Herbert Thomas

Chaire de Pal&mnthropologie et de Prghistoire, CollZge de France, 11, Place 1~4arcelin~Berthelot, 75231 Paris cedex 05, France

$evket $en

Cniuersite Pierre et Marie Curie, Laboratoire de Pal~ontologie des VerGbbr&, UA 720 CNRS, 4, Place

Jussieu, 75252 Paris cedex 05, prance

Jack Roger

Bureau de Recherches GMogiques et !Mini&es, O&am-la-Source B.P. n” 6009. 45KO OrlPans cedex 2, I*‘rance

Zaher Al-Sulaimani

Ministyv qf Petroleum and Minerals, Directorate General of Minerals. P.O. Box 551, Muuscatz Sultanate of O?mm

Received 7 November 1989

Revision received 16 March

1990 and accepted 12 April 1990

&words: Primates,

Anthropoidea,

Propliopithecidae, Moeripithecus, Oligocene, Sultanate of Oman

The discovery of Moeripithecus markgrafi Schlosser (Propliopithecidae, Anthropoidea, Primates), in the Ashawq Formation (Early Oligocene of Dhofar Province, Sultanate of Oman)

Several isolated teeth, a fragment of mandible with Ps-Ms, and four

associated upper cheekteeth (P-M’) of a large propliopithecid were

recovered from the early Oligocene locality of Taqah (Ashawq Formation,

Sultanate ofOman). All these teeth, as well as the paratype ofPropliopithecus ankeli, are here attributed to the Fayum anthropoid A4oeri,fCthccus markgmji Schlosser. This taxon was until recently known only from the holotype, a

fragment of mandible with M-M2,

Considering the striking differences in morphology between IW. markgraj and P. hneckeli (type species of the genus Pmpliopithecus), confirmed and

extended by the new material from Oman, we here conclude that Schlosser

was justified in recognizing the generic distinction between ,Woeripithecus and Prop1io.ithecu.r.

Journal ofHuman Boolution (1991) 20, 33-49

Introduction

A brief preliminary reconnaissance to the southern part of Dhofar Province (Sultanate of Oman) made in June 1987 led to the discovery of two fossiliferous localities-Thaytiniti and Taqah-at the base of a transgressive marine unit, the Ashawq Formation, which dates to the extreme base of the lower Oligocene (Thomas et al., 1989). Despite the limited quantity of sediments sampled at Taqah (only 15 kg), the locality yielded two isolated primate molars. One, a left R/12, was provisionally attributed to Catarrhini incertae sedis (cJ:

Oligopithecus savagei) the other, a right Ml(?), to Propliopithecidae (cJ: Aegyptopithecus)

(Thomas et al., 1988). During the second Omani-French Paleontology Expedition in March 1988, more than

2.8 tons of sediments sampled at Taqah were washed and screened, leading to the recovery of several dozen primate teeth as well as numerous tooth fragments. These remains, eight isolated teeth, a fragment of mandible with Pa-M3 and the canine root (TQ4) as well as four upper cheekteeth (P3-M2) belonging to a single individual (TQ3), indicate the existence of an anthropoid about the size of the largest anthropoids from the Qatrani Formation of the Fayum, Aegyptopithecus zeuxis and Propliopithecus ankeli (see Tables 1 and 2).

0047-2484/91/010033 + 17 $03.00/O 0 1991 Academic Press Limited

34 H. THOMAS ET AL.

Table 1 Dimensions of the lower teeth of the Propliopithecids (Taqah and Fayum)

Specimen C PS P4 Ml M2 M3 P,-M, Number I w 1 w 1 w 1 w 1 w lw length

Ailoeripithecus markgraj from Taqah

TQI 1 TQ6 TQlO

T&7 TQ8 TQ4

Type of hfoeripithecus markgraf’ SNM 12639 G

Paratype of P. ankeli’

DPC 5392 Right Left

Type of P. haeckeli’ SNM 12638

Right Left

P. chirobates’ mean A. reuais’ mean

7.1 5.1 5.8 3.8 5.8 3.9

5.7

59 7.1 4.4 5.5 5.5 5.2 5.9

4.8

6.7 5.1 5.0 5.9 4.2 5.2 5.3 5.4 5.2 5.7 6.2 5.6 20.4 6.6 5.1 4.5 5.8 4.2 5.2 5.2 5.5 5.2 5.6 6.4 5.6 19.8

4.4 2.9 4.2 3.5 3.4 4.0 4.7 4.6 4.9 4.7 5.1 4.0 17.4 4.2 3.2 3.5 3.8 4.8 4.8 4.8 4.8 5.1 4.1 17.2

5.5 4.0 5.5 3.5 3.8 4.1 5.0 4.8 5.4 5.4 5.9 4.9 18.87 6.7 4.6 6.3 3.9 4.5 4.8 5.9 5.3 6.8 6.5 8-2 6.4 21.6

5.0 or 5.7

4.7 5.1

5.0 6.7 5.0

5.0 6.2 5.0 21.0

5.3

1 Measurements from Kay el al., 1981. Measurements from Simons et al., 1987.

Despite excavation, washing and screening of almost 800 kg of sediment at Thaytiniti, no dental remains of large anthropoids have as yet been found there.

