FOSSIL BOVIDAE FROM THE LIMEWORKS QUARRY, 
MAKAP ANSGAT, POTGIETERSRUS 

By L. H. WELLs* and H. B. S. CooKE** 

ABsTRACT 

A short history of palaeontological work in the Makapan Valley is given and the setting 
of the fossil material is briefly described. Many hundreds of bovid cranial and dental 
fragments have been recovered, in addition to thousands of bones and bone fragments. 
The post-cranial material is not considered in the present account. The collection includes a 
number of species not separable from the living kudu, nyala, eland, Cape buffalo, blue 
duiker, mountain reedbuck, gemsbok, sassaby, brindled gnu and impala. Known extinct 
species include cf. Alcelaphus robustus, cf. Pelorocerus helmet and Oreotragus major. A new 
genus of aberrant Alcelaphine is described as M akapania broomi. Other new species are 
Cephalophus pricei, Redunca darti, Gazella gracilior, and Phenacotragus vanhoepeni. A feature 
of the assemblage is the tendency for species to be larger than their living counterparts. 
The whole described fauna of the deposits is discussed and it is considered that a lower 
Pleistocene (prob.ably Villafranchian) age is most likely. 

HISTORY 

The occurrence in the Makapan valley of consolidated bone-bearing cave deposits 
was first mentioned by Dart in 1925 and some specitnens were collect,ed by him 
two years 'later at a site now known as the "Limeworks Quarry". In 1937 Broom 
descri!bed a dwarf buffalo "Bos" makapani from a cement~ed bone breccia on the 
opposite side of rhe valley and this location has come to be ca·lled "Buffalo Cave". 
A year later Van Riet Lowe (1938) drew attention to further sealed cave deposits 
nea:r the head of the valley, adjoining an open cave associated with historic incidents. 
One of 1jhese deposits comprises banded layers of coloured ash "and bat guano, 
containing Middle Stone Age artifacts, and it was named the "Rainbow Cave"; 
the ovher deposit contains Earlier Stone Age (hand-axe) tools and shows a basal 
ash and guano layer which .led to the ·name "Cave of Hearths". Additio~al sites 
have also been found in the valley and their relative positions are shown in the 
accompanying map (Fig. 1). 

In 1945 and 1946 senior students and staff from the Department of Anatomy of 
the M~edical School in Johannesburg collected fossil material from dumps of waste 
at tthe Limeworks quarry. Several extinct species were recognised, some of them new, 
Cl!nd it was reaHsed that the fauna had affinities with that of the australopithecine· 
bearing cave breccias of the Sterkfontein ar~ea. Accounts war~ published describing 
some of this material ajnd dea1th with Primates (Broom and Jensen, 1946; Mollet, 
1947; Kitching, Wells and Westphal, 1948, Suidae (Dale and Tobiansky, 1947; 
Dale 1948) , Giraffidae (Cooke and Wells, 194 7) and a rodent mole (Broom 1948) . 

* Department of Anatomy, University of Cape Town, Rondebosch, C.P. 
** Department of Geology, University of the Witwatersrand, Johannesburg. 

1 



N 

0 .Y2 
I mile 
I 

j 
/. ·-· , -·-­-·-·-·-·-·-·-·-·-·-·-·-·-·-· . . DE' J-JOOP 

.[)£' 

KEY 
-·- FARM BOUNDARIES 
...-< STREAM VALLEYS 
=MAIN ROAD 
:::: FARM ROADS 

··"'"''·· CLIfF rACES 
• CAVE SITES 

rig. 1-Plan of Makapan valley showing location of main sites. (Based on survey by H. B. S. Cooke 1939, 1947 .) 



A paper on the Bovidae was presented to the Royal Society of South Africa by one 
of us (L.H.W.) in 1947, but it was withdrawn for revision in the light of a rapid 
increase in the quantity of material collected. 

Archaeologica~l excavations began at the Cave of Hearths in 1947 with the aid 
of funds provided by Dr Bernard P·rice and in the following year a human mandible 
was recovered (Dart 1948 e). The availability of a base camp in the area facilitated 
work on the waste dumps at the Limeworks quarry and in 1948 were recovered an 
australopithecine occiput (Australopithecus prometheusj Dart 1948 a, c) and, 
subsequently, an adolescent mandible (Dart 1948 b, d). Excavations at the Cave 
of Hearths stopped at the end of the 1949 winter season and were resumed only in 
1953 with funds made available by the Wenner-Gren Foundation. At the Limeworks, 
however, sorting of the dump material has been continued each winter season (except 
in 1952) and additional australopithecine fragments have been described (Dart 1949 
a, b, c; 1954; Bone, 1955). Some further accounts have appeared of baboons ('Broom 
and Hughes 1949; Kitching 195 2, 195 3) . New descr~ptions also include a chali­
cothere ·(George 1950), a klipspringer (Wells 1951), a hippopotamus (Kitching 
1951), ca·rnivores (Toerien 1952, 1955), hystricomorphs (Greenwood 1955) and 
hyracoids ( Churcher 1956). At present, the baboons are being re-considered by Freed.., 
man (1955) and the carnivores and Suidae are under review by Ewer. 

GEOLOGICAL SETTING 

The general physiography of the Makapan valley has been described by Barbour 
( 1949 a, b) and its geology by King ( 1951) with additional d·etails furnished by 
Van Riet Lowe (1938, 1943, 1948), Bosazza, Adie and Brenner (1946), Cooke 
( 195 2) , Dart ( 195 2) and Oakley ( 1954 a, b) . HaughhtJon ( 194 7) has discussed 
the cave breccias in general. King provides an excellent factua'l accoun~ of the 
sequence of deposits observed at each of the major sites. He gives a geological map 
of the upper part of the valley and sketch plan of the Limeworks site. While his 
descriptions cannot be questioned, his interpretations cf the successions and his 
correlations are subject to dispute. His conclusions differ from those of Barbour 
~nd have been criticised in some respects by Robinson (1952). 

Like those of the Sterkfontein area, the original caves (now lar.gely represented 
only by the cons·olidated fillings) were formed :by solution in do'lomitic iimestones 
of the Transvaal system, some of them selectively aiong old crush zones. Such caves 
tend to follow in depth the shallow southerly dip of the bedding planes of this 
formavion. Before the caviti.es became opened to the ~urface, travertines or drip­
stones of rather pure ca:lcium carbonate were formed on the floor, waHs a11d roof 
of the cavern. As an opening to rhe surface developed and grew in size, increasing 
amounts of detrital soi:l, wind ... blown dust and other material were carried in. Thinning 
of the retreating roof or lip of the cave ultimately led to the development, in many 
cases, of a covering of collapsed rock breccia or rubble. The detrital material has 

3 



been cemented to a greater or lesser degree by the infiltration of calcium carbonate.* 
At some loca:lities continued erosion has led to secondary cavernation of deposits 

of the first cycle, followed bz a subsequent cycle of infilling. Complete rernoval of 
the original cave roof and surface erosion of the already cemented cave deposit is 
characteristic of the older sites. As this general cycle is largely inherent in th~ process 
of cavern formation and fHling, it is clear that similC1r lithological sequences of 
detrital cave deposits may vary widely in age. The age is controlled to a great extent 
by ( 1) the moment at which a surface opening developed and ( 2) the rat·e of 
introduction of detritus, which is likely to begin slowly and increase in tempo. While 
the cave opening is small, the "cave climate" must exert a considerable influence on 
the character of the deposit but the detrital material introduced reflects in its physical 
state the impress of the external or regional climate. Tedious nineralogical and 
ana~ytical studies being pursued by Brain at the Transvaal Museum, Pretoria, 
are yielding most interesting resu'lts and promise to prov.~de a valuable tool for use in 
correlation together with the palaeontological evidence. Detailed results are 
unpublished but a brief note has appeared recently (Brain, Van Riet Lowe and 
Dart 1955). 

Except in the case of the Historic Cave, the deposits have come to light as a result 
of the activities of lime workers, who sought and mined the valuable basal travertine 
deposits. The travertine is almost pure calcium carbonate whereas the parent limestone 
is highly dolomitic and of little commercial value. In consequence of these mining 
activities, ~he original cave fillings were undercut by gaHeries and stopes so that 
access to the deposits at the present day is normally gained from these workings 
below the detrital filling. 

At the Makapan Limeworks, the lime-seekers fortun~tely discarded much of the 
material which lay above lihe pure basal travertine because it was contaminated with 
introduced sandy material. The waste dumps of sandy limestone and cemented detritus 
have thus proved a rich source of fossil bone, though many valuable specimeas must 
also have gone to the kilns. With very few exceptions the fossils previously described 
(and also nearly all those discussed in the present account) ha:ve come from the 
dumps at the Limeworks quarry and their horizons of derivation in the deposit cannot 
be determined precisely. Certain information can be deduced from the matrix material. 

A simplified plan of the L:meworks quarry area !s given in Figllre 2. A road 
separates the northern area of dumps and old kilns from the open quarry and 
underground galleries. Entrance to the workings i·s usuaHy gained from the western 
open cutting. The first part of the inner gallery js roofless and has a northwestern 
wall of crudely stratified red or pink sand with bones visible in the lower part. 

* King's suggestion of consolidation due to recrystallization of original calcite dust is here 
regarded as improbable. 

4 



TN 

A 

Brown breccia with lorqe blocks 

Brown brecc io with pebble bonds 

Coone pink breccia 
Grey mort 
Stratified pink or red breccia 
Travertine 
Dolomitic· limestone .......... , 

...... ' 

HUTS 

---

sea level 

Fence 
Op~n quarry 
Und~rground 

workings 

Dumps 

B 

Fig. 2-Simplified plan of the Makapan Limeworks site and a North-South section illustrating 
the geological succession of the fossil-bearing deposits. In the plan, 1 is the western quarry, 
2 the eastern quarry and 3 a quarry and cave known as the HHorse mandible Cave". 

(Based on surveys by H. B. S. Cooke 1939, 1947 and A. R. Hughes 1948.) 

5 



6 



The or~ginal floor travertine has been removed almost entirely leaving the old 
filling overhead (Figure 3). Resting on the old travertine was a manganous layer 
of mixed travertine and very fine pink sand. This give-s place to stratified pink or 
red breccia with occasional dark bone fragments and is then overlain, after a 
variable distance, by a grey layer of cemented marl rich in bone, sometimes actually 
in contact wirh the cave roof. This grey layer is considered to be the main fossil­
bearing horizon and its richness is probably due to the vet y slow rate of accumulation 
of introduced sediment. The floor of the old workings descends southward and 
the grey layer rises somewhat on the western wall so that it lies far above the 
present art:ficial floor. Eastwards, the grey layer dips sharply above the old floor 
and vanis'hes below a large cone of fallen brownish-red sandy breccia (Fig. 4). 

In this (southern) portion of the cave system, the grey layer is overlain by several 
feet of slight! y coarser and sandier pink breccia with a good fauna. Collapsed blocks 
of travertine, presumably fcrllen from a thinning roof, lie in the upper part of this 
zone. Overlying it is the coarser brownish-red sandy breccia which forms the recently 
collapsed cone. This material contains wea~hered pebbles and gravel from w~ich 
dubious artefacts of extremely primitive character have been recovered (Brain, Van 
Riet Lowe and Dart, 1955). Brain regards this brownish-red material as a part of 
the normal succession and hence as a direct continuation of the grey breccia. While 
this is proba:bly the case, the browner material might- possibly represent a later 
filled-in sink-hole. Detailed study of this question is awaited but, in the me2n time, 
the stratified pale pink breccia is regarded as locally the oldest; the coarser red 
breccia with dark bone is considered to be its direct successor, followed by the main 
fossiliferous grey layer, while the browner deposit is thought to be at least a little 
younger and possibly a good deal younger. However, the total time span covered is 
probably limited. The grey layer is rich in bone and is the source of most of the 
fossil material recovered; the red or pink zone contains only a modest qu1.ntity of 
material; the bvowner material is scarcely studied and there is considerable uncertainty 
regarding the reference of dump material to this zone. 

Fig. 3-View of the stratified pink and red sandy breccia with dark bones, resting on 
remnants of floor travertine, the mining of which prqduced the gallery. Photograph 
taken from dolomite hummock near centre of section AB of Fig. 2, looking towards A. 

