

Color varies from dark red black-brown to yellow and orange gray-brown in PPL


(x100) and dark red-brown in XP (x100). Composed of fine (<0.18 mm) el laths of plagioclase feldspar, randomly arranged in a dark, devitrified groundmass with fine opaques. An altered fine-grained basic igneous rock.

Low birefringent mineral (cf. kyanite): eq and el, sa–sr. Size = <0.52 mm; mode = 0.42 mm. Colorless to slightly brown in PPL. Two cleavages at 90° (one good, the other poor). High relief, low first-order birefringence white to light yellow, oblique extinction.

Very rare

Augite: eq, r. Size = 0.18 mm.Hornblende: el, sa. Size = 0.12–0.1 mm.Orthopyroxene: el, sr. Size = 0.65–0.5 mm.


Common


Few

Plagioclase feldsparWhite micaOpaques

Very few

Alkali feldspar

Rare

ChertTFs

Very rare

HornblendeAugite

MATRIX67%. Gray green-brown to orange-brown in PPL (x40) and dark brown to dark red and orange-brown in XP (x40). Homogeneous; optically inactive.

TFs rare, eq and el, r–wr. Size = <1.6 mm; mode = 0.45 mm. Orange and red-brown in PPL (x100) and dark orange and red-brown to dark-black brown in XP (x100). Composed of dark clay containing randomly oriented, fine-grained, eq and el, sa–r quartz, feldspar, mica, and chert inclusions, with abundant opaques. Clear to merging boundaries; high optical density; generally concordant with the groundmass, although the latter has no preferred orientation. These are clay pel-lets, which contain inclusions that are related to those in the fine fraction of the main body of this section.

VOIDS6%. Few voids, consisting of few macrovughs, common mesovughs, few mesovesicles, and very few microvughs and microvesicles. Single- to double-spaced; no preferred orientation.


FABRIC 11: Q UARTZ AND SCH IST

Kommos 98/46

INCLUSIONS28%. Eq and el, sa–sr, single- to double-spaced, no preferred orientation. Bimodal grain-size distribution; coarse fraction moderately well sorted.


Frequent

Monocrystalline quartz: eq and el, sa–r. Size = <0.55 mm; mode = 0.25 mm.

Common

Polycrystalline quartz and quartzite/cataclasite: eq and el, sa–sr. Size = <0.55 mm; mode = 0.35 mm. Ranging from grains composed of two or more large interlocking quartz crystals with undulose extinction, through inequigranular polycrystalline quartz grains with large eq quartz crystals in a very fine cataclastic quartz ma-trix, or grains with large quartz crystals grading into very fine quartz, to grains composed entirely of equigranular fine cataclastic quartz, which appears very chertlike but is clearly related to the others. These inclusions are strongly related to the quartz/mica schist fragments described below and some grains differ only in the absence of mica.

Quartz/mica schist fragments: eq and el, sa–sr. Size = <0.75 mm; mode = 0.35 mm. Range of metamorphic rock fragments containing predominantly quartz and micas. Can occur as polycrystalline quartz inclusion composed of two or more large interlocking quartz crystals with rare small white mica and/or biotite (which can be heavily altered), inequigranular and equigranular, almost cataclastic poly-crystalline quartz with varying amounts of randomly oriented mica laths and often opaques, and more ordered schist fragments composed of equigranular interlocking polycrystalline quartz interdispersed with aligned white mica or alternating bands of equigranular to inequigranular polycrystalline quartz and aligned biotite and white mica arranged in a medium to strong metamorphic fabric. These inclusions are related to the polycrystalline quartz and quartzite/cataclasite grains described above.

Few

Calcite: eq and el, r. Size = <0.95 mm; mode = 0.35 mm. Inclusions com-posed of very fine micrite, fine eq calcite crystals or larger (<0.1 mm) interlocking monocrystalline calcite crystals, or a combination of the two. May contain sparse eq and el monocrystalline quartz crystals (<0.08 mm), or less commonly biotite mica laths or zoisite. The micrite inclusions may be “muddy” and range in color from light yellow-gray to orange-brown in PPL (x100); they also contain opaques to a varying degree. This section also contains single crystals of monocrystalline calcite.