All the Taqah teeth, which form the basis for this paper, belong to Propliopithecidae and, as we shall see later, are attributed to Moeripithecus markgruj. The Ml (2) (TQl) found in 1987 at the same locality and also attributed to the same family (Thomas et al., 1988), seems to be distinct from the form described below. It is much smaller than the Ml of all known propliopithecids. According to Rasmussen & Simons (1988) this tooth bears “features that are resemblances to P. markgraj and Oligopithecus rather than to Aegyptopithecus”

Abbreviations: SNM = Staatliches Museum fiir Naturkunde, Stuttgart; CGM = Cairo Geological Museum; DPC = Duke Primate Center; TQ = Taqah locality (Oman) (field

Table 2 Dimensions of the upper teeth of the propliopithecids (Taqah and Fayum)

Specimen 1’ Ill’) P” p’ M’ W M2/1” P.-M2

NumbeI I w 1 w 1 IV 1 w 1 w I w 1 b Irngth

Moeri@thacus markgraj from Taqah

TQ5 4.7 4.1 TQ12 3.9 4.1 TQ3 4.6 6.8 3.8 7.1 5.1 7.6 4.9 8.0 (17.7)

Type of P. ankeli’ CGM 42847 3.6 7.5 3.9 7.4 5.8 7.8 6.6 8.7 1.6 7.4 20.8

P. chirobdes’ mean 3.2 3.7 4.7 6.6 5.1 7.3 A. zeuxis’ mean 4.75 7.2 3.9 6.9 .5.7 74 6.65 9.7 6.2 9.6

’ Measurements from Simon3 et al., 1987.

MOERXPITHECUS MARKGRAFI DISCOVERY IN ASHAWQ FORMATION 35

numbers). All the specimens are kept in the Natural History Museum at Muscat (Sultanate of Oman).

Systematics

Family Propliopithecidae Straus, 1961 Moer$ithecus Schlosser, 19 10

Emended diagnosis: Large propliopithecid, slightly smaller in size to Aegvptopithecus

teuxis; robust mandible with relatively deep mandibular ramus in adult; strong development of the lower premolars; very broad P 3, at least in males; reduced buccal cingulum on the lower molars. Differs from all other propliopithecids by the greatly Ilaring buccal margins of lower molar crowns, the hypoconulid closely approximating the entoconid, the M2 slightly shorter mesiodistally than M’, the relatively small hypocone, and the presence of a small “pericone” on the antero-internal corner of the lingual cingulum of Ml-M2. Differs from A. zeuxir by the very slight molar size increase posteriorly in the lower molars and by the M2 even shorter than M’, the absence of “beaded” condition of the inner cingula shelf of the upper molars and the absence of distal foveae on the lower molars. Differs from P. haeckeli mainly by its much larger size, the large and oval, obliquely oriented P3 and from P. chirobates by its larger size and the very strong development of the lower premolars.

Type species: Moeripithecus markgraj Schlosser, 19 10.

Emended diagnosis: Same as for the genus.

Synonymy:

Propliopithecus markgra$ (SCHLOSSER) - Simons, 1967. Propliopithecus haeckeli SCHLOSSER-Gingerich, 1978. Propliopithecus ankeli SIMONS et al., 1987 (Paratype only).

Holotype: a right mandibular corpus with Ml-2 (SNM 12639G) of a juvenile (see Simons, 1967 and Delson, 1975) or subadult individual (Kay et al., 1981).

Referred specimens:

Taqah: right mandible with Pz-Ms and the canine root (TQ4); right M, or M2 (TQ7): right Ph (TQ6); right Ms (TQ8); left P4 (TQIO); left lower canine (TQI 1); four isolated upper cheekteeth (P3-M2) belonging to a single individual (TQ3); upper central decidu- ous(3) incisor (TQ12); right upper lateral incisor (TQ5); fragment ofleft M’ or M” (TQ9). Fayum: lower jaw with right and left C-M:3 (DPC 5392).

Horizon and type locality: Jebel Q, t a rani Formation, Fayum, Egypt; type locality unknown.

Distribution: Ashawq Formation, locality of Taqah (Dhofar, Sultanate of Oman) and Jebel Qatrani Formation, Quarry V, 165 metre level (Fayum, Egypt).

Description of Taqah specimens

Mandible and lower dentition (Figures l-5, 8‘464-D)

Dental formula: ?. 1.2.3. Judging from its dimensions and robustness, the right mandible (TQ4) closely recalls

36 H. THOMAS ET AL.

those of Aegyptopithecus and Propliopithecus ankeli. Even though the base of the jaw is partly broken, one can estimate that its height below M2-Ms would be about 13-14 mm (Figure 2). The portion of the canine that is preserved has a cross-sectional area (length x width) of 27.0 mm2 just below the crown, which is large and suggests it is a male. The position of the canine alveolus shows that there was no diastema between it and the Pa.