Fig. 4-Cone of collapsed brown breccia, containing weathered pebbles and gravel in 
its lower portion. On the left of the picture is the downward-projecting solid mass of 

cave deposit shown in the centre of the section in Fig. 2. 

' 



CLAssiFICATION AND NoMENCLATURE 

Various classifications of the Bovidae have been put forward by numerous authors, 
notably Schlosser (1925), Schwarz (1937), Pilgrim (1939) and Simpson (1945). 
Simpson degrades the sub-families of most authors to the more reasonable rank of 
tribes, recognising thirteen and grouping them into five subfamilies as follows :-

Subfamily BOVINAE: Tribes 

Subfamily OEPHALOPHINAE: Tribe 

Subfamily Hl'PPOTRAGINAE: Tribes 

Subfamily ANTILOPINAE: Tribes 

Subfamily CAPRI1NAE: Tribes 

Strepsicerotini 
Boselaphini 
'Bovini 

Cephalophini 

Redunc~ni 
Hippotragini 
Alcelaphini 

Neotragini 
Antilop:ni 

Saigini 
Rupicaprini 
Ovibovini 
Caprini 

Although this classification may be subject to minor objections, it is simple, 
essentially reasonable and is adopted here. Neither the recent fauna of southern 
Africa nor the fossil assemblage from the Limeworks quarry includes any Caprinae* 
or Boselaphini. --Of the remaining eight tribes, at least one representative of each 
has so far been identified in this fossil assemblage. 

For t:he nomenclature of recent genera and species, Allen's "Checklist of African 
Mammals" ( 1939) is followed in the main, subject to such modification as is 
required by Simpson's classification. For the fossil species from Africa, the list given 
by Hopwood and Hollyfield ( 1954) serves as a useful guide. In descril,ing the 
teeth, the terminology used by Pilgrim (1939) is employed in preference to the Cope­
Osborn nomenclature. The abbreviations I, C, P and M will be used to indicate 
incisor, canine, premolar and molar teeth respectively~ the prefix D is used for 
decidious (milk) teeth and the prefixes R and L indicate the right and left jaws. 

In the paper communicated to the Royal Society of South Africa in 1947, :1 number 
of new names were prop?sed. Although these names were given in a circulated 
abstract, they were unaccompanied by any description and are here regarded as 
nomina nuda. The earlier names are entirely superseded in the present account. 

* Two extinct African genera (Pultiphagonides and Pelorocerus) assigned to the Caprinae by 
Hopwood and Hollyfield (1954) belong in reality to the Alcelaphini (vide infra). 

8 



SYSTEMATIC DESCRIPTION 

Subfamily BOVINAE 

Tribe Strepsicerotini 

S trepsiceros cf strepsiceros (Pallas) 

A number of partial upper and lower dentitions cannot be separated f1om the 
living kudu, though many of the teeth lie at or above the upper limit of size 
variation noted in living material. The best specime·n (M 2) * is a right maxillary 
fragment wi~h P4 - M 3 and part of P3 in advanced wear. The intact teeth measure 
82 mm, the length of the three molars being 69 mm. These measurements correspond 
exactly with those of a male kudu skull in the Transvaal Museum. The anterior and 
middle external folds of the molars are very reduced so that they lie practically 
flush with the face of the lobe behind them. In correspondingly worn teeth of the 
living kudu, these folds are also reduced, though not to quite ,rhe same degree. 
Simila·r reduction is a:lso apparent in a more worn third molar, 26 m·m in length, 
which occurs in a lmaxtHary fra:gment rogether with part of the second molar. 
A · well worn upper right dentition is a good deal larger, the three premolars 
having a combined length of 52 mm and the three molars a length of 77 mm 
compared with 44 mm and 71 mm in the largest living specimen studied. The folds 
are more norma'l in 'this specimen (M 3) . 

Also assigned to this species are: a partial palate with well-worn RP3 - RM1 and 
LP4 , LM1 and part of LM2 ; half a dozen isolated upper teeth; five ·mandibular 
fragments with one or two teeth in each; five loose lower molars; and a partial left 
lower jaw of a juvenile wivh DM·2- DM4 and parts of M1 and M2, the fatter ~till 
in alveolus. This last specimen (M 4) belongs to the collection made by Dart 
in 1927. 

The back half of a skuH, broken across in front of the orbits, shows no trace 
of horns and agrees closely with the skull of the female kudu, except that the width 
across the mastoid region is greater than in living material. This specimen (M 1111) 
is filled with firmly cemented fine bright pink silt a-nd the bone is unusually dark, 
suggesting that the specimen did not come from the usual grey breccia. 

A frontlet with a cavity for the right hom core and with a much damaged hind 
portion of the left core resembles the kudu in size and erection of the horns but the 
condition of the specimen (M 1112) prevents more exact comparison. 

A large fragment of ·rhe back part of a skull (M 555)., though badly damaged, 
agrees in all essential features with the corresponding portion of the skull of a recent 
kudu. It is therefore placed tentatively in this group. 

* Specimen number in the catalogues of the Bernard Price Institute for Palaeontological 
Research, Johannesburg. 

9 



Strepsiceros (Tragelaphus) cf angasi (Gray) 

Four partial upper dentitions, a loose RM2, two part~al lower jaws, a few loose 
lower molars and three fragments of hom core cannot at present be separated from 
the living nyala, though the fossil is somewhat larger. The best upper series is a right 
maxillary fragment with moderately worn M1 - M3, 55 mm in length (M 5), 
compared with 52 mm in the largest specimen available for comparison. A good 
lower left jaw has P4 - M3 in fairly early wear (M 6), the premolar being 16 mm 
long and the molar series 64 mm, compared witJh 14 mm and 58 mm in the 

living species. These dimensions thus exceed the observed upper limit of variation 
in the nyala and are comparable with typical material of the mountain nyala (T. 
buxtoni). The lower molars all show a moderately well-developed external basal 
pillar, strongest in M1 and weakest in M2. 

The three horn core fragments represent the tip, upper median and basal parts 
of at least two different horns. The somewhat triangular cross-section and spiral 
keel characteristic of this group are clearly shown. The siz.e of the basal portion 
is about normal for the nyala and smaller than in the mountain nyala. 

Arambourg (1941, 1947) described an extinct specie~ Tragelaphus nakuae from 
Omo which he regarded as related to the bongo. Both the teeth and the horns of 
the Orne form are larger than is the case with the Makapan material and specific 
identity seems improbable. 

Taurotragus cf oryx (Pallas) 

Remains not at present distinguishable from the existing eland are fairly plentiful, 
though a feature of the material is the youth of the individuals repre.sented. 

The most considerable specimen is a right maxillary fragment with DM2 - M2, 

lacking the outer walls of all the teeth except the last. Three other maxillae (two 
right and one left), carry DM2 - M1 and there are five fragments with milk teeth 
only. One partial right palate possesses M 1 and M 2 ; a dam~ged left maxilla 
(M 7) has P 4 (unworn) , M 1 and M 2

• There are also a piece of left maxilla 
with P 2 and P 3 in moderate wear and 'Some half dozen isolated upper molars in 
varying degrees of attrition. 

The upper permanent ·molars correspond exactly wit:h those of T. oryx in size and 
in the broadly rounded form of the medial lobes. In the isolated upper teeth the 
strong folds and weak ribs of the external wall are ideQtical in development with 
those of the living eland. A feature of the upper permanent molars which is not 
common in T. oryx is the presence in M 1 and sometirr1es in M 2 of a weak basal 
pillar extending as much as half the height of the crown. In the upper milk molars 
the basal pillar is rudimentary. 

Two left mandibular fragments, with DM2 - DM-± and DM,2 - M'2 respectively, 
can also be assigned to this species. A jaw fragment with P3 and P-± and two isolated 
unworn lower first molars may be included. 

10 



Tribe Bovini 

cf Syncerus caffer (Sparrman) 

A left maxillary fragment (1M 10) with DM2 
- M 1 in early wear is cons:derably 

larger than the remains ascribed to T aurotragus cf oryx, the three milk molars 

measuring 75 mm as compared with 64 mm in the eland. A right maxillary fragment 

carrie3 unworn DM3 and DM4 and there are a1so stx isolated milk molars and 

four first molars of similar character. All are practically unworn. These teeth are 

further distinguished from T aurotragus by prominent external ribs. DM4 has a well­

developed basal pillar, but this is not continued into a cingulum fold around the 

medial lobes as is the case in Hippotragus. 

AHowing for their very early state of wear, these teeth agree with those of the 

only . juvenile specimen of Syncerus caffer at our disposal. They are actuaUy some­
what larger than the recent dentition, in which the three n1ilk molars measure 66 mm. 

This is due at least partially to ~he less worn condition of the fossil teeth. The 

basal pillars of DM4 and M 1 are more slender, and that of M} possibly shorter, 

in the fossil than in the recent dentition; also there is no rudimentary basal pillar 

in DM3• These features may prove to be outside tlhe range of variation of S. caffer 
but the evidence on this point is at present too uncertain to justify specific distinction. 

A portion of a right mandible (M 15) carries M2 and a damaged and almost 

unworn M3, the combined length being 74 mm, which is 7 mm larger than -the 

biggest available living specimen. Two mandible fragments have DM4, unworn, in 
each. Another piece shows the crown of M1, still unerupted, and there is a fragment 

of jaw with a slightly worn first molar. Isolated low·er teeth include three complete 

fourth milk molars, a fragment of one, and a first molar, all unworn. All these 

· teeth are very simitlar in size and in structure ro the corresponding teeth in 

S. caffer, though on the whole they tend to be a little larger than in the specimens 

available for comparison. 

None of these teeth is large enough to be assigned to H omoioceras bainii (Seeley) 
or even to "Bubalus'' (? === Syncerus) andersoni Scott. ''Bos makapaani" Broom 

come3 from a different site in the area and is almost certainly much more recent 

in age than the Limeworks fossils; it is most unlikely that any of the Limeworks 

buffalo teeth belong to this dwarf species. 

Also probably representing the buffalo is a portion of an occiput (M 950) 

lacking the upper part of the braincase. 
As is the case with the other large species already considered, a feature of 

the buffalo material is the youth of the individuals represented. 

11 



Tribe Cephalophini 

Cephalophus pricei sp. -nov. 

Diagnosis: A cephalophine about thirty-five per cent lavger than Cephalophus 
natalensis and with relatively shorter premolars. 

Holotype: Left mandible with moderately worn cheek teeth P2- ·M3 (M 18; Fig. 5). 
Paratypes: Right mandible with P2 - M3 in early l\·ear (M I7; Fig. 5); right 

mandible with P2 - Ma in advanced wear (M 19; Fig. 5); left mandibular 
fragment with DM2 - M1 (M 9; Fig. 5); right horn core (1M 478; Fig. 6). 

Measurements (in mm): 

Holotype Para types 

M I8 M I7 M I9 M9 

DM2 length 6 
DM3 length 8 
DM4 length 14 

Series DM2 - DM4 28 

P2 length 5 7 5! 
P3 length 8 10 8 
P4 length IO I1 10! 

Series P2 - P4 23 28 24 

M1 lengcll II! 12 I2 I2 
M2 length I2 I4 I3 
M3 length I8 18 19 

Series M1 - Ma 42 44 44 

Ratio ,p X 100 
62.l 54! 55 

M 
2 

Jaw height at P4/'Mt 23 20 23 21 

Eighteen lower jaws and jaw fragments resemble closely the living duikers but 
the teeth are a good deal larger than in the biggest living South African specimens 
of Cephalophus natalensis (red duiker) or Sylvicapra grimmia (Cape duiker). 
The fossi'l material is somewhat larger even than the Bay Duiker of West Africa 
Cephalophus castaneus but is considerably smaller than the Yellow-backed Duiker 
C. sylvicultor. Comparative average measurements are given in the following tables: 

12 

• I 



Cepha- Cepha- Cepha-
Sylvicapra lophus lophus lophus 

Fossil grunnua natal ens is castaneus sylvicultor 

DM2 length 6. 6 5 7! 11 
DM3 length 8 7 61 

2 71 
2 13 

DM-t length 14 11! 91 
2 131 2 191 2 

Series DM2 - DM4 28 241 2 21 28t 431 2 

P2 length 6 61 5l 71 12 2 k 2 

P3 length 9 71 
2 

61 
2 8 12 

P-t length 101 2 8 8 9 121 2 

Series P2 - P4 251 2 22 20 241 2 361 2 

M1 length 12 9 8 91 
2 13 

M2 length 13 101. 2 10 12 171 2 

M3 length 181 2 131 2 13 15 23 

Series M1 - M3 431 2 33 31 361 2 531 2 

Ratio P X 100 
581 67 641 67 68 

M 
2 

Jaw height at P4/M1 22 16 15 20 29 

In the fossil form t:he average length of the cheek _tooth series is thirty-five 
per cent. greater than in C. natalensis and twenty-five per cent. greater than in 
Sylvicapra grimmia. In the living forms premolar length is close to two thirds that 
of the molars, but in the fossil the molars are proportionally more elongated, the 
third molar being particularly long. 

