AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 7221-37 (1987) Pliocene Hominid Partial Mandible From Tabarin, Baringo, Kenya STEVEN WARD AND ANDREW HILL Department of Sociology and Anthropology, Kent State University, Kent, Ohio 44242, and Human Anatomy Program, Northeastern Ohio Universities College of Medicine, Rootstown, Ohio 44272 (S. W.); Department of Anthropology, Yale Uniuersity, Box 2114, New Haven, Connecticut 06520 (A.H.) KEY WORDS Baringo, Chemeron, Hominid, Mandible, Pliocene ABSTRACT Sediments in the Tugen Hills, west of Lake Baringo, Kenya, form one of the best fossiliferous successions known in Africa spanning the period from 14 my to less than 4 my. Hominoid fossils have previously been recovered from a number of localities in the region. We describe here a new hominid mandible (KNM-TH 13150)from the site of Tabarin, in the Chemeron Formation. Isotopic determinations on a tuff below the fossiliferous horizon gives dates of 4.96 my and 5.25 my. The associated fauna is consistent with these results and independently suggests a minimum age for the specimen of 4.15 my. Although fragmentary, the preserved morphology of the Tabarin mandible is consistent with the diagnosis of the Pliocene hominid Australopithecus ufarensis. It can be distinguished from all other currently recognized hominoid taxa. Few hominoid fossils are known from Af- this paper, we describe the new specimen, its rica between the relatively rich samples from chronostratigraphic placement, make comearly Middle Miocene sites, such as in west- parisons with other hominoid samples', and ern Kenya up to 14 my in age, and those discuss problems of Chemeron stratigraphy. from 3.7 my onwards, which include austra- This is the first of a number of papers in lopithecines. This is unfortunate, especially which we will describe and review the anatsince new predictions of hominoid phyletic omy and context of hominoids from this area branching events based on DNA-DNA hy- and time interval. bridization studies (Sibley and Ahlquist, GEOLOGY 1984) predict a hominid divergence at about 7 my ago. Tabarin (Baringo Paleontological Research A number of the hominoid specimens that Project site: BPRP K077: UTM 36N ZR do come from this ten million year Miocene- 186844) is located about 2 km north of RonPliocene interval have been found in the area donin, northwest of Lake Baringo (Fig. 1). It of the Tugen Hills, west of Lake Baringo in forms a n outcrop of Chemeron Formation the northern Kenya Rift Valley. The Tugen sediments which are exposed in a half basin Hills, and rocks in the eastern foothills, form developed against the Saimo Fault of the one of the main recorded fossiliferous succes- main Tugen Hills scarp. They are part of a sions spanning this important time period (Bishop et al., 1971). Since 1981 the Baringo Received April 25, 1985; revision accepted June 27,1986. Paleontological Research Project (BPRP) has 'This sample includes all medium to large sized Proconsul worked in this area, refining the time caliin the National Museums of Kenya; all the Swapithe bration, reanalyzing the fauna, and collect- mandibles cus mandibles from Pakistan in the care of the Geological Survey of Pakistan, and some of the Indian specimens at the Museum ing additional fossils (Hill et al., 1985; Hill, of the Geological Survey of India in Calcutta; most of the manCurtis, and Drake, in press). As a result of dibles of Australopthecus afarensis from Hadar now in the Natthese efforts we have discovered a number of ural History Museum, Addis Ababa; LH-4, the holotype of Austrabpithecus afarensls from Laetoli, in the National Munew localities, including Tabarin, in the seums of Kenya; and Austrabptthecus africanus specimens in Chemeron Formation, from which has come the Department of Anatomy, University of the Witwatersrand, a hominoid partial mandible (Hill, 1985). In and in the Transvaal Museum, South Africa. 0 1987 ALAN R. LISS, INC. 22 S. WARD AND A. HILL discontinuous outcrop that strikes predominantly northsouth and is intermittently concealed by a piedmont regolith of lava boulders and finer material. A t the site, sediments lie on a flow of the Kaparaina Basalt, which separates them from the sediments of the Lukeino Formation beneath. The succession begins with approximately 20 m of fine to medium-grained sediments, which are overlain and in places baked by a n ignimbritic t d a c e o u s unit about 10 m thick. The upper part is a coarse pumice tuff, and pumiceous patches also occur at the - 0'30'N base. Next follows about 17 m of mainly yellow and orange silts and sands, succeeded by a dark brown, coarsely bedded ferruginous conglomerate containing poorly sorted igneous clasts. This is the fossiliferous unit Fig. 1. Map of the area to the west of Lake Baringo from which, judging by location and matrix, showing the Tugen Hills and the location of Tabarin. come many of the surface fossil specimens, including the hominid. There are other fossiliferous levels higher in the sequence of sediments, which continue towards the is important, as other Chemeron primates Saimo fault of the main Tugen Hills range. The history of description of the rocks come from the type area, and there is some which now constitute the Chemeron Forma- doubt as to their precise dating. Two are the tion has resulted in a relatively complicated type specimens of Paracolobus chemeroni, a situation (Hill, Curtis, and Drake, in press). complete skeleton, and of Parapapio barinRocks lying on basalts, subsequently named gensis (Leakey, 1969; Birchette, 1982; Eck the Kaparaina Basalts, in river valleys west and Jablonski, 1984); and another is a temof Lake Baringo were first named the Chem- poral of a hominid (Martyn, 1967; Tobias, eron beds by McCall et al. (1967). More de- 1967)(see below; also Ward and Hill, in preptailed work by Martyn (1967, 19691, resulted aration, a). In terms of stratigraphical position, interin his formalizing the unit and in allocating additional outcrops (the Kipcherere outcrops) nal succession, and faunal content, a number to the Formation, which he mapped north to of sites appear to correlate with Tabarin. OO"45'. Sediments resting on Kaparaina These include sites about 20 km farther Basalts were recorded north of this by Chap- south, such as Kibingor (BPRP KO01 = JM man (1971). Subsequently Pickford (1975a) 511) and the Kipcherere sites (BPRP K004allocated these and other exposures to the KO10 = J M 4891, which are part of the ChemChemeron Formation. To a t least the Barge- eron Formation sensu Martyn (1967; 1969). tyo River (approximately the latitude of Ya- Others nearer to Tabarin include Sagatia tya in Fig. l), the sediments forming this (BPRP KO75 = 2/232), a site close to Ronnorthern extension lie on the Kaparaina donin. North of the Bargetyo River the relaBasalt, and this along with lithological simi- tionships of rocks forming the Chemeron larities results in the correlation with the northern extension to those farther south are Chemeron (Kipcherere) exposures to the less clear. Sites in this northern area include south being little disputed. Further north, in BPRP KO37 (=2/211) (mistakenly recorded areas where the Kaparaina Basalt is absent, as 2 / 2 0 in Pickford et al., 19831, from which Pickford's reallocation of parts of the Kaper- has come a hominoid humerus (Pickford et yon Formation (McClenaghan, 1971; Chap- al., 1983) (see below). man, 1971; Pickford, 1975a) to the Chemeron RADIOMETRIC DATING unit has been questioned (Chapman and G. Curtis and R. Drake have performed W Brook, 1978; Chapman et al., 1978). Problems also exist in the relation of the Ar isotopic determinations on samples of the Kipcherere outcrops and their correlatives to tuffaceous unit underlying the main fossilithe Chemeron sequence in its type area. This ferous horizon. These give dates of 4.96 f 23 TABARIN MANDIBLE 0.03 my (on anorthoclase), and 5.25 f 0.04 my (on sanidine). The fossiliferous sediments are apparently comformable on