The premolars are striking because oftheir strong development, especially that of the Ps. The occlusal outline of the lower premolars is broadly oval (Figure 1). The long axis ofthe Ps is obliquely oriented (mesiobuccal to distolingual) which is due to an expanded distolingual area with respect to the axis of the toothrow. No trace of a buccal cingulum is visible on the two premolars. The P3 has dimensions larger than those of males of Aeg,@topithecus (cj DPC 1028) but not larger than CGM 40137 and DPC 1112. The Pa possesses a high protoconid, a very small paraconid and a slightly more developed distal tubercle. The latter emerges from a conspicuous cingulum which borders the entire lingual side of the crown. The long axis of the P+ is virtually parallel to that of the Ps. The protoconid and the metaconid are clearly separated. The bucco-lingually compressed trigonid and the talonid are clearly delimited by crests. There is no buccal cingulum. The first and the second molars have almost subrectangular occlusal outlines. The lengths of the first two molars are only about 6-8% greater than their breadths (Table 1). In their lengths, Mr is smaller than Mz which is smaller than MS, each tooth in this sequence being 6-10% longer than the one in front of it. The principal characteristic of the lower molars concerns the quite pronounced basal inflation, which imparts to the occlusal table of the crowns a ‘iconstricted” look. This is due essentially to the flaring buccal margin. This constricted morphology is enhanced by the absence of a distinct cingulum, which appears instead as a slight buccal swelling in the notch between the hypoconid and the protoconid. The degree of constriction of the crowns of Mi and Mz, as estimated from the ratio of intercuspal width to crown width at the talonid, is 0.46 for the Mi and 0.43 for the M2. The trigonid of both molars has no paraconid, and is very reduced whereas the talonid is elongated. In the MI the metaconid is located slightly to the rear of the line of the protoconid. The protoconid and the metaconid are higher than the talonid cusps. The oblique cristid connects with the posterior wall of the trigonid beneath the protoconid. The hypoconulid is well-developed, and situated towards the midline ofthe crown, but closer to the entoconid than to the hypoconid. The sulcus which separates the hypoconulid from the entoconid is, however, better developed than in the type of M. markgraj. The entoconid possesses a pre-entocristid joining the post-metacristid. There is not really a cristid joining the entoconid to the hypoconulid except in Ms. There is no clear distal fovea on Mi or M2, having been partly obliterated by small wear facets. There is no distal fovea on the MS.

Isolated lower teeth (Figures 5, 8B-D)

The sample consists of one left lower canine (TQ 11) ; one right P4 (TQ6); a left P1. (TQlO); a right Mi or M2 (TQ7) and a right Ms (TQ8).

The isolated lower cheekteeth scarcely differ from those in the mandible described above. The dimensions of the lower canine, (TQll), when compared with those of Aegytopithecus, suggest that it represents a male (Figure 8D).

The lower canine (TQll) is voluminous and has an oval occlusal outline. The cross- sectional area taken just below the crown is 29.0 mm2 (6.9 X 4.2), a value which is close to that measured in the canine of the mandible. Broken near the mid-height of the protoconid, this canine is characterized by the absence of a crest on the distal surface of the protoconid.

4OERIPITHECUS MARXGRAFI DISCOVERY IN ASHAWQ FORMATION 37

38 H. THOMAS ET AL.

The mesial crest of the protoconid is well developed and reaches the base of the crown, along which runs a cingulum that ends at the distal edge as a slight cingula swelling.

U@er dentition (Figures 1, 6 and 8&G)

Four associated cheekteeth, P3 to M’ (TQ3), clearly belong to a single individual (Figures 6 and 8F). They were collected in situ closely associated with each other, but not in an anatomical relationship, in a block of sediment less than 2 cm3. Considering the dimensions and proportions of these teeth, it is equally probable that they belong to the same species as the mandible. In fact, it is likely, although we cannot be certain, that these upper teeth belong to the same individual as the mandible. The two samples were found close together. They are the only two relatively complete specimens collected from 2.8 tons of sediments treated. The colour of the enamel in both is identical, whereas none of the other primate teeth from the site (several dozen isolated or fragmentary specimens) has the same colour. In addition, the stage of wear of the cusps in both sets of teeth is the same. Finally, in occlusion, there is a close fit between the specimens.

Upperpremolars (Figures 6 and SF). Both premolars are two-rooted. On P” the buccal root is traversed by a single pulp canal, even though the root originated from two roots, as evidenced by shallow valleys on the external and internal surfaces. The paracone of both premolars is well developed, that of P3 is better developed than that of the P’. The trenchant pre- and post-paracristae are also well developed. The protocones of P” and P’ are clearly less well developed. The occlusal outline ofthe P3 is triangular; that ofthe P” has a mesio-distal development of the lingual surface (the largest mesio-distal dimension is in the median part of the tooth). The post-protocrista of the P” is short and encloses a clear distal fovea. The pre-protocrista traverses the entire mesial surface thereby delimiting a less developed mesial fovea. These last features are rather similar in the P3, but the foveae are narrower. On this tooth the pre-protocrista reaches the buccal surface in front of the paracone where it forms a small stylar swelling. On P” there is a moderately well developed lingual cingulum. A comparable structure does not occur in P”. The buccal cingulum is absent in P“; it is barely visible in P3.

Upper molars (Figures 6 and SF). M’ and M2 are subequal in size. There is no size increase from front to back in the first two molars. The M’ has a trapezoidal occlusal outline. The

MOERIPITHECUS MARKGRAFI DISCOVERY tN ASHAWQ FORMATION 39

lingual margin is shorter mesiodistally than the buccal one. The occlusal outline ofthe M2

is similar to that ofM1 save for the fact that the buccal margin is oblique with respect to the

median axis of the tooth in M*. The position of the paracone with respect to the metacone is

more buccally placed in M* than it is in M’, so that there is a greater relative distance

between the paracone and the metacone.