The lower molars are distinguished from rhose of the gazelles by being much 
lower crowned and shortening very rapidly with wear. They have prominent, tapering, 
somewhat oblique external lobes which are rounded at the apex. The anterior and 
posterior internal folds are prominent and persistent but the middle fold soon 
disappears with wear. There are also well marked internal ribs. In P4 the medial 
cusp is prolonged forwards to unite with the anterior wing but is separated from the 
posterior wing by an open valley. This last feature, as well as the relatively low 
crown, distinguishes the tooth from that of Aepyceros.; in having the medial cusp 
united with the anterior wing it resembles the red rather than the Cape duiker. 
T·he anterior premolars, on . the other hand, are less compressed and sectorial than 
in C. natalensis and more nearly resemble those of S. grimmia. 

Also in the collection are three hom-cores of duiker-like character. The best 
$pecimen belongs to the right side (1M 478; Fig. 6). It lacks the extreme tip a.nd 

13 



-l 
·- J 

M 18 l 
~ 

M 17 

M 19 

M9 

M 21 

Fig. 5-Cephalophus pricei. M.18; holotype left mandible with P2 - M3 in moderate wear; 
M 17: para type right mandible with P2 - M3 in early wear; M 19: para type right mandible 
with P2 - M3 in advanced wear; M 9: paratype left mandibular fragment with DM2 - M1; 

M 21: referred partial maxilla with M1 - M3 in early wear. Natural t size. 

must have been about 85 mm long when complete. It is slightly oval at the base, 
the elongation being transverse to the axis of t:he skull. The greatest diameter 
is 18 mm and the least 15 mm. The largest red duiker hom core available is 80 mm 
long and 12 to 16 nun in diameter at the base; the Cape duiker has horn cores up 
to 90 mm in length with the base almost circular, having a diameter of 12-15 mm. 
The fossil core is somewhat stouter than in the red dufker but is very simi:lar in 
proportions and rises from the frontal bones in the san1e manner and at the same 
angle. Its resemblance to the horn core of Syfyicapra is less marked and this fact, 
together with the features of the lower fourth premolar, suggests that the new 
species be placed in Cephalophus rather tha~n Syrvicapra. 

l4 



L-·-·""'""· 
________ j 

Fig. 6- Cephalophus pricei. 
Paratype right horn core, M478, 

facial aspect. Natural size. 

Subfamily HIPPOTRAGINA·E 

Tribe Reduncini 

Probably a:lso to be referred to this 
new species is a partial rleft maxi·Ha with 
M1 - M3 in early wear (M 21; Fig. 5). 
The teeth are low - crowned, ~Vhe almost 
unworn third molar being ·only 19 mm 
hi~gh, compared with 25-30 mm in the very 
similar teeth of the gazelle Phenacotragus 
vanhoepeni to be described below. The 
outer wa~lls have strong folds and definite, 
though weak, ribs. The molar series is 
42 mm 1long, compared wit'h 27-28 mm in 
C. natalensis and 29.30 mm in S. grimmia. 
Four isolated rhird molars may also belong 
to this species. 

Jlhis new species has been named 1n 
honour of the late Dr Bernard Price. 

cf Cephalophus ( Guevei) caerulus 

(Hamilton-Smith) 

A fragment of right ·mandible with P -1 -

M2 in fairly advanced wear is indistinguish­
able from corre~pondi~ng teeth of the living 
blue duiker. An upper ~left third mo'lar, 
with part of the ·maxrlla, is also inseparable 
from thrs species. The fragments constitute 
the smallest bovid elements of the fauna. 

Remains of reduncines constitute the most abundant element in the coHection, 
fairly closely rivalled by the gazelle Phenacotragus. There are about fifty partial 
skulls and hom cores, twenty palates or maxillary fragments, thirty loose upper 
molars, fifty lower jaws or jaw fragments and twenty isolated lower molars. Two 
species are clearly pres·ent, one being very similar in size to the reedbuck R. arundinum 
but at least specifically distinct while the other is sufficiently close to the mountain 
reedbuck R. fulvorufula not to justify separation. The latter is relatively rare. It was. 
thought, at first, that vhe smaller horn cores and skull represented merely young 
individuals of the larger species but this possibility has been excluded for reasons 
which wiH be given below. 

15 -



I· 

I 
( 

['' . 
...... l 
0\ t 

[, 
j' 

r 
i 

0 Scm 
I I I I 

Fig. 7-Redunca darti. Holotype frontlet, M 446, facial and left lateral aspects. Two-thirds natural s1ze. 

/ 

_j 



Redunca darti sp. nov. 

Diagnosis: A reduncine the size of Redunca arundinum but possessing more massive 
horn cores with a distinct! y sigmoid profile. 

Holotype: A frontlet with well-preserved horns (M 446; Fig~ 7). 

Paratypes: A brain case and occiput with one hom core (M 447); a damaged sk~ll 
lacking braincase and horns but with palate and teeth (M 690; Fig. 8). 

The frontlet (M 446) belongs to an adult male animal and the horn core3 are 
nearly complete, only the extreme tip being missing in the left core and a few 
centimetres lacking on the right core (Fig. 7). Compared with the horn cores of 
R. arundinum, those of the fossil are massive and relatively short; their total length 
measured on the anterior curve cannot have exceeded 150 mm. At their bases the 
cores are oval in cross-section with a maximum diameter of 44 mm and minimum 
of 36 mm; the long axis is directed obliquely backward and laterally. 

The cores diverge at an angle of about 60° and rise abruptly from the frontals 
instead of sweeping back at an acute angle almost parallel to the skull surface as in 
R. arundinum. In profile they have a distinctly sigmoid curvature, the massive basal 
portion being convex anteriorly and the tapering upper portion concave anteriorly. 
The horns of R. arundium have a simple anterior concavit/ and those of R. fulYorufula 
are almost straight. 

The braincase (M 447) has the occiput and pterygoid region well-preserved. The 
block from which the specimen was recovered was broken across at the base of the 
hom core, which was almost detached and has had to be reinforced. In form it 
resembles the holotype frontlet very closely and specific ·:dentity appears certain. The 
occipital crest is higher above the foramen magnun1 than in R. arundinum, giving a 
slightly different appearance to the nuchal view. In the living species the braincase 
of the male has a maximum width of 73-81 mm; in this fossil specimen the breadth 
is 84 mm. In two other specimens, the width of the braincase is close to 80 mm. 

The skull lacks most of the braincase and the nasal region but the palatal aspect 
is almost complete (M 690; Fig. 8), except for the loss of the snout and some damage 
to the crests formed by the pterygoid process of the sphenoid and palatine bones. 
The left bulla is missing and there is ~orne damag~ in the mastoid region and to the 
zygomatic arches. The occiput and pterygoid regions agree in every way with the 
corresponding portions of the brain case specimen described above and there is no 
doubt that both belong to a single species. 

The upper aspect of the skull appears very incomplete owing to the lack of the 
braincase and nasal region. Parts of the malar and lacrimal bones are present 
on both sides. The anterior parts of the nasals are preserved and they show the 
flattening so characteristic of Redunca. In all dimens~ons, the skull agrees very 
closely with R. arundinum. Both sides of the palate carry M 1 - M 3 in moderate 
wear. The teeth are very similar to those of the reedbuck but are a little wider. 

17 



Fig. 8-Redunca darti. Paratype skull, M 690, palatal aspect. Natural size. 

18 



An excellent palate (M 24; Fig. 9) shows the entire set of cheek teeth on both 
sides, agreement with the teeth in the skull being most exact. Comparative measure­
ments are given below:-

Para type (M 690) Palate (M 24) 

Left Right Left Right 

P 2 length 7! 7 est 
P 3 length 9 9 
P 4 length 9! 91. 

2 

Series P 2 - P4 26 251. 2 

M 1 length 13! 14 13 13 
M 2 length 15 15 16 16 
M 3 length 17 17 171. 2 17 

Ser~es Ml - M3 451. 2 46 461. 2 46 

Ratio P X 100 
56 551 

M 2 

P2 is somewhat larger than in the reedbuck and tbe cheek teeth are relatively 
and absolute! y broader than in R. arundinum. In this respect there is some resemblance 
to the puku and, in fact, the fossil gives the impression of being in some respects 
intermediate between the puku and the reedbuck. The molar teeth have their inner 
lobes more rounded than in either puku or reedbuck, itJ. both of which there is a 
slight kink as if the lobe had been gently pinched n~ar its tip. In the plentiful 
additional material available no significant variation either in size or in pattern of 
the upper teeth has been observed. 

Four lower jaws have been selected from the abun~dant collection to illustrate 
the lower dentition of the new species. The teeth are closely similar to those of 
R. arundinum but, as with the uppers, they lie at the upper limit of size variation 
of the living form. The dimensions of these four specimens are given below and they 
are shown in Fig. 9. 

DM2 length 
DMs length 
DM4 length 

Series DM2 - DM4 

M. 27 M. 26 M. 25 M.948 
--'-

L. (juvenile) L. (early wear) R. (moderate) L. (advanced) 

6* 
9!* 

17 

32 

19 



M. 27 M. 26 M. 25 M.948 

L. (juvenile) L. (early wear) R. (moderate) L. (advanced) 

P2 length 61. 
2 7* 7 est 

P3 length 10 10* 8 
P4 length 12 111* 101 

2 2 

Series P2 - P-± 281 2 27 251 2 

Mt length 15 151 2 13* 111* 2 

M2 length 161 2 15 15* 
M3 length 21 191 2 21 

Series M1 - M3 53 47.1. 2 47 

Ratio P X 100 
54 57 54 

M 
He:ght of jaw at P4/Mt 24 26 24 26 

* overlapping 

Amongst the remaining material referable to this species is a good braincase with 
both the occiput and one hom core well preserved. The specimen (M 494) is 
curiously associated with a very good frontlet of the gazelle Phenacotragus vanhoepeni. 
Most of the remaining hom cores lack the extreme tips but it can be seen that the 
sharp curvature of the upper part of the core is characteristic. There are a few 
imperfect cores of juveniles and in these also the curvature is apparent. It is noted 
in the living R. arundinum that youthful specimens already possess the curvature 
exhibited at the tip in the adult form, though not always so markedly and it thus 
seems reasonable to exclude from R. darti the few shorter horn cores which correspond 
so closely in their almost straight character with the normal horns of R. fulvorufula. 

From the West African bohor R. redunca, the fossil species ~differs in the massiveness 
and anterior convexity of the horn cores, and also in the greater actual length of 
the molar series and relative reduction of the premolars in R. darti. The fossil species 
R. ancystrocera, described by Arambourg (1947) from the lower Pleistocene Omo 
beds of southern Abyssinia, has simply curved hom cores resembling those of R. 
arundinum but being considerably longer. The only other extinct reduncine recorded 
from South Africa is "Kobus" venterae Broom from Florisbad but this i:s larger 
than R. darti and has very widely divergent horns. 

In some respects the Makapan fossil resembles the Puku (Adenota vardoni) but 
these two species cannot be placed in the same genus unless the Puku is to be 
separated generically from the Kobs. Failing this, the fo:;·sil species is better retained 
in Redunca. The specific name is given in honour of Professor R. A. Dart, who 
initiated the exploration of this remarkable site. 

20 



M 25 

M 948 

I . 1 
l 
1 

Fig.9-Redunca darti. Referred dentitions. M 24: palate with cheek teeth in moderate wear; 
M 27: left mandibular fragment with DM2 - M1; M 26: left mandible with P2 - M3 in 
early wear; M 25: right mandible with P2 - M3 in moderate wear; M 948: left mandible 

with P3 - M3 in advanced wear. atural size. 