the surface of this dated tuff, so the hominoid is unlikely to be considerably younger than these isotopic dates. PALEONTOLOGY AND DATING An abundant and diverse fossil fauna comes from Tabarin (Table 1).The same fossiliferous horizon crops out at Sagatia, a site 1.5 km south, and additional species from there are also listed. The fauna includes aquatic animals, such as molluscs, fish, turtles, crocodiles, and hippopotami, along with others that might be found in the vicinity of a lake or river, proboscideans, equids, rhinoceroses, suids, bovids, carnivores, and primates. The fauna is compatible with the age suggested by the radiometric dates and in addition independently provides a n upper limit of 4.15 my. The most useful species for this are the suid Nyanzachoerus jaegeri, and the proboscideans, particularly A n a n c u s kenyensis. T A B L E 1. Fossil fauna from Tabarin and Sagatia Tabarin Mollusca Bivalvia Pisces Siluriformes cf. Clarias Reptilia Chelonia ’ Testudinidae Trionychidae Pelomedusidae Crocodilia Mammalia Primates Cerocopithecidae Papionini Hominidae A urtralopithecus cf. afarensis Carnivora Hyaenidae Felidae cf. Panthera Small species Mustelidae Enhydriodon cf campani Another species; family indet. Proboscidea Gomphotheriidae Anancus kenyensis Elephantidae cf Primelephas gomphotheroitlvs Deinotheriidae Deinotherium bozasi Perissodactyla Equidae Hipparion Rhinocerotidae Artiodactyl a Suidae Nyanzachoerus jaegeri Hippopotamidae Bovidae Several species Rodentia Hystricidae Hystrix Sagatia X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 24 S. WARD AND A. HILL Nyanzachoerus jaegeri is so far only known from sites older than 4.15 my. This age limit is given by the Mursi Formation a t the Omo, which is dated a t older than 4.15 my (corrected for new constants) (Brown and Lajoie, 1971; Harris and White, 1979). Among other independently dated sites that the species comes from is Aterir, which is likely to be over 4 my (Bishop et al., 1971; Harris and White, 1979). Ny. jaegeri is found in the Aramis Member and the Beearyada beds, which underlie the Kalaloo beds of the Sagantole Formation in the Middle Awash Complex (Kalb et al., 1982a, 198213,1982~). These units are beneath the Cindery Tuff dated at 3.84.0 my (Clark et al., 1984; Hall et al., 1984). There are sites just younger than 4 my where it is not found. For example, it is not known in the Hadar Formation (Harris and White, 1979; Johanson et al., 1982), nor in the Laetoli fauna at 3.7 my (Leakey et al., 1976; Harris and White, 1979). Despite assertions to the contrary in Coppens et al. (19781, Anancus kenyensis is not known later than 4 my. It is not found in the Hadar Formation (Johanson et al., 1982). In the Omo sequence it is only known for the Mursi Formation, dated a t older than 4.15 my (Brown and Lajoie, 1971; Beden, 1976). Mebrate and Kalb (1985, Fig. 3) are mistaken in recording it as part of the Laetoli fauna a t 3.7 my (Leakey et al., 19761, but it is found in the older Lower Laetolil Beds (J. Harris, personal communication). In the Middle Awash, all but one specimen comes from below the Cindery Tuff dated a t 3.8 and 4.0 my. Kalb reports a single specimen, which was not collected, in the Kalaloo beds of the Sagantole Formation, stratigraphically immediately above the Cindery Tuff horizon (Kalb et al., 1982~). Nothing else in the Tabarin fauna is so explicit, but all of it is consistent with the age suggested by the radiometric dating and by the suids and gomphothere. There are specimens of a n archaic elephant which are insufficient to identify exactly but are probably Primelephas gomphotheroides, the species found a t other Chemeron Formation sites. It is not known from Hadar or Laetoli; Maglio (1973) considers it typical of about 5 my. The cercopithecid specimen is only a single molar tooth in a fragment of mandible, but it can identified as papionine and there is nothing to prevent its being early Pliocene in age (MG Leakey personal communication). John Barry (personal communication) identifies the Enhydriodon as resembling the form found at Kanapoi; and the hyaenid is also similar to the one from Kanapoi, which also occurs in the Kaiso fauna. CHEMERON HOMINOIDS Two hominoids have previously been recorded from the Chemeron Formation. The first to be discovered was a hominid temporal bone from the Chemeron Formation in its type area (Martyn, 1967; Tobias, 1967). But this is not associated with the fauna that typifies the Chemeron a t the majority of sites, including Tabarin. It is likely to be considerably younger than this fauna and than the other specimens discussed here. Lavas above the sediments from which the hominid was recovered are dated a t 1.57 and 2.07 my (Hill et al., 1985). This will be discussed in more detail in a following publication (Ward and Hill, in preparation a). Pickford et al. (1983) described a proximal humerus fragment from a site in Baringo about 6.5 km north of Tabarin (BPRP KO37 = 2/211). They assigned it to the Hominoidea and suggested that several features of the specimen indicated hominid affinities. A study by Senut (1983)stresses that a familial allocation within the Hominoidea cannot be justified. Although it has been referred to the Chemeron Formation, the geological relationship of this site to the Tabarin sequence and sites farther south is not yet clear. However, it may be a similar age, as Nyanzache erus jaegeri is part of the associated fauna, and a n isotopic date of 5.07 my has been obtained on a tuff about 35 m beneath (Pickford et al., 1983). If there is only one species of large hominoid living in this area at this time period, and if the Tabarin specimen is indeed a hominid, then it would support the taxonomic suspicions of Pickford et al., (1983) regarding the humerus. THE NEW HOMINOID Preservation The Tabarin hominoid specimen (KNM-TH 13150;Fig. 2) was found by Kiptalam Cheboi. It was a surface find a few meters downslope from the outcrop of the ferruginous conglomeratic horizon. Matrix adhering to the specimen leaves no doubt that it derives from the conglomerate and is associated with the fauna from this unit. The partial mandible is a fragment of the right corpus, bearing the first and second permanent molars, as well as portions of the 25 TABARIN MANDIBLE A C B Fig. 2. The hominid specimen from Tabarin (KNM-TH 13150):(A) buccal, (Bj lingual, and (Cj occlusal views. alveoli of Pm4 and M3. The mandible is that of an adult hominoid, as can be determined by the state of molar wear, and by the condition of the third molar socket. The entire mesial wall of this alveolus is preserved, revealing a bifurcated root apex. From the state of the apical impressions left by the root, and from the presence of a distal interproximal wear facet on M 2 , it is evident that the third molar had completed its development. Of the two molars preserved in the specimen, MI is moderately worn, and a small portion of the protoconid has been lost me- sially. The second molar has sustained extensive damage to its enamel cap, obscuring details of its morphology. The crown is otherwise complete. The corpus is broken along a horizontal plane a t approximately the union of the alveolar process and the mandibular base. Below the distal root of the second molar, the corpus is complete, permitting measurement of its depth a t that point (Table 2). The lingual cortex is extensively damaged, but is in good condition under the second molar, where contours are well preserved. The alveolar TABLE 2. Mandibular metrics N Tabarin A. afarensis A. africanus S. siualensis P. africanus 6 3 5 4 Corpus height M2 j r Range 26.9 35.5 30.2 24.2 20.8 25.1-37.0 23.7-35.0 23.3-27.2 18.1-24.1 Alveolar crest breadth M2 jr Range Corpus X breadth M2 Range 11.2 14.0 14.9 12.5 08.4 19.8 23.7 23.1 18.7 14.5 18.8-28.5 22.0-25.1 16.3-19.1 11.1-19.2 - 12.1-17.5 14.9-15.0 11.1-14.1 08.1-11.0 - 26 S.W A R D A N D A.HILL A C D J E F Fig. 3. Subocclusal morphology drawn from lateral radiographs: (A) Tabarin (KNM-TH 13150);(B) Proconsul africanus (KNM-RU 1855);(C) Proconsul major (KNM-SO 396);(D)S i u a p ithecus indicus (AMNH 19413); (E) Siuupithecus sivalensis (GSP 4622); (F) A ustrnlopithecus afarensis (AL 400). See text for details. crest is intact on both the buccal and lingual below the lateral prominence, it is not possisides of the molars. There is some evidence ble to determine the relationships of the of osteolytic alveolar resorption around the oblique line. A fairly discernible upwardly concave line branches off the lateral promiroots of the first molar. nence and terminates on the alveolar crest Corpus morphology adjacent to the distal socket of Pm4. This The most striking feature of the Tabarin probably represents the line of buccinator specimen is its shallow corpus depth yet very attachment. In concert with a massive lateral promipronounced external contours. The lateral prominence is strongly developed and ap- nence, the buccal surface of the Tabarin manpears to have attained its maximum expres- dible shows pronounced hollowing below the sion adjacent to the contact between the first molar. This combination of relationships second and third molars. It terminates below imparts a strong surface relief to the corpus the distal root of MI. From these relation- and gives the impression of a strongly butships, and from the position of the fracture tressed corpus, despite its small size. Although it is extensively damaged, the a t the union of the ramus and the corpus, it can be concluded that the third molar was lingual cortex is sufficiently well preserved situated in a relatively posterior position, be- to reveal that there is only moderate relief hind the anterior edge of the ramus. Since beneath the second molar. The alveolar crest the external surface of the corpus is broken is robust above, and the lingual cortex be- 27 TABARIN MANDIBLE comes concave a t midcorpus. From what is preserved of the base, it appears that this part of the corpus was transversely thick. Occlusal morphology Although the first molar is worn, and the enamel on the occlusal surface of the second molar is badly weathered, useful metric and anatomical information is preserved (Table 3). From its wear configuration, as well as from fractured surfaces on its protoconid, it can be determined that the first molar has a thick enamel cap. On both molars, the trigonid is slightly broader than talonid, and the hypoconulid on MI is mesiodistally compressed. However, this compression is not the result of interproximal wear, which affects both molars at this contact. The metaconids are the largest cusps on each molar, followed by the protoconids. The buccal developmental grooves are well developed on each molar, and despite wear and erosion, it appears that both the protoconids and hypoconids were prominently expressed. This finding is reinforced by the strong bilobate appearance of the molar crowns when viewed from their buccal surfaces. Both molars show evidence of buccal flare, while the lingual surfaces of the crowns are vertical. As has been noted, the first molar crown is moderately worn, with wear being concentrated on the buccal cusps. A pair of hollowed out dentin islands mark the positions of the protoconid and hypoconid. These exposures were in the process of coalescing a t the time of death. They are bounded buccally by crescentic rims of enamel. Two other small dentin exposures are present: one a t the apex of the metaconid, the other on the hypoconulid. The metaconid is flattened by wear apically, and its lingual rim is preserved as a sharp edge. Despite its degree of wear, the metaconid is the tallest cusp on the crown. The entoconid has also been flattened by wear, but its dentin core has not been exposed. Evidence of wear on the second molar has been largely obliterated, but there is a small surface of exposed dentin on the protoconid. From the enamel tissue that remains, we believe it is unlikely that the second molar was as heavily worn as the first. Of the three interproximal contacts that are preserved, only the M1-Mz facet can provide any useful information concerning interproximal wear. At this contact, a pro- ,-- E 28 S. WARD AND A. HILL nounced interproximal facet has developed, squaring off the distal surface of the MI hypoconulid and incising the mesial fovea of the second molar. From what is preserved of the mesial surface of the first molar, a strongly developed interproximal facet is seen to be present, the result of contact with Pm4. Although the third molar is missing, observation of the contiguous occlusal surfaces of the first and second molars shows that a sinusoidal, or helicoidal, occlusal plane characterized the molars. Pitch reversal, or the “pas helicoide” (Ackermann, 1953) occurs on the talonid of Ma. Subocclusal morphology Lateral radiographs of KNM-TH 13150, in addition to examination of its broken surfaces, provide important information on alveolar process relationships and subocclusal anatomy (Fig. 3). Both molars have low, cynomorph pulp chambers. The pulp horns are obtuse and do not extend up into the dentinal mass of the crown. This probably indicates that the dentin horns under the enamel were likewise low and blunt, a factor that affects the process of dentin exposure on the occlusal surface (Martin, 1984). The root furcation under the crown of both molars is arch shaped, rather than narrow and cleft-like. The roots of the first molar are almost vertically inclined in lateral projection, and the distal root is thicker anteroposteriorly than the mesial. However, when measured buccolingually, the mesial root is broader than the distal. The same pattern holds true for the second molar, except that the roots are slightly more inclined distally. In both molars, the roots are of equivalent length when measured from the cervix. Both partially preserved sockets, the distal alveolus of the fourth premolar, and the mesial alveolus of the third molar, are sufficiently intact to determine basic subocclusal relationships. From what is preserved of the Pm4 socket, it is clear that this tooth had a pair of roots, mesiodistally aligned. The apex of the distal root is still present in its socket, and it is bifurcated a t its tip. At the point where the root is fractured, two root canals, lingual and buccal, are visible. The total length of this root is approximately the same as the roots of both molars. The mesial root socket of the third molar is exposed in the posterior break and is notable for its bifid apex. The socket is inclined distally with respect to the alveolar plane. In coronal projection, its axial alignment is strongly angled, with the apex inclined towards the buccal surface of the corpus and the cervix angled towards the lingual cortical plate. The molar roots occupy approximately 50% of total corpus depth. The position of the inferior alveolar canal is not clear from either the radiographs or direct inspection, so a precise measurement of alveolar process height is not possible. However, it is possible that the canal passes fairly close to, or even between, the bifurcated apices of the molar and premolar roots. This would indicate that the alveolar process-basal bone ratio is in the neighborhood of 50-55%. COMPARATIVE OBSERVATIONS Comparative material In order to identify the taxon to which the Tabarin mandible belongs, it is necessary to undertake phenetic comparisons with seven known fossil samples: the medium to large hominoids from the early Miocene of Kenya and Uganda (Andrews, 1978; Martin, 1981; Pilbeam, 1969); the South Asian Siuapithe cus sample (Pilbeam et al., 1980; Ward and Brown, 1985); Kenyapithecus from the middle to late Miocene of Kenya (Andrews and Walker, 1976; Pickford, 1982; Ishida et al., 1984); the Lothagam mandible (Patterson et al., 1970); the Lukeino molar (Pickford, 1975b; Corruccini and McHenry, 1980); the large A ustralopithecus afarensis sample from Laetoli and Hadar (Leakey et al., 1976; White, 1977; Johanson