A new reconstruction of RUD 77, a partial cranium ofDryopithecus brancoi from Rudabánya, Hungary

ANew Reconstruction of RUD 77, a Partial Craniumof Dryopithecus brancoi From Rudabanya, Hungary

LASZLO KORDOS1 AND DAVID R. BEGUN2*1Geological Institute of Hungary, H-1143 Budapest, Hungary2Department of Anthropology, University of Toronto,Toronto, Ontario M5S 3G3, Canada

KEY WORDS Miocene hominids; great apes; cranial restoration

ABSTRACT A newly reconstructed cranium (RUD 77) of the Miocenefossil hominoid Dryopithecus, formerly Rudapithecus (Kretzoi [1969] Symp.Biol. Hung. 9:3–11; Begun and Kordos [1993] J. Hum. Evol. 25:271–286) ispresented here. This specimen, from the late Miocene locality of Rudabanya,in northeastern Hungary, consists of portions of the neurocranium, face, andpostcanine dentition. Newly recovered portions of the parietal, occipital,temporal, zygomatic, and premaxillary bones, which are described here forthe first time, in association with previously described portions of thisspecimen (Kordos [1987] Ann. Hist. Nat. Mus. Natl. Hung. 79:77–88) makeRUD 77 among the most complete and well preserved neurocrania of anyMiocene hominoid. Detailed anatomical descriptions and measurements areprovided here, along with comparisons to other relatively complete Miocenehominoid cranial remains, and to living hominoids. While a more completephylogenetic analysis is in preparation based on the sample as a whole, it issuggested here that RUD 77 provides some additional evidence in support of aprevious hypothesis that Dryopithecus is more closely related to the Africanapes and humans than is Sivapithecus. Am J Phys Anthropol 103:277–294,1997. r 1997 Wiley-Liss, Inc.

Rudabanya is a late Miocene hominoidlocality in Hungary dating from zone MN 9(ca. 10 mya) (Kretzoi, 1975; Mein, 1990,1986; Bernor et al., 1987; Kordos, 1985,1991; Steininger, et al., 1990) from whichfossil hominoids have been recognized since1967. A number of fossil primates have beendescribed from Rudabanya (Kretzoi, 1975,1984). The larger specimens, previouslyattributed to the hominoids Rudapithecushungaricus and Bodvapithecus altipalatus(Kretzoi, 1975), are now attributed to Dryo-pithecus brancoi (Begun and Kordos, 1993).The smaller specimens, previously attrib-uted to Pliopithecus (Anapithecus) hernyaki(Kretzoi, 1975) and Rangwapithecus (Ataxo-pithecus) serus (Kretzoi, 1984), are now at-tributed to Anapithecus hernyaki (Kordos,1988; Begun, 1989; in part contra Begun,1988).

The geomorphology of the primate locali-ties at Rudabanya provides evidence of apeninsula extending into the Pannonianlake, a regional remnant basin of the earliergreater Paratethys sea system (Kordos, 1982,1991). The peninsula, composed of Triassiclimestones and dolomites, was covered withsmall, steep karstic valleys and other karstrelated landforms, most of which probablyformed during the early andmiddleMiocene.These paleovalleys, which include the local-ity from which most of the Rudabanya pri-

Contract grant sponsor: Natural Sciences and EngineeringResearch Council, Canada; contract grant sponsor: HungarianScientific Research Fund (OTKA); contract grant sponsor L.S.Leakey Foundation; contract grant sponsor: Foundation Fyssen.*Correspondence to: David Begun, Department of Anthropol-

ogy, University of Toronto, Toronto, Ontario M5S 3G3 Canada.E-mail: [email protected] 21 January 1995; revised 28 November 1996; ac-

cepted 17 March 1997.

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 103:277–294 (1997)

r 1997 WILEY-LISS, INC.

mates were excavated, locality II, typicallyconsist of alternating sequences of lacu-strine sediments, fine grained clastics (clays,marls, silts), lignites, and paleosols, repre-senting cyclic episodes of infilling and flood-ing of the karst depressions.In 1985 a partial cranium, RUD 77, was

discovered by site geologist G. Hernyak, andlater described (Kordos, 1987, 1991). Thisspecimen is more complete and less dis-torted than the remainder of the collection ofDryopithecus because it was protected by anetwork of roots from a tree found in growthposition in the overlying upper lignite layerfrom locality II. Two cubic meters of sedi-ment around the area from which RUD 77was recovered were collected by L. Kordos.This material was washed, sieved, andsorted, and produced, with the original find,101 cranial fragments and eight upper teeth.The bones are very fragile, with only thethicker cranial elements remaining. Thespecimen is highly fragmented but rela-tively undistorted. In light of new fossildiscoveries and additional information onRUD 77, a new reconstruction was com-pleted recently and is described here.

MATERIALS AND METHODS

Frontal and maxillary fragments, por-tions of the parietal bones, and much of thepostcanine dentitionwere initially assembledand described (Kordos, 1987). Subsequently,additional material from the frontal, tempo-ral, maxillary, sphenoid, and occipital boneswere identified and reconstructed. Conjoin-ing surfaces were identified by fracture con-tours and surface morphology and verifiedunder a stereomicroscope. Pieces were as-sembled with Ferrobond, a fast setting,highly stable glue soluble in acetone. Cra-nial contours were identified by surface mor-phology and maintained with tooth picksuntil a mold could be made. The specimenwas set in Plasticine and molded with DowCorning HS II RTV Silastic silicone follow-ing removal of the tooth picks. Casts weremade in dental stone and high resolutionepoxy resin. Measurements were made onthe original specimen with digital calipersand recorded to the nearest 0.1 mm. Con-tours were taken with a carpenter’s contourgauge. The specimen was compared to fossil

catarrhines from Kenya, Uganda, Turkey,Greece, China, India, Pakistan, Hungary,Austria, Germany, France, and Spain. Statis-tics on these data were calculated usingSigma Plot for Windows. The measurementsof the specimen are given in Table 1.

DESCRIPTIONPreservation

Premaxilla and maxilla. Detached por-tions of the alveolar and palatine processwith palatal surface and portions of thealveoli; a portion of the maxillo-premaxil-lary suture and the incisive foramen; por-tions of the outer bone table of the maxillarybody including the root of the zygomaticprocess and parts of the infratemporal fossa(Fig. 1).

Sphenoid. Detached portions of thegreater wing in the area of the infraorbitalfissure and lateral pterygoid plate.

Temporal. Portions of themandibular fos-sae and associated structures on both sides,the left side including most of the root of thezygomatic process, the superior half of theexternal auditory meatus, and a small partof the anterior portion of the mastoid pro-cess; parts of the petrous pyramid in thearea of the internal auditory meatus (bothsides), andmost of the tympanic tube (Fig. 2).

Occipital. Detached fragments of the nu-chal and occipital planes of the occipital

TABLE 1. Dimensions1 of RUD 77

Interorbital breadth 17.5Nasion-glabella 6.7Glabella-bregma chord 52.8Bregma-inion chord 65.7Glabella-inion chord 109.7Biparietal breadth 78 (min. est.)Nasion-maxilofrontale 9.8Lateral orbital pillar breadthTubercle 12.7Zygomaticofrontale 9.9

Orbital breadth 27.0Orbital height 30 (est.)Biorbital breadth (external) 84.3Biorbital breadth (internal) 73.0Superciliary arch thickness 4.6Articular eminence length 13.8Mandibular fossa-porion 14.1Supraglenoid gutter breadth 8.8

P3-6.6/10.3 RM1-8.8/10.4 LM2-9.8/10.9 RM3-9.2/10.4P4-6.4/10.6 LM1-9.0/— RM2-9.6/— LM3-9.5/10.81 In millimeters. Dental dimensions are length/breadth.

278 L. KORDOS AND D.R. BEGUN

squama including parts from asterion on bothsides; portions of the nuchal crest includingconjoining fragments in the area of inion.

Parietal. Conjoining fragmentsmaking upmost of the left parietal and a large part ofthe right; detached fragments from close tothe coronal and parietotemporal sutures(Figs. 3–5).

Frontal. Largely complete and attachedto the parietals; small parts missing fromthe orbital plates, the right orbital margin,and the squama superiorly and in the infra-temporal fossa.