Very few

White mica: el, sa–sr. Size = 0.25 mm; mode = 0.2 mm.

Rare

Plagioclase feldspar: eq and el, r. Size = <0.55 mm; mode = 0.3 mm. Can be slightly weathered and very rarely zoned.


Phyllite: el and eq, sr–r. Size = <0.5 mm; mode = 0.3 mm. Composed of fine biotite or white mica, very fine quartz, and opaques aligned in a strong fabric. Quartz may be very sparse; it may occur as small el lenses, form more continuous bands between the mica, or be ubiquitously interdispersed. There are two main types of phyllite, one biotite-rich and another white mica-rich. One inclusion is composed half of biotite mica-rich phyllite and half of fine, roughly equigranular polycrystalline quartz. This suggests a link between the phyllite inclusions and the polycrystalline quartz/cataclasite and schist fragments in this section.

TFs: eq, r–wr. Size = <0.4 mm; mode = 0.15 mm.

Very rare

Microcline: eq and el, sr. Size = 0.35–0.25 mm.Biotite mica: el, sa–sr. Size = <0.2 mm; mode = 0.15 mm.Feldspar and biotite rock fragments: eq, r. Size = <0.45 mm; mode = 0.2 mm.

Composed of plagioclase and alkali feldspars with biotite mica laths. Polycrystalline quartz and calcite rock fragment: eq, sr. Size = 0.45 mm. Com-

posed of monocrystalline calcite crystals and fine brown rhombic mineral grains contained within a large interlocking polycrystalline quartz grain.

Altered igneous rock fragment: el, sr. Size = 0.45 mm. Consisting of fine (<0.08 mm) el low to medium birefringent crystals (cf. sillimanite) with white mica and biotite, randomly oriented in an altered matrix containing quartz and a high-relief mineral (cf. zoisite).

Igneous rock fragment: el, sr. Size = 0.40 mm. Composed of alkali feldspar with granophyric texture together with quartz and biotite. A fine-grained acid igneous rock (cf. microgranite).

Epidote: eq, sa. Size = 0.24 mm.Quartz, feldspar, and zoisite rock fragment: eq, sr. Size = 0.40 mm. Composed

of roughly equigranular quartz, alkali feldspar, and micrite crystals, with some as-sociated high-relief blue and yellow-blue crystals (cf. zoisite).

Quartz and plagioclase feldspar rock fragment: el, sr. Size = 0.2 mm.Opaques: eq and el, sa–r. Size = <0.25 mm; mode = 0.15 mm. Heavily altered

rock fragments.


Frequent


Common

White mica

Few

Biotite micaCalciteOpaques

Very few

SchistCataclasite/polycrystalline quartz


Rare

HornblendePhyllite

Very rare

MicroclineZoisitePlagioclase feldspar

MATRIX69%. Gray green-brown to red and orange green-brown in PPL (x40) and yellow gray-brown to dark red-brown in XP (x40). Homogeneous; optically inactive; noncalcareous.

TFs rare, eq, r–wr. Size = <0.4 mm; mode = 0.15 mm. Color varies from gray green-brown to orange and gray red-brown in PPL (x100) and dark red-brown to dark black-brown in XP (x100). Composed of dark fine clay with sparse very fine (<0.02 mm) quartz and mica. Can have varying numbers of opaques. Boundaries clear to diffuse; neutral to very high optical density; concordant to discordant with the groundmass.

VOIDS3%. Few voids, consisting of very rare macroplanar voids, very rare macrovughs, common mesovughs and mesovesicles, and few microvesicles. Generally open-spaced; no preferred orientation.

FABRIC 12: ALKALI FELDSPAR

Kommos 98/57

INCLUSIONS18%. Eq and el, sa–sr. Bimodal grain-size distribution; coarse fraction poorly sorted.


Frequent

Feldspar: eq and el, a–r. Size = <1.75 mm; mode = 0.32 mm. Simple twinned to untwinned, usually unaltered. Can have associated opaques.

Very few

TFs: eq and el, wr–r. Size = <1.5 mm; mode = 0.35 mm.