21 



Redunca cf fulYorufula ( Afzelius) 

In addition to seven partial horn cores, the collection includes a frontlet with 
part of the braincase and both horns (M 465) , all of which are very similar to the 
living R. fulYorufula. The right core of the frontlet is complete, except for the 
extreme tip, but the left core lacks the top one third. The horns are nearly straight 
and cannot be assigned to R. darti, especially as sutures suggest an adult individual. 
The deep sculpturing of the frontal bone below the supraorbital foramen is notable 
in the specimen (Fig. 10) and is characteristic of Redunca. Similar features of the 

Fig. lG-Redunca cf fulvorufula. Frontlet, M 465, facial aspect. Natural size. 

22 



face are shown in another specimen (M 507), which is a hornless female. The 
braincase and occiput are well preserved and agree closely with material of the living 
species; possibly the fossil is a little larger than is usual. 

The pres·ence of a reduncine which is not R. darti is confirmed by a piece of 
right maxilla and palate with P 4 - M 3 in moderate wear (M 28). The follow~ng 
are the lengths of the teeth: P 4 === 8! mm; M 1 === 12! mm; M 2 === 12! mm; 
M 3 === 13 mm; P 4 

- M 3 === 46! mm; M 1 
- M 3 === 38 mm. These dimensions 

and the pattern of the teeth conform closely to R. fulYorufula. There are also two 
isolated upper third molars which appear to belong to this species. 

Tribe Hippotragini 

cf Oryx gaz.ella (Linnaeus) lc 

Five specimens in the collect~on are hippotragine in form and cannot be separated 
at present from the living gemsbok. The best specimen is a portion of left maxilla 
with M 2 and M 3 in moderate wear (M 34). The teeth are respectively 30 mm 
and 29 mm long, which is slightly larger than in iiving examples. An internal 
basal pillar rises to the crown in M 2 and half way up the side in M 3• Two worn 
upper right third molars also belong here and a 1noderately worn right lower first 
molar (M 8) 30! mm long and 29 mm high agrees also with lower teeth of the 
gemsbok. 

Tribe Alcelaphini 

Damaliscus sp. ( aff albifrons) 

A frontlet with the bases of the two horn cores (M 781 ; Fig. 11) indicates the 
presence of a Damaliscus which resembles the living blesbok but d~splays features 
which might be regarded as excluding it from that species. In the fossil the bases 
of the horn cores are strongly oval and measure 33 ·mm and 45 mm across their 
minor and ·major axes compared with 31 by 43 mm in a large Hving blesbok. In 
the fossil form tile better preserved (right) core shows a rapid upward taper and 
at a he~ght 75 mm above the base of the rugose core, th~ dimensions are reduced to 
17 mm by 29 mm; in a typical living specimen at the same height, these dimensions 
are 18 mm by 31 mm; thus the base is a little larger and the taper is more abrupt 
in the fossil than in the living form. In the living species the axis of elongation 
makes an angle with the median line of the skull very close to 4 5° ; in the fossil 
this line is more strongly transverse and approaches 60 °. The spiral twisting of 
the core is sharper in the fossil than in the living form. The pedicle at the· base 
of the horn is also stouter and less elevated in the fossil specimen. The inference 
is that the horns of the fossil resemble those of the blesbok but are shorter, taper 
and twist ·more abruptly and stand on a wider, stouter and low·er .frontlet. 

A partial braincase and occiput (M 510) is typically ~Damaliscine in form but 
is rather more massive, having a very heavy occipital condyle and higher occiput 

23 



than the living blesbok. As the frontlet suggests a more robust skull, the occiput 
probably belongs to the same species. 

Two upper third molar teeth resemble those of the blesbok very closely and are 
comparable in size, as also is a right lower jaw with P 4 (erupting below a fragment of 
DM4), M1and M2 being in early wear (M 778). 

cf Gorgon taurinus (BurcheH) 

A fair number of partial dentitions and isolated upper and lower cheek teeth 
agree very closely with those of the living brindled gnu. The best specimen is a partial 
palate with well worn LM1, LM2

, RM2 and RM3 (:t\1 857). At the front lobe 
of M 2 the palate is 58 mm wide and the lengths of the individual teeth are : 
LM1 === 21! mm; LM2 === 25! mm; RM2 === 26 mtn; RM3 === 22! mm. 

The best lower jaw (M 858) belongs to the right side and there is a less complete 

I 

l 
l 
I 

I 
! 

-~- ~-J 
Fig. 11-Damaliscus sp. (aff albifrons). Frontlet, M 781, facial aspect. Natural size. 

24 



fragment of the left side which is probably its mate. In both specimens P3 - M3 are 
preserved and M1 - M3 has a total length of 69t mm. 

Five other partial lower jaws, three partial upper dentitions, a dozen loose lower 
molars and three dozen isolated upper molars may also be assigned to this species. 
There are, in addition, a number of juvenile dentitions which cannot with Cef~ainty 
be distinguished from those of Makapania broomi, which is described below. 

cf Alcelaphus robustus Cooke 

Two upper third molars and one lower molar are typically alcelaphine in 
structure but are larger than the living species. They are shown in Fig. 12 and rna y _ 
be compared w:th similar specimens from the Vaal river gravels which were named 
A lcelaphus robustus (Cooke 1949). 

cf Pelorocerus helmei (Lyle) 

Three well-worn upper molars of alcelaphine type (Fi6. 12) are even larger than 
those tentatively placed in Alcelaphus robustus and are closely comparable with 
teeth previously referred by the present authors to Pelorocerus helmei (Lyle). 

r ----­

' I 
l 
t 

I 

- --s 

M548 

I 

~ 
I 
I 

. .. 

-· . j 

M 76~ ·,j 
- - "<- -- j 

~-·· . -... >-~.,:·:·~· - ········ ·-·····~·: ..... ~:-~-.. ... :.__:___>··~J 

Fig. 12-LEFT: cf Alcelaphus robustus. Two left upper third molars (M775, M776) and a lower 
.right first molar (M 777) RIGHT: cf Pelorocerus helmei. An upper right third molar (M 548), 

a ·qamaged upper right second molar (M 764) and an upper left first molar (M 765). 
Natural size. · · 

25 



Hoffman (1953) provides confirmatory evidence for the size of the teeth of that 
species but indicates that a related form, wh:ch he regards as generically separate 
and calls Lunatoceras mirum, has almost identical teeth. The reference of the scanty 
Makapan material must, therefore, be very tentative at this stage. 

A left mandibular fragment with the three milk molars in advanced wear (which 
is one of the specimens recovered by Professor Dart in 1927) clearly belongs to 
this group. DM4 measures 30 mm in length, con1pared with 25 mm for an equally 
worn tooth of a large Gorgon taurinus, a ratio comparable with that between Gorgon 
and Pelorocerus in adult third molars. 

Hopwood and Hollyfield (1954) have erroneously tncluded Pelorocerus among 
the Caprinae. It is true rhat in the form of the horn-cores P elorocerus is quite similar 
to the ex:t~nct genus Pultiphagonides from Olduvai, which Hopwood and Hollyfield 
also regard -as caprine. Through Dr Hopwood's kindness we have been able to 
examine the type skull of Pultiphagonides african!-ls Hopwood in the British Museum 
(Nat. Hist.) ; on dental characters we consider it to be unquestionably a:lcelaphine. 
There appears therefore no good reason to depart from the view that Pelorocerus is 
also alcelaphine. 

MAKAP ANIA gen. nov. 

Alcelaph~ni of medium to large size, with horns arising close to the posterior orbital 
margin and directed strongly laterally; teeth hypsodont; upper molars with typically 
alcelaphine outer ribs and folds but medial 'lobes more V-shaped and resembling 
those of T ragelaphus; lower molars alcelaphine in structure but with ~lateral lobes 
more V-shaped. Genotype Makapania broomi sp. nov. 

Makapania broomi sp. nov. 

Diagnosis: A Makapania approximately the size of Gorgon taurinus. 

H olotype: Partial skull with horn cores and part of palate bearing LM3 and part 
of LM2 (M 162; Figs. 13, 14). 

Paratypes: (a) Slightly crushed snout (M 163; Fig. 15); 
(b) Crushed snout and associated portions of mandibles (~M 651 A, 

B, C, Fig. 16); 

(c) Portion of right mandible with P2 - M3 (M 655; Fig. 16). 
The holotype specimen consists of the anterior portion of the brain case, with 

parts of both hom cores, and the hinder portion of the palate. The facial region 
is damaged. 

A low transverse elevation on the roof of the brain case represents the horn­
pedicle of Alcelaphus. The hom cores, which spring from the extremities of --,this 
elevation, are at first directed somewhat upwards and then pass almost horizonta:lly 
outwards in the transverse plane; they appear to have been upcurved towards their 
tips. This form differs both from Alcelaphus and from Connochaetes in having the 

26 



root of the hom core very close to rhe posterior borx:ler of the orbit. The preserved 
portion of the more complete (right) hom core measures approximately 17 em in 
length, and probably at least 10 em of the tip is missing. At its base the core is 
oval in cross-section, measuring 62 mm in transverse diameter and 48 mm in 
anteroposterior diameter; at 15 em from rhe base it is nearly circular in section, 
measuring 3 2 x 3 3 mm. 

The bony palate is unusual in having its median portion (between the pterygoid 
plates) prolonged further posteriorly than the lateral portions. L1M 3 and the hinder 
portion of :UM2 and of RM3 in early wear are preserved. The breadth of the palate 
between the third molar teeth is approximately 60 mm. The intact LM3 is 28 mm 
long and 18 mm broad. 

The peculiar formation of the palate and also the pattern of the teeth in the 
holotype make it possible to identify with this species a crushed snout with 
LP3 - M3 in early wear (M 163). Another crushed snout (M 652A) with LM2 

and M 3 in early wear also clear! y belongs to this sp~cies and is associated in a 
single block with damaged left and right mandibles. The right mandible (M 652C) 
has the outer lobes of the molars rather damaged and only the root of P4 remains. 
The left mandible (M 652B) has LP3 - M3 well preserved, the premolars being 
unworn. The upper and lower dentitions are thus definitely associated and it is then 
possible to identify as belonging to this species a well worn right mandible with 
P2 - Ms in advanced wear (M 655, Fig. 16). The dimensions of the teeth in these 
specimens are given below. 

Measurements of upper and lower cheek teeth (in mm). 

UPPER LOWER 

M 162 M163 M652A T.M.t M652B M652C M655 

p2 Length 11 p2 Length 10 
Breadth 10 Breadth 6 

p3 Length 12 14 p3 Length 15 12 
Breadth 13-l 2 13 Breadth 8 9 

P4 Length 13 14 p4 Length 16 15 
Breadth 15 15 Breadth 10 11 

Series P2 - P4 39 37 

Ml Length 21 25 M1 Length 24 22* 19 
Breadth 19 19 Breadth 12 c12* 13 

M2 Length 26 26? 27 M 2 Length 26 251. 
") 

25 
Breadth 19 18 19 Breadth 12 c12* 15 

M3 Length 28 31 24+ M3 Length 34? 34 36 
Breadth 18 19 15+ Breadth 111. 

") 
12 14 

Series Ml - M3 78 84 81 80 

t Specimen in Transvaal Museum * Dan1aged 

27 



N 
00 

FF'""'"':""'·--:---~""""~~.-

r 
r: 
t;, 

L 
f. 

r:_, 

I. 

X t 

. y2 

Fig. 13-Makapania broomi. Palatal aspect of holotype skull fragment (M 162). One half natural size. On the left are 
shown the left third molar and partial second molar, natural size. 



N 
\0 

.. 

o , ' . .10 em 
I ea F?3 .k+a es · '. e=J 

Fig. 14- .1!akapania broon1i. Dorsal aspect of bolo type skull fragment (M 162). One half naturai size. 

i 

,J 

j 



i 
~ 

-·---~· 

Fig. 15-Makapania broomi. Slightly crushed snout of paratype, M 163, palatal aspect. 
atural size. 