Zygomatic. Much of the left zygomatic,joined to the frontal, but lacking most of thetemporal process, the infraorbital portion,and small pieces from the maxillary process(Fig. 3–5).

Dentition. Left P3 to M3, right M1 to M3

(Fig. 1). The specimen is housed in thecollections of the Geological Museum of theHungarian Geological Institute, Budapest.

AnatomyPalate and dentition (fig. 1). The speci-men is an adult with well worn dentition

and completely obliterated cranial sutures.The maxilla is highly fragmented, but sev-eral important anatomical features can bediscerned. A portion of the maxillary pala-tine process is preserved attached to a por-tion of the alveolar process, which is itselfattached to the lingual roots of the first andsecond molars. Near the midline it is thickin cross section. In the region of the molars,the alveolar process is shallow lingually. Asmall portion of the palatal process of thepremaxilla preserves part of the incisiveforamen, which was small compared to ex-tant primates with fenestrated palates (pro-simians, monkeys, and hylobatids) and verysimilar in size to that of RUD 12, a D.brancoi female hemipalate (see below). Asmall portion of the right lateral incisoralveolus in the same fragment indicatesrelatively vertically implanted incisorsplaced well medial to the molar row. A smallportion of canine alveolus from the left sideindicates a canine root medial to the mostlateral extent of the root of the P3. The rootof the zygomatic process of the maxilla,preserved on a separate fragment, was thickanteroposteriorly. It is difficult to positionrelative to the dentition, but appears to havebeen high and probably between M1 and M2,judging from a small portion of alveolarprocess surface bone that fits onto the me-sial buccal root of the M1. No other conjoin-ing or identifiable fragments of maxilla arepreserved.All the teeth are high crowned with long,

slender buccal roots and thicker, bilaterallycompressed lingual roots. The premolars areoval and very close to the same size. The P3is slightly longer buccally than lingually, butis not truly triangular, as it is in RUD 44, alarge male of the same species identified assuch from its caninemorphology (see below).The metacone is slightly larger than theprotocone, but the cusps cannot be said to beheteromorphic, as they are in early Miocenehominoids. There is a broad, relatively nar-row mesial fovea separated from a larger,more elongated distal fovea by a subtlecrest, probably a worn medial metacrista.The cusps are placed slightly more mesiallythan distally on the crown (though this ispartly due to wear), resulting in premetacris-tae and preprotocristae that are somewhat

Fig. 1. Occlusal view of the maxilla and dentition ofRUD 77.

279CRANIUM OF DRYOPITHECUS BRANCOI

shorter than the postmetacrista and postpro-tocrista. The postprotocrista is slightly flaredlingually to form the distolingual corner ofan expanded talon. The P4 is slightly longer

with a larger, more elongated talon, andmore centralized cusps (along the buccaland lingual margins). The precrista andpostcrista of both cusps are roughly equal inlength, as are the buccal and lingual mar-

Fig. 2. The temporomandibular fossa and associated structures in RUD 77 in lateral (left) and inferior(right) views. a, entoglenoid process; b, postglenoid process; c, anterior wall of the auditory canal; d,external auditory meatus.

Fig. 3. Lateral views of RUD 77. Note the welldeveloped ridge in the temporal fossa visible on the leftside (bottom photo).

Fig. 4. Superior view of RUD 77.


gins of the tooth, giving it a more rectangu-lar shape than the P3. In other details it isquite similar to the P3. Both premolars havethree roots (two buccal, one lingual).The molars have a simple occlusal mor-

phology. The protocones and hypocones ofboth M1s are worn down to the roots, thoughasymmetrically, with the protocone beingmore worn on the left side and the hypoconemore worn on the right. Only the mesialmost portion of the protocone and the prepro-tocrista remain on the right (nothing re-mains of the left), while only a part of theposthypocrista near the middle of the distaledge of the left M1 remains (the hypocone iscompletely worn away on the right). Thebuccal cusps are also worn, but quite dis-tinct, with only very small dentine pits attheir tips. Among the lingual cusps, themetacone is larger than the paracone, andthere is a well formed buccal notch betweenthe two. On both cusps, the mesial crista areshorter than the distal crista, which contrib-utes to the formation of a very short, narrowmesial fovea (bounded buccally by a short

preparacrista and a very small paraconule)and a more elongated, large distal fovea.The trigon basin, though larger than thefovea, is nevertheless short bucally com-pared to theM2. The transverse crista demar-cating the mesial and distal edges of thetrigon are worn but still well marked.The M2s are also well worn lingually, but

more symmetrically than the M1s. On bothM2s the protocone is gone, but slightly moreof the hypocone is preserved on the left sidethan on the right. Bucally, the paracone isnoticeably larger than the metacone. Thepreparacrista ends at a small, lingually dis-placed paraconule, which is somewhat largerthan on M1 and contributes to the formationof a somewhat more elongated mesial foveathan on the M1. The crista proportions arethe same as on the M1 but all are moreelongatedmesiodistally, resulting in broaderfovea and a broader trigon. The transversecristae are also well marked, though worn.The M3s preserve more occlusal anatomy

than the more anterior molars. Dentine pitsare exposed on the protocones only, with theright one being more worn than the left. ForM3s, their occlusal morphology is relativelysimple, though somewhat more complicatedthan the anterior molars. The protocone andparacone are roughly equal in size, followedby the hypocone, with the metacone beingthe smallest. This gives the teeth a slightlydistolingually tapered shape, though less sothan in RUD 45 and RUD 140, associatedwith the male palate RUD 44. Preparaco-nules and preprotoconules are present, theformer roughly the same size as on the M2s.Transverse cristae are not distinct on eitherM3, despite less wear than on the anteriormolars, resulting in a more or less continu-ous, flat occlusal surface from the mesial todistal ends of the tooth lingual to the apicesof the buccal cusps. This flat surface undu-lates slightly from the presence of indistinctaccessory cristae descending from the buccalcusps. RUD 45 and 140, though less worn,have a similar pattern of a flat occlusalsurface, small accessory cusps, and less dis-tinct crista than the more anterior molarsassociated with the same individual.In all three molars the cusps are peripher-

alized, that is, they are buccolingually com-pressed and their apices occur at the mar-

Fig. 5. Frontal view of RUD 77.


gins of the crowns. In all details, the dentitionof RUD 77 is extremely similar to that ofRUD 12 and other isolated teeth from Ruda-banya attributed to D. brancoi. It is alsovery similar to RUD 44 and associated teeth(RUD 45 and RUD 140), with minor differ-ences in premolar morphology that are prob-ably related to sexual dimorphism (see be-low).The anterior molars of RUD 77 have a

distinctive pattern of occlusal wear com-pared to other fossil hominoid taxa, but onethat is reminiscent of other specimens fromRudabanya. As noted above, the M1 and M2

are strongly worn lingually, with wear ex-tending well up the roots. This pattern issimilar to that seen in RUD 6, the only otherDryopithecus molar from the site exhibitinga similar degree of wear. One possible expla-nation for this pattern of extreme but local-ized wear is damage to the crown surfacerelated to caries. This is an intriguing possi-bility in light of the fact that another indi-vidual, RUD 44, has large bilateral crowncaries on the P3s and a smaller caries involv-ing the distal margin of the M2 and themesial margin of the M3 (RUD 45). We knowof no other Miocene hominoid sample withthis frequency of caries or potential caries,nor have caries been identified in any teethof Dryopithecus other than those attributedto Dryopithecus from Rudabanya. If thesewear features are related to caries forma-tion, this may have interesting implicationsfor reconstructing certain aspects of the diet,ecology, and health of D. brancoi. Finally,the pattern of dentine exposure in RUD 77suggests that the enamel was relatively thinand dentine penetrance (the degree to whichthe dentine horns penetrate the enamel cap)relatively high, compared to formswith thickenamel and low penetrance (Sivapithecus,Ouranopithecus, Australopithecus).

