New discoveries and interpretations of hominid fossils and artifacts from Vindija Cave, Croatia

41 Pages·2003·2.05 MB·English

JournalofHumanEvolution46(2004)25–65
New discoveries and interpretations of hominid fossils and
artifacts from Vindija Cave, Croatia
James C.M. Aherna*, Ivor Karavanic´b, Maja Paunovic´c, Ivor Jankovic´d,
Fred H. Smithe
aDepartmentofAnthropology,UniversityofWyoming,Box3431,Room123,AnthropologyBld.,Laramie,WY82071-3431,USA
bDepartmentofArchaeology,FacultyofPhilosophy,UniversityofZagreb,IvanaLucˇicˇa3,10000Zagreb,Croatia
cInstituteofQuaternaryPalaeontologyandGeology,CroatianAcademyofSciencesandArts,Zagreb,Croatia
dInstituteforAnthropologicalResearch,Amruseva8,10000Zagreb,Croatia
eDepartmentofSociology/Anthropology,LoyolaUniversity,Chicago,IL60626,USA
Received30January2003;accepted30September2003
Abstract
Beginning with excavations during the 1970s, Vindija Cave (Croatia) has yielded significant Middle and Upper
Paleolithic fossil and archaeological finds. We report on seven recently identified hominid fossils, a newly associated
partial hominid cranial vault from level G , nine possible bone retouchers, and a revised interpretation of the
3
Mousterian artifact assemblage from the site. This new information reinforces our knowledge of the complex
biocultural phenomena revealed in unit G and earlier deposits at Vindija. Six of the new hominid fossils derive from
stratigraphic units G and I, while one lacks exact provenience. All specimens preserving diagnostic anatomy are
fromNeandertals.Oneofthepostcranialremains,aradiusfragmentwhichexhibitsNeandertal-likeanatomy,comes
from level G and is congruent with the previously established association of Neandertals with an early Upper
1
Paleolithic industry at the site. The partial cranial vault represents the most complete Neandertal from Vindija. The
possible retouchers derive from unit G. Our analysis of these artifacts suggests that both percussion and pressure
techniquesmayhavebeenusedbyNeandertalsinthefinalstageoftoolproduction(retouching).
ThispaperalsopresentsarevisionoftheartifactanalysisforlateMousterianlevelG .Weseparatedrawmaterialsinto
3
twomaingroupsduetothedifferingwaysthatthematerialsfractureandthedifferingmorphologyofthedebitage.Theuse
of raw material in level G is different from earlier Middle Paleolithic levels at Vindija. This indicates that the G late
3 3
NeandertalsweremakingchoicesregardingsourcematerialsomewhatmoreliketheUpperPaleolithicpeopleatthesite.
Wheninterpretedwithinalargerregionalframework,theVindijaarchaeologicalandhominidfossilremainsdemonstratea
complex,mosaicpatternofbioculturalchangeintheLatePleistoceneofsouth-centralEurope.
(cid:1)2003PublishedbyElsevierLtd.
Keywords:Vindija;Neandertal;MiddlePaleolithic;Retouchers;LithicAnalysis;Lithics;HumanEvolution;South-CentralEurope
* Correspondingauthor.Tel.:[phone];fax:[phone]
E-mailaddresses:[email protected](J.C.M.Ahern),ikaravan@mudrac.ffzg.hr(I.Karavanic´),[email protected]
(M.Paunovic´),[email protected](I.Jankovic´),[email protected](F.H.Smith).
0047-2484/04/$-seefrontmatter(cid:1)2003PublishedbyElsevierLtd.
doi:10.1016/j.jhevol.2003.09.010
26 J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65
Introduction its importance in a letter dated 1926 (cited in
Vukovic, 1949).
