Fig. 3. High-powered photomicrograph ofthe mononuclear infiltrate about a smallblood vessel in the cerebellum of an agam-maglobulinemic chicken; this is similar tolesions in control animals except for theabsence of plasma cells.

the development of EAE. In that theintensity of the disease was equivalentin the bursectomized and irradiatedanimals to that in both the controlirradiated and unirradiated chicks, itcould not ibe demonstrated that circu-lating antibody was protective. Com-plement-fixing antibodies to brain anti-gens were not measured. Our studyconfirms the findings of Jankovic thatthymectomy in the newly hatched chickinhibits the ability of the animal tomanifest EAE, and it adds support tothe contention that EAE is a manifes-tation of delayed hypersensitivity.

MICHAEL E. BLAWM. D. COOPERR. A. GOOD

Department of Pediatrics and Divisionof Neurology, University ofMinnesota Medical School,Minineapolis 55455

References

1. B. H. Waksman, Int. Arch. Allergy 14 suppl.1, 1 (1959).

2. P. Y. Paterson, J. Exp. Med. 111, 119 (1960);S. H. Stone, Science 134, 619 (1961).

3. B. H. Waksman, J. Infect. Dis. 99, 258(1956).

4. P. Y. Paterson, J. Immunol. 78, 472 (1957).5. A. Szenberg and N. Warner, Brit. Med. Bull.

23, 30 (1967).6. P. Y. Paterson and S. M. Harwin, J. Exp.

Med. 117, 755 (1963).7. S. H. Appel and M. B. Bornstein, ibid.

119, 303 (1964).8. B. D. Jankovi6 and M. Ilvaneski, Int. Arch.

Allergy 23, 188 (1963).9. A. J. Chaflin, 0. Smithies, R. K. Meyers,

J. Iunnutunol. 97, 693 (1966); B. D. Jankovicand K. Isakovi6, Natuire 211, 202 (1966).

10. M. D. Cooper, R. D. A. Peterson, R. A.Good, Natutre 205, 143 (1965); M. D. Cooper,R. D. A. Peterson, M. A. South, R. A. Good,J. Exp. Med. 123, 75 (1966).

1' September 1967

1200

Immunological Time Scale forHominid Evolution

Abstract. Several workers have ob-served that there is an extremely closeimmunological resemnblance between theserum albumins of apes and mant. Ourstudies with the quiantitative micro-complement fixation method confirmthis observation. To explain the close-ness of the resemblance, previous work-ers suggested that there has been aslowing down of albumin evolutionsince the time of divergence of apesand man. Recent evidence, however,indicates that the albumin moleculehas evolved at a steady rate. Hence,we suggest that apes and man have amore recent common ancestry thanis usually supposed. Ouir calcuilationslead to the suggestion that, if man andOld World monkeys last shared a coin-mon ancestor 30 million years ago,then mani and African apes shared acommoni ancestor 5 million years ago,that is, in the Pliocene era.

It is generally agreed that the Afri-can apes are our closest living rela-tives. However, the time of origin ofa distinct hominid lineage has been asubject of controversy for over 100years (1). The absence of an adequatefossil record has forced students ofhominid evolution to evaluate the phy-logenetic significance of anatomical andbehavioral characteristics in the livingprimate species in order to attempt asolution to that controversy. The na-ture of the problem is such, however,that no definitive answer has yet beengiven. Current estimates range from adate in the late Pliocene (2) to onein the late Oligocene or early Miocene(3) for the origin of the hominids.This great range (4 million to 30 mil-lion years) effectively negates any mean-ingful discussion of the nature of ourpre-Australopithecine ancestors, for theearly dates bring us near to a primitiveprosimian stock, while the late oneswould suggest that a common ancestorfor man and the African apes mightwell resemble a small chimpanzee.One solution to this question lies

in the measurement of the degree ofgenetic relationship which exists be-tween man and his closest living rela-tives. As it has recently become clearthat the structure of proteins closelyreflects that of genes, it is to be ex-pected that quantitative comparativestudies of protein structure should aidin providing this measure of geneticrelationship (4).