Neurocranium. The temporal bones arehighly fragmented, but large articular por-tions of both temporal bones, most of the lefttympanic plate, and fragments of both pe-trous pyramids provide a number of impor-tant anatomical details (Figs. 2, 6, 7). Themandibular fossa, preserved on the left side,is more strongly concave anteroposteriorlythan mediolaterally (Fig. 2). It is anteropos-

teriorly constricted, and bordered anteriorlyby a broad, saddle-shaped (transversely con-cave and anteroposteriorly convex) articulareminence, and posteriorly by a small, poste-riorly inclined postarticular eminence (Fig.2). On the right side, a very large, broad,inferiorly directed entoglenoid process ispreserved, missing only the tip. On the leftside the entoglenoid process is more seri-ously damaged, exposing extensive pneuma-tization of this process. On the right side abroad, well formed articular tubercle projectsinferiorly from the base of the root of thezygomatic process (Fig. 2). A separate frag-ment originally posterior and lateral to thearticular fragment includes the large, semi-circular superior portion of the externalauditory meatus, posterior to the postgle-noid process, and posterior to that the ante-rior end of the mastoid crest. With these twofragments conjoined, it is possible to recon-struct a broad supraglenoid gutter with alaterally flared lateral margin superior tothe mandibular fossa (Fig. 2).Damage in the area posterior and medial

to the postglenoid process suggests thatthere was fusion of the articular and tym-panic portions of the temporal. The tympano-squamosal suture, which would be present ifthese two portions were fused, is not visible,and is replaced by a broken bone surfaceexposing pneumatization deep to the postgle-noid process. The breakage extends somedistance medial to the postglenoid process.More medially, between the postglenoid pro-cess and the entoglenoid process, the ante-rior face of the petrotympanic fissure ispreserved intact. If the articular and tym-panic portions were not fused, the tympano-squamosal suture would likely have beenpreserved more laterally, to a point distal tothe postglenoid process. The tympanosqua-mosal suture is typically not preserved whenthe tympanic and articular portions sepa-rate in taxa in which the suture fuses, but istypically present in taxa in which the twoportions do not fuse, as in Sivapithecus andPongo among the great apes (Ward andBrown, 1986). Finally, damage on the endo-cranial side of the articular and petrousportions of the temporal indicate extensivepneumatization of the articular eminence,


the root of the zygomatic process, and thepostglenoid process.Two small fragments preserve parts of the

endocranial and internal surfaces of thepetrous pyramids in the region around theinternal auditory meatus and the fenestra(Fig. 6), and a third preserves the basicra-nial surface of the left tympanic plate and aportion of the wall of the tympanic cavity(Fig. 7). The left dorsal surface preserves alarge internal auditory meatus, posterior,superior, and lateral to which there is ashallow subarcuate fossa (Fig. 6a). Posteriorand lateral to the meatus there is a promi-nence that probably accommodates the pos-terior semicircular canal within the bone.Inferior to this prominence there is a smallnotch corresponding to the aqueduct of thevestibule (Fig. 6a). Posterior and lateral tothe prominence and the aqueduct is a con-cave surface representing a small portion ofthe cerebellar surface (Fig. 6a). Superiorand lateral to the prominence and the cer-ebellar surface is a shallow groove runningposterolaterally that probably represents aportion of the superior petrosal sinus. Thedorsal surface on the right side is less com-pletely preserved, and is essentially a mir-ror image of the homologous anatomicalfeatures described for the left side. On bothspecimens, the distance between the inter-nal auditory meatus and the cerebellar sur-face is relatively short, suggesting that thecomplete petrous pyramid was mediolater-ally compact (Fig. 6a).The internal aspects of two pieces of the

petrous pyramid just described preserve de-tails of the morphology of a portion of thewall of the tympanic cavity (Fig. 6b). Bothspecimens preserve the cochlear and vestibu-lar fenestrae, the promontory, the medialhalf of the canal for the facial nerve, andpart of the posterior wall of the carotid canal(Fig. 6b). Posterolaterally, both specimenspreserve evidence of mastoid pneumatiza-tion. CT scans reveal a typical great apemorphology of the bony labyrinth (Spoor,1996).The last temporal fragment preserves a

few details of the left tympanic plate (Fig.7a). Anteriomedially a moderate eustachianprocess is preserved. Lateral and slightlyposterior to it is the posterior surface of the

petrotympanic fissure. Most laterally is adamaged, pneumatized region that corre-sponds to the area posterior to the postgle-noid process in the vicinity of the tympano-squamosal suture (Fig. 7a). The medialextent of the tympanic plate is preservedmostly intact (Fig. 7a). It is relatively broadand flat, with a concavity more pronouncedmedially than laterally. The concavity sepa-rates the eustachian process from the raisedmedial end of the tympanic crest, whichrises anterior andmedial to a well developedstyloid pit. The tympanic crest is well devel-oped, and courses posterolaterally (Fig. 7a).Posterior to the lateral most extent of thetympanic crest is the anteromedial half ofthe stylomastoid foramen. Judging from theorientation of the petrotympanic fissure sur-face, which normally faces anteriorly tomeetthe tegmen tympani, the tympanic plate inRUD 77 must have been oriented morevertically than horizontally (Fig. 7a).A portion of the tympanic wall contribu-

tion of the tympanic plate is also preserved

Fig. 6. Drawings of the endocranial (A) and lateral(B) views of the left petrous pyramid of RUD 77.Approximately six times natural size.


on this specimen, deep to the basicranialsurface described above (Fig. 7b). The mostprominent feature in this view is the tym-panic sulcus, which is very well preserved(Fig. 7b). Lateral to it is a broken, pneuma-tized area that corresponds to the base of theexternal acoustic meatus. Medially, a smallgroove marks the probably position of theauditory canal.The orbital and neurocranial regions are

well preserved (Figs. 3–5). The nearly com-plete supraorbital region is preserved fromvery close to the frontozygomatic suture onthe right to the left frontozygomatic suture,to which is attached most of the left zygo-matic bone. The most prominent feature ofthe supraorbital region is glabella, which ispreserved intact at nasion (the nasal bones

are missing). At glabella, the interorbitalspace is convex transversely and superoinfe-riorly, contributing to a mild projection ofglabella (Figs. 3, 5). Superior to glabellathere is a shallow but distinct transversedepression separating the orbital portion ofthe frontal bone in the midline from thesquamous portion (Fig. 3). Continuous fromglabella are short, low, rounded ridges, ori-ented posterolaterally. The left ridge, whichis more distinct that the right, is no longerpalpable beyond about 25% of the distanceacross the breadth of the orbit. At this pointthis ridge is separated from the orbitalmargin by a shallow groove running parallelto the ridge. The superior orbital margins onboth sides are formed from thickened, hori-zontally oriented ridges. The superior sur-faces of these ridges face superoanteriorly,and are separated from the anterior tempo-ral lines by a shallow but distinct groove.Based on comparisons to other primates, webelieve that the mild ridges connected toglabella are supraorbital tori, and the thick-ened ridges above each orbit are supracili-ary arches. The groove superior to glabellaand the ridges is the supratoral sulcus.Comparisons of these structures to theirhomologues in other primates are providedin Begun (1994, 1995).The supraciliary arches, which are supero-

inferiorly compressed and anteriorly project-ing, merge bilaterally with the surface of thezygomatic process of the frontal close to thefrontozygomatic sutures. In this region oneach side the superior surfaces face moresuperiorly. These surfaces, between the lat-eral third or so of the orbital margins andthe anterior temporal lines, are roughenedby transversely oriented grooves, which fadelaterally and disappear at the frontozygo-matic suture.The lateral orbital surface continues infe-

rior to the frontozygomatic suture, beingpreserved on the left side on the zygomaticbone (Fig. 5). The margin is thicker at thefrontozygomatic suture than more inferi-orly, but it does not form a sharp edge as inmany primates. A bend in the margin about1 cm superior to the level of jugale and justabove the level of the most superior zygo-maticofacial foramen marks the beginningthe medial curvature of the orbit toward the

Fig. 7. Drawings of the basicranial (A) and tympanic(B) surfaces of the left tympanic plate of RUD 77.Approximately six times natural size.