and White, 1980; White and Johanson, 1982); and Austral@ pithecus africanus from Makapansgat and from Sterkfontein, South Africa (Tobias, 1980). In addition there is a mandible from Buluk, north Kenya (Leakey and Walker, 1985). There are taxonomic problems with the Lukeino molar, and the Lothagam mandible is also difficult to classify taxonomically, for the same reasons the Tabarin mandible is. However, the Lothagam specimen is usually attributed to Australopithe cus (Patterson et al., 1970; Howell, 1978), and we will support that practice here. The early Miocene hominoid material from Kenya and Uganda spans the time period from about 23 my to about 14 my. The specimen from Buluk is about 17 my old. The South Asian Sivapithecus is known from 11.8-7.2 my (Barry, in press). Kenyapithecus is dated at around 15-14 my. Australopithecus afarensis ranges from 3.7-2.8 my. 29 TABARIN MANDIBLE The isolated tooth from the Lukeino Formation, also in the Baringo area about 2.5 km north of Tabarin, is the oldest hominoid specimen that might be a hominid. It was tentatively referred to the Hominidae by Andrews (in Pickford, 1975b). This tooth is older than the Tabarin specimen. Recent determinations have refined age estimates to about 5.6-6.2 my (Hill et al., 1985). The fragment of hominid mandible with a single tooth from Lothagam 1 is bracketed radiometrically between 8.5 my and 3.8 my (corrected for new constants) and paleontoloeical estimates focus on 5.5 mv (Patterson etal., 1970; Maglio, 1973; Harris and White, 1979). Corpus morphology The strongly sculpted contours on the buccal surface of the Tabarin mandible are similar to a t least two Kenyapithecus mandibles from Kaloma (KNM-MJ 5) (Pickford, 1982) and Nachola (KNM-BG 91514) (Ishida et al., 19841, a mandible fragment from Buluk (KNM-WS 124) (Leakey and Walker, 1983, and to all A. afarensis mandibles (White et al., 1982). These specimens all have a prominently expressed lateral prominence, narrow 1 2 to moderately broad extramolar sulcus, and pronounced hollowing of the corpus under the premolars and first two molars. The Tabarin specimen differs in these features from most equivalently sized Siuapithe cus mandibles from India and Pakistan (Fig. 4). These latter specimens have a broader extramolar sulcus and a transversely broader corpus. The lateral prominences of the Siuapithecus mandibles, while pronounced, are not as strongly sculpted as in the Tabarin mandible but tend to have greater definition of the buccinator line. Despite extensive size diversity, few Pre consul mandibles bear any resemblance to the Tabarin mandible. Of those that do, most are large. For example, KNM-LG 452, a female F! major mandible (Martin, 1981), exhibits a strongly sculpted lateral corpus but is considerably larger than KNM-TH 13150. A smaller mandible from Rusinga (KNM-RU 1678) is metrically similar to the Tabarin mandible with respect to corpus depth under the second molar but shows considerably less surface relief, including a weakly inflated lateral prominence and oblique line. Finally, the mandibular corpora of all Proconsul species are transversely narrow, while those of 3 4 a 0 J 0 Fig. 4. Transverse and coronal sections of the mandibular corpus: ( A l , B1) the Tabarin mandible (KNM-TH 13150);(A2, B2) Proconsul africanus (KNM-RU 7290);(A3, B3) Siuapithecus siualensis (GSP 4622);(A4, B4) Australopithecus afarensis (AL 198). Buccal surface is to the left; lingual to the right. 30 S. WARD AND A. HILL hominids and the late Miocene hominoids from South Asia are transversely broad (Fig. 4). The Lothagam mandible (KNM-LT 329) also differs from the Tabarin fragment in terms of corpus relief. In the Lothagam specimen, the lateral prominence is weakly developed, the extramolar sulcus is very narrow, and there is little evidence of hollowing under the first and second molars. Finally, the oblique line is only faintly visible, and no useful details concerning its topography can be discerned. Lateral corpus contours preserved on KNMTH 13150 are quite similar to the general pattern described for the Laetoli and Hadar mandibles (White et al., 1981). This pattern involves hollowing of the corpus anterior to the lateral prominence and a high origin of the ramus, with consequent narrowing of the extra molar sulcus. Both the Tabarin mandible, as well as the mandibles of A. afarensis, are distinguished from the South African Australopithecus africanus mandibles in the extent of corpus inflation and in the size of the extra molar sulcus. A. africanus mandibles are characterized by diminished surface relief (MLD 18, MLD 40, Sts 7, Sts 36, Sts 52b), a lower origin of the ramus at its junction with the corpus, and a broader extramolar sulcus (MLD 40, Sts 7, Sts 36, Sts 52b) (see White, 1977). The lingual surface of the hominoid mandibular corpus has few useful diagnostic features. As noted previously, the Tabarin mandible shows bulging of the alveolar process into the oral cavity, with a longitudinal concave furrow below the level of the alveolar canal. This furrow is the posterior continuation of the intertoral sulcus that extends posteriorly from the genial fossa in the symphyseal region. Early Miocene hominoids lack a n intertoral sulcus, and they also lack a n inferior symphyseal torus. Except for large individuals, such as male l? major (KNM-SO 396) there is very little relief on the lingual side of the corpus. In addition, no Proconsul mandibles evince the internal twisting of the alveolar process that produces the characteristic bulging of its internal surface. The Sivapithecus mandibles from India and Pakistan all have a n intertoral sulcus, and on some larger specimens, the mylohyoid line is discernible. The Lothagam mandible has a pronounced intertoral sulcus, a very well-defined mylohyoid line, and a deeply excavated submandibular fossa. The extent of internal surface relief in the Lothagam specimen in fact approaches the pattern that characterizes Homo. However, a long, matrix-filled crack in the lingual cortical plate of this specimen may indicate a certain amount of buckling that could increase the surface relief. The Hadar and Laetoli mandibles show variations in lingual corpus relief, but all conform to one basic pattern. Within this pattern, shallow mandibles (AL 145, AL 288) tend to have a large transverse dimension which is associated with less internal relief. Larger specimens (AL 277, AL 333w-60)have more clearly defined intertoral sulci and mylohyoid attachments. However, all the A. afarensis mandibles show pronounced buccal-lingual twisting of their alveolar processes, and where preserved, a n undercut area under the third molars, the subalveolar fossa of Weidenreich (1936). White (1977)has described the medial surface of the mandibular corpus of A. africanus is some detail and noted that it presents, on one hand, a pattern of contours intermediate between that of many early Homo specimens from East Africa and, on the other, “robust” australopithecines (MLD 18, MLD 40, Sts 52b). Occlusal morphology Despite wear and damage, occlusal anatomy of the two molars present in the Tabarin mandible are clearly distinct from the P r e consul lower molar pattern of buccal cingula, thin enamel, and separated, conical cusps. Thus the only hominoid dental samples that provide useful comparisons are the Kenyapithecus specimens, Siuapithecus, the Lothagam first molar, the Lukeino molar, and the Laetoli-Hadar sample. Direct comparison between the few Kenyapithecus mandibular molars and the Tabarin teeth is difficult, due to wear and damage to much of the published material. From the observations that are possible, the Tabarin molars share with Kenyapithecus thickened enamel, reduction of cingula, low, blunt cusps, and less certainly, accumulation of differential molar wear. In these features, both Tabarin and the Kenyapithecus teeth from West Kenya and Nachola are similar to Siuapithecus molars from South Asia. However, there are also differences. A