HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE AND A

REVISION OF THE GENUS IN AFRICA!

Epouarp L. Bong

Department of Vertebrate Palaeontology, University of Louvain*

and

RONALD SINGER?

Department of Anatomy, University of Chicago, U.S.A.

(With 24 tables, 16 text figures and 13 plates)

CONTENTS

PAGE

INTRODUCTION : : : : : : ; é ; b : : 3-270

General : : 2 : : ; ‘ : : : : 270

The Larmsrannnes deposits : 5 ; j : : i : : a GIy)

Localization of the sites . : : ; : . : 0 : : of) Ga

History of the discoveries : f : E é : A : : ai) Aaa)

Geology : : : ; : : : : : 279

Faunal assemblage com Langebaanweg : : : : : : : : . 280

Methods and nomenclature . : 3 280

GEOGRAPHICAL DISTRIBUTION OF AFRICAN SITES WHERE HIPPARIONIDS HAVE BEEN RECORDED 281

Morocco. ‘ : : : s 4 ‘ j , : 0 és - 281

Algeria : : : : é : : : : é : . é . 281

Tunisia : : a: : : j : : ; : ; : : 71/282

Fayum : 5 : ; : ; f : 4 : é : . Tn23o

Ethiopia. ‘ 4 i ‘ ! : Y ‘ : : : : 22)

Central Africa. : : : : : : , : . : : . 282

East Africa j y ‘ ‘ : 3 : : 4 : : ; 23

Uganda : : 5: : : : : : : é : : ee 2O3)

Kenya 3 : : : : ; : ‘ : : 6 : P2o3)

Tanganyika . . ; : ; : : d : : : ; . 283

South Africa : 2 F ‘ . : ; : 4 ; : : M203

Cape Province : : : : : ; ; 2 : : 6 ay Chae

Orange Free State : ; : : : : 4 : : : . 283

Transvaal. : : : : We 233

SUMMARY OF PUBLISHED GEOLOGICAL EVIDENCE FROM AFRICAN HIPPARIONID SITES . 16 283

Miocene . : ; : : 3 ‘ : : : : , : W283

Pliocene : : : : : : s : F ; : p i . 286

Pleistocene . : : , . : : : : ‘ : : 21) 286)

North ie : : : : : : : : ; : : . 286

1 The cost of publication of this paper has been met, in part, by grants from the ‘Fondation

Universitaire’ of Belgium, and the Wenner-Gren Foundation for Anthropological Research,

Inc., New York.

’ Present address, Facultés Universitaires, Namur, Belgium.

’ Honorary Curator of Physical Anthropology, South African Museum, Cape Town.

273

Ann. S. Afr. Mus. 48 (16), 1965: 273-397, 24 tab., 16 figs., 13 pls.

274 ANNALS OF THE SOUTH AFRICAN MUSEUM

East Africa

South Africa . 2 : : 5 A : :

LISTS OF FAUNAL ASSEMBLAGES AT THE AFRICAN HIPPARIONID SITES

Individual hipparionid sites

Oued el Akrech

Oued el Hammam and Mercena

Ain Hanech

Garet Ichkeul

Tozeur :

Wadi Natrun

Omo

Koro Toro

Kaiso

Kanam.

Eyasi_ .

Laetolil (South Serengeti

Olduvai

Vaal River women apenrell

Cornelia (Uitzoek) :

Kromdraai :

Bolt’s Farm

Makapansgat :

Elandsfontein (Hopefield)

Synopsis (Table 1) F 4 : : : : : :

COMMENTARY ON THE PUBLISHED FAUNAL ASSOCIATIONS AT HIPPARIONID SITES IN AFRICA

Miocene

Pliocene

Pleistocene .

North INP :

East and Central Africa

South Africa .

Conclusions based upon the Eanall ond peological evidencen in the iRicnataine

CHRONOLOGICAL RANGE OF Hipparion

Upper limit

Lower limit :

GENERAL DESCRIPTION OF EUElinearion TEETH

Upper dentition

Lower dentition

Summary of the diamoristic of Euerarin teeth

Upper dentition

Lower dentition ; : :

REVIEW OF ADDITIONAL ENAMEL ELEMENTS (cies rps) OF THE LOWER TEETH

Description

Ectostylid .

Ptychostylid .

Protostylid

Hypostylid

Entostylid

Frequency

Ectostylid

Ptychostylid .

Protostylid

Hypostylid

Evolutionary history of the sty ‘lids

Evolutionary significance of the cones and stylids

Ecological considerations

‘Taxonomy based on stylids

REVIEW OF PUBLISHED AFRICAN Heparin CRANIAL MATERIAL

287

290

292

293

294

295

296

297

2098

298

299

300

301

303

302

314

316

317

320

321

322

323

325

326

327

328

329

330

331

332

333

335

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

Miocene

Oued el Harner end Niateeat

Camp Berteaux

Other sites

Discussion on the fica chamcten ios Ennparion afftercorsp Avengers, Toson

Enamel plications

Protocone :

The double knot

Stylids

Hypsodonty

Pliocene :

Mascara :

St. Arnaud Cumeicsy

Ain el Hadj Baba .