In both molars the hypocone, which is much smaller than the protocone and originates

from the cingulum, is relatively poorly developed. In the Taqah molars the occlusal surface

6.0

5.0

4.0

8.0

7.0

6.0

5.0

4.0

8.0

70

6.0

5.0

W I I I I I I I

-

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

M2

- TQ4

X

W I I I I I I I I

4.0 4.5 5.0 5.5 6.0 6.5 70 7.5

W I 1 I I I

4.0 6!0 1 I

4.5 5.0 5.5 6.5 7.0 75

+ OMAN A P chirobotes 8

A A. zeuxis 8 o P. chirobates 9

. A. zeuxis _I!

X “P.ankeii” (Paratype)

H = P. haeckeli ( Type ) M= M. markgrafi ( Type)

Figure 3. Bivariate plots (in mm) oflower tooth length (L) against lower tooth width (W), after Simons rl al. (1987) (N.6.: scale modified). In the original version of this figure, Simons ef al. (1987: Figure 5) inverted the data points relating to the P. ankeii paratype. The data points relating to the other propliopithecids are correctly plotted. In our figure we plot the correct position of the data points of the P. nnkeli paratype.

40 H. THOMAS ET AL.

Figure 4. Moeripithecus markgraj: right upper lateral incisor TQ5, lingual view

of the hypocone, although set apart from that of the protocone, is closer to it than is the case in Aegypto@thecus. In addition, the hypocone is slightly lingual to the protocone. The cingulum bordering the mesial and lingual margins of the protocone does not have the beaded morphology characteristic ofilegyptopithecus. There is no buccal cingulum on M’. In contrast, on M’ there is a very small buccal cingulum running along the base of the paracone and swelling at the confluence of the pre-metacrista and the post-paracrista. The well developed crista obliqua (post-protocrista + hypometacrista) reaches the base of the

0' 5mm Figure 5. Moeripithecus markgraji: left lower canine TQl 1 (male), lingual view.

MOERIPITHECUS MARKGRAFI DISCOVERY IN ASHAWQ FORMATION 41

t ’ a ’ B i 0 5mm

Figure 6. Moorripithecus markgraj: right upper dentition (P”-M2) of TQ3, isolated teeth of the same individual.

metacone. Another crest begins at the paracone and, extending lingually towards the protocone, crosses the pre-protocrista delimiting a very narrow anterior fovea. Neither paraconule nor metaconule are discernible. There is a distinct but small or rudimentary “pericone” on the antero-internal corner of the lingual cingulum in both molars.

Upper incisors (TQ5 and TQl2) (Figures # and 8$ G)

The right upper lateral incisor (TQ5) has a basically triangular occlusal outline. The cross- sectional area measured just below the crown is 14.3 mm2 (4.6 X 3.1). The mesial crest of the protocone is well developed, reaching the base ofthe crown and extending back into the disto-lingual cingulum. A small wear facet occurs on the distal surface of the protocone, but it does not affect the cingulum. The distal crest of the protocone joins the base of the crown and the cingulum via a small distal tubercle.

One upper central deciduous(?) incisor (TQ12) is tentatively referred to the same species. It is triangular in occlusal outline, has a curved incisive margin and a moderately well developed lingual cingulum.

Comparisons and discussion

The Taqah anthropoid, represented principally by a mandible with Ps to Ms (TQ4) and by four associated upper cheekteeth (TQ3), d oes not belong to the Parapithecidae. It differs from all known Fayum parapithecids and from Biretia piueteaui from Bir el Ater (Algeria) (de Bonis et ul., 1988) in the following characters:

- its large size, close to that of AeQptopithecus zeuxis and Propliopithecus nnkeli;

- the loss of PZ and (P’?); - heteromorphy of Ps.4 which contrasts with the structure and homomorphy of the two

rear lower premolars of parapithecids; - the absence of constriction between the trigonid and the talonid; - the absence of large bulbous isolated conules in the upper cheekteeth; - the absence of a paraconule in the upper premolars; - a relatively deep mandibular body; - lack of a sulcus between the protoconid and the metaconid in P4-Mt: the two cusps

are joined by a crest.

42 H. THOMAS ET AL

P4

IO W I I I 1 I 1 I I

5.0 6.5 7.0 %5 8.0 8.5 9.0 9.5

6.0

55

5.0

6.5 M2

6.0

5.5

5.0 * TQ3

w I I I I I I

6.0 65 7.0 25 0.0 85 9.0 9.5

Sr OMAN X P. ankeli

l A.zeuxis 0 P. chirobates

Figure 7. Bivariate plots (in mm) of upper tooth length (L) against upper tooth width (xv), after Simons et ul. (1987). (N.B.: scale modified.)

In contrast, the dental characters of the Taqah anthropoid closely recall those of the second group of early anthropoids from the Fayum, the propliopithecids, notably in the loss of P2, the premolar heteromorphy, the sectorial honing Pa (these three characters are probably linked), their bulbous-cusped, bunodont molars, the hypoconulid placed more or less in the midline of the tooth, and the relatively deep mandibular body.