particularly useful comparison involves the Tabarin mandible and GSP 4622, a S. siualensis mandible from the Potwar Plateau in Pakistan. TABARIN MANDIBLE Both specimens are almost identical in depth dimensions beneath the second molar, and making minimal allowances for interproxima1 wear, are very similar in first and second molar mesiodistal length. Other similarities include thickened enamel and differential tooth wear. However, they differ in the manner in which occlusal wear accumulates. In the Tabarin mandible, which shows slightly less wear on MI than does GSP 4622, a dent i n exposure has appeared on the metaconid, while the dentin horn of the metaconid of the small Siuapithecus mandible is not yet exposed. Also, the Tabarin M1 metaconid has been flattened apically, while metaconid wear on the M1 of GSP 4622 has accumulated on the buccal flanks of the cusp. These differences in the topography of metaconid wear may indicate concomitant differences in mandibular trajectories during the terminal phases of the occlusal contact sequence. It may also suggest differences in the extent of maxillary-mandibular canine interlock in the two taxa, although Kay (1981) asserts that canine interlock does not limit mandibular movement in Siuapithecus. Our interpretation of molar macrowear in all currently recognized species of Siuapithecus and Australopithecus leads us to conclude that protoconid and metaconid attrition accumulates differently in the two genera. Siuapithecus metaconids sustain wear in a manner quite similar to Kenyapithecus, with this cusp generally configured as a tall, angular peak as the protoconid is progressively flattened. It is tempting to impute the pattern of metaconid flattening characteristic of hominids at least partially to canine reduction, but there is at present insufficient evidence to support this functional interpretation. Other differences between the Tabarin teeth and small Siuapithecus molars include greater buccal flare in the African specimen, which mediates broader crown dimensions in molars of equivalent length, evidence of greater cusp lobation in the Tabarin teeth, a distinct hypoconulid in the Tabarin MI, and apparent expansion of the M2 trigonid. The Lukeino molar, which is probably a n MI, is excellently preserved but is slightly smaller than the Tabarin first molar. While the worn condition of the Tabarin MI. precludes a detailed morphologic comparison, the Lukeino molar shares with Tabarin buccal flare and pronounced cusp lobation. It is doubtful that repeatable crown component measurements can be made on the Tabarin 31 teeth, but it does appear that both first molars share a common pattern of cusp size ratios, with the metaconid being the largest cusp, followed by the protoconid hypoconid, entoconid, and hypoconulid. This pattern seems to be in accord with the data on hominid molar cusp areas reported by Wood et al. (1983). These workers showed that this pattern of cusp size relationships was characteristic of South and East African Australopzthecus, while the areas of the protoconid and metaconid are reversed in East African early Homo specimens. The data reported by Wood and his colleagues (Wood and Abbot, 1983; Wood et al., 1983) also shows a general reduction in the size of the trigonid cusps and enlargement of talonid elements. This trend does not appear evident in the Tabarin first molar, which is likely to possess the primitive condition for hominids. A comparison of first molar morphology in the Tabarin mandible, the Lothagam mandible (KNM-LT 3251, and the combined Sterkfonteinhfakapansgat sample from South Africa generally confirms the conclusions of White et al. (1981), Kimbel et al. (1985) and White (1985) that a basic and fundamental pattern of differences separate the South and East African “gracile” australopithecines at the species level. In the Tabarin-Lothagam comparison, both first molars share a swollen hypoconid and a pronounced bilobate buccal cusp configuration. The hypoconulid is mesially appressed in both teeth, but is more strongly expressed as a distinct element in the Tabarin mandible. The degree of occlusal wear as determined by dentin exposure is similar in both molars, with the Lothagam molar showing slightly more exposed dentin. In both teeth the dentin islands are deeply incised, with lingually concave crescentic rims of enamel delimiting the exposed dentin. Metaconid and entoconid wear contours are also congruent in the two molars, especially apical flattening of the former. Features of first lower molar anatomy and wear described for the Tabarin and Lothagam mandibles are concordant with the A. afarensis first molar sample (White et al., 1981, White, 1985, Johanson, 1985). Buccal molar flare associated with bulbous buccal cusps is a consistent feature in A. afarensis, as is squaring off of the distal margin of MI. This trait is due largely to mesial appression of the hypoconulid in this species. Apical flattening of the metaconid is characteristic of 32 S. WARD AND A. HILL A. afarensis lower first and second molars. In later stages of wear, the inflated protoconid and hypoconid mediate the development of a bicrescentic rim that circumscribes a deeply incised dentin exposure (LH 4, AL 198-1, AL 400-lb). The South African lower molar sample from Sterkfontein and Makapansgat is distinguishable from the Pliocene East African specimens in several respects. Cusp relief is generally lower, imparting an inflated, massive appearance to the entire crown. As noted by White et al. (1981) the bilobate buccal configuration that is diagnostic of A. afarensis is either attenuated or lacking entirely. There also seems to be a considerable amount of anatomical heterogeneity in A. africanus lower molars, and some discussion has been devoted to the very robust molars of MLD 2, leading to speculation that it may have affnities with A. robustus (see Tobias 1980 for details). MLD 2 has massive cusps, very prominent metaconids, and strongly expressed protostylids. Protostylids are also prominent on some other Transvaal hominids, especially Sts 18. In addition, A. africanus molars are generally broader than the Pliocene hominid lower molars from East Africa. It is likely that these anatomic features are implicated in the pattern of molar wear accumulation that differentiates South and East African early hominids. Molar attrition in A. africanus begins in the typical hominoid fashion, with faceting appearing early on the buccal cusps, followed by exposure of small dentin islands on the protoconid and hypoconid. At a slightly later stage, and in the absence of pathological occlusion, wear is spread more evenly over the entire crown, while in A. afarensis, Tabarin and Lothagam, wear is concentrated for a longer period on the buccal half of the occlusal surface. The pattern of early molar wear that characterizes A. africanus is rather similar to that of A. robustus (SK 6, SK 843, SK 1587, SK 23, and others), which shows a generalized flattening of the crown in early stages of wear, with dentin esposures occurring after much of the original crown relief has been lost (Wallace, 1978). Radiographically determined enamel thickness data recently reported by Sperber (1985) indicates that the enamel caps of A. robustus are on average thicker than those of A. africanus. This, coupled with differences in cusp disposition and fissure patterns, may account for differences in the time required to expose substantial areas of dentin in the two species. Dental metrics Length and breadth dimensions for the Tabarin molars and a comparative sample of Miocene hominoid and Plio-Pleistocene hominid molars are presented in Table 3. The Tabarin M1 falls within the A. afarensis sample range for length, but below the Hadar-Laetoli range for breadth. It is quite similar in both dimensions to the Lothagam MI, and larger than the Lukeino molar. When compared to the A. africanus molar sample, the Tabarin mandible proves to be considerably smaller, falling