Few

Igneous rock fragments, two types: 1. eq and el, sr–wr. Size = <2.4 mm; mode = 0.25 mm. Fine-grained igneous rock composed of phenocrysts of large-zoned plagioclase or twinned and untwinned alkali feldspar, in a glassy matrix containing fine el, randomly oriented plagioclase and alkali feldspar laths, quartz, altered ferromagnesian minerals, and opaques. Some smaller grains do not contain phenocrysts, and the fine feldspar laths may have a weak preferred orientation.


One grain contains a phenocryst of biotite. This is a volcanic rock of intermediate composition (cf. andesite or trachyte). 2. eq and el, sa. Size = <0.45 mm; mode = 0.3 mm. Composed of well-formed, lath-shaped, twinned and untwinned alkali feldspar crystals (<0.15 mm) enclosed by one or more large interlocking quartz crystals with even extinction. Usually have associated opaque altered minerals; one inclusion has small epidote crystals. Feldspars may be aligned or randomly oriented. Volcanic rock of acid/intermediate composition.

Rare

Phyllite: el, sa. Size = <0.65 mm; mode = 0.4 mm. Composed of strongly aligned biotite mica with some white mica and fine eq quartz crystals (<0.02 mm) or sparse, well-aligned, very long white mica crystals surrounded by very fine quartz.

Quartzite, cataclasite, quartz, and mica schist: el, sa. Size = <0.95 mm; mode = 0.45 mm. Ranging from grains composed of coarse (<0.35 mm) interlocking quartz crystals with rare very fine, sparsely distributed biotite and white mica, through more inequigranular, cataclastic grains containing fine white mica and altered biotite, to grains with well-aligned, equigranular, interlocking quartz grains with sparse micas oriented with the metamorphic fabric.

Micrite: eq and el, sr. Size = <0.34 mm; mode = 0.16 mm.Opaques: eq and el, sr–r. Size = <2.05 mm; mode = 0.3 mm.

Very rare

Perthite: eq, sa. Size = 0.35 mm.Biotite mica: el, sa. Size = 0.45 mm.Microfossil (sponge spicule): el, a. Size = 0.18 mm.Monocrystalline calcite: eq, r. Size = 0.15 mm.Amphibolite: el, sa. Size = 0.2 mm. Composed of four crystals (0.16–0.1 mm)

of light yellow pleochroic mineral with medium relief. Two good cleavages at 120°. Oblique extinction and bright yellow anomalous interference colors (cf. altered hornblende).


Dominant

Alkali feldspar

Few

White micaOpaques

Very few

Monocrystalline quartzCalcite

Rare

Biotite micaPolycrystalline quartzIntermediate igneous rock fragments


Very rare

Plagioclase feldsparSponge spiculesPhyllite

MATRIX79%. Yellow- and orange-brown to gray green-brown in PPL (x40) and orange-brown to slightly red green-brown in XP (x40). Homogeneous; optically slightly active to moderately active.

TFs very few, eq and el, wr–r. Size = <1.5 mm; mode = 0.35 mm. Orange and red-brown to dark brown in PPL (x100) and dark orange- and red-brown to dark gray-brown in XP (x100). The most common type is a dark red and brown noncalcareous clay containing many inclusions that also occur in the main body of the section: these include large untwinned alkali feldspar crystals, finer lath-shaped twinned and untwinned alkali feldspar crystals, quartz and alkali feldspar intermediate rock fragments (as described above), biotite mica, monocrystalline quartz, opaques, el white micas, and epidotes, all randomly arranged and ranging from poorly to well packed. Boundaries vary from sharp to merging; high optical density; concordant to discordant with the groundmass.

There are also some less abundant, less distinct TFs in this section, including several eq, r features with merging boundaries and low to neutral optical density. Composed of a fine clay, not dissimilar to the main body of the sample, but with fewer inclusions. Could therefore be classified as a textual depletion feature.

This sample also contains one large, highly calcareous, micritic TF, el, r. Size = 1.5 mm. Contains abundant fine micas and opaques, as well as a large polycrystal-line quartz grain and a phyllite fragment. Clear to diffuse boundaries; moderate optical density; somewhat discordant.