Vindija Cave, located in northwestern Croatia, Excavations by S. Vukovic´ began in 1928 and
hasyieldednumeroushominidfossilsandarchaeo- continued, with many interruptions, for the next
logical remains. Most of the Vindija discoveries thirty years (Vukovic´, 1949, 1950; Karavanic´,
span the crucial period in Europe of 25–45 ka, 1995).Vukovic´’sexcavationswerebothinsideand
which saw the appearance of modern humans, the in front of the cave and were mostly limited to
Middle to Upper Paleolithic transition, and the the upper levels of the deposits (Vukovic´, 1949,
disappearance of Neandertals. Vindija Cave has 1950). He identified three Paleolithic cultural
contributed significantly to our understanding periods:Mousterian,Aurignacian,andMagdalenien
of the complexity of these events. The latest (which he later called Gravettian [Vukovic´, 1970])
Mousterian-associated hominid fossils from the and also published an analysis of the Mesolithic
site exhibit a mosaic of Neandertal and early archaeology from the site (Vukovic´, 1961).
modern human anatomy which suggests at least Vukovic´ recognizedasmanyassevenstratigraphic
somedegreeofregionalevolutionarycontinuityin units at Vindija, but these cannot be correlated
Europe(SmithandRanyard,1980;Wolpoffetal., with the stratigraphy reported from more recent
1981; Smith, 1984, 1994; Frayer et al., 1993). Also excavations.
significantisthefactthatNeandertalfossilsdating Excavations at Vindija, under the direction of
to less than 30 ka are associated with the earliest M.Malez,beganinJuly1974andcontinuedevery
Upper Paleolithic artifacts from the site (Smith season through 1986. During that time period,
andAhern,1994;Karavanic´,1995;Karavanic´ and approximately 60 hominid specimens associated
Smith, 1998). Since systematic excavations at the with Mousterian or earliest Upper Paleolithic
site ended in 1986, research has focused on identi- industries and 45 specimens associated with
fication and analysis of previously overlooked Epigravettian artifacts were discovered. Further-
hominid fossils and artifacts (e.g., Smith and more,extensivearchaeologicalandfaunalremains
Ahern, 1994), continued analysis of the fossil and were collected. Numerous publications have dealt
archaeological samples (e.g., Trinkaus and Smith, withthefindingsfromVindija(Malez,1975,1980;
1995; Karavanic´ and Smith, 1998; Ahern et al., Malez and Rukavina, 1979; Malez et al., 1980;
2002), and the site’s chronology (Karavanic´ et al., Wolpoff et al., 1981; Malez and Ullrich, 1982;
1998; Smith et al., 1999; Wild et al., 2001). In Smith et al., 1985, 1999; Smith and Ahern, 1994;
this paper, we report on additional fossil and Karavanic´, 1995; Trinkaus and Smith, 1995;
archaeological discoveries and interpretations that Karavanic´ et al., 1998; Wild et al., 2001; Ahern
enhance our understanding of the biology and et al., 2002).
behavior of the Vindija hominids and of the
mosaic pattern of biocultural change in the Late Stratigraphy and chronology
Pleistocene of south-central Europe.
VindijaCaveislocatedonthesouthwestsideof
History of excavation Krizˇnjak Peak at an elevation of 275 m above sea
level (Malez et al., 1980). Krizˇnjak Peak is part of
ThefirstpublishedmentionofVindijaCavewas theRavnaGora,whichisasoutheasternextension
madebyD.Hirc(Hirc,1878)inPrirodniZemljopis of the Alps. The cave itself is a single chamber
Hrvatske (Natural Geography of Croatia). Hirc measuring approximately 50 m deep, 28 m wide,
reported the discovery of 38 bone fragments and andmorethan10minheight.AccordingtoMalez
20 pottery sherds from the cave. He further cited and colleagues (Malez et al., 1984), the cave was
mention of the cave in an 1801 note found at the formedinUpperBadenFormationsandstoneand
parish in Donja Voc´a. Gorjanovic´-Kramberger, limestone during the Middle Pliocene as the result
leader of the Krapina excavations (1899–1905), of tectonic folding and cracking of lake basin
was aware of the cave and made mention of deposits. The subterranean cavern was given its
J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65 27
Fig.1.TheVindijastratigraphicsequence.*Vi13.8and284havesolidprovenience.Provenienceasshownforfossilsdenotedwitha
“+”isprobable,butnotcertain.Otherspecimensdescribedinthetextarenotshownbecausetheyeitherlackprovenience(Vi11.48,
13.9)orareproveniencedtojust“G”(Vi11.47).†Seetextforinterpretationoftheradiocarbondates.