Proteins appear to evolve over time,as do the organisms of which they area part. Thus, we may speak of thecommon ancestor of, for example, thehuman and chimpanzee serum albuminmolecules, this ancestral molecule be-ing present in the common ancestor ofman and the chimpanzee. From thetime that the human and chimpanzeelineages separated, their albumins havehad the opportunity of evolving inde-pendently until today they are recog-nizably different, 'but homologously re-lated, molecules. Such homologies maybe studied by immunological techniques,the magnitude of the immunologicalcross-reaction serving as a measure ofthe degree of structural similarity 'be-tween the two kinds of albumin (5).The immunological methods used in

this investigation were similar to thosedescribed earlier (5, 6). Serum sampleswere obtained from all the living generaof apes and from six representativegenera of Old World monkeys andstored at -10°C (7). Albumin waspurified from individual chimpanzee,gibbon, and human serums by themethod of Hoch and Chanutin (8).Groups of three or four rabbits wereimmunized by three courses of injec-tions with each of the purified al-bumins. The antiserums were testedfor purity 'by immunodiffusion, immu-noelectrophoresis, and microcomple-ment fixation (MC'F) with whole serumand purified albumin. Antibodies tocomponents of serum other than al-bumin were always detectable with thefirst two methods, but they were toolow in concentration to interfere witlthe MC'F analysis of the cross-reac-tions discussed below (9). Pooled anti-serums were made by mixing the in-dividual antiserums in reciprocal pro-portion to their MC'F titers (10). Thedegrees of cross-reaction were expressedquantitatively as the index of dissimi-larity or immunological distance (ID)that is, the relative concentration ofantiserum required to produce a com-plement fixation curve whose peak wasas high as that given by the homologousalbumin.

These antiserums were used to ob-tain the data summarized in Table I.Some of these results have alreadybeen published (5, 6). With the anti-serum pool prepared against humanserum albumin, the albumins of theAfrican apes (gorilla and chimpanzee)reacted more strongly than those ofthe Asiatic apes (orang, siamang, andgibbon). The antiserum pool directedagainst chimpanzee (Pan troglodytes)

SCIENCE, VOL. 158

on

Aug

ust 1

2, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Aug

ust 1

2, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Aug

ust 1

2, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Aug

ust 1

2, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

albumin showed the albumin of thepygmy chimpanzee to be immunologi-cally identical to that of the homolo-gous species. Human and gorilla al-bumins reacted somewhat less well butmore strongly than did the albuminsof the Asiatic apes. The antiserum pooldirected against gibbon (Hylobates lar)albumin reacted most strongly withthat of the siamang. The albumins ofthe other apes and man were ap-preciably less reactive.

Conclusions albout genetic relation-ships among the albumins of apes andman may be drawn from these data.The albumins of the orang, gorilla,chimpanzee, and man stand as a unitrelative to those of the gibbon andsiamang. This is evident from the dataobtained with the antiserum to Hylo-bates albumin (Table 1). The close re-

lationship between the albumins of thegibbon and siamang is consistent withthe fact that these two genera of apesare usually placed together in a separatefamily or subfamily (Hylobatinae). Itis also evident that the albumins ofthe gorilla, chimpanzee, and man standas a unit relative to the albumins ofthe Asiatic apes. Moreover, the al-bumins of the gorilla, chimpanzee, andman appear to be equidistantly related;the gorilla and chimpanzee albuminsare no closer to each other than eitheris to human albumin. These relation-ships are consistent with those sug-gested by Goodman on the basis ofqualitative immunodiffusion data (11).