VOIDS3%. Very few voids, consisting of rare macrovughs, common mesovughs, few el mesovughs, and few mesovesicles. Crude alignment of inclusions and voids with the margins of the section.

Patrick S. Quinn

Universit y Col lege Londoninstitute of archaeolog y31–34 gordon squarelondon wc1h 0p yunited kingdom

patr ick.quinn@uc l .ac .uk

hesperia 80 (201 1)Online Supplement

APPENDIX 2

ELEMENTAL CONCENTRATIONS MEASURED BY INAA IN LATE MINOAN TRANSPORT JARS FROM KOMMOS

An appendix to “A World of Goods: Transport Jars and Commodity Ex- change at the Late Bronze Age Harbor of Kommos, Crete,” by P. M. Day, P. S. Quinn, J. B. Rutter, and V. Kilikoglou: Hesperia 80 (2011), pp. 511– 558; http://dx.doi.org/10.2972/hesperia.80.4.0511

© The Amer i c an Sc hoo l o f C l a s s i c a l S tud i e s a t Athens h t t p : / /dx .do i . o rg /10 .2972/he spe r i a . 80 .4 .0511 . app2

Sample Number Sm Lu U Yb As Sb Ca Na La Ce Th Cr Hf Cs Sc Rb Fe Ta Co Eu

Kommos 98/1 5.27 0.38 2.9 2.86 7.7 1.0 7.72 0.97 29.0 63.1 9.87 321.0 4.28 5.70 22.17 70 6.12 1.1 35.43 1.18

Kommos 98/2 5.61 0.35 4.6 2.76 12.3 1.3 7.32 1.59 30.9 59.1 9.70 313.0 4.04 7.84 21.45 110 5.88 1.0 34.62 1.13

Kommos 98/3 4.46 0.32 2.4 2.52 4.2 0.6 3.11 1.25 23.7 56.8 9.12 348.0 4.33 6.29 18.98 110 5.12 0.8 24.04 0.94

Kommos 98/4 4.33 0.36 2.7 2.47 5.4 0.7 7.19 0.86 24.8 53.5 9.46 485.0 4.38 8.75 21.94 110 6.18 0.8 41.80 0.98

Kommos 98/5 4.80 0.36 2.3 2.68 4.2 0.6 7.41 0.76 30.3 64.7 10.50 353.0 4.17 6.41 20.45 110 6.02 1.2 39.04 1.11

Kommos 98/6 9.09 0.53 4.7 4.11 32.8 8.0 5.00 0.51 54.5 120.7 21.40 200.0 7.14 12.10 21.59 160 7.14 1.8 20.28 1.81

Kommos 98/7 8.11 0.50 3.8 3.86 6.3 1.6 4.18 0.75 47.3 98.6 13.90 90.0 6.08 7.77 19.50 130 5.25 1.6 26.77 1.74

Kommos 98/8 9.57 0.60 4.3 4.76 13.2 1.6 1.00 0.52 53.0 114.9 15.20 115.0 7.37 9.39 19.38 150 5.28 1.8 33.79 1.96

Kommos 98/9 5.09 0.34 2.7 2.81 8.3 0.8 4.51 0.56 27.9 66.0 10.30 283.0 4.71 2.81 18.71 50 5.72 1.2 31.48 1.11

Kommos 98/10 5.14 0.34 2.7 2.77 10.5 0.7 7.69 0.81 28.3 60.1 9.33 386.0 3.97 2.72 21.17 60 5.89 1.1 34.02 1.26

Kommos 98/11 4.37 0.32 2.7 2.72 4.2 0.7 4.69 0.72 25.8 54.9 10.80 357.0 4.18 7.87 20.88 130 5.86 1.0 32.73 1.04

Kommos 98/12 4.25 0.29 3.2 2.31 10.5 1.3 4.77 0.22 33.0 54.5 9.62 112.0 5.79 3.21 12.97 40 3.51 1.2 7.09 0.89

Kommos 98/13 4.78 0.34 3.3 2.59 5.4 0.7 5.14 1.02 26.2 59.1 10.40 405.0 4.02 5.94 20.19 120 5.78 1.1 33.50 1.03