Since 1967, Simons has suggested that Moeripithecus is a synonym of Propliopithecus and that the Propliopithecidae of the Fayum comprise two genera only, Propliupithecus and Aegyptopithecus, which are sometimes themselves considered as synonyms (Szalay & Delson, 1979; Swindler & Olshan, 1982; Andrews, 1985). According to the former concept, these two genera contain five species, four belonging to Propliopithecus (P. markgraj, P. haeckeli, P. chirobates and P. ankeli) and one to Aegyptopithecus (A. zeuxis). The first two species are known with certainty only from the holotypes, both of which were collected at the

MOERIPITHECUS MARKGRAFI DISCOVERY IN ASHAWQ FORMATION

- Ab

0’ IOmm

Figure 8. Mueripith~cus mnrkgrajfrom Taqah (Ashawq Formation, Oman): A. right mandible with Ps-MY, TQ4: (a) medial view, (b) lateral view, (c) occlusal view; B. right PJ, TQ6; C. right MS, TQ8; D. left lower canine, TQI 1; E. right upper lateral incisor, TQ5; F. right upper dentition with P”-M”, TQ3: G. upper central deciduous(?) incisor. TQl2.

beginning of the century from unknown parts of the Jebel Qatrani Formation. P. ankeli,

discovered and described more recently (Simons et al., 1987), came from Fayum Quarry V, the 165 m level ofthe Jebel Qatrani Formation. As for the species P. chirobates, it represents the best known among the four species, and has been found in Quarries I and M, approximately 80 to 85 m above Quarry V.

Two new genera, Catopithecus and Proteopithecus, have recently been named by Simons (1989), from a new Fayum site at about the 47 m level ofJebe1 Qatrani. Placed by him in a new subfamily, the Oligopithecinae, within the Propliopithecidae, the two genera appear to be related to Oligopithecus. Catopithecus and Proteopithecus differ from all other known propliopithecids by their smaller size. Unlike the “propliopithecines”, Catopithecus browni is characterized by the presence of a paraconid on M ,, the less bulbous cusps and sharper crests on upper premolars and molars, and by the relatively much smaller hypocone.

Proteopitheczzs s_ylzGze differs from the “propliopithecines” principally by well developed cingula which surround the upper molars and by the presence of a small mesostyle nodule (Simons, 1989).

Comparisons with Aegyptopithecus zeuxis The Taqah propliopithecid differs from ilegyptopithecus principally by having a very slight molar size increase from front to rear: M1 is 7.3% shorter (mesio-distal length) than My. For the two anterior upper molars, the size increase is even less (5.3% for the width) or the opposite (M’ slightly shorter than M’) (Figure 7). In =le@topithecus, M2 is markedly larqer

44 H. THOMAS ET AL.

than M’ and Mi averages 17% shorter (m-d) than M2 (Fleagle & Kay, 1983: Table 1). The ratio of the upper molar dimensions from Taqah, and thus their occlusal outline, immediately distinguishes them from those of Aegyptopithecus. The M’ from Taqah is 49% broader (bucco-lingual breadth) than it is long (m-d length). According to Kay et al. (1981), this ratio is only 40% in Propliopithecus and 30% in Aeg,@topithecus. Additionally, the Taqah first and second upper molars bear relatively small hypocones. This is in marked contrast to the great distolingual development of this cusp in molars of Aegvptopithecus in which the occlusal outline of the molars is consequently much more rhomboidal. The Taqah upper molars do not have the “beaded” morphology of the prominent inner cingula shelf which is so characteristic of Aeuptopithecus.

The Taqah lower molars differ from those of Aeaptopithecus by the absence of a true buccal cingulum, which is well developed in ilegyptopithecus zeuxis, and by the absence of clear distal foveae in the molars. The molar cusps are situated closer to the lingual and labial margins of the crowns than in A. zeuxis. The crests are better developed than in A. Zeuxis and the cusps are less-broad based and “puffy”. However, the degree of constriction of the molar crowns (Ml and My of TQ4), as estimated from the ratio of intercuspal talonid width to maximum crown width at the talonid (0.46 for the Ml and 0.43 for the Ms), is very close to that of A. teuxis (0.46 for the M 1 and 0.47 for the M2) (Kay et al., 1981: 309).

Although the height of the mandible below Mz-Ma is comparable to the smallest specimen ofAeoptopithecus zeuxis, the buccal swelling of the mandibular body below MT-Ms is, however, less pronounced in M. markgraji from Taqah.

Comparison with Propliopithecus haeckeli In the morphology and proportions of the teeth, the Taqah form is clearly different from Propliofiithecus haeckeli, the type species of the genus, of which only the lower dentition (C-Ms) is known. The holotype of P. haeckeli belongs to a male according to some authors (Simons et al., 1987)) or a female according to others (Kay et al., 1981; Fleagle & Kay, 1983). The principal difference between P. haeckeli and the Taqah species lies in the morphology of the anterior lower premolar (Pa), which may reflect sexual dimorphism. The strong development of the P3 contrasts dramatically with the considerably smaller dimensions of this tooth in P. haeckeli. By comparison with the Taqah specimen, the Ps of P. haeckeli is much less sectorial in occlusal outlme, the crown is nearly round and is low, and the long axis of the tooth is almost at right angles to the long axis of the toothrow, rather than oblique. As far as the Taqah molars are concerned, the main difference from those of P. haeckeli is the absence of the prominent buccal cingulum. A buccal cingulum is also present on P4 of P. haeckeli, which is lacking in the Taqah specimen, but this may not be relevant given population differences in cingulum development in other primates.