Table 1 also shows that with eachantiserum pool the Old World monkeyalbumins gave markedly weaker reac-

tions (mean ID = 2.3) than thosegiven by any hominoid albumin. Thesize of the immunological gap betweenthe albumins of hominoids and OldWorld monkeys is illustrated by thefact that, at an antiserum concentra-tion where all hominoid albumins givestrong reactions, Old World monkeyalbumins give no reaction with anti-serums to hominoid albumins. Thus,the albumins of all the living apes are

much more similar to each other thanany of them is to nonhominoid al-bumins.The phylogenetic significance of the

above findings, however, is not un-

equivocal. For example, at least twoexplanations are possible for the ex-

tremely close structural similarity ofape and human albumins. On the one

hand, albumin evolution may have beenretarded in the ape, human, or bothlineages since their separation. Good-man (11) and Hafleigh and Williams

I DECEMBER 1967

OUGOCENE

CommonAncestor

a

30

MIOCENE PLIOCENE, -"-HPLEISTOCENE:

Hylobates

Symphalangus

Pongo

Homo

i. Pan

tGorilla

Old World- U i h -

I 120 10

TIME (Millions of Years)

Monkeys

0

Fig. 1. Times of divergence between the various hominoids, as estimated from im-munological data. The time of divergence of hominoids and Old World monkeys isassumed to be 30 million years.

(12) have tended to favor this explana-tion. Alternatively, the close molecularrelationship may reflect a more recentcommon ancestry between ourselves andthe living apes than is generally sup-

posed, albumin evolution having pro-

ceeded at the usual rate for primates.Rates of albumin evolution in pri-

mates have recently been investigated(6). The evidence appears to rule outthe first explanation. No conservatismwas detected in the evolution of any

particular hominoid albumins relative

to any other, nor, indeed, in the evolu-tion of hominoid albumin relative tothose of any other primate group. Al-bumin evolution appears to have pro-

ceeded to the same extent in the vari-ous ape, human, and monkey lineages.Therefore, it seems likely that apes andman share a more recent common an-

cestry than is usually supposed.The data presented above enable us

to calculate how recently this commonancestor lived. We have recently shownthat albumin evolution in primates is

Table 1. Reactivity of various primate albumins with antiserums prepared against hominoidalbumins. Antiserum to Honio albumin was a pool obtained by mixing three antiserums inreciprocal proportion to their MC'F titers. The three antiserums and their titers were 54(1/7000), 6F (1/11,000), and Hafleigh and Williams' pool II (1/5000). Titer of pool equaled1/7000. Data taken from (5) and (6). Antiserum to Pan albumin was a pool obtained by mix-ing four antiserums [92 (1/5000), 102 (1/3000), 972 (1/2500), 99, (1/2500)] in reciprocal pro-portions to their MC'F titers. Titer of pool eaualed 1/3000. Some of these data are taken from(6). Antiserum to Hylobates albumin was a pool obtained by mixing three antiserums [824(1/1500), 834 (1/4500) and 844 (1/1900)] in reciprocal proportion to their MC'F titers. Titer ofpool equaled 1/2000. The six species of Old World monkeys tested were Macaca mulatta, Papiopapio, Cercocebus galeritus, Cercopithecus aethiops, Colobus polykomos, and Presbytis entellus.

Index of dissimilaritySpecies of albumin Antiserum Antiserum Antiserum

to Homo to Pan to Hylobates

Homninoidea (apes and man)Homo sapiens (man) 1.0 1.09 1.29Pan troglodytes (chimpanzee) 1.14 1.00 1.40Pan paniscus (pygmy chimpanzee) 1.14 1.00 1.40Gorilla gorilla (gorilla) 1.09 1.17 1.31Pongo pygmaeus (orang-utan) 1.22 1.24 1.29Symphalangus syndactylus (siamang) 1.30 1.25 1.07Hylobates lar (gibbon) 1.28 1.25 1.00

Cercopithecoidea (Old World monkeys)Six species (mean ± S.D.) 2.46 ± .16 2.22 ± .27 2.29 ± .10

1201

f

I

a remarkably regular process (6). Line-ages of equal time depth show very sim-ilar degrees of change in their al-bumins. The degrees of change shownwould therefore seem to be a functionof time, and a mathematical relation-ship between ID and the time ofdivergence of any two species mustexist. Thus, albumin molecules canserve as an evolutionary clock or datingdevice. The calibration of that clock,that is, the elucidation of the relation-ship between ID and time, would allowus to calculate the time of divergencebetween apes and man (13).