well below the length and breadth dimensions reported for the Sterkfontein and Makapansgat australopithecines. When compared to r( africanus and S. siualensis, the Tabarin MI is larger in both dimensions but is approximately the same size as the S. sivalensis MI. The Tabarin second molar falls within the low end of the A. afarensis length racge, but falls below the Hadar-Laetoli range for breadth. When compared to A. africanus, the Tabarin M2 falls well below the South African sample ranges for both length and breadth. It is slightly longer than the second molar of S. sivalensis, but falls within the breadth range for this species. When compared to the second molars of I.! africanus, the Tabarin M2 is larger in both dimensions. Conversion of the dental metric data in Table 3 to crown shape indices is shown graphically in Figure 5. The Tabarin M1 falls into a “hominid” cluster with A. afarensis, A. africanus, and the Lothagam mandible. The smaller Lukeino molar falls below this group at a shape index of 99, while the two Miocene hominoid species fall well above it. Figure 5b shows crown shape indices for the same specimens and samples as presented in Figure 5a, with the exception of the Lukeino molar. The crown shape index for the Lothagam M2 was calculated from measurements taken from the roots of this tooth, which are broken quite close to the cervical line of the missing crown. However, the value shown here should be regarded only as a n estimate. The Tabarin M2 yields a crown-shape index of 118, which most closely approximates the crown shape index of P africanus. A. afarenis and A. africanus fall below a value of 110, as does the estimated Lothagam index. The S. sivalensis value is slightly over 110. 33 TABARIN MANDIBLE - M, SHAPE INDEX P africanus * S indicus Lukeino Lothagam A Tabarin A afarensis 0 -- I I A africanus - M, SHAPE INDEX P africanus S sivalensis Lothagam w 0 Tabarin A afarensis - 1 A africanus 1 a 1 - 4 I I I I I I I I I 90 95 100 105 110 115 120 125 130 Fig, 5. Molar crown shape indices. See text for discussion. These data show that the Tabarin first molar falls within currently known Pliocene hominid ranges for length, breadth, and shape, while the second molar does not. The Tabarin M2 is transversely narrow, and this accounts for its high shape index. Postmortem loss of enamel on both its buccal and lingual surfaces may partially account for its small buccolingual diameter, since the contours of the crown show evidence of the buccal expansion that characterizes the mandibular molars of A. afarensis. In addition, White (1985) has noted that there are no small individuals represented in the 3.7-my Laetoli sample, and that this bias may skew the distribution of dental metrics and crown shape indices in Pliocene hominids. It is therefore uncertain what effect a small A. afarensis M2 would have on combined sample metrics. Finally, it is possible that buccal expansion of Pliocene hominid lower molars began as a trend prior to 3.7 my. Subocclusal morphology Since the subocclusal morphology of KNMTH 13150 is undistorted, it proves to be a useful source of phenetic information (Fig. 3). Lateral periapical radiographs were made on mandibles of medium- to large-sized Proconsul (I? africanus, €? nyanzae, ?? major), on samples of Sivapithecus, Australopithecus afarensis, and A. africanus. This radiographic survey demonstrates that the Tabarin mandible differs markedly from the Proconsul specimens in its enamel thickness, pulp chamber morphology and volume, root furcation pattern, root canal volume, and the topography of the alveolar process. The Proconsul molars vary in enamel thickness (Martin 1985), but they share a common pattern of capacious pulp chambers with tall and acute pulp horns. The pulp chamber floors are quite thin, and the roots diverge away from the furcation a t a pronounced angle. As noted above, the Tabarin molars have low, vertically compressed pulp chambers. The pulp chamber floor rises to a high position, almost touching the ceiling and markedly constricting the chamber a t this point. Where the root canals in Pre consul mandibular molars tend to be wide, even in adult specimens, they are very narrow in the Tabarin teeth. Proconsul molar roots are relatively longer than those in the Tabarin mandible, and as a consequence, the alveolar canal takes a lower course through 34 S. WARD AND A. HILL the corpus than it does in KNM-TH 13150. The Tabarin and Proconsul molars share a pattern of symmetrical root lengths from the distal root of Pm4 through the mesial root of M3. When the subocclusal anatomy of the Tabarin mandible is compared to the South Asian Sivapithecus sample, considerably more similarities emerge than was the case in the Tabarin-Proconsul comparison. Both Tabarin and the two best sampled Sivapithecus species, S. indicus and S. sivalensis, share thick molar enamel, low pulp chambers with a n intrusive and thick floor, low, blunt pulp horns, and narrow root canals. Also similar is an anteroposteriorly thickened distal molar root, and roots diverging from the furcation a t a small angle. Finally, both Tabarin and Asian Sivapithecus share a pattern of bifid root apices. There are also some notable differences. The most consistent subocclusal feature that segregates the Tabarin specimen from all known Sivapithecus mandibles is a pronounced change in root lengths between the first and second molars. S. indicus and S. sivalensis always show this shift in molar root lengths; in fact, it is one of their most diagnostic features. The Tabarin mandible has a distal Pm4 root, as well as first molar roots that are equivalent in length to the second molar roots, and to the mesial root of M3. There are also differences in the transverse dimensions of the molar roots. Sivapithecus mesial molar roots are equal in buccolingual breadth to their distal counterparts, or are slightly narrower. Conversely, in the Tabarin mandible the mesial molar roots are always broader than the distal. Although there are differences in surface contours that distinguish the Tabarin and Lothagam mandibles, they share a common subocclusal pattern. The Lothagam mandible preserves the roots of the first and second molars and the mesial roots of the third molar. Although the Lothagam MI is larger than that of Tabarin, the pulp chamber in both is low and constricted. The floor of the pulp chamber in the Lothagam MI contacts its ceiling. This severe vertical compression is in part the result of secondary dentin deposition in response to heavy occlusal wear. There are in fact no notable differences in the subocclusal morphology of these two mandibles. While it appears that the roots of the Lothagam MI are shorter than those of the second and third molars, this difference is the result of MI having been partially extruded from its socket during fossilization. Other similarities include bifid root apices and small root canals and a pattern of root angulation not found in any earlier hominoid. Both the Lothagam and Tabarin mandibles share a complex arrangement of molar root orientation in which the mesial root, which is always broader than its distal counterpart, is more vertically implanted than the distal. The latter root is axially aligned in frontal projection, with its apex slanting towards the buccal cortical plate of the alveolar process. This combination of broad and vertically oriented mesial roots, along with narrower, angled distal roots, produces a serrate root system when the molars are viewed lingually. The only other hominoid taxon with this subocclusal molar pattern is Australopithecus afarensis. Although they span a large size range, the mandibles from Laetoli and Hadar share a common pattern of subocclusal anatomy, a pattern that they also share with the Tabarin mandible. Despite the extensive size variation, the Hadar mandibles have low, vertically constricted pulp chambers, narrow root canals, roots of equal length throughout the premolar-molar series, bifid