30 



The most conspicuous feature of the upper molars is the pronounced V-shape of 
their medial lobes, in which respect they differ from all living alcelaphines. Super­
fic~ally the teeth resemble those of Tragelaphus but are distinguished by being much 
more hypsodont; the height of the crown of M 2 in specimen M 652 is 49 mm and 
of M 3 is 56 mm. The extemal folds are strongly marked and the ribs moderately 
developed, giving ~he outer face of the teeth a typically alcelaphine appearance. As 
the teeth are worn down, they increase in width much less rapidly than do those of 
Tragelaphus; the groove between the medial lobes is wider and shallower and there 
is a large central enamel island. There is no basal pill2 r rising from the cingulum. 
In ~the rear enamel island, a posterior spur is sometimes present but in the anterior 
one this is unusual. The upper premolars are small in proportion to the molars, as in 
Connochaetes and Gorgon, but are relatively a little larger than in the gnus. 

Similarly, the lower molars differ from those of liv:ng Alcelaphines in having 
angular lateral lobes. They are distinguished frotn tho~·.e of Tragelaphus by being 
more hypsodont (M3 in M 652 having a crown height of 57 mm) a~nd by possessing 
a rudimentary antero-lateral fold or flange. As in the upper molars there is no 
basal pillar. P 4 is hypsodont and agrees essentially with P 4 in Alcelaphus, 
Connochaetes and Gorgon; P3 is also hypsodont but of simpler form; P2 is a simple 
compressed, sectorial tooth. 

In addition to the type material, there are several other specimens which contribute 
information regarding the structure of the skull and horns of Makapania. A partial 
brain case, with the occipital region rather eroded (M 499) possesses the posterior 
portion of the basal part of each horn core and its generic reference is quite certain. 
The ·lower edge of the occipital condyles and the basilar tubercles are also preserved. 
Another specimen (M 952) comprises an occiput with part of the brain case and it 
seems to agree closely in dimensions and character with the back of the previous 
specimen. This fragment displays a sharp posterior ridge above the foramen magnum, 
a feature possessed very markedly al~o by Damaliscus and, to a lesser degree, by 
Alcelaphus. 

Three other pa:r:tial skulls a~nd twenty fragments of horn cores attest to the 
typical character of the holotype specimen. One fragment (M 487) comes from near 
the tip of the core and indicates a somewhat gnu-like shape in this region, though 
the curvature is less marked than in either Connochaetes gnou or Gorgon taurinus. 

Five maxillary fragments with teeth in various stages of wear and some twenty 
isolated upper molars confirm the features shown in the type. These teeth tend to be 
slightly larger than those of G. taurinus. Seven lower jaw fragments and more than 
a dozen isolated lower molars also agree wtth the paratypes but, as the latter are 
either very worn or relatively unworn, a good specimeu with LP4 - M2 in normal 
wear is figured here (•M 962; Fig. 16) togetiher with an upper and a lower series 
of what is believed to be the milk dentition of Makapania. 

31 



\N 
N 

M 95 

M 953 

·M 652B 

M 655 

Fig. 16--Makapania broomi. M 954: referred juvenile palate with milk teeth; M 953: referred partial juvenile mandible with LDM4 - M2; 
M 652B: paratype right mandible with P3 - M3 in early wear; M 65 5: paratype right mandible with P2 - M3 in moderately advanced wear. 

Natural size. 



A damaged brain case of a very young individual (M 512) has part of the base 
of the left horn preserved. This specimen resembles fairly closely a gnu skull of 
similar age but the frontoparietal sutures are different and would fit well with the 
probable state in Makapania at thi3 early age. 

In the collections of the Transvaal Museum, Pretoria, there is a maxillary fragment 
from the Makapan limeworks, possessing LP2 - M 2 in early wear, which c!early 
belongs to Makapania broomi. The same species has also been observed in a private 
collection from deposits in the Sterkfontein area. 

By the form of its horn cores this antelope could be tegarded as falling between 
the kongoni (Alcelaphus cokei) and its allies on the one hand and Gorgon taurinus 
on the other, but it differs from both in ha vi·ag the horns arising close to the 
orbits. It might therefore be considered as an offshoot from the common ancestral 
stock of these two types which has escaped specialisation in the origin of the hom 
core, but has developed its own "pseudo-tragelaphine" specialisation in the molar 
teeth. This form cannot therefore be assigned to either Alcelaphus or Connochaetes, 
nor can it be placed in any of the described extinct genera, Pelorocerus (van 
Hoepen, 1932), Parmularius (Hopwood, 1934), or Parestigorgon (Dietrich, 1950). 
The upper molars somewhat resemble those assigned by Dietrich ( 1950) to 
Praedamalis deturi, but the character of the associated hom core is wholly different. 

Pultiphagonides africanus Hopwood has horn cores of similar character but shorter. 
It has been mentioned above that Pultiphagonides on its dental characters is certainly 
an alcelaphine; its molars are almost indistinguishable from those of Alcelaphus. The 
Makapa·n fossil with its very specialised molars cannot be ass~igned to the 
Olduvai genus. It has therefore seemed justifiable to create a new genus for which 
the name MakaPa.ttia appears appropriate; the specific name has been cho~en in 
memory of ·the ~late Dr ·Robert Broom who contributed so much to our knowledge 
of the Quaternary mammals of South Africa. 

Subfam:ly ANTILOPINAE 

Tribe Neotragini 

Oreotragus major Wells 

The collection includes a number of jaws, teeth and horn cores of neotragine type, 
apparently belonging to a single species. The best representative specimens comprise 
a fron~let with both horn cores (M 476; Fig. 18), an entire palate wi1Jh P2 

- M 3 

on ·borh sides (M 951; Fig. 17) and two left mandibles ·wicil a·ll the cheek teeth 
(M 997 and M 998; Fig. 17). The teeth of the palate are rather worn but a ~left 
maxilla displays a:H the teeth in early wear (M949; Fig. 17). 

These remains agree in all essential characters wibh the genus Oreotragus but 
their dimensions exceed by fifte~n to twenty per cent the limits of the liv:ng 
0. oreotragus. A large exci·nct species 0. major was founded upon a skuH from a read 
breccia deposit at another site in the Makapan Valley (Wells 1951) and the present 
collection conforms sufficient closely to be referred to the same speci~s. 

33 



w 
~ 

M -.9 

M ···9~·· t 

Fig. 17-0reotragus 1najor. M 651: palatal view of somewhat crushed type skull from the Swartkrans (Makapan) site; 
M 949: referred partial left maxilla with p2 - M3 in early wear; M 951: referred palate with cheek teeth in moderately 
advance·q wear; M 997: left mandible with P2 - Ma in fairly early wear; M 998: left mandible with P2 - M3 in moderately 

advanced wear. Natural size. 



Fig. 18-0reotragus major. Frontlet, M 476, facial aspect. atural size. 

In the type skull only the lower part of the right horr. core was present, the left 
hom hav:ng been lost. The frontlet poss~esses both horns and the distance from 
the centre of each core to the median suture is 30 mm, exactly the same as in the 
type; the horn cores are slightly stouter but three isolated cores agree more exactly 
with the type. The heights of the left and right cores are approximately 43 and 
46 mm respectively. 

The teeth of the type skull have not been figured and the palatal aspect of the type 
is illustrated in Fig. 17 for comparison with selected specimens from the Limeworks 
depo.;it. The following are the relevant measurements (in mm) : . 

Type Palate 
(left) (left) 

p2 Length 9-! 
Breadth 7-! 

p3 Length 10 9! 
Breadth 8 7-! 

p4 Length 8 8 
Breadth 8 7-! 

Series p2 - P4 27 

35 

M951 
(right) 

9-! 
8 
9 
8 
8 
7-! 

26! 2 

M949 
(left) 

9 
6.l-2 

9 
7 
8 
7 

26 

M999 
(right) 

&\-2 

7 

Oreotragus 
or eo tragus 
(average) 

22j 



Oreotragus 
Type Palate M951 M949 M999 or eo tragus 
(left) (left) (right) (left) (right) (average) 

Ml Length lOl 2 
9.1 

2 9 9 10-! 9 
Breadth 9.1 

2 8 8 8 8 71_ 
2 

M2 Length 12 11.1 2 11.1 2 11 12 9.1 2 

Breadth 9-! 8! 2 8! 2 8.1 2 8! 2 8 
M 3 Length Ill_ 

2 10 9.1 
2 9!est 10! 10 

Breadth 8 8 8 7.1 
2 

71_ 
2 7-! 

Series M l - M 3 34 31 30 29.1 2 33 28! 2 

Ratio P X 100 
87 88! BE- 79 

M 

The relatively large size of M2 in the fossil material as compared with the living 
species is noteworthy. 

In the type skull, only the second and third lobes of the left 'M3 represent the 
lower dentition. The breadth of the second lobe is 6 mm and the length of the 
fragment is 9 mm, giving an estimated length for t~e complete third molar of 
about 14 mm. In shape and size the teeth of the two dozen partial lower jaws 
conform to the expected dimensions. The lengths of the teeth in the two most 
complete left lower jaws are given below (in mm) :-

M 997 6.\ 9 9.1 25 l 9 10 13 32 78 2 2 
I 

M 998 5 8-! lOl 24 10 10-! 14 34! I 69! 2 

l 
Oreotragus 4 7.! 9 
oreotragus 2 20! 8 9-! 12! 1 30 68-! 

It would appear that vhe type skull lies close to the upper limit of size variation 
of the species or, alternatively, that the material from the Limeworks deposit 
represents a slight! y smaller variety or race. 

Tribe Antilopini 

Aepyceros cf melampus (Lichtenstein) 

A portion of hom core (M 654) approximately 8 em in length, shows the 
characteristic ringing of the impala core (Fig. 19) . As far as can be determined 
from the curvature it belongs to the central part of the hom. It is larger than the cores 

36 



in any specimens available but may not lie outside the conceivable limits of 
variation. A partial right lower jaw (M 759) with well worn Mt - M3 and a 
fragment of P 4 is ind:stinguishable from living material, as also is a juvenile left 
lower jaw with DM3 - M3 in early wear (M 758). A number of isolated upper 
teeth probably belong to this species. 

+- · "' 

Fig. 19-Aepyceros cf melampus. Fragment of horn core (M 654) 

Gazella gracilior sp. nov. 

Diagnosis: A gazelle somewhat resembling G. rufifrons but with more delicate horns 
and having Mt relatively larger. 

H olotype: A frontlet with horn cores, damaged at tlhe tips (M 773; Fig. 20). 

Paratype: A left lower jaw fragment with P4 (damaged) - M3 (M 767; Fig. 21). 

. The frontlet comprises part of the parietal bone, the horn cores and the fronta:ls 
as far as the supraorbital foramen qJ? each side. The horns are fairly close together, 
the basal separation being 22 - 23 mm, and they rises steeply from the brain case. 
The heights of the right and left cores as preserved are approximately 60 mm and 
70 mm respectively, measured on the inner s:de; the probable height when complete 
is about 10 - 11 em. The hor:ns curve gently outwards and backwards, the cQndition 
of the extreme tip being uncertain. In cross-section, the cores are oval with antero­
posterior elongation. The greatest and least diameters at the base are 21! mm and 
16! mm for the left hom and 22 mm and 17! mm for the right horn; 4 em higher 
up, the measurements are 15! mm and 13 mm on both cores. An isolated broken 
core (M 772) is closely similar in shape and in basal dimensions but tapers more 
rapidly upwards. 

In the British Mus·eum (~Natural History) is an inr.eresting specimen from this 
locality, taken over by Dr K. P. Oakley for exhib~tion. It comprises a right horn 
core thrust into the marrow cavity of a broken long bone (Fig. 20). 'Dhe tip is 

37 



w 
00 

r -.. ··-···~·.-~-

Fig. 20-Gazella gracilior. Holotype frontlet, M 773, facial aspect (centre) and left lateral aspect (right); on the left 1s 
a photograph of a cast of the referred specimen in the British Museum (Natural History.) Natural size. 



missing and the length of core preserved is 61 mm. The greatest and least diameters 
.at the base are 25 nun and 19 mm respectively and at the broken tip are 17 nun and 
12 mm. It is thus slightly larger than the holotype but is otherwise closely simJar. 

On the brain case, the suture between the frontals lies on a ridge which passes 
into a raised area on the anterior edge of the parietal. The horn base above the 
supraorbital foramen is elevated from the surface of the frontals ~ 

In the living African gazelles, the horn bases of males are normally oval in 
~ection and those of females are more nearly circular. In G. granti, in which the 
.cores of the male are greatly elongated in section, the females also have rather oval 
horn cores; but this is exceptional and, in any case, the fossil has horns far too 
short to belong to the G. granti female. There is thus good reas.on to consider 
the fossJ frontlet as belonging to a male animal. 