modern opening during the Upper Pliocene as the MalezandRukavina(1979)dividedtheVindija
Sˇokot creek cut the valley that Vindija Cave now deposits into 14 stratigraphic units, labeled A
overlooks. Deposition in the cave probably began through N (Fig. 1). A, B, and C are Holocene,
in the Early Pleistocene (Malez et al., 1984). while D through N are Pleistocene. Three of these
28 J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65
Table1
RadiocarbondatesfromtheVindijaCavea
Layer Sample 14Cage(ka) Lab.no. Reference
E Cavebear 18,500(cid:2)300 Z-2447 Obelicetal.,1994
F Charcoal 24,000(cid:2)3300 Z-612 Srdocetal.,1984
F Charcoal 29,700(cid:2)2000 Z-613 Srdocetal.,1984
F Charcoal 27,000(cid:2)600 Z-551 Srdocetal.,1984
F/d/d Cavebear 26,600(cid:2)930 Z-2433 Obelicetal.,1994
G Cavebear 18,280(cid:2)440 Z-2432 Obelicetal.,1994
1
G Cavebear 33,000(cid:2)400 ETH-12714 Karavanic,1995
1
G Cavebear 46,800+2300/(cid:3)1800 VERA-1428 Wildetal.,2001
1
G Neandertal 29,080(cid:2)400 OxA-8296 Smithetal.,1999
1
G Neandertal 28,020(cid:2)360 OxA-8295 Smithetal.,1999
1
G Neandertal >42,000 Ua-13873 Kringsetal.,2000
3
H Cavebear 33,400+2000/(cid:3)1600 VRI-1125 Wildetal.,2001
I Cavebear 37,000(cid:2)600 VERA-0109 Wildetal.,2001
J Cavebear 34,700(cid:2)500 VERA-0105 Wildetal.,2001
aAdaptedfromWildetal.,2001.
units, F, G, and K, are further subdivided into Smithetal.,1985;Wolpoff,1999).Themostrecent
multiple levels (Wolpoff et al., 1981). Although Vindija hominid sample is from level D and is
some of the cave sediments were affected by cryo- associated with a late Upper Paleolithic industry.
turbation and/or by ice-wedging, the stratigraphic These level D hominids are clearly anatomically
sequence was based on the remaining undisturbed modern humans (Wolpoff et al., 1981).
part (Malez and Rukavina, 1979; Malez and Directradiometricdatingofanimalandhuman
Ullrich, 1982; Karavanic´ and Smith, 1998; boneshasbeenconductedatdifferentlaboratories
Paunovic´ et al., 2001). in the attempt to better define the chronology of
Malez’s excavations yielded 4 to 5 stratigraphi- the Neandertal fossil and cultural remains from
cally distinct hominid samples associated with Vindija (Karavanic´ et al., 1998; Rink et al., 1999;
Middle (level G and, as reported in this paper, Smith et al., 1999; Wild et al., 2001). A compi-
3
unit I) and Upper (levels G and F and unit D) lationofall14CdataobtainedfortheVindijaCave
1 d
Paleolithic industries. The largest of these samples is given in the Table 1. The direct AMS radio-
isfromlevelG .Morphologically,thesefossilsare carbon dating of two Neandertal specimens (Vi
3
attributabletoNeandertals,butmany(e.g.,Smith, 207 and 208) from level G yielded dates of
1
1982, 1984, 1994; Wolpoff et al., 1981; Frayer 29,080(cid:2)400 yrs BP and 28,020(cid:2)360 yrs BP,
et al., 1993; Wolpoff, 1999) have argued for respectively (Smith et al., 1999). However, non-
their evolutionary intermediacy between earlier destructive (cid:1)-ray spectrometry and electron spin
Neandertals, such as those from nearby Krapina, resonance (ESR) dating of level G has yielded
1