Kommos 98/14 4.75 0.40 2.8 2.98 6.3 0.5 7.08 0.55 27.8 62.2 11.10 395.0 4.00 1.33 25.27 60 7.42 1.2 41.73 1.13

Kommos 98/15 4.82 0.32 2.5 2.52 5.9 0.8 3.76 1.40 26.7 62.5 9.27 362.0 4.91 5.34 16.52 100 4.87 1.2 27.63 1.16

Kommos 98/16 4.77 0.34 2.1 2.65 5.8 0.6 4.65 0.95 22.3 56.0 8.17 427.0 3.91 3.12 25.43 70 6.66 1.4 44.49 1.17

Kommos 98/17 4.84 0.35 2.6 2.86 6.0 0.9 4.47 1.01 26.0 56.2 8.57 290.0 3.96 6.15 19.32 90 5.29 1.0 30.34 1.13

Kommos 98/18 5.49 0.36 2.7 2.88 4.9 0.6 5.28 0.72 31.3 68.7 10.70 358.0 4.52 6.21 19.52 110 5.53 1.3 33.39 1.20

Kommos 98/19 5.11 0.35 2.7 2.71 14.6 0.8 5.19 0.85 27.7 61.5 9.86 508.0 4.80 2.04 20.47 50 5.84 1.2 31.10 1.15

Kommos 98/20 5.51 0.39 2.9 3.03 9.3 0.6 3.17 1.19 30.0 68.9 11.30 377.0 4.92 1.92 18.12 60 5.28 1.3 30.65 1.16

Kommos 98/21 4.31 0.28 1.8 2.19 7.3 0.5 6.85 0.62 22.6 54.3 9.04 1166.0 1.45 4.90 19.92 80 6.48 0.9 49.91 1.02

Kommos 98/22 5.16 0.38 2.9 2.85 7.9 0.6 7.67 0.88 28.1 62.5 10.20 360.0 4.19 5.31 21.79 60 6.16 1.1 35.14 1.22

Kommos 98/23 5.79 0.37 2.8 2.91 11.4 0.6 4.51 0.74 31.8 70.0 12.00 419.0 5.07 2.88 19.94 60 5.77 1.1 35.58 1.35

Kommos 98/24 5.08 0.37 2.6 2.78 12.4 0.9 7.08 0.64 28.4 62.6 10.70 408.0 4.07 6.33 21.84 90 6.05 1.0 38.64 1.19

Kommos 98/25 4.19 0.36 2.3 2.58 9.1 0.5 3.87 0.67 22.9 55.4 10.00 792.0 3.87 1.17 24.21 40 7.10 1.1 51.68 0.96

Kommos 98/26 4.20 0.33 2.4 2.61 12.1 0.5 7.65 0.76 22.0 51.4 10.30 460.0 4.07 1.40 24.04 30 6.68 1.1 43.93 1.03

Kommos 98/27 4.64 0.35 2.4 2.73 5.7 0.6 8.11 0.73 27.3 63.6 11.20 406.0 3.89 3.04 22.71 30 6.54 1.1 42.09 1.10

Kommos 98/28 4.53 0.32 2.2 2.45 8.9 0.9 5.57 1.01 22.9 51.2 7.88 306.0 4.22 6.03 17.79 90 5.39 1.1 27.72 1.07

Kommos 98/29 5.00 0.35 2.7 2.79 5.0 0.8 8.40 1.16 26.3 56.3 8.20 293.0 3.99 5.83 23.32 80 6.21 1.2 31.53 1.28

Kommos 98/30 4.72 0.34 2.3 2.61 4.4 0.8 8.12 0.90 26.0 57.8 9.90 404.0 3.54 8.34 21.71 110 5.96 1.0 37.62 1.09

Kommos 98/31 9.73 0.44 4.4 3.51 13.1 1.2 1.14 0.70 65.5 107.0 16.90 110.0 8.39 7.26 18.04 130 5.30 1.9 14.25 1.82

Kommos 98/32 4.36 0.29 2.1 2.28 6.0 0.7 7.27 0.69 23.7 53.2 9.20 322.0 3.68 5.60 19.05 90 5.71 0.9 34.64 0.98