maxilla. Not enough of the inferior portion ofthe orbit is preserved, however, to confi-dently reconstruct the shape of the orbit.The lateral orbital margin is vertical supe-rior to this point, which is more typical ofprimates with orbits that are broader thantall.The anterior temporal lines are strongly

developed, especially for a hominoid of thissize (Fig. 4). From the frontozygomatic tu-bercle they run a nearly straight courseposteromedially toward the midline. Abouthalf way across the orbits, they turn sharplyposteriorly. At this point, they change fromraised, prominent ridges marking the ante-rior and superior boundaries of the infratem-poral fossa, to rounded, more subtly definedridges on the distal portions of the frontalsquama and the parietal bones (Fig. 4). Attheir closest to each other, which is slightlyposterior to midcranial length, the temporallines are about 9.8 mm apart. The frontaltrigone, bounded bilaterally by the anteriortemporal lines and anteriorly by the supraor-bital ridges, is biconvex, but more curvedanteroposteriorly than transversely. The in-terorbital space is broad with small airsinuses inferior to nasion. In frontal view itis mildly hourglass shaped, being broader atthe superior orbital margins and at the mostinferiorly preserved end than between thetwo, suggesting that the orbital space wouldhave become broader distally. Two distinctsurfaces are apparent on the nasal processesof the frontal bones. The orbital surfacesface laterally into the orbits. The facialsurface, between the rounded orbital mar-gins and the nasal sutures, face anteriorlyand laterally. These are slightly concavetransversely, and probably mark the supe-rior most portion of the lacrimal grooves.The frontonasal suture is V shaped and

strongly acutely angled. The bone is about 8mm thick between the surface for the nasalbones on the frontal and the endocranialsurface of the frontal, and is not extensivelypneumatized. However, bilaterally to thecentral portion of the nasal suture are welldeveloped sinuses that excavate the interor-bital region from the point at which it ispreserved to the level of nasion. The platesof bone that remain, on the facial and endo-cranial sides, are thin. The left side, which is

slightly more complete than the right, pre-serves two small loculi separated by a ridge.A frontal radiograph of RUD 77 (Fig. 8)reveals the low position and minimal exten-sion of the frontal sinuses.The frontozygomatic suture, preserved on

the left, is oriented nearly vertically. At thefrontozygomatic suture the anterior tempo-ral lines are thickened into well definedtubercles. From the frontozygomatic tu-bercles, in addition to the strongly anteriortemporal lines, strongly developed crestsrun inferiorly and posteriorly (Figs. 3 and 8).On the left side, where more of the infratem-poral fossa is preserved, this crest forms astrong buttress, triangular in cross section,with inferolateral and superomedial facesseparated by a ridge. This buttress is alsovisible in the radiograph (Fig. 8). The ridgeand the superomedial face are roughened forthe attachment of the anterior fibers oftemporalis. This prominent buttress sepa-rates the infratemporal fossa into two sec-tions. The superomedial portion is relativelybroad and shallow. The inferolateral portionis very deep and narrower, forming a largesulcus oriented vertically and slightly medi-ally, and bounded laterally by the temporalprocess of the zygomatic bone (Figs. 3 and 4).

Fig. 8. Radiograph in frontal view of RUD 77. Openarrows indicate the small frontal sinuses placed inferiorto nasion (N). Also note the buttressed lateral orbitalpillar running inferiorly and slightly medially from thefrontozygomatic tubercle (closed arrows). See text.


The lateral orbital pillar (frontal processof the zygomatic bone) is broad and thefacial surface convex and oriented anterolat-erally. The body of the zygomatic bone ispierced by at least four very small zygomati-cofacial foramina. It is convex superiorly butconcave inferiorly, indicating the presence ofa depression lateral to the zygomaticomaxil-lary suture. A small portion of the temporalprocess of the zygomatic is preserved, arch-ing posteriorly with little lateral flare, andenclosing a deep, broad sulcus for anteriorand lateral portion of temporalis. Jugale isjust below the level of the inferior orbitalmargin.The parietals have low, rounded, but

clearlymarked superior temporal lines sepa-rated by a roughened bare area midsagitally(Fig. 4). The coronal contour is broad and thepostorbital constriction appears to have beenrelatively slight. The infratemporal surfacewas broad anteroposteriorly. The parietalsare mostly preserved close to the midline,but enough is preserved laterally and inferi-orly to say that the cranium was fairlybroad, probably with a biparietal breadththat was greater than the biorbital breadth.In this reconstruction, 20 to 21 mm aremissing along the midsagittal chord be-tween the most posterior edge of the mainset of conjoined fragments and a smaller setof conjoined fragments with inion and aportion of the nuchal torus. This smallersection can be positioned with confidence,however, using three anatomical features(inion, the nuchal torus, and the superiortemporal line), and by aligning the ectocra-nial contours (Figs. 3 and 4). It is, in fact,fairly clear how this smaller, posterior piecemust be placed. The anterior end of thetemporal line of the posterior piece can bealigned with the posterior end of the tempo-ral line of the anterior fragment. There is aslightly medialward convexity to this line onboth sections, so that in superior view theycan be aligned to create a smooth contour tothe temporal line. In posterior view, the endsof the temporal lines serve to position thefragments relative to one another, and boththe plane of the bare area between thetemporal lines and the nuchal crest serve toposition the smaller fragment relative to theanteroposterior and transverse axes. In the

final analysis, though the two pieces are notconnected, the presence of so many clearlymarked anatomical features makes it pos-sible to fit these two pieces of neurocraniumwith confidence (Figs. 3, 4).The angle between the occipital and nu-

chal planes is relatively wide. The nuchaltorus is not strongly developed, and is dam-aged at inion. A small suprainiac fossa indi-cates the position of inion, and correspondswith the midsagittal contour, reinforcing theplacement of anterior and posterior cranialfragments. The torus, which is better pre-served on the left side, is slightly raisedsuperior to the occipital surface, and projectsmore strongly posterior to the nuchal sur-face. Only about 10 mm of the length ispreserved, but the nuchal torus does archwith a mild anterosuperiorly oriented con-vexity. The bare area is relatively broadposteriorly, and marked by fine pitting andstriations that distinguish it from the por-tions of the parietal bones originally deep tothe deep fascia of temporalis (Fig. 4).Without the inferior orbital margins or a

completely secure basis for positioning thetemporal fragments on the neurocranial andfacial portions of the specimen, it is difficultto orient this specimen in lateral view. How-ever, a few nontraditional characters givesome indication of a possible orientation ofthis specimen. In most catarrhines the or-bital margins are vertical, or they are curvedsymmetrically relative to a vertical axis, inlateral view. The anterior most parts of thesuperior orbital plates are also usually ori-ented nearly horizontally. While there issome error in positioning RUD 77 based onthese landmarks, an approximation can bemade. On this basis, it is unlikely that thefrontal squama was as vertically oriented asin humans or orangs, or as horizontallyoriented as in African apes or australopithe-cines. Posteriorly, the neurocranium curvedinferiorly toward inion, and the nuchal planewas oriented posterosuperiorly. The neuro-cranium overall has a long superior cranialcontour. Figure 9 presents a comparison ofcranial lengths in some hominoids relativeto orbital size, a good correlate of overallbody size in anthropoids (Aiello and Wood,1994). RUD 77 is indistinguishable fromAfrican apes while Pongo and Hylobates


have shorter crania, as does Proconsul. En-docranially, the frontal poles are well pre-served, narrow, and inferiorly projecting.Posteriorly, small portions of the occipitallobe and cerebellar impressions are pre-served either side of midline.A sharp frontalcrest is preserved, continuous with a wellmarked superior sagittal sulcus.