and early modern Europeans. The six speci- significantly older dates (Karavanic´ et al., 1998;
mens from level G represent the latest Neander- Rink et al., 1999). The older dates should be
1
tals in Europe (Smith and Ahern, 1994; Smith regarded as suspect because of the relatively low
et al., 1999). The ten specimens from level quantity of uranium in the measured pieces and
F are associated with an Aurignacianlike in- thehighstandarderrorsof(cid:1)-raydating(Churchill
dustry (Karavanic´, 1995; Wolpoff et al., 1981). and Smith, 2000). Given radiometric dates for
Their morphology is enigmatic, resembling both other levels in the cave, the presence of Upper
Neandertals and modern humans in the limited Paleolithic elements in the G archaeological
1
anatomy that is preserved (Wolpoff et al., 1981; assemblage, and inconsistent (cid:1)-ray results, the
J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65 29
radiocarbon—especially AMS-dates should be Level G is of special importance in that it
1
givenpriority(Smithetal.,1999;Wildetal.,2001). yielded Neandertal fossils, both Mousterian and
Nevertheless,itmustbenotedthattheapplication Upper Paleolithic lithic artifacts and early Upper
ofthe14Cmethodtoanimalbonesfromthissiteis Paleolithic bone points. This level represents a
not straightforward because of the poor preser- warmer period, likely the Podhradem interstadial,
vation state of the collagen in most specimens. andconsistsofacharacteristicreddish-brownclay.
Recently at VERA laboratory in Vienna, AMS Thecolorofthisclayissodistinctthatitspresence
dating was attempted for cave bear bone samples within bone specimens has been used confidently
from units F through H of the cave sediments, to determine their provenience (Smith and Ahern,
and,unfortunately,thecollagenofallsampleswas 1994; Karavanic´ and Smith, 1998). Three of the
already highly degraded (Wild et al., 2001). Thus, five radiocarbon dates for level G are consistent
1
the ages derived from AMS measurements on with a Podhradem date. The two remaining
Vindija animal bone should be treated with dates correspond either to the Lower Wu¨rm
caution:theyoungagesobtainedonsamplesfrom (46,800+2300/(cid:3)1800 yrs BP, see Table 1) or
layers H, I and J may be due to an insufficient Middle Wu¨rm (18,280(cid:2)440 yrs BP, see Table 1)
cleanup of the samples (Wild et al., 2001), or due stadials.Thespuriousdates,althoughpossiblythe
tobio-and/orcryoturbationsregisteredintheLate result of contamination, highlight the possibility
Pleistocene deposits of the cave (Paunovic´ et al., that G collections include materials mixed from
1
2001).Also,theageforonecavebearbonesample underlying and/or overlying levels via cryotur-
fromlevelG of33,000(cid:2)400yrsBPwasobtained bation in parts of the cave. Such a scenario has
1
at ETH Zu¨rich, while at the same time in the cave been suggested to explain the mix of Neandertal
bearspecimenfromlevelG ,sufficientcollagenfor fossils, Mousterian lithics, and Upper Paleolithic
3
an age determination was not present (Karavanic´, bone points (Stringer, 1982; Kozlowski, 1996;
1995). Montet-White, 1996; Zilha˜o and d’Errico, 1999).