Utique (northern Tunisia)

St. Donat ela

Pleistocene :

North een

@Oranv pe

Ain Jourdel

Beni Foudda

Wadi Natrun

Central Africa

Koro Toro

East Africa

Omo: i:

Olduvai Gueluciae unpublished data)

South Serengeti ;

Lake Eyasi

Lake Albert

South Africa .

Namaqualand

Christiana

Transvaal sanirailopinediae cave 5 janesaction

Cornelia (Uitzoek)

Note on Hipparion incisors

A SUMMARY OF PUBLISHED POSTCRANIAL REMAINS OF AFRICAN Hibparion

Unpublished data.

THE ECOLOGICAL ADAPTATIONS OF Ethane

Ecological and functional significance of aoe featuresh

Overall dimensions of the skeleton :

Relative dimensions of cheek teeth and incisors .

The development of the preorbital fossa

The isolated protocone

Enamel plications .

Hypsodonty .

The double knot

Total tooth structure

The extremities

The Upper Miocene and Blemrocene biotones nonineiedl by Mriean fipoonionids

Effects of bionomical conditions on the evolution and the nea of Agiean

hipparionids ; : é

UNPUBLISHED MATERIAL FROM THE VAAL RIVER DEPOSITS, SOUTH AFRICA

Sydney-on-Vaal and Pniel :

MATERIAL FROM LANGEBAANWEG, CAPE ER OVINGE! SOUTH AFRICA.

Description of Hipparion teeth

Milk dentition

Lower permanent dentinon

276 ANNALS OF THE SOUTH AFRICAN MUSEUM

General description : : 3 ; . : 4 : 2 6 370

Upper permanent dentition . : : . c : : < 5 BP

General dimensions (mm) of P?— Me ; ; : : : : : 5 RIF

Description of selected individual teeth : . c : 6 BIG

Appearance of the tooth sectioned just above the aittdldlis : : ; - 375

Typical features of the upper teeth . : ; . : : : 5 Sl

Postcranial equid remains. . ; . : : 370

ORIGIN AND DIFFERENTIATION OF AFRICAN HIPPARIONIDS . : : : : gE

TAXONOMIC STATUS OF THE AFRICAN HIPPARIONIDS . : : : : : 5 S387)

Discussion . ; . : . : 5 5 c : : : : - 390

ACKNOWLEDGEMENTS. : 5 : : : . : : : 5 7392

SUMMARY. : : ; 4 : : : : : : : . 5 308

REFERENCES . : 3 : ; : ; : : : 5 : : - 303

INTRODUCTION

GENERAL

The discovery of the genus Hipparion de Christol 1832 in Africa was first

mentioned as late as 1878 (Pomel) and then again in 1884 (Thomas), but it

was not properly described before 1897 (Pomel). These early papers dealt with

specimens found in the Maghreb, North Africa. Additional finds were then

made in the Oran and Constantine areas, and, somewhat later, specimens

were discovered in Tunisia, in the Nile Valley at Wadi Natrun, in Ethiopia,

in the Kaiso Beds of Uganda, and at Cornelia (Uitzoek) in South Africa. The

scanty stratigraphical data and faunal associations of these discoveries led to

an overmultiplication of genera and species, based mostly on isolated single

teeth.

Subsequently more complete discoveries and more detailed descriptions

were published. In 1937 Hopwood presented some conclusions of the Oldoway

Expedition in a monograph on the fossil Equidae. Dietrich (1942) described

an important hipparionid! series from South Serengeti, to which Arambourg

(1947) added important collections from Omo (Ethiopia). The more extensive

material gave rise to a more unified knowledge of the African Hipparion. It

enabled Gromova (1952), in her revision of the genus Hipparion, to point out

the characteristic features of the African forms, as well as to sketch their

general evolutionary and migratory patterns within Africa.

The lack of stratigraphical correlations and insufficient knowledge of

conditions of exposure and of faunal associations at the various sites where

hipparionids have been discovered, only permitted attempts at relative dating

of the various sites and their fossil horizons. Furthermore, these sites belong

exclusively to Upper Pliocene and Pleistocene epochs. Therefore, it was

generally accepted that Hipparion had appeared in America in Pliocene times

1 The term ‘hipparionid’ refers in a general sense to all forms of African tridactyl equidae.

Previously Dietrich (1942, p. 94) erected the sub-family ‘Hippariinae’ for the same purpose

but without taxonomic definition. Furthermore he also used (1942, p. 94) ‘Hipparion de Christol,

7.w.S.’ in a non-generic sense.