COMPARISONS TO OTHER FOSSILHOMINOIDS

A less complete reconstruction of RUD 77has been compared previously to other fossiland living hominoids (Kordos, 1987, 1991).Here we focus on comparisons based on thenewly reconstructed portions of the speci-men, and to specimens not previously com-pared to RUD 77. These comparisons aredirected at establishing the sex of RUD 77and the basis for its attribution to D. bran-coi. A more comprehensive analysis includ-ing details of the functional anatomy andthe phylogenetic relations of D. brancoi willbe based on the entire sample from Ruda-banya, and is in preparation.On the basis of similarities in the denti-

tion of RUD 77 to other specimens from

Rudabanya, this specimen can be attributedto D. brancoi, previously identified at Ruda-banya on the basis of a large sample ofspecimens (Begun and Kordos, 1993). Theteeth are morphologically very similar to allupper molars and premolars attributed toDryopithecus from Rudabanya. They arevery close in size to the postcanine dentitionof the female hemipalate RUD 12 attributedto a female on the basis of its unambiguouslyfemale canine morphology (Begun, 1987,1992a; Kelley, 1995) (Table 2). The postca-nine dentition of RUD 77 is slightly smallerin most dimensions than that of RUD 15,which is also attributed to a female based oncanine morphology (Begun, 1987) (Table 2).RUD 77 is similar to but smaller and moregracile than RUD 44, another partial cra-nium from Rudabanya attributed to a maleof the same species, based on canine mor-phology (Andrews and Martin, 1987; Begun,1987; Kordos, 1991). There is no evidence ofany other great ape taxon at Rudabanya.Despite the absence of a canine, there can

be little doubt that RUD 77 is a female. Asnoted above, a small portion of the alveolusof the canine is present indicating a medialposition of the canine root relative to the P3,as is typical of the small canines of femaleanthropoids. As noted earlier, the P3, inbeingmore rectangular rather than triangu-lar, is morphologically more similar to twoknown females (based on the associatedcanines), and unlike either the male fromRudabanya with an associated canine, orother males of Dryopithecus with associatedcanines and P3s (D. laietanus and D. crusa-fonti [see below]). The small dental size ofRUD 77 is also a strong indicator of the sexof this specimen (Table 2). In primates inwhich evidence of postcanine dimorphism islacking, size cannot be considered a reliableindicator of sex. However, at Rudabanya thesample consists of a number of males andfemales that are identifiable as such on thebasis of preserved, associated canines. RUD77 very clearly falls among the females bothin morphology and size (Table 2). The denti-tion of RUD 77 is the same size or smallerthan a number of specimens attributed tofemales (RUD 12 and 15), and much smallerthan specimens attributed to males (RUD 7and 44) (Table 2).

Fig. 9. Box plot of selected hominoid cranial lengths(glabella-inion chord) over orbital size [(orbitalbreadth 1 orbital height)/2]. Boxes enclose 25th and75th percentiles, and vertical lines represent the medi-ans. Horizontal lines represent the 10th and 90th percen-tiles, and open circles the 5th and 95th percentiles. Notethe long cranium of RUD 77 and the short cranium ofPongo, Hylobates and Proconsul. See text for discussion.


Compared to other cranial material ofDryopithecus, RUD 77 differs from amaleD.laietanus from Can Llobateres, CLl 18000(Begun and Moya-Sola, 1992; Moya-Sola etal., 1992; Moya-Sola and Kohler, 1993, 1995;Begun, 1994, 1995) in having a convex fron-tal trigone and glabella, a narrower, moredeeply excavated zygomatic temporal sur-face, less well developed anterior temporallines, slightly less extensive frontal sinus, ahigher nasion relative to the superior orbitalmargin, a flat interorbital surface betweennasion and glabella, more superiorly facingfrontal zygomatic process, a more convexmalar surface, a slightly lower zygomaticroot, and smaller zygomaticofacial foramina.Dentally the two specimens are very similar,though in RUD 77 the P3 is more symmetri-cal and closer in size and shape to P4, andthe M2 is largest in RUD 77, while in CLl18000 the M3 is the largest tooth. Moredetailed morphological differences in dentalmorphology between D. brancoi and both D.laietanus and D. crusafonti have been pub-lished elsewhere (Begun, 1992b). Both speci-mens are close in size, dentally and cranially(interorbital breadth, orbital breadth, or-bital margin thickness, etc.), despite the factthat the Can Llobateres specimen is a male(based on canine morphology) and RUD 77 afemale (see above). The upper dentition ofD.crusafonti is larger than that ofD. laietanus,and closer in size to that of male D. brancoi(RUD 44) (Begun, 1992b).A number of the differences noted above

between RUD 77 and CLl 18000 are notdiagnostic of species differences between D.brancoi andD. laietanus. There is overlap inmorphology when other specimens attrib-uted to D. laietanus and D. brancoi areincluded in the comparisons. Premolar het-

eromorphy, which distinguishes CLl 18000from RUD 77, is also found in the male RUD44, and is probably associated with sexualdimorphism rather than taxonomy (seeabove). RUD 44 is also closer to CLl 18000than to RUD 77 in anterior temporal linedevelopment, palatal depth, and probablyfrontal zygomatic process orientation, sug-gesting that these differences are also amatter of sexual dimorphism and not oftaxonomy. In contrast, in the individualrepresented by RUD 44, M2 was the largesttooth, like RUD 77 but unlike CLl 18000.RUD 44 is associated with an isolated M3,RUD 140, that is smaller than the M2 ofRUD 44 (Kordos and Begun, submitted, a).The mesial interstitial facet of RUD 140matches the distal facet on theM2 of RUD 44perfectly (Kordos and Begun, submitted, a).M2/M3 size ratio may in fact be a differencefrom D. laietanus of taxonomic significance.Other potential species specific differ-

ences also include zygomatic root height,malar surface flattening (intermediate inRUD 44), nasion position, glabellar and fron-tal trigone morphology, and overall size.Zygomatic root height is difficult to assess inD. brancoi because most specimens preservethe zygomatic process of the maxilla onlyvery close to the origin. Its position on themaxilla is clearly relatively high, as in mod-ern great apes, and not very close to thealveolarmargin, as inmost early andmiddleMiocene forms. As measured on the singlespecimen on which it is reasonably wellpreserved, the zygomatic root height is lowerthan in CLl 18000, but both are well withinthe ranges of extant great apes (Begun,1994, Fig. 9). The difference may be speciesspecific, though the small sample size pre-cludes a definitive judgment.Much the sameis true for the difference in malar surfacemorphology. A diversity of malar surfaceshapes from flat to more convex can be foundin Pan and Pongo, so that the differencebetween RUD 77 and CLl 18000 may not besignificant. As noted above, D. laietanusmales, identified on the basis of caninemorphology, are dentally smaller than D.brancoi males. While there are no femalecanines found in association with other teethin D. laietanus, there are many very smallpremolars andmolars ofD. laietanus, includ-

TABLE 2. Dental size1 comparisons among male andfemale Dryopithecus brancoi and RUD 77

P3 P4 M1 M2 M3

D. brancoimalesRUD 44 9.7 9.0 10.2 11.7 10.6RUD 7 — 9.4 11.0 — —

D. brancoi femalesRUD 12 7.9 8.5 9.9 — —RUD 15 9.0 8.6 10.1 10.5 —RUD 77 8.4 8.5 9.6 10.4 9.4

1 (Length 1 breadth)/2, in millimeters.


ing those of the type specimen, which aresmaller than any D. brancoi postcaninetooth. There is thus abundant evidence of asmall form of D. laietanus, significantlysmaller than males of that taxon or thanmales and females of D. brancoi, and almostcertainly attributable to females.The morphology of RUD 77 conforms to a

pattern of morphological diversity thatserves to distinguish among the currentlyrecognized species of Dryopithecus. Thoughthe sample sizes are small, this morphologi-cal patterning and a corresponding tempo-ral and geographic clustering of Dryopithe-cus localities support the view thatD. brancoiis distinct from other species ofDryopithecus(Begun and Kordos, 1993).RUD 77 has been compared to other homi-

noids elsewhere (Kordos, 1987, 1988, 1991;Begun, 1994, 1995; Begun andKordos, 1997).The dental sample of Dryopithecus has alsobeen compared extensively to other homi-noids. The dentition of RUD 77 is morpho-logically quite consistent with other Dryopi-thecus fromRudabanya, and, with its heavilyworn molars, contributes relatively few newcomparative data. Thus, we focus on newcomparisons based on the newly identifiedportions of this specimen described here,mainly from the posterior neurocranial andtemporal regions. We focus primarily oncomparisons to other taxa known from theseanatomical regions, which excludes a largenumber of otherwise reasonably well knownfossil hominoids. The entire sample of D.brancoi will be compared to all known fossilhominoids elsewhere (Kordos and Begun, inpreparation).The most complete Miocene hominoid cra-