When the radiometric dates and other chrono- However, none of the lithics from G exhibit
1
logically pertinent evidence (e.g., fauna, sedimen- modification that would be unequivocally sugges-
tology)aretakentogether,itispossibletoprovide tive of vertical movement through cryoturbation
a reasonable, albeit not definite, chronology for (Karavanic´ and Smith, 1998, 2000). Furthermore,
the Vindija sequence. A U/Th date of 114,000 ka the Aurignacian-type split-based bone point and
(Wild et al., 1987/1988) combined with fauna and virtually all the hominid remains from the level
sediments similar to those from Krapina indicate were excavated from non-cryoturbated areas
thatunitKislikelyRiss-Wu¨rminterglacialinage. (Wolpoff et al., 1981). Finally, the direct AMS
Units H and F and levels G and G represent radiocarbon dates for two G Neandertal speci-
4 3 1
colder periods, while G and G represent warmer mens places them well within the time range of
1 5
periods (Malez et al., 1984; Wolpoff et al., both the Podhradem interstadial and the central
1981). Based on correlation with sites from the European Aurignacian. Although contemporary,
MoravianKarst(Wolpoffetal.,1981;Smithetal., the exact circumstances of the Neandertal associ-
1985), the single AMS radiocarbon date of ation with the early Upper Paleolithic are not
>42,000 (Krings et al., 2000; see Table 1), and a known. Unit F, with its partly Neandertal-like
U/Th date of cave bear bone of 41,000+1000/ fossils and Aurignacian artifacts (in F and F ),
d/d d
(cid:3)900 yrs BP (Wild et al., 2001), the important dates to the Middle Wu¨rm stadial based on its
hominid-fossil bearing level G correlates to the cold-adapted fauna and associated radiocarbon
3
Lower Wu¨rm stadial (following the Moravian dates (see Table 1). Compared to the case of
Karst sequence: Musil and Valoch, 1966; Valoch, level G , the possibility that cryoturbation caused
1
1968). Radiocarbon dates for the Lower Wu¨rm the association of Neandertal-like fossils with
stadiallevel(7a)atKulnarangefrom38,600+920/ Aurignacian artifacts is stronger for F and
d/d
(cid:3)800 yrs BP to 45,660+2850/(cid:3)2200 yrs BP F . Yet, given the evidence from G , such an
d 1
(Valoch, 1977/1978, 1981). association in the lower F unit would not be
30 J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65
Table2
NewlyassociatedandnewlyidentifiedVindijacranialspecimens
Primary Otherspecimen Newinventory Description Level Discoveryyear
specimen numbers numbers (yearidentified)
number
284-230- 284a 11.19 Rightsupraorbitalsegmentattached Gb 1976–1978(allfour
3
255-256 230a 11.22 toalargeportionofcentraland associatedin1996)
255a 11.37 posteriorfrontoparietalfragment
256a 11.35
11.47 – 11.47 Frontalsquamafragmentwith Gc 1976(1996)
temporalline
11.48 – 11.48 Posteriorfrontalsquamafragment Unknown 1974(1996)
11.49 – 11.49 Frontalsquamafragmentwith Gc 1978(1996)
3
temporalline
11.52 11.52 Partialrightmandibularramus Ic 1975(2002)
13.7 – 13.7 Partialleftilium Gc 1978(2002)
3
13.8 – 13.8 Leftproximalradiusshaft Gb 1974(2002)
1
13.10 – 13.10 Inferiorangleofrightscapula Ic 1975(2002)
aPreviouslydescribedbyWolpoffetal.(1981).
bProvenienceconsidereddefinite.
cProbableprovenience,basedonspecimenlabelonly.
unprecedented. Given the limited anatomy pre- as “13”. Following the same pattern, a similar
servedbytheunitFfossils,itisalsonotpossibleto system, using instead “21,” “22,” “23,” has been
rule out the possibility that the hominids were applied to all Epigravettian or more recent
early moderns that retained some Neandertal-like human finds (Units D and C). A full listing of the
features. AurignacianandMousterianassociatedhominids’
inventory numbers (old and new) will be provided
in the forthcoming catalogue (Rabeder et al., in
Note on the new numbering system press). The possible hominid specimens found
amongfaunalassemblagesareprovisionallyinven-
In order to avoid confusion caused by the toried as “33,” with the number to the right of
discovery of additional hominid remains found decimal designating the individual specimen. For
during excavations between 1981 and 1986 or “33” specimens that are eventually identified as
recentlyamongthefaunalcollections,theInstitute hominid, their inventory numbers will change to
of Quaternary Paleontology and Geology insti- “11,” “12,” or “13,” with the specimen number
tuted a new numbering system during the late given to the right of the decimal.