Parallelism of the two horns is exhibited by G. cuyieri, G. leptoceras, G. rufifrons, 
G. spekei, G. thomsoni, G. dama and G. granti. The two last named are much 
too large and the horns of the Dama Gazelle curve too ~harply backwards. Speke's 
·Gazelle is of comparable size, though rather larger, but the sharp backward arching 
.of its horns serve to eliminate it. Of rhe remaining four, the following are average 
dimensions for the core bases : 

G. cuyzert 32 x 22 mm 
G. leptoceras 26 x 19 mm 
G. rufifrons 27 x 19 mm 
G. thomsoni 32 x 22 tnm 

AH are thus somewhat larger than the fossil frontlet. Both G. rufifrons and G. 
thomsoni show ridging of the fronto-parietal region and the fossil form resembles 
these two species more closely than it does any other African gazelle; the fossil 
<:learly had fairly short horns and is more like G. rufifrons in this respect than like 
the long-horned G. thomsoni. 

M767 

M:766 

Fig. 21-Gazella grae1lior. M 767: paratype left mandibular fragment with P4- M3 in moderate 
wear; M 768: referred left mandibular fragment with P4 - M in moderate wear; M 766: left 
mandible with P3 -- M3 in advance(\ wear; M 771: maxillary fragment with RM2 - M3 in 

moderate wear. Natural size. 

39 



Four lower dentitions (all of the left side) are gazelle-like in form and of the 

correct size to be associated with the frontlet described above. One of these (M 767; 

Fig. 21) has been selected as a para type of the species. It has the greater part of 

P4 preserved and the whole of M1 - M3 in moderate wear. A similarly worn 

fragment (M 7 68; Fig. 21) has P 4 intact but lacks M3 and a specimen in advance 

wear (M 776; Fig. 21) has P3 - M3 and the root of P2 preserved. The measurements 

of these specimens are given below (in nun):-

P2 Length 

Breadth 

Ps Length 

Breadth 

P4 Length 

Breath 

Series P2 - P4 

M1 Length 

Breadth 

M2 Length 

Breadth 

Ms Length · 

Breadth 

Series M1 - Ms 

Ratio ·P X 100 

M 

Height of jaw at P4jM1 

Para type 

M 767 

9 est 

5 

12 

6! 
13 

6.1 
2 

17.1 2 

6.1 
2 

421 2 

17 

M 768 

10 

5 

12 
6 

13.1 2 

6.1 
2 

19 est 

M 766 

5!est 

3 est 

7 
4 

7! 
5 

20 est 

10 

7 
12 

8 

18 
7 

40 

50 

18 

A right maxillary fragment (M 771; Fig. 21) with M2 and M3 and also an 

isolated RM3 show typically gazelle-like teeth which may be provisionally assigned 

to this species. In the fragment, M 2 is 13! mtn long and 9! mm broad and M 3 

is 14! ·mm long, 9! mm broad and 18 mm high; the isolated RM3 is 14 mm 1long and 

9 nun broad. The table overleaf gives average figures for the dimensions of tee~h in 

the living gazelles of comparable size. 

40 



G. cuvieri G. leptocera~ G. rufifrons G. thomsoni Fossil 

P2 Length 51 
2 

51 
2 5 5 51? 

2" 
Breadth 31 

2 3 3 3 3 ? 
P3 Length 8 71 

2 7! 7 7 ? 
Breadth 4 . 31 

2 4 4 4 
P4 Length 41 

2 4 4 5 5 

Series P2 - P4 221 2 21! 201 2 20 21!? 

·M1 Length 11 101 2 10 10 11 
Breadth 6 6 6 61 

2 
61 

2 
M2 Length 131 2 13 121 2 121 2 13 

Breadth 6l-2 
61 

2 7 71 
2 7 

M3 Length 17 17l 2 181 2 181 2 18 
Breadth 6 6 7 61 

2 7 

Series M 1 - M3 411 2 41 41 41 42 

Ratio P X 100 
54 52.1 50 49 51 

M 2 

Height of jaw at P4jM1 17 17 18 19 18 

M 2 Length 14 12! 121 2 13 13! 
Breadth 8 8! 91 

2 101 2 
91 

2 

M 3 Length 13 13 14.1 2 15 14 
Breadth 7 7! 9 9 91 

2 

Height 18 17 15 16 18 

The closest comparison, both in dime·nsions and in crown pattern, is with 

G. rufifrons and G. thomsoni. Although it is not emphasised by the average figures, 

comparison of individual dentitions having cheek teeth of similar overall length shows 

that Mt in the fossil is consistently larger than in G. rufifrons or in G. thomsoni. In 

th:s respect it approximates to G. cuvieri but differs sharply in having broader teeth. 

It thus seems clear that the fossil cannot be identified specifically with the living 

African gazelles. 

Of the fossil species hitherto described from southern and East Africa whose horns 

are known, none is comparable with G. gracilior. The teevh previously described as 

G. wellsi (Cooke 1949) come close to those in the present collection but identity 

cannot yet be established. The teeth of G. wellsi appear to be somewhat more 

hypsodont than in G. gracilior. 

41 



it 

Scm 
I E?3 E?3 I 

Fig. 22-Phenacotragus vanhoepeni. Holotype frontlet, M 599, facial and left laterial aspects. Two-thirds natl.lral size. 



Phenacotragus ~~anhoepeni sp. nov. 

Diagnosis: A Phenacotragus with horns more massive, rising more vertically from 
the frontal bone and more compressed laterally than in P. recki Sdhwarz. 

Holotype: Frontlet with both horn cores slightly damaged at the tips (M 599; 
Fig. 22). 

P aratypes: (a) Complete brain case with parts of the frontals and lower part of 
right horn core (M 406). 

(b) ·Damaged skull with palate but lacking roof of brain case (M 598; 
Fig. 23). 

There are a large number of partial skulls and portions of horn cores which belong 
to an antelope of the gazelle group but differ from the living African species. Many 
partial upper and lower jaws and teeth belong to a medium sized gazelle and one 
partial skull furnishes an anatomical link between the upper dentition and the 
horn cores. 

In the holotype frontlet (M 599; Fig. 22) the horn cores rise almost vertically 
from the brain case and their basal porfons diverge slightly. They are sharply 
recurved backwards, with the terminal portion deflected outwards and upwards. 
The length of the more complete core, measured along the anterior curve, is 16 em 
as preserved and is estimated as 20 - 22 em when intact. The basal separation 
between the cores is 18 mm. The cores are compressed laterally, the greatest (antero­
posterior) diameter being 4 2 mm and the least (transverse) diameter 29 mm; 5 em up 
from the base the corresponding dimensions are 36 mm and 22 mm; 10 em from 
the base they are 28 mm and 17 mm, and at the broken t ~p are 17! mm and 12!- mm. 
The fifty or more other horn cores and fragments confirm that the features shown 
in the holotype are reasonably typical; the range of variation is small. 

As in Gazella gracilior, the suture between the frontals runs along a marked 
ridge and widens into an ·elevated area at the front of the parietal. 

These cores do not compare at all well with those of any of the exist~ng African 
species of Gazelle, nor with those of t1he extinct species G. helmoedi van Hoepen. 
They correspond fairly closely with the horn cores of the extinct Phenaco-tragus recki, 
described by Schwarz (1937) from the Oldoway beds of Tanganyika, and can 
reasonably be referred to the same genus. The Makapansgat specimens indicate an 
animal with horns more massive than those of P. recki and the horns differ also in 
rising more steeply from the frontal bone, in having a longer straight lower portion 
and in being laterally more compressed. These differences are regarded as of 
specific value. 

The paratype brain case (M 406) provides information regarding the back part 
of the skull, wh:ch is a little larger than in P. recki. The occipital and basal parts 
of the brain case are in a good state of preservation hut vhe parietal is very slight! y 
crushed. The nuchal aspect is almost identical with that of the living impala but the 

43 



Fig. 23- Phenacotragus vanhoepeni. Palatal aspect of paratype skull, M 598. Natural stze. 



basilar part of the occipital bone in the fossil is stouter and longer w:th much more 
strong! y developed basilar tubercles. The bulges of the parietal behind the parietal 

eminence in the impala are lacking in the fossil form, giving a different profile to the 
brain case. The portion of the occipital bone lying above the occipital crest is wider 
in the fossJ than in the impala. 

The para type skull (M 598) has sufficient of the back of the brain case preserved 

for it to be virtuaH y certain that it belongs to the same species as the brain case just 
described. The interest of the specimen attaches prima·ri'ly to the palate (Fig. 23) 
which possesses M 1 - M 3 of both sides. The teeth are smaller · than those of the 

impala or the springbok, though similar in general pattern. They agree closely with 
the palate of P. recki figured originally by Schwarz (1937) but in the Makapan 
specimen the lateral folds are sharper. From this palate it is possible to identify some 

fourteen partial upper dentitions and a dozen loose upper molars as belonging 
to the same species. Four of these from the right side have been selected to illustrate 
different stages of wear; their measurements (in mm) are given below, together 

with those of the paratype palate, and they are illustrated in Fig. 24. 

Eearly Early Moderate Advanced 
Para type wear wear wear wear 
M 598 IM 615 IM 611 ·M 609 M 612 P. recki 

p2 Length 6! 61 
2 

Breadth 6 51 
2 

p3 Length 8 10 71 
2 

Breadth 6 61 
2 ~ 2 

p4 Length 8 8! 9 7! 81 
2 

Breadth 7 71 
2 7! 9 8 

Ser~es P2 - P4 22! 221 2 

M 1 Length 14 14 14 131 2 111 2 13! 
Breadth 11 10 9! 101 2 11! 111 2 

M 2 Leng~h 161 2 16 16 15 15! 141 2 

Breadth 11 10 10 10! 111 2 101 2 

M 3 Length 14 14 15 17 141 2 

Breadth 10 9 9! 11 10 

Series M 1 - M 3 44! 44 43! 44 42! 

Ratio P X 100 53 
M 

45 



It is not possible to associate the lower dentition with the uppers with any great 

degree of certainty but there are two dozen lower jaws or fragments and a dozen 

loose lower molars which have the requisite size and form. These may be accepted 

with some confidence as representing the lower dentition of P. yanhoepeni. They are 

typically gazelline in form, resembling the lower teetJh of the spring buck very closei y 

except that the fossil possesses well developed second and third premolars, whereas in 

the springbuck the second premolar is usually lacking and even the third premolar 

is vestigial. Six dentitions have been selected to illustrate various stages of wear. 

Their measurements (in mm) are given below and four are illustrated in Fig. 24. 

M 533 M496 
Very Early Early 

(R) (L) 

{DM2 5?) 
3? 

(DM3 9 ) 
4 

4 
3 
8 
4 

P'2 Length 
Breadth 

P3 Length 
Breadth 

P4 Length 
Breadth 

(DM415!) 
6 

10 
5 

Series P2 - P4 

M1 Length 
Breadth 

M2 Length 
Breadth 

M 3 Length 
Breadth 

Series M1 - M 3 

Ratio P X 100 

M 

Height of jaw 
at P4jM1 

14 
6! 

20 

22 

13 
7 

15! 
6! 

21 
6! 

491. 2 

18 

M502 M42 M38 M35 
Moderate Moderate Moderate Advanced 

(L) (L) (L) (L) 

8! 
41. 

2 

9 
6 

12! 
7! 

15 
8 

21~ 
7! 

49 

20 

root 6 
?3 

root 7 
?4 
8 
5 

21est 

111. 2 
7 

15 
8 

22 
7 

481. 2 

43 

20 

root 5! 
?31. . 2 

root 6~ 
?4 
9 
6 

21est 

11 
7! 

14 
7! 

23! 
7! 

481. 2 

43 

19 

4! 
3! 
61. 

2 
4 
9 
5 

20 

11 
7 

13! 
8 

22! 
8 

47 

421. 2 

21 

The specific name is chosen in honour of Dr E. C. N. van Hoepen, who first 

recogn:sed the occurence of a true gazelle (G. helmoedi) in ~he Quaternary of South 

Africa. 