Comparison with P. chirobates The differences between the Taqah form and the three other large propliopithecids are, by contrast, more subtle. It differs from P. chirobates by having a much larger and more robust mandible, probably also by the large and robust lower canine (judging from the canine root), by the strong development of the lower premolars (especially the broad Ps), by the presence of a buccal inflation at the base of M,-Ms and by the absence of a buccal cingulum in the lower molars.

MOERIPITHECUS MARKGRAFI DISCOVERY IN ASHAWQ FORMATION 45

Comparison with P. ankeli nnd M. mar-kg&

The lower dentition of the Taqah propliopithecid is extremely close to that of the paratype

of P. ankeli, despite its less pronounced buccal swelling of the corpus below M2-M3, the

greater development of its Ps, the presence in this tooth of a small, more developed distal

tubercle, the more rectangular occlusal outline of the first two molars (which are

proportionately slightly shorter and broader in P. ankeli) and the greater development of

crests than in P. ankeli. The similarities concern the following characters: presence of

buccal inflation of the lower molars (related to the flaring buccal surface), the enlargement

of the lower premolars (in particular the Ps), obliquity of the long axis of the Ps with respect

to the axis of the tooth row, the M2 only slightly larger than the MI, the weak or indistinct

cingula on the molars, the indistinct distal fovea on Ml-M2 and its absence on MS and the

absence of buccal cingula on the premolars (a buccal cingulum on PI was specified in the

diagnosis of P. ankeli by Simons et al., 1987, but noted as missing in their description on

p. 144). The differences between the two specimens are so minor that we conclude that the

two mandibles belong to the same tax-on, even at the species level.

Even though the upper teeth in the holotype of P. ankeli are very worn, leaving almost no

possibility for making valuable observations regarding their occlusal morphology, the

following differences between P. ankeli and the Taqah specimen can be determined:

- The upper premolars of P. ankeli would, according to Simons et al. (1987), possess

three roots. The Taqah propliopithecids possess only two roots.

- The M2 of P. ankeli is larger than the M’ because of its buccolingual development,

although the size increase from first to second upper molars is less pronounced than it is in

‘4. eeuxis. In contrast, only a slight increase in width can be observed between the first and

second upper molars of the Taqah form. As far as the mesio-distal length of the Taqah

upper molars is concerned, M2 is slightly shorter than R/I’, a feature which appears to be

unique among propliopithecids. All the other characters of the upper dentition, in

particular the occlusal outline of the cheekteeth, are comparable, even though, according

to Simons et al. (1987)) the occlusal outline of the upper premolars of P. ankeli are “broadly

oval or rectangular . rather than the somewhat triangular outline typical of other

species”. In fact, the oclusal outline is not clear, because the teeth are heavily worn.

~ The upper molars bear a large hypocone and a continuous lingual cingulum which

runs from the base of the hypocone around the protocone. The Oman upper molars are

quite distinctive in having a pericone and a relatively small hypocone. The pericone does

not occur in any of the large propliopithecids of the Fayum, but is known in Apidium

pkiomense (where it is well developed), in Hoangkonius stehlini (Gingerich, 1977) and perhaps

on M2 in Proteo@thecus sylaiae (Simons, 1989).

According to Simons et al. (1987), P. ankeli does not show any special resemblance to

P. markgrgfi. For these authors the first two lower molars in the holotype right mandible

(SNM 12639G) of P. markgraf, briefly described by Schlosser (1910) under the name

Illoen’pitkecus markgraji, differ from those of P. ankeli in their more developed crests, the

position of the hypoconulid very close to the entoconid and by the smaller relative distance

between protoconid and metaconid and between hypoconid and entoconid, giving the

crown a very “constricted” look, evidently related to the pronounced flaring margins of the

crowns (Figure 9). Even though the synonymy of the two genera Propliopithecus and

Moeripithecus has been maintained by Simons since 1967, it is these characters that have

46 ETAL. H. THOMAS

5mm

generally been applied by authors who accept the specific distinction of “Propliopithecus”

markgraj.

However the supposedly distinctive characters of the first two molars in the paratype mandible of P. ankeli, which are used to distinguish this specimen from the holotype of “Propliopithecus” markgraj (Simons et al., 1987), are insufficient for establishing their separation at the specific level. Indeed, quite the opposite is the case. These two molars are astonishingly similar to those of “Propliopithecus”markgraj, especially ifone considers, as we do here, that almost all the differences cited by Simons et al. (1987) may be related to the degree of wear of the cusps of P. ankeli; the two molars of “Propliopithecus” markgraj are virtually in an unworn condition. It is clear that all intercuspal distances tend to increase gradually with wear because the buccal and lingual cusps are located on convex flanks of the crown. As far as the crests are concerned, however, they tend to reduce or even to disappear with wear. The degree of proximity of the hypoconulid-entoconid, as estimated from the ratio of the hypoconulid-entoconid distance divided by the intercuspal talonid width (hypoconulid-entoconid), is 0.25 for the Mr ofP. ankeli and 0.21 for the same tooth in “Pro@iopithecus” markgraj. This slight difference, which is nevertheless visible to the naked eye, could also be due in part to differential wear, in which case it has little significance.