Kommos 98/33 7.48 0.41 4.0 3.33 5.2 0.4 13.08 0.96 38.1 82.7 7.98 101.0 7.22 1.78 16.84 40 4.59 2.0 23.29 1.64

Kommos 98/34 7.96 0.42 2.7 3.40 11.6 0.4 9.54 0.75 37.8 89.7 9.50 121.0 8.31 0.54 19.44 30 6.43 2.0 26.12 1.87

Kommos 98/35 7.65 0.39 2.9 3.18 7.7 0.4 12.81 0.39 37.9 86.0 8.31 101.0 6.50 1.26 16.77 30 5.42 2.0 21.64 1.71

Kommos 98/36 8.42 0.46 3.1 3.83 13.4 0.3 10.91 1.17 43.4 92.8 9.03 119.0 8.91 0.70 20.33 20 5.68 2.3 25.85 1.95

Kommos 98/37 8.06 0.43 3.0 3.27 3.9 0.3 11.94 0.41 41.1 93.5 9.06 111.0 7.18 1.29 17.41 30 5.57 1.8 23.21 1.76

Kommos 98/38 6.14 0.37 4.5 2.96 5.0 1.5 2.90 1.96 44.1 93.5 19.10 237.0 7.60 8.34 11.02 130 3.70 1.5 14.35 1.32

Kommos 98/39 7.09 0.39 3.0 3.06 7.4 0.3 11.90 0.55 33.6 78.2 7.43 102.0 7.75 0.60 16.52 20 5.48 1.7 23.11 1.67

Kommos 98/40 7.27 0.38 2.4 3.00 14.7 0.4 12.81 0.50 34.8 80.9 8.10 107.0 6.81 0.89 17.57 20 5.86 1.8 24.10 1.71

Kommos 98/41 6.89 0.34 2.3 2.87 3.0 0.3 11.02 0.55 34.2 79.3 7.29 111.0 6.81 1.87 17.59 50 5.60 1.8 24.98 1.74

Kommos 98/42 6.75 0.38 2.2 2.96 6.5 nd 9.79 0.66 32.1 73.5 7.03 119.0 7.72 1.52 17.95 50 5.78 1.5 26.28 1.70

Kommos 98/43 6.70 0.37 2.3 2.98 6.8 0.4 12.40 0.38 33.7 78.4 7.65 105.0 8.23 1.84 16.44 40 5.34 1.8 22.36 1.59

Kommos 98/44 6.74 0.34 2.3 2.80 5.6 0.4 8.73 0.78 33.4 76.2 7.42 96.5 7.48 1.39 16.12 40 5.05 1.7 21.06 1.60

Sample Number Sm Lu U Yb As Sb Ca Na La Ce Th Cr Hf Cs Sc Rb Fe Ta Co Eu

Kommos 98/45 7.10 0.41 3.0 3.09 3.5 0.3 8.05 1.02 34.6 77.7 6.93 115.0 7.23 nd 19.29 40 6.08 1.8 31.43 1.87

Kommos 98/46 3.91 0.27 1.8 2.12 7.9 0.7 3.63 1.30 21.0 45.0 6.91 435.0 3.53 6.41 16.90 70 4.37 0.8 30.21 0.99

Kommos 98/47 6.53 0.35 2.1 2.86 6.6 0.4 11.01 0.34 34.0 82.3 8.03 99.5 7.48 1.60 16.16 30 5.23 1.8 22.06 1.74

Kommos 98/48 6.82 0.34 2.8 2.80 6.6 0.5 11.79 1.12 35.2 76.8 7.14 117.0 5.75 1.59 14.50 40 3.84 1.7 15.43 1.53

Kommos 98/49 7.65 0.40 3.1 3.22 6.9 nd 10.54 1.12 39.8 87.0 7.99 105.0 5.90 nd 18.10 40 5.03 2.1 26.59 1.86

Kommos 98/50 8.17 0.41 3.0 3.41 nd 0.3 14.44 1.13 41.5 91.2 8.37 107.0 6.48 1.88 17.25 40 4.68 2.3 17.12 1.81