molar root apices, molar roots broader mesially than distally, vertically implanted mesial roots, and bucally inclined distal roots, producing a pronounced serrate molar root pattern (Ward et al., 1982). All of these characters are particularly well shown in LH-4, the type specimen of A. afarensis. In these features it is not possible to distinguish the Tabarin and Lothagam mandibles from A. afarensis mandibles. The subocclusal pattern of Australopithecus africanus mandibles from Makapansgat and Sterkfontein is somewhat more variable than that presented by the Pliocene specimens from East Africa. Several of the South African specimens, but particularly MLD 18, show some similarities to Australopithecus robustus. These similarities are manifested in transversely broad and massive roots, tall and projecting pulp horns, and absence or weak development of the serrate molar root configuration. A. africanus molars also show evidence towards a n increase in root length, which also characterizes A. robustus, and A. boisei. TABARIN MANDIBLE 35 erecting a new taxon to accommodate it. When new material is available from TaAs the recent history of the Ramapithecus barin, or if A. afarensis is revised, then the issue has amply demonstrated, accurate tax- attribution may change. onomic assignment of fragmentary hominoid mandibles can be difficult without a suffiACKNOWLEDGMENTS cient sample of synchronic specimens that This work forms part of the Baringo Pacollectively preserve all major parts of the leontological Research Project, based a t Harfacial skeleton. Our assessment of the morvard and Yale Universities, and carried out phologic characters preserved in KNM-TH jointly with the National Museums of Kenya. 13150, coupled with a trait-by-trait comparaIt was financed by grants from the National tive analysis of all known major Miocene hominoid and Plio-Pleistocene hominid col- Science Foundation to David Pilbeam, John lections, directs us to the conclusion that it Barry and Kay Behrensmeyer (# BNS 81is a hominid mandible. The specimen is 408181, and to David Pilbeam, John Barry, clearly not attributable to any currently rec- and Andrew Hill (# BNS 84-075751, and by a ognized Miocene hominoid taxon, but does grant from the Louise Brown Foundation to satisfy, in the parts preserved, the diagnosis Andrew Hill and David Pilbeam. Steven of A. afarensis developed by Johanson et al. Ward acknowledges support from the Na(1979). The hypodigm of A. afarensis, as con- tional Science Foundation (# BNS 84-08126) structed by its describers, in fact embraces a and from the Foundation for Research into capacious envelope of mandibular size and the Origins of Man. For their help with the shape. It is sufficiently accommodating to field work a t Tabarin we thank Kiptalam contain both the Tabarin and Lothagam Cheboi, John Kimengich, and Sally Mcmandibles. Although the validity of the Jo- Brearty. Richard Leakey and the National hanson et al. diagnosis for A. afarensis is Museums of Kenya provided much logistical currently a matter of contention (Tobias, support, and we are grateful to the Govern1980; Olson, 19851 we believe that it accu- ment of the Republic of Kenya for research rately defines the spectrum of Pliocene hom- permission. Mary Leakey kindly allowed us to examine the Laetoli hominid material; inoid mandibular variation. To this point in our discussion, we have Richard Leakey gave permission to examine excluded the possibility that the Tabarin the Lothagam mandible; Donald Johanson mandible might represent a n unknown gave access to the Hadar material a t the taxon. However, different criteria apply in Cleveland Museum of Natural History; simply identifying a particular specimen as Philip Tobias permitted the examination of a known taxon than in describing a new specimens a t the University of the Witwataxon on the basis of a particular specimen: tersrand, as did Elisabeth Vrba a t the Transa hypodigm of a new species should show vaal Museum. We are grateful to Garniss diagnostic evidence of belonging to a partic- Curtis and Robert Drake for the results of ular family if it is to be referred to it; a n isotopic age determinations which they perindividual specimen need not. The Tabarin formed. We thank John Barry, Bobbie Brown, mandible accords with descriptive criteria for Jay Kelley, and David Pilbeam for their the Hominidae, even if it cannot be shown to helpful comments on the manuscript. Sally fulfill all definitive diagnostic criteria. This McBrearty drew Figure 1; Figures 3 and 4 are by Bobbie Brown. was the view taken in Hill (1985). Although the Tabarin mandible is incomLITERATURE CITED plete, its preserved features are concordant Ackermann, F (1953) Le Mecanisme des Machoires. Paris: with the Johanson et a1 (1978) diagnosis of Masson. Australopithecus afarensis; for these reasons, Andrews, P (1978) A revision of the Miocene hominoidea we assign KNM-TH 13150 to Australopitheof east Africa. Bull. Br. Mus. Nat. Hist. A. 30235-224. cus cf. afarensis. In suggesting a close affin- Andrews, P, and Walker, AC (1976) The primate and other fauna from Fort Ternan, Kenya. In GL Isaac and ity with this species, we are merely indicatER McCown (eds): Human Origins: Louis Leakey and ing that it is very similar to members of that the East African Evidence. Berkeley, CA: Benjamin. species’ hypodigm, is not similar to those of pp. 279-304. any other species yet known, and shows no Barry, J (in press) A review of the chronology of Siwalik characters of sufficient difference to justify horninoids. Proc. Xth Cong. Int. Primat. Soc, Nairobi. TAXONOMIC STATUS OF THE TABARIN MANDIBLE 36 S.WARD AND A. HILL Beden, M (1976) Proboscideans from the Omo Group formations. In Y Coppens, FC Howell, GL Isaac and REF Leakey (eds): Earliest Man and Environments in the Lake Rudolf Basin. Chicago: University of Chicago Press, pp. 193-208. Birchette, MG (1982) The postcranial skeleton of Paracolobus chemeroni. Ph.D. thesis, Harvard University. Bishop, WW, Chapman, GR, Hill, A, and Miller, JA (1971) Succession of Cainozoic vertebrate assemblages from the northern Kenya Rift Valley. Nature 233:389394. Brown, FH (1982)Tulu Bor Tuff at Koobi Fora correlated with the Sidi Hakoma Tuff at Hadar. Nature 300:631633. Brown, FH, and Lajoie, KR (1971)Radiometric age determinations on PLiocenePleistocene formations in the lower Omo Basin, Ethiopia. Nature 229:483-485. Chapman, GR (1971) The geological evolution of the northern Kamasia Hills, Baringo District, Kenya. Unpublished Ph.D. thesis, University of London. Chapman, GR, and Brook, M (1978)Chronostratigraphy of the Baringo Basin, Kenya Rift Valley. In WW Bishop (ed): Geological Background to Evolution in Africa. London: Scottish Academic Press, Geological Society of London, pp. 207-223. Chapman, GR, Lippard, SJ,and Martyn, JE (1978)Stratigraphy and structure of the Kamasia Range, Kenya Rift Valley. J. Geo. Soc. Lond. 135265-281. Clark, JD, Asfaw, B, Assefa, G, Harris JWK, Kurashina H, Walter, RC, White, TD, and Williams MAJ (1984) Palaeoanthropological discoveries in the Middle Awash Valley, Ethiopia. Nature 307:423-428. Coppens, Y, Maglio, VJ, Madden, CT, and Beden, M (1978)Proboscidea. In VJ Maglio and HBS Cooke (eds): Evolution of African Mammals. Cambridge: Harvard University Press, pp. 336-367. Corruccini, RS, and McHenry, HM (1980) Cladometric analysis of Pliocene hominoids. J. Hum. Evol. 9:209221. Eck, GG, and Jablonski, NG (1984) A reassessment of the taxonomic status and phyletic relationships of Papio baringensis and P a p 6 quadratirostris (Primates: Cercopithecidae). Am. J. Phys. Anthrop. 