All the characters described for P. ankeli are precisely those that occur on the Mr and MZ of the Taqah mandible: in the latter specimen the hypoconulid seems to be placed even more in the mid-line position. The morphology of the two anterior lower molars from Taqah, by comparison with other propliopithecids, is also amazingly similar to that of “Propliopithecus” markgraj and P. ankeli, especially in the high degree of constriction of the crowns. The same slight buccal swelling in the notch between the hypoconid and the protonid is observed in “P.” markgraj. The ratio of intercuspal talonid width to crown width at the talonid for Mt and MT ofTQ4 (0.46 for the Mt and 0.43 for the M,) is close to that of ‘P. ” markgrq’i (0.4 for the Mr and 0.38 for the Mg: these ratios determined on casts) (0.43 according to Kay et al., 1981: 304).

Conclusions

Taking into account all the observations concerning the lower dentition, we propose to attribute the Taqah form, as well as the paratype of P. ankeli, to Moeripithecus markgraj.

Considering the important morphological divergence between this species and Propliopithecus haeckeli, confirmed and extended by the new finds from Oman, we consider it justified to recognize the generic distinction between Moeripithecus and Propliopithecus.

MOERIPITHECUS MARKGRAFI DISCOVERY IN ASHAWQ FORMATION 47

Until the new specimens of Moeripithecus were found, there might have been a case for declaring these two genera synonyms. With the additional fossils from the Fayum and from

Oman, however, it is clear that Schlosser (1910) was correct in concluding that there were

two distinct genera in the small sample available to him. It is clear that we could opt for a more conservative approach and not resurrect the genus

?Iloeripithecus. However, as the upper dentition from Taqah clearly belongs to the lower dentition (most probably the same individual) and as it constitutes the best evidence for distinguishing the Taqah material from all other known propliopithecids, the former of which is characterized by some primitive features such as relatively small hypocone, upper molars much broader buccolingually and narrower mesiodistally, and presence of a pericone on M’ and M’, we conclude here that Moeripithecus is a valid genus. Moeripithecus is definitively not congeneric with Aegyptopithecus (which is derived at least for the characters mentioned above), contrary to the prediction of Kay et al. (1981). Although the large size of Ail. markgruji, the robustness of its lower premolars and the basal inflation of the lower molars could be interpreted as synapomorphies linking 1V. markgraj to ilegy;btopithecu.s,

M. markgraji is probably not directly in the ancestry of Aegyptopithecus, a conclusion already suggested by Simons et al. (1987: 46) for their new species P. ankeli.

Concerning this new species, it is surprising that Simons et al. (1987) chose a maxilla fragment bearing deeply worn teeth for the type of P. ankrli, when they had at their disposal an almost complete lower jaw. This choice is even less understandable when one considers that the type-specimens for all other propliopithecids are mandibles. In making this choice, the authors put even themselves in the embarrassing position ofhaving to define the species almost solely on the basis of the paratype. The species P. ankeli is therefore in reality based on the paratype and not on the holotype.

Having attributed the paratype of P. ankeli to h1. markgraj, we are obliged to maintain the taxon P. ankeli, which is based on a maxillary type specimen. However, because of the difficulty of redefining the taxon due to the extremely worn teeth that it contains, the identity of the species will probably remain doubtful. While certain characters can be used for determining its relationships (reduced size increase from front to back of the upper molars, large size and oval occlusal outline of the upper premolars, large hypocone, complete lingual cingulum and apparent lack of a pericone*), it cannot easily be claimed that it belongs to the same species as the paratype of P. ankeli on the basis that the two specimens co-occur in the same quarry and that they occlude “very well”.t As Simons et al.

(1987) themselves pointed out regarding the type of P. ankeli, “the occlusal surfaces of the molars are too heavily worn to make many determinations of crown anatomy”, which indicates that the holotype is inadequate for definitive diagnosis. Therefore we suggest that P. ankrli is a nomen z~anum (see Mones, 1989). Accordingly we assign the type specimen of P. ankeli to an indeterminate propliopithecid.

The locality from which the type specimen of 111. markgraj was collected by Richard Markgraf is not known. Gingerich ( 1978)) b asing his opinion on historical considerations, thought that the specimen, as well as the type specimen of Propliopithecus haeckeli (the type species of the genus), could have come from the Upper Fossil Wood Zone of the Fayum. According to Kay et al. (1981), h owever, no other specimens attributed with certainty to Al. ntarkgrafi‘or to P. haeckeli have been found in the Upper Fossil W’ood Zone, despite years

* An anonymous referee stated that there is no tracr of a pcricone in molars of the hotot) pv of P. on/w/i. t The same anonymous referre supplied this IICW information concerning the twlot~ pc and parat) pc of

P. mkrii.

48 H. THOMAS ET AL.

of intensive research in the deposits of the Qatrani Formation. This suggests that these specimens must have been collected in the basal sequence of the formation. Only a few specimens collected in the intermediate level “quarry” could be attributed to P. haeckeli

(Kay et al., 1981).