46 



~ 
'..J 

- --· _,.... ... ._.~."7""'-..._--:"'"'----:-.··-··-::-~· ··-r:---:::--··n----......... •. ~- ...,._ ... ~·------· w._._. - ....... ~----v-·'(>···-~--

M -615 

M 533 

J · 

M 35 

Fig. 24-Phenacotragus vanhoepeni. REFERRED RIGHT UPPER DENTITIONS: M 615: p2 - M2 in early wear; M 611: p-1 - M3 in early 
wear; M 609: p3 - M3 in moderate wear; M 612: p4 - M3 in advanced wear. REFERRED LOWER DENTITIONS: M 533: RDM3 - M1 ; 

M 496: LP2 - M3 in early wear; M 502: LP3 · M'3 in moderate wear; M 35: LP2 - M3 in advanced wear. Natural size. 



DISCUSSION 

Large species 

It has been noted that, in several of the species not distinguishable from living 
forms, the mean size of the fo:;sil remains a{Proaches the upper size limit of the 
existing type. This is most apparent w:th the impala and Tragelaphus, and is 
suggested also for the buffalo, kudu and oryx. It seems possible that all rhese, 
and perhaps a:lso the klipspringer ( Oreotragus ma}or), represent large forms ancestral 
to the smaller existing species. Such a tendency for decrease in size of species during 
the Quaternary has been observed in other parts of ·the world. 

EnYironment 

The Mcrkapan Limeworks fauna includes no fewer than eighteen bovld species. 
Of these eight ca.n be regarded as certainly extinct; three of them are identified 
wit~h forms already known elsewhere while the other five are new. Three species 
belong to extinct genera (Phenacotragus, Makapania and Pelorocerus or Lunatoceras) 
and one to a genus not now represented in southern Africa (Gazella). 

This great wealth of species is clearly related to the physiographic setting of the 
site in a scrub-covered mountain mass, with extensive open or bush-covered plains 
at no great distance. It is instructive in this regard to compare the Makapan list of 
species with those from two sites belonging to a much more recent per:od (late 
Gamblian or Makalian*) - the Border Cave, Ingwavuma, (Codke, Malan and 
Wells 1945) and the Wonderwerk Cave, Kuruman (Malan and Cooke 1941; Malan 
and Wells 1943). The Border Cave assemblage repres·ents a broadly comparable 
range of environments under a climatic regime similar to that of the Makapan site 
to-day; that from Wonderwerk belongs to an area of appreciably lower rainfall. 
While the Border Cave fauna includes kudu, impala, and reedbuck, but lacks eland, . 
Oryx, and Damaliscus, ·n that of Wonderwerk the three last named are present 
(together with "Pelorocerus helmei'') but the former three are lacking. The presence 
in the Makapan fauna of an Oryx (albeit scantily represented) indicates that at some 
stage in the accumulation of the deposit vhe climate was appreciably drier than it 
ts now. 

In point of numbers the Makapan fauna is dominated by four species, Redunca 
darti, Phenacotragus 'Yanhoepni, Makapania broomi, and the wildebeest. It is remark­
able rhat three of these four are extinct and two belong ~no extinct genera. The less 
abundant species fall into two group3, those which are moderately well represented 
and those which are notably rare. In the former may be placed the buffalo, eland, 
and kudu, the supposed nyala, the mountain reedbuck, and three extinct species, 
Cephalophu, pricei, Oreotragus major, and Gazella gracilior. The la~ter group includes 

* Using these terms purely in a stratigraphic sense and without implying any sort of climatic 
correspondence. 

48 



the blue duiker, Oryx, impala, clle somewhat problematical Damaliscus, and the two 
large extinct species "Alcelaphus robustus" and ''Pelorocerus helmei". 

Although none of these three groupings can be said to consist entirely of species 
characterising a particular environment, the total impression conveyed by the bovid 
fauna suggests a moderately well watered, bushy valley opening on to nearby plains 
which, at no great distance, were decidedly dry. 

Age and Affinities 

· The only pre-Gamblian bovid fauna hitherto described from South Africa is that 
of the Vaal River deposits (Cooke 1949) . This is essentially a plains assemblage, 
though it includes both "steppe" (Damaliscus, springbok, eland) and "savannah" 
(impala, sable, kudu) species. The Vaal River fauna includes "Pelorocerus", Alce­
laphus robustus, a Gazella seemingly different from the Makapan species, and a giant 
buffalo (Homoioceras bainii), which has not been identified at Makapan. Including 
the other mammalian groups, the Vaal River assemblage as a whole resembles that 
of Olduvai in Tanganyika (~Leakey 19511) and is probably of Kanjeran age in the 
main but includes Kamasian elements. 

The Makapan Bovidae give the immediate impression of a more archaic chara~ter 
than those of the Vaal River gravels, and the Makapan deposit may confidently 
be regarded as older than Kanjeran. On the other hand, the large number of species 
not distinguishable from living forms seems to exclude an age greater than Kageran, * 
but it must be remembered ~hat more complete knowledge would probably diminish 
rather than increase this number. As between a Kageran or Kamasian age, the 
bovid fauna offers little helpful guidance. A more exact a:ssessment of 'bhe age of the 
~eposit can therefore be reached only by taking into account the whole of the 
described fauna. A valuable preliminary attempt in this direction has been made by 
Ewer (1956). 'Her arguments in favour of a Kageran age are based on faunal lists 
for southern Africa in which data for ~he Bovidae (and some other important 
elements) were not available. It is not surprising, as she herself anticipated, that her 
conclusions may need some modification in the light of additional evidence. 

Apart from the Bovidae here considered, the fauna of the Makapan Limeworks 
quarry includes a varied assortment of other mammals. Some of these have be~ 
described during the past ten years by different authors but tjhere are still a number of 
elments in the fauna which have not hitherto been record·ed. The following table 
constitutes as complete a list of the whole fauna as it is pos'S'ible to present at thi's 
stage. In the case of species first named from this site, the author's name and date 
serves as a reference; in othe·r cas·es, the reference to ~he aut:hority for recording the 
species is given in square brackets after the name of the specific taxonomist. New 

* typified faunally by the-· deposits at Omo, Kaiso and Kanam. 

49 



records are indicated by an asterisk ( *) following the item and definitely extinct 
species are shown by an obelus ( t) before it. 

INSECTIVORA 

Gen. et sp. indet. * 

PRIMATES 

t Simopithecus darti (Broom and Jensen 1946) [Freedman, 1955] 
( == Dinopithecus sp. (Broom and Hughes 1949); == Brachygnathopithecus peppercorni 
Kitching 1952; == Gorgopithecus wellsi Kitching 1953) 

t Parapapio broomi Jones [Mollet 1946; Freedman 1955] ( == P. makapani Broom and 
Hughes 1949) 

t Parapapio jonesi Broom [Kitching, Wells and Westphal, 1948; Freedman 1955] 
t Cercopithecoides williamsi Mollet 1946 [Freedman 1955] 
t Au.stralopithecus prometheus Dart 1948 

LAGOMORPHA 

Leporidae indet. * 

RoDENTIA 

Muridae indet* 
Hystrix c£ africae-australis Peters [Greenwood 1955] 

t Hystrix major Greenwood 1955 
t Xenohystrix crassidens Greenwood 1955 
t Gypsorhychus makapani Broom 1948 

CARNIVORA 

Canis sp.* 
Vulpes sp. * 
Cynictis sp.* 

t Crocuta c£ bre'Yirostris Aymard [Toerien 1952] 
t Hyaena makapani Toerien 1952 

Felidae indet* 
t Machaerodus darti Toerien 1955t 

PROBOSCIDEA 

Gen. et sp. indet* 

HYRACOIDEA 

t Proca'Yia trans'Yaalensis Shaw [Churcher 1956] 
t Procavia antiqua Broom [ Churcher 1956] 

PERISSODACTYLA 

t S tylohipparion sp. * 
t Metaschiz.otherium transvaalensis George 1950 
? Ceratotherium sp.* 

Diceros cf bicornis* 

+ Equals Therailurus barlowi (Broom) according to Ewer (This Journal below] 

50 



ARTIODACTYLA 

t Potamochoeroides hypsodon Dale 1948 
t ((Pronotochoerus" shawi Dale 1948 
t N otochoerus sp. * 

Hippopotamus cf amphibius Linnaeus [ == H. makapanensis Kitching 1951] 
Giraffa cf camelopardalis Linnaeus [Cooke and Wells 1947) 

t Griquatherium cf cingulatum Haughton [Cooke & Wells 1947) 
Strepsiceros cf strepsiceros (Pallas)* 
Strepsiceros (Tragelaphus) cf angasi (Gray)* 
T aurotragus cf oryx (Pallas)* 
cf Syncerus caffer (Sparrman) * 

t Cephalophus pricei sp.nov.* 
cf Cephalophus (Guevei) caerulus (Hamilton Smith)* 

t Redunca darti sp.nov. * 
Redunca cf fulvorufula (Afzelius) * 
cf Oryx gazella (Linnaeus) * 
Damaliscus sp. ( aff. albifrons) * 
cf Gorgon taurinus (Burchell)* 

t cf Alcelaphus robustus Cooke* 
t cf Pelorocerus helmei (Lyle)* 
t Makapania broomi gen. et sp.nov.* 
t Oreotragus major Wells* 

Aepyceros- cf melampus (Lichtenstein)* 
t Gazella gracilior sp.nov. * 
t Phenacotragus vanhoepeni sp.nov. * 

In some of the groups listed above (e.g. L4gomorpha) ~he species have not yet 

been worked out and it is impossible to indicate whether they are extinct or not. 

Allowing a minimum of one species in each of these cases, the total faunal list 

compr:ses not fewer than forty-eight species. Apart from the Bovidae there are 

30 species, of which 19 are certainly extinct and 5 cannot be separated specifically 

from living forms; the position of the remaining 6 is "doubtful". In rhe case of 

the 18 1Bovidae, eight are extinct and 9 cannot be distinguished specifically from 

living species, although in several cases this statement needs to be qualified owing 

to the fragmentary character of the remains; 1 species (the Damaliscus) is probably 

extinct and may be placed in the "doubtful" category. 

Of the 41 adequately determined species 27 are extinct and 14 cannot be separated 

from living forms. Ewer (1956) gives for the Australopithecine deposits as a whole 

a figure of 12% for the species "not more ~han sub-specifically distinct" from the 

living forms. For the Ma!kapan Limeworks the indicated percentage is now 34%, 

which affects the conclusions materially. For the percentage of genera which are 

extinct, Ewer gives 53% for the Australopithecine deposits as a whole; the present 

Makapan figure is 55% without the iBovidae but only 41 % when they are included. 

T 'his latter figure is comparable witth 1Ewees 40% each for Omo and Olduvai. 

51 



As Ewer's figures for the Transvaal limestone caves did not include any records 
of Bovidae, h~r conclusions regarding the probable relative ages of the sites are not 
affected at present. The view that Makapan Limeworks and the Sterkfontein type 
site are the oldest and are very close to each other in time may later be tested 
by a study of the Bovidae; ~he Kromdraai and Swartkrans sites are clearly somewhat 
younger. One cannot avoid concluding that when the 1Bovidae at Sterkfontein, Krom­
draai and Swartkrans are studied, the apparent percentage of extinct genera will 
decrease as it has at Makapan and will come more into line with those for Omo and 
Olduvai, where the Bovidae are already included. 

According to Hopwood and to Macinnes (in Leakey 1951), !!he faunas of Omo, 
Kanam and Kaiso typify the Kageran stage; those of Olduvai Beds I and II, 
Laetolil and Rawi characterise the 1Kamasian stage; and Olduvai Bed N, Olorgesailie 
and Kanjera represent the Kanjeran stage. The table below shows the extinct genera 
represented at Makapan and in one or other of these stages. 

MAKAPAN 

Simopithecus 

"Machaerodus" 

S tylohipparion 

M etaschizotherium 

Pronotochoerus 

Notochoerus 

Phenacotragus 

KAGERAN 

? 

X 

X 

X 

X 

KAMAS IAN 

X X 

X X 

X 

X X 

While the statistical value of such a tabulation is slender, it suggests that the 
Makapan Limeworks deposit may be partly of Kageran age. The best fit would be if 
it belongs to the latest Kageran or earliest Kamasian. In terms of the ~European time 
scale this is most probably Upper Villafranchian. It has been mentioned above rt:hat 
at Makapan a relatively dry climate during at least part of the period represented 
by this deposit can be inferred. It would probably be rash, however, to equate this 
too narrowly with the late Kageran dry phase or "Kageran-Kamasian interpluvial" 
of East Africa (vide du T oit, 194 7) . 