nial specimen is the female skull of Procon-sul (KNM-RU 7290) (Le Gros Clark andLeakey, 1951; Walker et al., 1983; Walkerand Pickford, 1983). Elsewhere, compari-sons have been made between Dryopithecusin general and Proconsul, and these areconsistent with the comparison of RUD 77and KNM RU 7290 (Kordos, 1987, 1991;Begun, 1994, 1995). For example, the den-tal, maxillary, and frontal morphology ofthese two genera have been compared indetail, and it is not necessary to repeat thesecomparisons here. However, it is now pos-sible to expand these comparisons to include

details of occipital, temporal, and overallneurocranial morphology.Compared to KNM RU 7290, RUD 77 has

a less strongly marked external occipitalprotuberance, a more posteriorly orientednuchal plane, and an anteroposteriorlybroader temporal fossa. The neurocraniumis absolutely longer and broader and moreelongated relative to breadth in RUD 77(Fig. 9). The orbits and the teeth of RUD 77are slightly larger than in KNM-RU 7290,which indicates that the body size of RUD 77was very likely to have been somewhatgreater than that of KNM-RU 7290. Theneurocranial dimensions of RUD 77, how-ever, are considerably larger, indicating thatthe relative brain size of RUD 77 is largerthan that of KNM-RU 7290 (Kordos andBegun, submitted, b).The articular and petrous portions of the

temporal bones of KNM RU 7290 are notpreserved as well as those of RUD 77 butsome important differences are neverthelessapparent. KNM-RU 7290 has a very largesubarcuate fossa (Ward, personal communi-cation), as is the case of most other primatesother than great apes and Dryopithecus(Moya-Sola and Kohler, 1995; Begun, 1995).A few interesting exceptions to this patternexist (Spoor and Leakey, 1996), but these areprobably due to the parallel occurrence of areduction in the size of the subarcuate fossain these forms and in hominids (Spoor andLeakey, 1996). The glenoid fossa in KNM-RU7290 is smaller and anteroposteriorly morenarrow than in RUD 77, which is consistentoverall with its smaller size. The postglenoidand entoglenoid processes are also smallerand less projecting in KNM-RU 7290. Thesedifferences, although consistent with thesize difference, are also consistent with otherdifferences in cranial robusticity betweenthe two specimens, suggesting functionaldifferences in dietary adaptations as well.RUD 77 has much more strongly developedcrests for the temporalis muscle periorbit-ally and neurocranially than is the case inKNM-RU 7290, in which these crests arepoorly developed. All of these differencessuggest a more strongly developed mastica-tory apparatus in RUD 77 compared toKNM-RU 7290. That size may not have


much to do with it is suggested by thepresence of strongly developed ectocranialcresting similar to that of RUD 77 in thecraniumofTurkanapithecus (KNM-WK16950),which is actually somewhat smaller thanKNM-RU 7290. KNM-WK 16950 also has anassociated mandible with a more robustcorpus and an anteroposteriorly larger con-dyle than KNM-RU 7290, suggestive of amore powerful chewing apparatus. It shouldbe noted, however, that KNM-WK 16950,like RUD 77, has a more strongly worndentition, which probably indicates olderindividuals than KNM-RU 7290, in whichthe M3s appear to have just erupted, andthis may explain some but not all of thedifferences between the KNM-RU 7290 onthe one hand and KNMWK 16950 and RUD77 on the other.ContinuingwithTurkanapithecus, the new

data of RUD 77 reveal a number of addi-tional characteristics to add to a long list ofdifferences between Turkanapithecus andDryopithecus. In contrast to the compari-sons made between Dryopithecus and Pro-consul, fewer detailed comparisons have beenmade to Turkanapithecus. A comprehensivelist of differences between the two taxa isthus provided here. RUD 77 differs from theprobably male specimen of Turkanapithe-cus, KNM-WK 16950 (Leakey et al., 1988a)in being larger and in having a longer andrelatively larger P4 compared to P3, no lin-gual cingulum, no mesiobuccal cingulum,less P3 cusp heteromorphy, lower P3 para-cone in buccal view, with mesial and distalcrests of equal length and diverging from thecusp aspices at a greater angle, relativelylarger M1 compared to M2, reduced meta-cones relative to paracones on M2–3, buccallyconcave postcanine tooth rows, larger orbitswith less anteriorly projecting superior or-bital margins, smaller interorbital breadth,better developed supraorbital tori and sul-cus, more vertical frontal squama, moreprojecting glabella, less postorbital constric-tion, a longer neurocranium relative tobreadth, a smaller articular tubercle, and amore anteroposteriorly convex articular emi-nence. RUD 77 also lacks the distinctivewear pattern of the molars of KNM-WK16950, which produces a complicated net-work of crests on the entire occlusal surface

rather than the deep notches confined to thelingual cusps of RUD 77.RUD 77 differs from an Afropithecusmale

(KNM-WK 16840) (Leakey et al., 1988b) inmany of the same dental and cranial fea-tures as Proconsul. Kordos (1991) and Be-gun (1994) include some comparisons offrontal morphology between RUD 77 andKNM-WK 16840. Little has been publishedon comparisons of other aspects of the faceor the teeth. RUD 77 has longer premolarsand molars, P4 longer than P3, relativelylarger M1 compared to M2, reduced cingula,less P3 cusp heteromorphy, lower P3 para-cone in buccal view, with mesial and distalcrests of equal length and diverging from thecusp apex at a greater angle, reduced meta-cones relative to paracones onM2–3, a smallerincisive foramen, shallower palate in themolar region, more superiorly placed zygo-matic root, and a thicker lateral orbitalpillar than KNM-WK 16840. RUD 77 isfurther distinguished in lacking the lowcrowned, low cusped, strongly flared postca-nine teeth of KNM-WK 16840, and in beingsmaller in overall facial dimensions withlessmesiodistally compressedmolar trigons,relatively larger orbits, narrower interor-bital breadth, less strongly developed tempo-ral lines that do not meet to form a sagittalcrest, a broad, convex frontal trigone, andrelatively and absolutely less postorbitalconstriction. KNM-WK 16840 is intermedi-ate in cingulum development between Pro-consul and RUD 77, while another specimenattributed toAfropithecus (KNM-WK17102)lacks cingula on the M2–3, as in RUD 77.RUD 77 differs from male Sivapithecus

(GSP 15000) (Pilbeam, 1982; Ward andBrown, 1986) in a large number of charac-ters described in several publications (Kor-dos, 1987, 1988, 1991; Begun, 1994). Despitethe substantial differences in cranial anddental morphology, RUD 77 and GSP 15000are similar in premolar and molar propor-tions and in lackingmolar cingula. Both alsoshare a number of characters of the facefound in other hominids (Kordos, 1988, 1991,Begun, 1994).RUD 77 has also been compared exten-

sively to Ouranopithecus male (XIR-1) (deBonis et al., 1990; Kordos, 1991; Begun,1994, 1995). As noted in Begun (1995), RUD