Regarding the hypothesis of Kay et al. (1981), it is interesting to note that the paratype mandible of P. ankeli, attributed here to Al. markgraji, came from Quarry V. Although we now have, for the first time, a good idea of what the dentition of M. markgraj looks like, we think that it is still premature to attempt to elucidate the relationships between Aeepopithecus, Moeripithecus and the two species of Propliopithecus.

At least part of the difficulties in determining the character states and synapomorphies are due to the lack of knowledge that exists concerning P. chirobates, P. haeckeli and M. markgraji. It is also difficult to take into account intraspecifically variable traits such as those related to sexual dimorphism or individual variation (i.e., Ps morphology, premolar development, cingulum development, molar proportions and molar size differentials). Obviously further knowledge on cranial and postcranial anatomy is necessary to provide evidence to support special relationships, as was done by Kay et al. (1981), on the basis of outgroup comparisons with parapithecids and Oligopithecus.

Acknowledgements

We are grateful for the very helpful advice, discussions and comments of our colleagues E. Delson, D. Russell, J. Schwartz, B. Senut and, in particular M. Pickford, who also translated the original manuscript. We thank A. Rosenberger and two anonymous reviewers for their advice and comments. Particular thanks must be expressed to the Centro Studi Ricerche Ligabue (Venice) which provided valuable assistance. We are also anxious to thank Mrs Elyane Molin, Mrs Michele Mannu and M. Denis Serrette for technical assistance and MS Elizabeth A. Liebman for the drawings. Finally we gratefully acknowledge the help of M. Mohammed bin Hussain bin Kassim, Director General of Minerals, Ministry of Petroleum and Minerals, who generously provided scientific and logistic support in Oman.

References

Andrews, P. (1985). Family group systematics and evolution among catarrhine primates. In (E. D&on, Ed.)

Ancestors: The Hard Euidence, pp. 14-22. New York, Alan R. Liss, Inc.

Bonis, L. de, Jaeger, J. J., Coiffait, B. & Coiffait, P. E. (1988). D&ouverte du plus ancien primate catarrhinien

connu dans l’Eoc?ne supCrieur d’Afrique du Nord. C.R. Acad. Sci. Paris, 306, II, 929-934. Fleagle, J. G. & Kay R. F. (1983). N ew interpretations of the phyletic position of Oligocene hominoids. In Newel

Interpretetiom of Ape and Human Ancestry (R. L. Ciochon & R. S. Corruccini, Eds), pp. 181-210. New York,

Plenum Publ. Corp. Gingerich, P. D. (1977). Radiation of Eocene Adapidae in Europe. G&bias, MCm. sp. 1, 165-182.

Gingerich, P. D. (1978). The Stuttgart collection of Oligocene primates from the Fayum Province of Egypt.

Paliiont. 2. 52, 82-92. Kay, R. F., Fleagle, J. G. & Simons, E. L. (1981). A revision of the Oligocene apes of the Fayum Province, Egypt.

Am. J. ~+Ls. Anthrofi. 55, 293-322. Mows, A. (1989). Nomen dubium us. nomen vanum.Jour~ul of Vertebrate Paleontology, 9 (Z), 232-234.

Rasmussen, D. T. & Simons, E. L. (1988). New specimens of Oligopithecus sauagei, Early Oligocene primate from

the Fayum, Egypt. Folia Primatol. 51, 182-208. Schlosser, M. (1910). ijber einige fossile Saiigetiere au dem OligocHn van Agypten. Zool. Anz. 35, 500-508.

MOERIPITHECUS MARKGRAFI DISCOVERY IN ASHAWQ FORMATION 49

Simons, E. I,. (1967). Review of the ph?letic interrelationships of Oligocene and Miocene Old World

Anthropoidea. Coil. int. du C.N.R.S. n” 163. ProhlPmes actuels de Paltontoloeie iEvolution des Vert&rdsi.

Paris 6111 Juin 1966, 597-602. /J \

Simons, E. L. (1989). Description of two genera and species of Late Eocene Anthropoidea from Egypt. Proc. A’&.

Acad. Sci. L’SA. 86, 9956-9960. Simons, E. L., Rasmussen, D. T. & Geho, D. I,. (1987). A new species of Propli@Uzecus from the Fayum, Egypt.

Am. ,J. ,~hg~. Anthrafml. 73, 139%147.

Swind&r. D. R. & Olshan, A. F. (1982). Molar size sequence in old bborld monkeys. po/ia Primalol. 39, 201-212.

Szalay, F. S. rYr Delson. E. ( 1979). /+~olutionary h&w of the Primafe.r. New York. Academic Press.

Thomas, H., Roger, J., Se;, S. & Al-S 1 _ ” LI almani, Z. (1988). DCcouvcrtc des plus anciens “Anthropoi’des” du

continent araho-africain et d’un Primate tarsiiforme dans I’OligocPne du Sultanat d’Oman. C.R. Acad. .C Pa’arir, 306, II, 823-829.

Thomas, H., Roger, J.; Sen, S., Roudillon-de-Grissac, C. & Al-Sulaimani, Z. (1989). Dicouverte de \‘ertPlx+s

fossiles dans l’Oligoci_ne inf&eur du Dhofar (Sultanat d’Oman). &&IA, 22 (I), 101-120.