This relationship is likely to be rathersimple. If the amino acid sequences ofproteins also evolve at steady rates,and there is evidence that they oftendo (14), then the relationship betweenID and time of divergence should beof the same form as the relationshipbetween ID and structural difference(number of amino acid replacements).Direct evidence for a simple correla-tion 'between immunological cross-re-activity and structural relatedness isavailable from complement fixationstudies of hemoglobins (15, 16) andcytochromes c (17) of known aminoacid sequence. Indirect evidence forsuch a correlation is provided by thecorrespondence between cross-reactivityand phylogenetic relatedness which hasbeen demonstrated for a variety ofproteins (15, 18).

It appears likely that log ID is ap-proximately proportional to the timeof divergence (T) of any two species,that is, log ID = kT, where k is aconstant. This relationship is evidentfrom several sets of MC'F data ob-tained with various purified dehydro-genases of fishes, amphilbians, reptiles,and birds whose times of divergencecan be estimated from the fossil record(15, 19). Let us suppose that a similarrelationship is appropriate for albuminevolution in primates (for an extendeddiscussion of this, see 20).

Although the primate fossil record isfragmentary, it does, in combinationwith the available immunological evi-dence, provide sufficient evidence tosuggest that the lineages leading to theliving hominoids and Old World mon-keys split about 30 million years ago(21). That is, the ID of 2.3 which isthe mean ID observed between the al-bumins of hominoids and Old Worldmonkeys corresponds to a T value ofabout 30 in the above equation. Iflog 2.3 = k x 30, then k = 0.012.Since the mean ID bet'ween the al-bumins of man and the African apes

1202

is 1.13, the time of divergence of manfrom the African apes is log 1.13 divid-ed by 0.012, that is, 5 million years.Proceeding similarly, we calculate thatthe lineage leading to the orang separat-ed from that leading to the Africanapes 8 million years ago, and that thetime of divergence of the gibbon andsiamang lineage from that leading tothe other apes and man is 10 millionyears (Fig. 1).There are, of course, at this stage

in our investigation, uncertainties inthese calculations. We may possilbly bemaking erroneous assumptions about(i) the time of divergence of apes andOld World monkeys, and (ii) the natureof the relationship between ID and timeof divergence. We feel, however, thatthese possilble errors are unlikely to beof sufficient magnitude to invalidate theconclusion that apes and man divergedmuch more recently than did the apesand Old World monkeys. In our opin-ion, the albumin data definitely favorthose who have postulated that manand the African apes shared a com-mon ancestor in the Pliocene.

If the view that man and the Africanapes share a Pliocene ancestor andthat all the living Hominoidea derivefrom a late Miocene (22) form is cor-rect, a number of the problems thathave troubled students of this groupare resolved. The many features ofmorphology, particularly in the thoraxand upper limbs, which man and theliving apes share in varying degrees,but which were not present in theMiocene apes, such as Dryopithecus(Proconsul), Limnopithecus, and Plio-pithecus (23), are then seen as due torecent common ancestry and not, asgenerally accepted, to parallel or con-vergent evolution (24).We suggest that the living apes and

man descended from a small memberof the widespread Miocene dryopithe-cines, which became uniquely success-ful due to the development of thelocomotor-feeding adaptation known asbrachiation. The adaptive success ofthis development and the subsequentradiation of the group possessing itmay have made this group the onlysurviving lineage of the many apes pres-ent throughout the tropical and sub-tropical Miocene forests of the OldWorld. Possibly the African membersof this radiation, in the Middle Plio-cene (due perhaps to pressure fromthe developing Cercopithecinae), beganvarying degrees of adaptation to a ter-restrial existence. The gorilla, chim-panzee, and man appear to be the

three survivors of this later radiation.According to this hypothesis, some 3million years are allowed for thedevelopment of bipedalism to the ex-tent seen in the earliest fossil hominid,Australopithecus.The concept of an immunological