77 and XIR-1 differ primarily in charactersrelated to size, but are otherwise similar indental proportions and facial morphology, asare the samples ofOuranopithecus andDryo-pithecus more generally. These taxa areprobably more closely related to each otherthan they are to other hominids (Begun,1995; Begun and Kordos, 1997).RUD 77 has also been compared in detail

to a female Lufengpithecus (PA 677) (Wu etal., 1986; Kordos, 1988; Schwartz, 1990). Aswith Sivapithecus and Ouranopithecus, thedifferences from RUD 77 overall are lessthan with early and middle Miocene forms.RUD 77 has not been compared in detail

to Oreopithecus. The dentition of Oreopithe-cus is well known (Hurzeler, 1951; Harrison,1986), and quite distinct overall in crownand occlusalmorphology fromRUD77. How-ever, there are some similarities. Like RUD77, Oreopithecus has relatively more elon-gated molars and premolars than earlyMiocene hominoids, but has more stronglydeveloped molar cingula than RUD 77.M1/M2 proportions are similar to RUD 77,but M3 is the largest molar in Oreopithecus,whereas M2 is largest in RUD 77. The root ofthe zygomatic process of the maxilla inOreopithecus (IGF 11778) is lower on thealveolar process than it seems to have beenin RUD 77 or other specimens from Ruda-banya. The interorbital space lacks a frontalsinus, and the glabella is not projecting, as itdoes, albeit mildly, in RUD 77. In Oreopithe-cus the lateral orbital pillars are narrowerand more rounded transversely. There aretwo small zygomaticofacial foramina supe-rior to the inferior orbital margin, as in RUD77. The anterior temporal lines, which aremore strongly developed than in RUD 77,converge more strongly toward the midlineand do so more anteriorly on the frontal inIGF 11778 than on RUD 77. The anteriortemporal lines probably met very far anteri-orly on this skull, forming a sagittal crestwith a very anterior origin. No sagittal crestis present on RUD 77. There are also indica-tions on this specimen of a very deep tempo-ral fossa.Jugale and the root of the temporal pro-

cess of the zygomatic bone are also pre-served inOreopithecus, and differ from RUD77. In Oreopithecus the temporal process is

deep superoinferiorly, and strongly concavemedially, suggesting a relatively morestrongly developed masseter muscle thanthat of RUD 77. The notch at jugale is moreacute, and the temporal process runs moreposterosuperiorly than on RUD 77. Jugale isalso lower on the face inOreopithecus.The neurocranium is severely crushed on

IGF 11778 but a number of comparisons cannevertheless be made with RUD 77. A welldeveloped sagittal crest is apparent, con-nected posteriorly to a very strongly devel-oped nuchal crest, forming together a com-pound temporonuchal crest surrounding verydeep concavities on the occipital and pari-etal bones for the temporalis muscle. Thisextreme ectocranial cresting is very differ-ent from RUD 77, and is also much morestrongly developed than on male specimensof Dryopithecus (CLl 18000, RUD 44). To-gether with indications of strong postorbitalconstriction, this morphology implies a rela-tively much smaller endocranial cavity onIGF 11778 as well.Portions of the articular and basicranial

aspects of the temporal bones are also pre-served on IGF 11778. Like RUD 77 there is abroad articular eminence and a stronglydeveloped, projecting entoglenoid process.Although distorted, the mandibular fossaappears to have been somewhat deeper thanthat of RUD 77, and the articular tuberclemore prominent. Unlike RUD 77, the poste-rior surface of the postglenoid process ispreserved intact, indicating that there wasno fusion of the articular and tympanicportions of the temporal bone in this region.The comparison of RUD 77 to IGF 11778 is

complicated to some extent by the fact thatRUD 77 is a female and IGF 11778 a male.IGF 11778 was probably a bigger individualthan RUD 77, judging from the size of thedentition and the associated mandible, butwas probably not larger thanRUD44.Wherethe same comparisons are possible, most ofthe differences between RUD 77 and IGF11778 also distinguish RUD 44 from IGF11778. While size may have played somerole in the differences between RUD 77 andIGF 11778, most of the differences are re-lated to fundamental dissimilarities in over-all cranial morphology between the Dryopi-thecus andOreopithecus.


Finally, RUD77 differs fromAustralopithe-cus afarensis in being smaller with thinnerenamel, less molarized premolars, relativelysmaller molars, thicker lateral orbital pil-lars, more convex malar surface, verticalfrontozygomatic suture,more strongly devel-oped anterior temporal lines, less stronglydeveloped supraorbital tori and supratoralsulcus, a shorter palate anteriorly with alarger incisive foramen, shorter incisive ca-nal, a shorter and more vertical premaxilla,no groove between the glenoid fossa and thepostglenoid tubercle, a moremedially placedpostglenoid tubercle, a less pneumatizedtemporal squama in the suprameatal re-gion, more vertical frontal squama, and ahigher inion. Detailed occlusal differences inthe molar region are similar to those de-scribed for Ouranopithecus and Sivapithe-cus. The same postcanine morphological dif-ferences distinguish Australopithecusanamensis from RUD 77 (Leakey et al.,1995). RUD 77 also differs from Ardipithe-cus ramidus (White et al., 1994) in havingsmaller postcanine teeth with occlusal mor-phology differences similar to A. afarensis,and less temporal pneumatization. Judgingfrom the descriptions in White et al. (1994),Ardipithecusmay also differ from RUD 77 inhaving a shallower temporomandibular joint,and a more weakly projecting entoglenoidprocess. However, RUD 77, like other greatapes, is closer to A. ramidus than to A.afarensis in P3 crown morphology, enamalthickness, and possibly in having a moremedially placed postglenoid process (Whiteet al., 1994).

SUMMARY AND CONCLUSIONS

RUD 77 is the most completely preservedcranium of Dryopithecus, and one of onlytwo reasonably well preserved and rela-tively undistorted neurocranial specimensof anyMiocene hominoid. New aspects of theanatomy of this specimen, a new reconstruc-tion of RUD 77, and new comparisons toother hominoids have been presented here.A restoration of the female skull of D. bran-coi based on a number of specimens fromRudabanya has been published in Kordos(1987). Similar composite or hypotheticalrestorations of Ouranopithecus and Austra-lopithecus have also been published (de Bo-

nis andKoufos, 1993; Kimbel et al., 1984), ashas a restoration of Proconsul based onKNM-RU 7290 (Walker et al., 1983). Giventhe new data presented here, a revisedcomposite restoration of a female D. brancoiskull is presented in Figure 10. Like theother restorations noted previously, thisshould be regarded as somewhat hypotheti-cal, illustrative of general as opposed todetailed aspects of cranial form, and as astarting point for a more refined, more de-tailed restoration based on future discover-ies. A comparison of this restoration withrestorations of other fossil hominoids andlateral views of living hominids is presentedin Figure 11.New characters preserved on RUD 77

tend to reinforce previous suggestions thatDryopithecus is more closely related to Afri-can apes and humans than is Sivapithecus(Kordos, 1987; Dean and Delson, 1992; Be-gun, 1994, 1995; Begun and Kordos, 1997).These newly identified characters includeanteroposteriorly elongated temporal fos-sae, elongated neurocrania, deeper glenoidfossae, less prominent articular tubercles,

Fig. 10. Restoration of a female D. brancoi in lateralview. The mandibular corpus and lower teeth are basedon RUD 17, and the ramus on RUD 2. The maxilla andpremaxilla are based on RUD 77, RUD 12 and RUD 15.The condylar process has been omitted to expose theTMJ.


more prominent entoglenoid processes, lessconstricted and elongated postglenoid pro-cesses, and possible partial fusion of thearticular and tympanic portions of the tem-poral. RUD 77 also preserves a frontal sinusbelow nasion, distinct supraorbital tori, aninflated glabella, and supratoral sulci. Thesecharacters, previously identified on RUD 44,also serve to distinguish Dryopithecus andtheAfrican apes and humans fromSivapithe-cus (Begun, 1994).

ACKNOWLEDGMENTS

We thank Peter Andrews, Louis de Bonis,David Dean, Eric Delson, Clark Howell, JayKelley, Tim White, and two anonymous re-viewers for comments on an earlier versionof this paper.

LITERATURE CITED

Aiello LC and Wood BA (1994) Cranial variables aspredictors of hominine bodymass.Am. J. Phys. Anthro-pol. 95:409–426.

Andrews P and Martin L (1987) Cladistic relationshipsof extant and fossil hominoids. J. Hum. Evol. 16:101–118.

Begun DR (1987) A review of the genus Dryopithecus.Ph.D. thesis, University of Pennsylvania.

Begun DR (1988) Catarrhine phalanges from the LateMiocene (Vallesian) of Rudabanya, Hungary. J. Hum.Evol. 17:413–438.

Begun DR (1989) A large pliopithecine molar fromGermany and some notes on the Pliopithecinae. FoliaPrimatol. 52:156–166.

Begun DR (1992a) Miocene fossil hominids and thechimp-human clade. Science 257:1929–1933.

Begun DR (1992b) Dryopithecus crusafonti sp. nov., anew Miocene hominoid species from Can Ponsic(Northeastern Spain).Am. J. Phys.Anthropol. 87:291–310.

Begun DR (1994) Relations among the great apes andhumans: New interpretations based on the fossil greatapeDryopithecus. Yrbk. Phys. Anthropol. 37:11–63.