1990s. The system has only been applied to homi- Inordertoavoidconfusion,wewillusetheold
nid remains from unit F and below, which are inventorynumbersforallearlierdescribedVindija
Aurignacian and Mousterian in age (the Epi- specimens (see Wolpoff et al., 1981; Malez and
gravettian remains from unit D are excluded from Ullrich,1982;Smithetal.,1985,SmithandAhern,
this system). This new system uses “11” for all 1994), while the new specimens are referred to
hominid cranial remains with the number to the under the new numbering system. Old and new
right of decimal designating the individual speci- inventory numbers and other details for all of the
men (e.g., “11.48”). Hominid dental remains are Vindija specimens described here are given in
designated as “12”, and the postcranial parts Table 2.
J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65 31
Fig.2.Thepartialcalotte,Vindija284-230-255-256,innormafrontalis:(A)naturallyarticulated;(B)symmetricallyreconstructed;and
(C)artisticreconstruction.Scaleis1cm.
32 J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65
Fig.3.SuperiorviewofVindija284-230-255-256,apartialcalotte:(A)naturallyarticulatedand(B)symmetricallyreconstructed.Scale
is1cm.
J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65 33
Fig.4.Lateralviewofthepartialcalotte,Vi284-230-255-256.Thesupraorbitalpiece,Vi284,isnotvisibleinthisview.Scaleis1cm.
New hominid fossil associations and specimens cranial remains comprise a large portion of left
ilium (Vi 13.7), a section of left proximal radial
This study reports on a newly associated an- diaphysis (Vi 13.8), and a portion of right scapula
teriorcranialvaultandsevenundescribedhominid (13.10).WiththeexceptionsofVi284-230-255-256
fossilsfromVindijaCave(seeTable2).Thenewly and Vi 13.8, provenience is based on the excava-
associated partial vault, Vi 284-230-255-256, con- tionlabelswrittenonthespecimensandshouldbe
sists of four previously described specimens from regarded as tentative (see Table 2). The exact
level G (Wolpoff et al., 1981). When articulated, provenience of Vi 284-230-255-256 is known from
3
these four fossils comprise the most complete excavationwitnesses,whileVi13.8’sattributionto
cranialvaultfromVindija,morecompletethanthe G is based on the characteristic red clay embed-
1
Vi 261-277-278 frontal published by Malez and ded in the fossil. Vi 11.52 and Vi 13.10 are of
colleagues (Malez et al., 1980; see also Smith and particular interest since they tentatively come
Ranyard,1980;Wolpoffetal.,1981).Inhisexami- from unit I, and thus represent the oldest known
nation of the Vindija faunal remains, T. D. White hominid fossils from Vindija.
recognized numerous specimens as possibly homi-
nid. Of these, we describe here four cranial speci- Vi 284-230-255-256: a newly associated partial
mens and three postcranial specimens that are cranial vault
clearly hominid. Two of the new specimens are
small frontal fragments preserving the temporal This newly associated cranial vault specimen
line (Vi 11.47, Vi 11.49). The other two fossils (Figs. 2–4) comprises four previously described
consistofposteriorportionsofthefrontalsquama fossils from level G . Vindija 284, a supraorbital
3
(Vi 11.48) and mandible (Vi 11.52). The post- segment, articulates with the inferomedial frontal
34 J.C.M.Ahernetal./JournalofHumanEvolution46(2004)25–65
squama Vi 255, which articulates with Vi 256, a frontal Vi 227. Very little bilateral frontal bossing
section of superolateral frontal squama. Vi 256 is present, with the squama gradually curving
articulates with the partial parietal Vi 230. These towardthecoronalsuture.Wolpoffandcolleagues
four specimens comprise what is a single partial (Wolpoff et al., 1981) report that Vi 256’s squama
vault and supraorbital region, designated as Vi is thin relative to that in the Krapina sample,
284-230-255-256. Wolpoff et al. (1981) recognized although the difference is not significant. Vi 256’s
the probable association between the frontal frag- thickness (5.4 mm at the lateral eminence) is
ments Vi 255 and Vi 256 with the partial left average for the Vindija sample (Wolpoff et al.,
parietal, Vi 230. Vi 284, a right lateral supra- 1981).