time scale is a logical extension of thefinding that the immunological prop-erties of serum albumin have under-gone regular changes during primateevolution. The utility of that conceptcan be tested in several ways. First,extensive immunological studies of pro-teins of known amino acid sequenceare needed in order to make sure ofthe nature of the relationship betweencross-reactivity and degree of sequenceresemblance. Second, albumin evolutionshould be investigated by the use ofimmunological methods in other mam-malian groups, such as the ungulates,where an extensive fossil record isavailable. Third, a search must be madefor other primate proteins that exhibitconstant evolutionary rates. This willpermit an independent calculation ofthe time of origin of hominids.

VINCENT M. SARICHDepartments of Anthropology andBiochemistry, University of California,Berkeley

ALLAN C. WILSONDepartment of Biochemistry,University of California, Berkeley

References and Notes

1. T. H. Huxley, Evidence as to Man's Place inNature (Williams and Northgate, London,1863).

2. S. L. Washburn and D. A. Hamburg, inPrimlate Behavior, 1. DeVore, Ed. (Holt,Rinehart and Winston, New York, 1965), pp.1-15.

3. A. H. Schultz, Yerkes Newsletter 3 (1), 15(1966); J. Beuttner-Janusch, Origins of Man(Wiley and Sons, New York, 1966); B.Campbell, Human Evolution (Aldine, Chi-cago, 1966); D. R. Pilbeam, Science J.3 (2), 47 (1967); E. L. Simons and D. R.Pilbeam, Folia Primat. 3, 81 (1965); L. S. B.Leakey, Nature 213, 155 (1967).

4. Preliminary studies of nucleic acids and ofhemoglobin and myoglobin similarities amongprimates have been published (B. H. Hoyer,E. T. Bolton, B. J. McCarthy, R. B. Roberts,in Evolving Genes and Proteins, V. Brysonand H. Vogel, Eds. (Academic Press, NewYork, 1965), pp. 581-90; R. L. Hill andJ. Buettner-Janusch, Fed. Proc. 23, 1236(1964); P. C. Hudgins, C. M. Whorton,T. Tomoyoshi, A. J. Riopelle, Nature 212,693 (1966)).

5. V. M. Sarich and A. C. Wilson, Science 154,1563 (1966).

6. , Proc. Nat. Acad. Sci. U.S. 58, 142(1967).

7. The sources of these serums have alreadybeen acknowledged (5, 6).

8. H. Hoch and A. Chanutin, Arch. Biochem.Biophys. 51, 271 (1954).

9. See footnote 21 in 5 for a detailed discus-sion.

10. We define the MC'F titer of an antiserum asthat concentration at which the homologousantigen will fix 75 percent of the availablecomplement at the peak of the C'F.

11. M. Goodman, Human Biol. 33, 131 (1961);ibid. 34, 104 (1962); ibid. 35, 377 (1963);Amer. J. Phys. Anthropol. 26, 225 (1967).

SCIENCE, VOL. 158

12. A. S. Hafleigh and C. A. Williams, Jr.,Science 151, 1530 (1966).

13. Zuckerkandl and Pauling have pointed outthe possibility of using proteins as evolu-tionary clocks (see 14).

14. E. Zuckerkandl and L. Pauling, in Horizonsin Biochemistry, M. Kasha and B. Pullman,Eds. (Academic Press, New York, 1962), p.189; in Evolving Genes and Proteins, V.Bryson and H. J. Vogel, Eds. (AcademicPress, New York, 1965), p. 97; R. F. Doo-little and B. Blomback, Nature 202, 1471964); E. Margoliash and A. Schejter, Ad-van. Protein Chem. 21, 113 (1966); J.Buettner-Janusch and R. L. Hill, Science147, 836 (1965); L. F. Smith, Amer. J. Med.40, 662 (1966); J. Derancourt, A. S. Lebor,E. Zuckerkandl, Bull. Soc. Chim. Biol.49, 577 (1967).