Begun DR (1995) Late Miocene European orang-utans,gorillas, humans, or none of the above? J. Hum. Evol.29:169–180.

Begun DR and Kordos L (1993) Revision ofDryopithecusbrancoi SCHLOSSER 1901 based on the fossil homi-noid material from Rudabanya. J. Hum. Evol. 25:271–286.

Begun DR and Kordos L (1997) Phyletic affinities andfunctional convergence in Dryopithecus and otherMiocene and living hominids. In DR Begun, CVWard,and MD Rose (eds.): Function, Phylogeny and Fossils:Miocene Hominoid Origins and Adaptations. NewYork: Plenum, pp. 291–316.

Begun DR and Moya-Sola S (1992) A new partial cra-nium of Dryopithecus laietanus from Can Llobateres.Am. J. Phys. Anthropol. Suppl. 14:47.

Bernor RL, Brunet M, Ginsburg, L, Mein P, Pickford M,Rogl F, Sen S, Steininger F, and Thomas H (1987) Aconsideration of some major topics concerning oldworld Miocene mammalian chronology, migrationsand paleogeography. Geobios 20:431–439.

de Bonis L, Bouvrain G, Geraad D, and Koufos G (1990)New hominoid skull material from the late Miocene ofMacedonia in northern Greece. Nature 345:712–714.

de Bonis L and Koufos G (1993) The face and mandibleof Ouranopithecus macedoniensis: Description of newspecimens and comparisons. J. Hum. Evol. 24:469–491.

Dean D and Delson E (1992) Second gorilla or thirdchimp? Nature 359:676–677.

Harrison T (1986) A reassessment of the phylogeneticrelationships of Oreopithecus bambolii Gervais. J.Hum. Evol. 15:541–584.

Hurzeler J (1951) Contribution a l’etude de la dentitionde lait d’Oreopithecus bambolii Gervais. Ecol. Geol.Helv. 44:404–441.

Kelley J (1995) Sex determination in Miocene catar-rhine primates. Am. J. Phys. Anthropol. 96:391–417.

Kimbel WH, Johanson DC, and Rak Y (1994) The firstskull and other discoveries ofAustralopithecus afaren-sis at Hadar, Ethiopia. Nature 368:449–451.

Kimbel WH, White TD, and Johanson DC (1984) Cra-nial morphology of Australopithecus afarensis: A com-parative study based on a composite reconstruction ofthe adult skull. Am. J. Phys. Anthropol. 64:337–388.

Kordos L (1982) The prehominid locality of Rudabanya(NE Hungary) and its neighbourhood: A palaeogeo-

Fig. 11. Lateral views of selected hominoid craniadrawn to the same length. From top to bottom: left,Pongo, Gorilla, Pan paniscus, Pan troglodytes; right,Australopithecus afarensis, Proconsul, Ouranopithecus,Dryopithecus brancoi. A. afarensis from Kimbel et al.(1994).Proconsul fromWalker et al. (1983).Ouranopithe-cus modified from de Bonis and Koufos (1993) based onthe authors’ personal observations.


graphic reconstruction. M. All Foldt. Evi. Jel. 1980-ROL:395–406.

Kordos L (1985) Environmental reconstruction for pre-hominids of Rudabanya, N.-E. Hungary. Schr. zurUr-und Fruhgeschichte. 41:82–85.

Kordos L (1987) Description and reconstruction of theskull ofRudapithecus hungaricusKRETZOI (Mamma-lia). Ann. Hist. Nat. Mus. Natl. Hung. 79:77–88.

Kordos L (1988) Comparison of early primate skullsfrom Rudabanya and China. Anthropol. Hung. 20:9–22.

Kordos L (1991) Le Rudapithecus hungaricus de Ruda-banya (Hongrie). L’Anthropologie (Paris) 95:343–362.

Kordos L and Begun DR (submitted, a) Primates fromRudabanya: Allocation of specimens to individuals,sex and age categories.

Kordos L and Begun DR (submitted, b) Dryopithecuscranial capacity reveals stasis in early hominid en-cephalization.

Kretzoi M (1969) Geschichte der Primaten und derHominization. Symp. Biol. Hung. 9:3–11.

Kretzoi M (1975) New ramapithecines and Pliopithecusfrom the lower Pliocene of Rudabanya in north-eastern Hungary. Nature 257:578–581.

Kretzoi M (1984) New hominoid form from Rudabanya.Anthropol. Kozlem. 28:91–96.

Le Gros Clark WE and Leakey LSB (1951) The MioceneHominoidea of East Africa. Fossil Mammals of EastAfrica. British Museum (Natural History) 1:1–117.

Leakey RE, Leakey MG, andWalkerA (1988a) Morphol-ogy of Turkanapithecus kalakolensis from Kenya. Am.J. Phys. Anthropol. 76:277–288.

Leakey RE, Leakey MG, andWalkerA (1988b) Morphol-ogy of Afropithecus turkanensis from Kenya. Am. J.Phys. Anthropol. 76:289–307.

Leakey MG, Feibel CS, McDougall I, and Walker A(1995) New four-million-year-old hominid species fromKanapoi andAllia Bay, Kenya. Nature 376:565–571.

Mein P (1986) Chronological succession of hominoids inthe European Neogene. In JG Else and PC Lee (eds.):Primate Evolution. Cambridge: Cambridge Univer-sity Press, pp. 58–70.

Mein P (1990) Updating of the MN zones. In EHLindsay, V Fahlbusch, and P Mein (eds.): EuropeanNeogeneMammal Chronology. NewYork: Plenum, pp.73–90.

Moya-Sola S, Gibert J, and Begun DR (1992) Newspecimens of Dryopithecus laietanus from the VallesPenedes Catalonia, Spain. Abstracts of the 24th Con-gress of the International Primatological Society, pp.261.

Moya-Sola S and Kohler M (1993) Recent discoveries ofDryopithecus shed new light on evolution of greatapes. Nature 365:543–545.

Moya-Sola S and Kohler M (1995) New partial craniumof Dryopithecus Lartet, 1863 (Hominoidea, Primates)from the upper Miocene of Can Llobateres, Barcelona,Spain. J. Hum Evol. 29:101–139.

Pilbeam DR (1982) New hominoid skull material fromthe Miocene of Pakistan. Nature 295:232–234.

Schwartz JH (1990) Lufengpithecus and its potentialrelationship to the orang-utan clade. J. Hum. Evol.19:591–605.

Spoor F (1996) The ancestral morphology of the hominidbony labyrinth: the evidence from Dryopithecus. Am.J. Phys. Anthropol. Suppl. 22:219.

Spoor F, and LeakeyM (1996)Absence of the subarcuatefossa in cercopithecids. J. Hum. Evol. 569–575.

Steininger FF, Bernor RL, and Fahlbusch V (1990)EuropeanNeogenemarine/continental chronologic cor-relations. In EH Lindsay, V Fahlsbusch and P Mein(eds.): European Neogene Mammal Chronology. NewYork: Plenum Press, pp. 15–46.

WalkerAand PickfordM (1983) New Postcranial Fossilsof Proconsul africanus and Proconsul nyanzae. In RLCiochon and RS Corruccini (eds.): New Interpreta-tions of Ape and HumanAncestry. New York: Plenum,pp. 325–351.

Walker, A, Falk D, Smith R, and Pickford M (1983) Theskull of Proconsul africanus: Reconstruction and cra-nial capacity. Nature 305:525–527.

Ward SC and Brown B (1986) The facial skeleton ofSivapithecus. In DR Swindler and J Erwin (eds.):Comparative Primate Biology, vol. 1: Systematics,Evolution, and Anatomy. New York: Alan R. Liss, pp.413–454.

White TD, Suwa G, and Asfaw B (1994) Australopithe-cus ramidus, a new species of early hominid fromAramis, Ethiopia. Nature 371:306–312.

Wu R, Xu Q, and Lu Q (1986) Relationship betweenLufeng Sivapithecus and Ramapithecus and theirphylogenetic position. ActaAnthropol. Sinica 5:1–30.


Documents

A new reconstruction of RUD 77, a partial cranium ofDryopithecus brancoi from Rudabánya, Hungary