65:109-134. Hall, CM, Walter, RC, Westgate, JA, and York, D (1984) Geochronology, stratigraphy and geochemistry of Cindery Tuff in the Pliocene hominid-bearing sediments of the Middle Awash, Ethiopia. Nature 308126-31. Harris, JM, and White, TD (1979)Evolution of the PlioPleistocene African Suidae. Trans. Am. Phil. Soc.69:l128. Hill, A (1985)Early hominid from Baringo, Kenya. Nature 315:222-224. Hill, A, Curtis, G, and Drake, R (in press) Sedimentary stratigraphy of the Tugen Hills, Baringo District, Kenya. In R Renaut, L Frostick, and JJ Tiercelin (eds): Sedimentation in the African Rift System. Oxford: Blackwells, Geological Society of London. Hill, A, Drake, R, Tauxe, L, Mohaghan, M, Barry, JC, Behrensmeyer, AK, Curtis, G, Fine Jacobs, B, Jacobs, L, Johnson, N, and Pilbeam, DR (1985) Neogene paleontology and geochronology of the Baringo Basin, Kenya. J. Hum. Evol. 14:759-773. Howell, FC (1978) Hominidae. In VJ Maglio and HBS Cooke (eds): Evolution of African Mammals. Cambridge: Harvard University Press, pp. 154-248. Ishida, H, Pickford, M, Nakaya, H, and Nakano, Y (1980) Fossil anthropoids from Nachola and Samburu Hills, Samburu District, Kenya. African Study Monographs (Sumlementarv Issue 2) DD. 73-85. Johanson, DC, and White T (1980) On the status of Australopithecus afarensis. Science 207:1104-1105. Johanson, DC, White, T, and Coppens, Y (1978) A new species of the genus Australopithecus (Primates: Hominidae) from the Pliocene of eastern Africa. Kirtlandia 28:1-14. Johanson, DC, Taieb, M, and Coppens, Y (1982)Pliocene hominids from the Hadar Formation, Ethiopia (19731977): Stratigraphic, chronologic, and paleoenvironmental contexts, with notes on hominid morphology and systematics. Am. J. Phys. Anthrop. 573373-402. Kalb, JE, Jolly, CJ, Mebrate, A, Tebedge, S, Smart, C, Oswald, EB, Cramer, D, Whitehead, P, Wood, CB, Conroy GC,Adefris, T, Sperling, L, and Kana, B (1982a) Fossil mammals and artefacts from the Middle Awash Valley, Ethiopia. Nature 298:25-29. Kalb, JE,Jolly, CJ, Tebedge, S, Mebrate, A, Smart, C, Oswald, EB, Whitehead, P, Wood, CB, Adefris, T, and Rawn-Schatzinger, V (1982b) Vertebrate faunas from the Middle Awash Valley, Afar, Ethiopia. J. Vert. Paleontol. 2(2):237-258. Kalb, JE, Oswald, EB, Mebrate, A, Tebedge, S, and Jolly, CJ (1982~)Stratigraphy of the Awash Group, Middle Awash Valley, Afar, Ethiopia. Newsl. Stratigr. 11(3):95-127. Kay, RF (1981) The nutcrackers-a new theory of the adaptations of the Ramapithecinae. Am. J. Phys. Anthrop. 55141-152. Kimbel, WH, White, TD, and Johanson, DC (1985)Craniodental morphology of the hominids from Hadar and Laetoli: Evidence of Parunthropus and Homo in the mid-Pliocene of Eastern Africa? In E Delson (ed)Ancestors: The Hard Euidence. New York: Alan R Liss pp. 120-137. Leakey, MD, Hay, RL, Curtis, GH, Drake, RE, Jackes, MK, and White, TD (1976) Fossil hominids from the Laetolil Beds. Nature 262:460-466. Leakey, REF (1969) New Cercopithecidae from the Chemeron Beds of Lake Baringo, Kenya. In LSB Leakey (ed): Fossil Vertebrates of Africa, Vol. 1. New York: Academic F’ress, pp. 53-70. Leakey, REF, and Walker, AC (1985) New higher primates from the early Miocene of Buluk, Kenya. Nature 318:173-174. Maglio, VJ (1973) Origin and Evolution of the Elephantidae. Trans. Am. Phil. SOC.63t1-144. Martin, LB (1981) New specimens of Proconsul from Koru, Kenya. J. Hum. Evol. 10:139-150. Martin, LB (1984) The relationships of the Later Miocene Hominoidea. Unpublished Ph.D. thesis, University of London. Martin, LB (1985) Significance of enamel thickness in hominoid evolution. Nature 3141260-263, Martyn, JE (1967) Pleistocene deposits and new fossil localities in Kenya. Nature 215:476-477. Martyn, J E (1969)The geological history of the country between Lake Baringo and the Kerio River, Baringo District, Kenya. Unpublished Ph.D. thesis, University of London. McCall, GJH, Baker, BH, and Walsh, J (1967)Late Tertiary and Quaternary sediments of the Kenya Rift Valley. In WW Bishop fed): Background to Evolution in Africa. Chicago: University of Chicago Press, pp. 191-220. McClenaghan, MP (1971) Geology of the Ribkwo area, Baringo District, Kenya. Unpublished Ph.D. thesis, University of London. Mebrate, A, and Kalb, JE (1985) Anancinae (Proboscidea: Gomphotheriidae) from the Middle Awash Valley, TABARIN MANDIBLE Afar, Ethiopia. J. Vert. Paleontol. 5r93-102. Olson, TR (1985) Cranial morphology and systematics of the Hadar Formation hominids and “ Australopithecus africanus” In E Delson (ed): Ancestors: The Hard Evidence. New York: Alan R. Liss, pp, 102-119. Patterson, B, Behrensmeyer, AK, and Sill, WD (1970) Geology and fauna of a new Pliocene locality in northwestern Kenya. Nature 226:918-921. Pickford, M (1975a) Stratigraphy and palaeoecology of five late Cainozoic formations i n the Kenya Rift Valley. Unpublished Ph.D. thesis, University of London. Pickford, M (1975b) Late Miocene sediments and fossils from the northern Kenya Rift Valley. Nature 256:279284. Pickford, M (1982) New higher primate fossils from the middle Miocene deposits a t Majiwa and Kaloma, western Kenya. Am. J. Phys. Anthrop 58:l-19. Pickford, M, Johanson, DC, Lovejoy, CO, White, TD, and Aronson, JL (1983) A hominoid humeral fragment from the Pliocene of Kenya. Am. J. Phys. Anthrop. 60r337346. Pilbeam, DR (1969) Tertiary pongidae of east Africa: evolutionary relationships and taxonomy. Peabody Mus. Nat. Hist. (Yale Univ.), Bull. 31:l-185. Pilbeam, DR, Rose, MD, Badgley, C, and Lipshutz, B (1980) Miocene hominoids from Pakistan. Postilla 18I:l-94. Schmitt, TJ, and Nairn, AEM (1984) Interpretations of the magnetostratigraphy of the Hadar hominid site, Ethiopia. Nature 309t704-706. Senut, B (1983) Quelques remarques a propos d‘un humerus d’hominoide pliocene provenant de Chemeron (bassin du lac Baringo, Kenya). Folia Primatologia 41:267-276. Sibley, CG, and Ahlquist, J E (1984) The phylogeny of hominoid primates, a s indicated by DNA-DNA-hybridisation. J. Molec. Evol. 20:2-15. Sperber GH (1985) Comparative primate dental enamel thickness: A radiological study. I n PV Tobias (ed) Hominid EuolutioK Past, Present, and Future. New York: Alan R Liss pp. 443-454. Tobias, PV (1967) Pleistocene deposits and new fossil localities in Kenya. Nature 215:478-480. Tobias, PV (1980) Austrulopithecus ufurensis and A ustrulopithecus africanus: critique and alternative hy- 37 pothesis. Palaeontol. Afr. 23:l-17. Wallace, J (1978) Evolutionary trends in the early hominid dentition. In C Jolly (ed) Early Hominids ofAfrica London: Duckworth pp. 285-310. Ward, S, Johanson, DC, and Coppens, Y (1982) Subocclusal morphology and alveolar process relationships of hominid gnathic elements from the Hadar Formation: 1974-1977 collections. Am. J. Phys. Anthrop. 57r605630. Ward, S, and Brown, B (1985) The facial skeleton of Siuapithecus indicus. In DR Swindler (ed): Comparative Biology of Primates. New York: Alan R. Liss (in press). Ward, S, and Hill, (in preparation, a) A The hominid temporal from Chemeron, Baringo, Kenya. (To be submitted to Am. J. Phys. Anthrop.) Ward, S, and Hill, A (in preparation, b) The hominid mandible from Lothagam, Kenya. (To be submitted to Am. J. Phys. Anthrop.) Weidenreich, F (1936) The mandibles of Sinanthropus pekznenszs: A comparative study. Paleontol Sinica Series D 7:l-162. White, TD (1977) New fossil hominids from Laetoli, Tanzania. Am. J. Phys. Anthrop. 46t197-230. White, TD (1977) New fossil hominids from Laetoli, Tanzania. Am. J. Phys. Anthrop. 46,197-230, White, T D (1985) The hominids of Hadar and Laetoli: An element-hy-element comparison of the dental samples. In E Delson (ed): Ancestors: The Hard Evidence. New York: Alan R. Liss, pp. 138-152. White, TD, and Johanson, DC (1982) Pliocene homini; mandibles from the Hadar Formation, Ethiopia: 197’ 1977 collections. Am. J. Phys. Anthrop. 57501-544 White, TD, Johanson, DC, and Kimbel, WH (1981) Australopithecus africanus: its phyletic position reconsidered. S. Afr. J. Sci. 77r445-470. Wood, BA, Abbot, SA (1983) Analysis of the dental morphology of Plio-Pleistocene hominids I. Mandibular molars-crown area measurements and morphological traits. J. Anat. 136:197-219. Wood, BA, Abbot, SA, and Graham, SA (1983) Analysis of the dental morphology of Plio-Pleistocene hominids 11. Mandibular molars-study of cusp areas, fissure pattern and cross sectional shape of the crown. J. Anat. 137r287-314.