15. A. C. Wilson, N. 0. Kaplan, L. Levine, A.Pesce, M. Reichlin, W. S. Allison, Fed. Proc.23, 1257 (1964).

16. M. Reichlin, M. Hay, L. Levine, Imminno-chemistry 1, 21 (1964); M. Reichlin et al., J.Mol. Biol. 17, 18 (1966).

17. M. Reichlin, personal communication; E.Margoliash, M. Reichlin. A. Nisonoff, inConformation of Biopolymers, G. N. Rama-chandran. Ed. (Academic Press, New York,1967), vol. 1, p. 253.

18. C. A. Leone, Ed., Taxoniomic Biochemistryand Serology (Ronald Press, New York,1964); C. A. Williams, Jr., in Evolutionaryanid Genetic Biology of Primates, J. Buettner-Janusch, Ed. (Academic Press, New York,1964), vol. 2, p. 25.

19. S. N. Salthe and N. 0. Kaplan, Evolution 20,617 (1966).

20. V. M. Sarich, in Human Evoluttion 1968,S. L. Washburn, Ed. (Holt, Rinehart, Win-ston, New York, in press).

21. This date will probably appear relativelyrecent to many. There is, however, certainlyno fossil material to contradict it. The firstfossil Old World monkeys (Mesopithecius,Prohylobates) are found in the middle-to-early Miocene of Africa [E. L. Simons, inEvolutionary and Genetic Biology of Pri-mates, J. Buettner-Janusch, Ed. (AcademicPress, New York, 1963), vol. 1, pp. 65-129;Nature 205, 135 (1965)]. On the other hand,it can fairly reliably be assumed that allliving primates derive from a common an-cestral form living no earlier than the lateCretaceous. At least one of the living pro-simians, Tarsiuts, represents a lineage alreadydistinct in the early Eocene (see Simons). Asthere is no reason to believe that any otherprosimian group shows closer relationship tothe higher primates, these suborders (An-thropoidea and Prosimii) were already dis-tinct no less than 50 and no more than ap-proximately 70 million years ago. Althoughno upper limit on the time of origin of aseparate Old World monkey line is avail-able, there would probably be general agree-ment with the suggestion, which is entirelyconsistent with the immunological evidence,that appreciable periods of common an-cestry characterize both the Catarrhini (apes,man, Old World monkeys) after their sepa-ration from the ancestors of the New Worldmonkeys and the living Anthropoidea aftertheir separation from the Prosimii. To con-tain these periods within the 50-to-70 millionyear limits set above would seem to requirethat a reasonable upper limit on the timeof divergence of apes and Old World monkeysbe set at 30 to 45 million years. We feel thatthe lower end of this time scale is moreprobable. If this were not so, the relation-ship between ID and time of divergence forprimates as a whole would be extremely com-plex. We feel that on immunological groundsthis is unlikely (20).

22. B. M. Funnell, Quiart. J. Geol. Soc. London120s, 179 (1964).

23. J. R. Napier and R. P. Davis, Fossil Mam-mals of Africa, No. 16 (British Museum Nat-ural History, London, 1959); W. E. LeGrosClark, Fossil Evidence for Human Evolution(Univ. of Chicago Press, Chicago, ed. 2,1964); F. Ankel, Folia Primatol. 3, 263(1965); S. L. Washburn, Classification andHuman Evolution (Aldine, Chicago, 1963),pp. 190-203.

24. The availability of antiserums to Pan andHylobates albumins allows a partial answerto a caveat that might be made against theseconclusions. It could be argued that the simi-larities among, for example, hominoid albu-

1 DECEMBER 19f7

mins are a result of parallel evolution andnot recent common ancestry. That is, im-munologically similar albumin structures haveevolved in genetically separated lineages. Itcan be seen that this would, in the case ofapes and man. require a most improbableset of coincidences, for the ID's of otherapes and man obtained with antiserums togibbon albumin are very nearly equal (ob-viously except for the siamang), indicatingthese albumins have changed to much thesame degree since their separation from thegibbon; yet the antiserums to human andchimpanzee albumins show that human,chimpanzee, gorilla, and orang albumins arequite different from one another. To sup-port the idea of parallelism, then, one wouldhave to postulate that gibbon albumin evolvedin parallel to, those of the other apes untilthe radiation of these lineages began, andthen diverged from all of these (clearly gib-bon albumin cannot have evolved in parallel

ordinarily considered detrimental to cellpears to be adaptive.