orbital, articulates with Vi 255. If any of the Vi 284 preserves a small section of temporal
articulations appear artificial, that between the notch, running from a point just inferior to fron-
parietal 230 and frontal 256 is most problematic. totemporale to the most posterior point of the
Noclearalignmentcanbemadebetweenthesetwo frontomalar suture. It is preserved posteriorly
along their coronal sutural surfaces. Although all as far as the posterior break, 14.5 mm from
of the specimens differ in their coloration, this is frontomalare temporale, and 26 mm behind the
not out of the ordinary for the Vindija fossils, anteriorsupraorbitalborder(Wolpoffetal.,1981).
given the clear articulation of vastly differently The inferior margin of the temporal line is pre-
coloredspecimensfromthesite(e.g.,Vi275andVi served.Thesurfaceofthefossaisoneofthemore
261).AllofthearticulationsamongVi255,Vi256, rugose of the Vindija specimens. This reflects the
Vi 284, and Vi 230 have been eroded and abraded general gracility of the temporal fossa in Vindija
to varying degrees. Associations among the four sample rather than a particularly rugose temporal
specimens are based on similarities in form and fossa on Vi 284. Compared to Krapina specimens
size, as well as articulation along adjoined pieces. such as Kr 37.6, Kr 27/28, and Kr 37.1, Vi 284’s
The four specimens combine to form the most temporal fossa surface is fairly smooth. The
completeoftheVindijacranialspecimens,andthe frontomalar suture is preserved in its entirety, but
conjoined vault is comparable in preservation to none of the articulation with the sphenoid is
the Krapina frontal specimens Kr 4 and Kr 27/28 preserved. Along with Vi 202, Vi 284 has the
(see Radovcˇic´ et al., 1988; Ahern, 1998). Wolpoff largestfrontomalarsuturewhenmeasuredbyarea.
and colleagues (Wolpoff et al., 1981) gave brief Thisisofinterestsincethespecimen,insomeother
and separate descriptions of Vi 284 and Vi 230- respects (e.g., supraorbital torus morphology, see
255-256, while we provide here a more detailed below), is fairly gracile. When viewed from the
description and analysis of these fossils as the front, the sutural surface appears notched, with a
partial vault of a single individual. distinct lateral overhang (Fig. 2). This overhang is
The supraorbital portion, Vi 284, does not distinct from that seen in Krapina 28, which
preserve any of the frontal squama. Vi 255 is an exhibits a bulbous thickening above and lateral to
inferomedialsectionofsquamaandposteriorsinus frontomalare temporale that is a continuation of
wall,whileVi256isaposterolateralsquamalpiece thetorus.Nosuchbulbousdevelopmentispresent
fromtheleftside.Abroadsagittaltorusispresent onVi284.Theoverhangustandisactuallyangled
in the supraglabellar region. This torus continues slightly more horizontally than the torus at this
superiorly as a slight swelling and is still palpable point.
at the level of the superior break. Like Vi 260, Vi Vi 284 preserves the right lateral part of the
262,andsomerobustmodernhumans,thefrontal specimens, Vi 284 visibly has even less of a supra-
squama is fairly vertical relative to the anterior toral sulcus than the subadult Vi 279. In this
face of the supraorbital torus (Wolpoff et al., manner, Vi 284-230-255-256 differs dramatically
1981). The squama arises from a very shallow from most other Neandertals, whose supraorbital
supratoral sulcus. However, the sulcus is more tori are much more projecting. Thickness at mid-
pronounced and verticality of the squama is sig- orbit cannot be preciesely measured (contra
nificantly less than that exhibited by the juvenile Wolpoffetal.,1981)butitcanbelaterally(Table3).