Kommos 98/51 7.30 0.40 4.0 3.07 6.6 1.5 10.01 0.81 41.1 82.6 10.80 128.0 4.70 4.31 12.31 70 3.48 1.2 9.66 1.50

Kommos 98/52 6.68 0.36 2.7 2.85 7.5 0.4 12.70 0.40 34.7 75.9 7.23 95.4 6.61 1.52 14.57 40 4.60 1.5 19.26 1.48

Kommos 98/53 7.98 0.41 3.3 3.24 5.6 0.4 13.79 1.27 41.0 89.8 8.32 107.0 6.39 2.04 17.50 40 4.76 2.0 16.40 1.73

Kommos 98/54 7.22 0.45 3.3 3.51 6.9 1.2 5.13 0.64 40.5 91.6 12.60 132.0 6.42 7.06 17.52 110 4.59 1.5 23.40 1.50

Kommos 98/55 7.44 0.45 2.5 3.79 9.2 0.6 13.20 0.40 42.1 82.6 9.97 536.0 5.12 3.48 16.41 60 4.60 1.3 32.39 1.66

Kommos 98/56 6.36 0.34 3.5 2.66 17.4 0.7 15.65 0.38 29.2 67.5 6.29 142.0 4.42 1.68 13.22 40 4.94 1.4 21.19 1.51

Kommos 98/57 8.80 0.54 4.6 4.65 14.3 0.9 1.95 1.69 69.3 141.6 27.40 156.0 9.96 8.75 15.42 140 4.88 2.6 20.10 1.15

Kommos 98/58 4.54 0.30 1.4 2.41 3.9 0.3 11.73 0.25 24.9 46.1 5.11 3223.0 2.40 1.10 10.25 10 2.67 0.6 23.58 0.99

Kommos 98/59 9.16 0.53 5.8 3.92 19.7 1.2 10.17 0.46 45.1 98.3 9.18 190.0 6.18 2.31 20.00 50 7.03 2.0 34.12 2.23

Kommos 98/60 6.08 0.36 2.1 3.00 7.8 0.4 8.40 0.35 30.1 76.6 9.35 109.0 9.04 0.97 12.06 30 4.11 1.6 19.24 1.32

Kommos 98/61 5.60 0.35 2.5 2.91 9.3 0.7 4.36 1.02 30.6 69.3 11.40 268.0 4.72 6.85 19.58 120 5.57 1.1 27.74 1.16

Kommos 98/62 5.95 0.30 3.3 2.44 9.3 0.6 10.70 0.60 29.7 65.7 7.81 206.0 4.02 2.31 15.51 40 4.53 1.3 20.39 1.34

Kommos 98/63 6.26 0.39 2.2 3.31 4.6 0.4 11.04 0.34 31.8 78.0 8.21 108.0 9.55 1.66 12.27 30 4.06 1.5 25.48 1.41

Kommos 98/64 3.94 0.27 1.5 2.10 8.7 0.4 6.14 0.31 18.4 51.5 6.94 79.5 7.81 1.23 10.12 30 3.39 1.3 12.17 0.87

Kommos 98/65 5.77 0.37 1.8 3.11 7.3 0.5 13.29 0.31 32.2 67.1 6.92 2991.0 3.31 1.32 14.23 30 3.68 0.9 31.39 1.39

Kommos 98/66 7.91 0.41 2.5 3.26 12.9 0.3 12.89 0.34 39.6 93.5 8.90 119.0 7.58 0.85 17.90 30 5.73 2.0 24.34 1.92

Kommos 98/67 7.93 0.46 2.0 3.81 6.3 0.5 10.60 0.56 41.6 94.0 8.90 110.0 9.92 1.95 13.70 40 4.47 1.5 26.50 1.79

Kommos 98/68 4.58 0.27 4.1 1.94 9.7 0.7 19.12 0.32 21.6 43.3 4.30 114.0 3.32 1.47 10.64 40 4.07 1.0 17.09 1.06

Kommos 98/69 6.70 0.35 3.3 2.55 6.2 0.6 7.36 0.51 34.2 78.0 8.61 647.0 4.80 2.50 17.32 60 5.65 1.8 29.57 1.57