While studying the effects of addi-tional endogenous heat in animals, weimplanted aseptic, electric heat sourcescovered with medical-grade siliconerubber in the abdominal cavities ofsheep. Two of these devices (5.1 cmwide, 15.1 cm long, and 0.64 cm thick)were implanted end to end in the dorsalabdominal cavity and extended from therenal artery caudally to the aortic bi-furcation. After recovery of the ani-mals, voltage (direct-current) was ap-plied to each heater through insulatedwire leads from an external electroniccircuit capable of sensing the heater-coil temperature and maintaining itwithin ± 0.05°C. Another ewe withidentical implants to which no heatwas applied served as a control.The temperature at the surface of

the heater was calculated from themeasured temperature of the heater coiland a predetermined conductance fac-tor for the silicone rubber coveringeach heater.

Power input was continuously re-corded, and calculated temperatures atthe heater surfaces were tabulated. Datafrom one of these experiments areplotted in Fig. 1. The two strip heaterswere designated anterior and posterior,according to their positions relative toeach other in the dorsal abdominalcavity. The initial surface temperature

to all four nonhylobatid lineages at the sametime). We see that to reconcile the immuno-logical data with many of the current viewsconcerning hominoid evolution (3) we areforced to postulate a remarkable series ofcoincidences. In the same way those whowould attribute the morphological similari-ties in the trunk and upper limbs, among allthe living apes and man, to parallelism mustmake a similar appeal to coincidence.

25. Supported by a PHS predoctoral fellowshipto V.M.S. (1-Fl-GM-30,454-01) and an NSFgrant to A.C.W. (GB-6420). Portions of thiswork have been reported at meetings of theAmerican Association of Physical Anthro-pologists (8 April 1966 and 24 April 1967)and to the AAAS (30 December 1965). Wethank N. Arnheim, J. Gerhart, T. Jukes, G.Sensabaugh, and S. L. Washburn for helpfuldiscussion.

26 September 1967 .

structure. The vascularization thus ap-

of the anterior heater (T.a) was 45°C;that of the posterior heater (Tsp) was42°C. Both temperatures were in thenoxious range (1). As heating con-tinued, a decline in both temperaturesoccurred in spite of the increase inpower to each heater required to keepheater-coil temperatures constant.

TseC45-44 -

4342-4 1 -

2 30-Ej20 -

a

< O-

AWATTS2I

0-

T\T--

- - -- - - - - -- - -TspAE.B.F. ANT.

,------------------------AE.BS POST.IF

*&WATTS ANT.

A&,WATTS POST.

4 8 2 16 20 24 28DAYS

Fig. 1. Differential heating of anteriorand posterior dorsal abdominal heat-exchangers in a sheep. Ta, Surface tem-perature of anterior heater; T, surfacetemperature of posterior heater; Awatts,increase in power necessary to maintainthe heater coil at a constant temperature;AE.B.F., increase in effective blood flowequivalent to increase in power (0.85 as-sumed to the specific heat of blood).

1203

Visceral Tissue Vascularization: An Adaptive Response toHigh Temperature

Abstract. Electrical heat sources implanted in the abdominal cavities of sheepwere heated to give initial temperatures of 420 and 45°C at the surfaces of theheaters. During 18 days of constant heating, a vascularized connective-tissueenvelope encapsulated the heat sources, and the temperatures at the surfaces ofthe heaters declined 0.80 and 1.8°C, respectively. The degree of vascularizationand the magnitude of the decrease in the surface temperature appeared to berelated to the proximity of the tissue's initial temperature to 45°C, a temperature