If, as appears to be the case, the Makapan Limeworks fauna is essentially 
Villafranchian, it is of interest that the Bovidae are mostly referable to genera still 
living. This feature is also apparent in ~he assemblage from Omo (which may be a 
little older) . The inference is that the African 1Bovidae were complete! y differentiated 
to the generic level at the beginning of the Pleistocene. Unfortunately the:r late 
Tertiary history in this continent is still unknown, but the Villafranchian ~Bovidae 
of India are so different from the African forms that a considerable period of :..eparate 
d:versification in Africa must be regarded as highly probable. 

52 



AcKNOWLEDGEMENTs 

The authors wish to express ~heir indebtedness to Professor R. A. Dart for placing 
the material at their disposal for description; to Mr A. R. Hughes for his energy in 
collecting most of the ·ma:terial and for taking the excellent photographs here 
reproduced; to Mr. James Kitching for his field work and for his skill in cleaning 
the specimens from their hard matrix and preparing them for study; to Dr A. S. 
Brink for providing laboratory facilities and otJher help; and to ·Dr S. 'H. Haughton 
for scientific advice and ed:torial assistance. Finally, the second author wishes to 
record his debt to the Royal Society, the Nuffield Foundation and the Wenner­
Gren Foundation for financial aid which made possible a study visit to England, 
Europe and East Africa during which comparat:ve data were collected that have 
been of considerable aid in assessing the Makapan material. 

REFERENCES 

ALLEN, G. M., 1939. A checklist of African mammals. Bull. Mus. Comp. Zoo!. Harvard, 
83: 1-763. 

ARAMBOURG, C., 1941. An til opes nouvelle du Pleistocene ancien de l'Omo ( Abyssine) . 
Bull. Mus. Hist. Nat. Paris (2) 13: 343. 

ARAMBOURG, C., 1947. Contribution a !'etude geologique et paleontologique du bassin 
du lac Rodolphe et de la basse vallee de l'Omo. Deuxieme partie, Paleontologie. 
Mission Scient. Omo 1932-1933 I Geol.-Anthrop.: 418-432. 

BARBOUR, G. B., 1949a Ape or man? An incomplete chapter of human ancestry from 
South Africa. 0 hio J. Sci. 49: 129-14 5. 

BARBOUR, G. B., 1949b Makapansgat. Scient. Monthly, 69: 141-147. 
BONE, E., 1955. Une clavicule et un nouveau fragment mandibulaire d'Australopithecus 

prometheus. P alaeont. A fr. 3 : 87-101. 
BOSAZZA, V. L., ADIE, -R. J. and BRENNER, S., 1946. Man and the great Kalahari 

desert. J. Natal Univ. Coli. Scientific Soc. 5: 1-9. 
BRAIN, C. K., VAN RIET LOWE, C. and DART R. A., 1955. Kafuan stone artefacts 

in the post-Australopithecine breccia at Makapansgat. Nature 175: 16. 
BROOM, R., 1937. Notices of a few more new fossil mammals from the caves of the 

Transvaal. Ann. Mag. Nat. Hist. (10) 20: 510-511. 
BROOM, R., 1948. The giant rodent mole, Gypsorhychus. Ann. Transv. Mus., 21: 47-49. 
BROOM, R. and HUGHES, A. R., 1949. Notes on the fossil baboons of the Makapan 

Caves. S. Afr. Sci. 2: 194-196. 
BROOM, R. and JENSEN, J.S., 1946. A new fossil baboon from the caves at Potgieters­

rust. Ann. Transv. Mus. 20: 337-340. 
CHURCHER, C. S., 1956. The fossil hyracoidea of the Transvaal and Taungs deposits. 

Ann. Transv. Mus. 22: 477-501. 
COOKE, H. B. S., 1949. Fossil mammals of the Vaal River deposits. Mem. Geol. Sur'Y. 

S.Afr., 35 (3): 117 pp. 
COOKE, H. B. S., 1952. Quaternary events in South Africa. Proc. First Pan-Afr. Congress 

on Prehistory, 1947. Oxford: Blackwell: 26-36. 

53 



COOKE, H. B. S., MALAN, B. D. and WELLS, L. H., 1945. Fossil Man in the 
Lebombo Mountains, South Africa: the 'Border Cave', Ingwavuma District, Zululand. 
Man 1945 (3): 6-13. 

COOKE, H. B. S. and WELLS, L. H., 1947. Fossil mammals from the Makapan valley, 
Potgietersrust III Giraffidae. S. A fr.J. Sci. 43: 232-23 5. 

DALE, M. M., 1948. New fossil Suidae from the Limeworks Quarry, Makapansgat, 
Potgietersrust. S. A fr. Sci., 2: 114-116. 

DALE, M. M. and TOBIANSKY, D., 1947. Fossil mammals from the Makapan valley, 
Potgietersrust II Suidae. S. Afr.J. Sci. 43: 304. 

DART, R. A., 1925. A note on Makapansgat: a site of early human occupation. 
S. Afr.J. Sci., 22: 454. 

DART, R. A., 1948a. An Australopithecus from the Central Transvaal. S. Afr. Sci., 
1: 200-201. 

DART, R. A. 1948b. An adolescent promethean australopithecine mandible from 
Makapansgat. S. Afr. Sci., 2: 73-75. 

DART, R. A. 1948c. The Makapansgat proto-human Australopithecus prometheus. Amer. 
J. phys. Anthrop. (N.S.) 6: 259-283. 

DART, R. A., 1948d. The adolescent mandible of Australopithecus prometheus. Amer.J. 
phys. Anthrop. (N.S.) 6: 391-411. 

DART, A. R. 1948e. The first human mandible from the Cave of Hearths, Makapansgat. 
S. Afr. archaeol. Bull, 3: 96-98. 

DART, R. A., 1949a. The cranio-facial fragment of Australopithecus prometheus. Amer. J. 
phys. Anthrop. (N.S.) 7: 187-213. 

DART, R. A., 1946b. Innominate fragments of Australopithecus promethus. Amer.J. phys. 
Anthrop. (N.S.) 7: 301-333. 

DART, R. A., 1949c. A second adult palate of Australopithecus prometheus. Amer.J. phys. 
Anthrop. (N.S.) 7: 335-338. 

DART, R. A., 1952. Faunal and climatic fluctuations in the Makapansgat valley: their 
relation to the geological age and promethean status of Australopithecus. Proc. First 
Pan-Afr. Congress on Prehistory, 1947. Oxford: Blackwell, 96-106. 

DART, R. A., 1954. The adult female lower jaw from Makapansgat. Nature, 173: 286. 
DIETRICH, W. 0., 1950. Fossile Antilopen und Rinder Aquatorialafrikas. (Material der 

Kohl-Larsen'schen Expeditionen). Palaeontographica 99, A: 30-32. 
DU TOIT, A. L., 1947. Palaeolithic environments in Kenya and the Union - a contrast. 

S. A fr. archaeol. Bull. 2: 28-40. 
EWER, R. F., 1956. The dating of the Australopithecinae: faunal evidence. S. Afr. 

archaeol. Bull, 11: 41-45. 
FREEDMAN, L., 1955. The fossil Cercopithecoidea of South Africa. Thesis for the 

degree of Ph.D., University of Cape Town (awaiting publication). 
GEORGE, M., 1950. A chalicothere from the Limeworks Quarry of the Makapan Valley, 

Potgietersrust District. S. Afr. J. Sci., 46: 241-242. 
GREENWOOD, M., 1955. Fossil hystricoidea from the Makapan Valley, Transvaal. 

Palaeont. Afr. 3: 77-85. 
HAUGHTON, S. H., 1947. Notes on the Australopithecine-bearing rocks of the Union 

of South Africa. Trans. geol. Soc. S. Afr. 50: 55-59. 
HOFFMAN, A. C., 1953. The fossil alcelaphines of South Africa - genera Peloroceras, 

Lunatoceras and Alcelaphus. Nav. Nas. Mus. Bloemfontein 1 (3): 41-56. 
HOPWOOD, A. T., 1934. New fossil mammals from Olduvai, Tanganyika Territory. 

Ann. Mag. nat. Hist. (10) 14: 546-550. 

5 



_HOPWOOD, A. T. and HOLLYFIELD, J. P., 1954. An annotated bibliography of the 
fossil mammals of Africa (1742-1950). Fossil Mammals of Africa, 8: 194 pp. 

KITCHING, J. W., 1951. A new species of Hippopotamus from Potgietersrust. S. Afr. 
]. Sci., 47: 209. 

KITCHING, J. W., 1952. A new type of fossil baboon. Brachygnathopithecus peppercorni. 
gen. et sp.nov. S. Afr. J. Sci. 49: 15-17. 

KITCHING, J. W., 1953. A new species of fossil baboon from Potgietersrust. S. Afr. ].Sci. 
50: 66-69. 

KITCHING, J. W., WELLS, L. H. and WESTPHAL, E., 1948. Fossil cercopithecid 
primates from the Limeworks Quarry, Makapansgat, Potgietersrust. S.Afr.Sci., 1: 171-172. 

KING, L. C., 1951. The geology of the Makapan and other caves. Trans. Roy. Soc. S. Afr., 
33: 121-151. 

LEAKEY, L. S. B., 1951. Olduvai Gorge. Cambridge: Cambridge University Press. "The 
Olduvai fauna" by A. T. Hopwood, 20-30 and D. G. Macinnes 31-33. 

MALAN, B. D. and COOKE, H. B. S., 1941. A preliminary account of the Wonderwerk 
Cave, Kuruman District. S. Afr. J. Sci. 37: 300-312. 

MOLLET, 0. D. v. d. S., 1947. Fossil mammals from the Makapan valley, Potgietersrust I. 
Primates. S. Afr. J. Sci., 43: 295-303. 

OAKLEY, K. P., 1954a. The dating of the Australopithecinae of Africa. Amer. ]. phys. 
Anthrop. (N.S.) 12: 9-28. 

OAKLEY, K. P., 1954b. Study tour of early hominid sites in southern Africa, 1953. 
S. Afr. archaeol. Bull. 9: 75-87. 

PILGRIM, G. E., 1939. The fossil Bovidae of India. Palaeontologia Indica (N.S.) 26 
(1): 356 pp. 

ROBINSON, J. T., 1952. The Australopithecine-bearing deposits of the Sterkfontein area. 
Ann. Transv. Mus. 22: 1-19. 

SCHLOSSER, M. 1925. in ZITTEL, K. A. von. Textbook of Palaeontology London: 
Macmillan and Co.: 14-20. 

SCHWARZ, E. 1937. Die fossilen Antilopen von Oldoway. Wiss. Ergebn. Oldoway­
Exped. 1913. (N.F.) 4: 8-104. 

SIMPSON, G. G., 1945. The principles of classification and a classification of mammals. 
Bull. A mer. Mus. nat. Hist., 85: 350 pp. 

TOERIEN, M. J., 1952. The fossil hyenas of the Makapansgat Valley. S. Afr. ].Sci., 48: 
293-300. 

TOERIEN, M. J., 1955. A sabre-tooth cat from the Makapansgat Valley. Palaeont. Afr. 
3: 43-46. 

VAN HOEPEN, E. C. N., 1932. Voorlopige beskrywing van Vrystaatse soogdiere. Palaeont. 
Navors. nas. Mus. Bloemfontein, 2: 63-65. 

VAN RIET LOWE, C., 1938. The Makapan caves: an archaeological note. S. Afr. ]. Sci., 
35: 371-381. 

VAN RIET LOWE, C., 1943. Further notes on the Makapan caves. S. Afr. ]. Sci., 40: 
289-295. 

VAN RIET LOWE, C., 1948. Cave breccias in the Makapan Valley. Roy.Soc.S.Afr.Spec. 
Pub[. Robert Broom Comm. Vol.: 127-131. 

WELLS, L. H., 1951. A large fossil ldipspringer from Potgietersrust. S. Afr. J. Sci. 47: 
167-168. 

WELLS, L. H. and MALAN, B. D., 1943. A further report on the Wonderwerk Cave, 
Kuruman District. S. Afr. ]. Sci. 40: 258-270. 

55