Kommos 98/70 5.99 0.31 3.1 2.44 5.7 0.4 17.77 0.24 33.8 69.0 8.93 77.9 2.83 3.10 10.43 60 2.90 0.8 9.53 1.29

Kommos 98/71 5.58 0.32 3.6 2.35 14.2 0.6 18.35 0.30 26.5 55.5 5.04 125.0 4.03 nd 12.25 40 4.56 1.2 20.45 1.36

Kommos 98/72 6.61 0.36 4.2 2.64 22.9 1.1 16.08 0.42 31.8 67.6 6.09 152.0 4.82 0.69 15.96 20 5.91 1.7 24.01 1.67

Kommos 98/73 7.03 0.39 4.4 2.97 7.0 1.6 9.39 0.45 38.2 78.4 10.60 127.0 4.72 3.90 12.39 70 3.48 1.2 8.72 1.43

Kommos 98/74 5.17 0.29 4.1 2.17 11.2 0.6 19.58 0.36 24.2 50.7 4.56 112.0 3.49 1.38 10.95 30 4.14 1.0 18.29 1.24

Kommos 98/75 6.04 0.35 4.1 2.65 15.2 0.8 18.60 0.32 28.2 58.3 5.92 128.0 4.07 1.46 12.79 30 4.55 1.3 20.28 1.44

Kommos 98/76 6.95 0.36 2.8 2.82 6.4 0.4 14.95 0.39 35.3 80.9 7.73 102.0 5.93 1.61 14.93 40 4.83 1.7 21.18 1.53

Kommos 98/77 4.40 0.29 2.4 2.39 9.2 0.7 12.02 0.46 22.8 54.7 8.85 1025.0 3.12 4.96 15.70 80 4.53 0.9 27.89 0.98

Kommos 98/78 6.01 0.32 3.7 2.65 13.7 0.7 19.05 0.27 26.7 60.7 5.61 144.0 4.07 1.57 13.21 30 4.83 1.3 21.43 1.44

Kommos 98/79 5.89 0.29 3.0 2.51 4.7 0.5 18.13 0.22 31.6 66.0 8.92 79.4 2.56 3.20 10.21 60 3.10 0.7 10.52 1.19

Kommos 98/80 5.15 0.29 3.8 2.25 10.5 0.8 20.58 0.34 23.9 50.2 4.35 111.0 3.18 1.34 10.77 30 3.91 1.1 17.89 1.18

Kommos 98/81 4.86 0.28 2.8 2.15 7.7 nd 17.46 0.15 27.8 55.5 6.86 118.0 4.18 1.54 10.99 30 3.24 1.2 10.27 1.12

Kommos 98/82 3.83 0.26 1.6 2.09 6.3 0.2 19.77 0.28 19.7 41.3 4.70 64.6 5.80 0.88 7.77 20 2.54 0.8 11.22 0.96

Kommos 98/83 6.84 0.35 3.1 2.81 5.2 0.6 18.74 0.24 38.3 76.7 10.50 99.2 3.37 4.41 11.99 80 3.31 1.0 11.18 1.50

Kommos 98/84 7.06 0.38 3.6 2.93 8.7 1.1 11.32 0.58 39.2 78.8 10.80 116.0 4.84 3.56 11.96 60 3.48 1.1 8.90 1.55

Kommos 98/85 7.84 0.44 2.7 3.71 10.4 0.5 5.75 0.32 37.8 97.1 9.85 140.0 9.30 1.54 14.99 30 4.76 1.6 18.44 1.87

Kommos 98/86 5.06 0.29 4.4 2.08 12.1 0.8 18.87 0.39 23.9 49.5 4.72 119.0 3.28 1.39 11.33 40 4.21 1.0 18.43 1.29

Kommos 98/87 4.98 0.33 2.8 2.58 5.2 0.7 6.53 1.11 27.7 59.1 9.72 344.0 4.66 4.76 18.43 80 5.16 1.1 29.45 1.17

Kommos 98/88 5.21 0.33 2.7 2.70 7.4 0.6 3.23 1.24 28.2 58.9 10.00 302.0 4.70 2.85 17.32 60 4.97 1.0 28.72 1.20

Documents

A WORLD OF GOODS