As the present authors have come to the conclusion that all African forms belong to a

single genus Hipparion, the general term ‘hipparionid’ corresponds to the taxonomic one.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE P2717)

and had migrated to Eurasia where the differentiation of the ‘Pontian fauna’

had taken place, and that only subsequently, as late as Pliocene times, Hipparion

had reached Africa as a kind of refuge area. This view was still supported

less than ten years ago when material was discovered in lacustrine limestone

at Mascara (Arambourg, 1956).

Recently Hipparion material was discovered in Algeria in undoubted

Upper Miocene beds. This led Arambourg (1959) to propose the relationships

between the several African hipparionids, and to reduce drastically the former

several genera and species to a more sober and unified taxonomy of three

different groups, namely H. africanum, H. sitifense and Stylohipparion sp. These

formed a chronological succession from the Upper Miocene through the

Upper Pliocene to the ‘Villafranchian’ or early Pleistocene times. This classifi-

cation was more satisfactory and had definite merits and advantages which will

be commented on later in this paper.

THE LANGEBAANWEG DEPOSITS

Recent discoveries (Singer, 1961) of a fairly extensive series of Hipparion

teeth at the Langebaanweg fossil sites, from which Stegolophodon sp. had pre-

viously been reported (Singer & Hooijer, 1958), necessitated a thorough study

and comparative review of the African hipparionids. The unique and uniform

nature of the Langebaanweg material adds materially to our knowledge of

the African hipparionids and also provides new evidence for revising the

relationships between the several forms already recognized in the literature.

Thus the purpose of this paper is twofold:

(a) to record a description and diagnosis of the Hipparion remains from

Langebaanweg, and

(6) to review and revise the taxonomy of the African hipparionids.

Localization of the sites

The Langebaanweg fossil sites are situated, like the ‘Elandsfontein’ (Hope-

field) site, in the vast Sandveld of the Western Cape, and are located about

75 miles (120 km.) northwest of Cape Town (fig. 1) and 12 miles (19 km.)

from the late Middle Pleistocene—early Upper Pleistocene site at ‘Elands-

fontein’. The former are 150-200 feet (46-62 metres) above sea-level, while

the latter are situated at 300 feet (Singer, 1961).

The fossils are mainly located in and around two quarries being commer-

cially exploited for phosphates by the African Metals Corporation (AMCOR).

The “Varswater’ site is situated on the farm ‘Langberg Suid’, and the ‘Baard’s’

site is on the farm ‘Remainder of Langberg’, the two being subdivisions of

‘Langberg’ (fig. 2).

EN story of the discoveries

The first hipparionid teeth were recovered from Baard’s Quarry in 1958.

In 1961, Mr. Robin Warren, an employee of AMCOR, recovered two excel-

278 ANNALS OF THE SOUTH AFRICAN MUSEUM

LA NGEBAANWEG

KEN ‘

SALD ANHAg SOR ERE ED)

BAY

Elandsfontein

Sout ay

° CAPE TOWN

English miles

Fic. 1. Map indicating the position of Langebaanweg relative to Cape Town.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 279

lently preserved milk molars at Varswater. Thereafter the remainder of the

specimens were found at various sites in the two quarries.

Geology

The geology of the more than 4 square miles of fossiliferous deposits is

not easy to interpret. There is little with which to compare it, as virtually

nothing is known of the Quaternary geology of the Cape Province.

The surface siliceous sands, possibly aeolian, forms a discontinuous cover

of 4 to 5 feet, which may extend to more than 15 feet in a few places. Below

this layer, at both quarries, there occurs a Dorcasia-type caliche, discontinuous

and stratigraphically above the phosphates. There is also ferricrete, younger

than the phosphates and intimately associated with them.

The Dorcasia-type surface limestone also occurs on the solid calcrete

dunes at ‘Elandsfontein’, at Saldanha Bay, at False Bay (Singer and Fuller,

1962) and elsewhere. The distribution of the caliche is patchy, and is

considerably stripped by erosion.

At Varswater the phosphate, a variety of collophane, occurs as the cement-

ing material of the consolidated sand, probably as the result of diagenetic

replacement of interstitial limes. There are three main layers which vary

laterally in thickness within a range of 3 feet to 6 feet.

At Baards, the phosphate occurs as nodular to bouldery phoscrete (up to

5 feet thick). The phoscrete is a hard, compact mass of phosphatized sand

consisting of sand grains which have been cemented and partially corroded

by an amorphous calcium phosphate cement. It overlies the phosphatized

nodular sands, sandy clays, and clayey sand. The phoscrete and nodular

phosphate sands, represent replacement of older shelly sands by phosphate

solutions, the phosphate of which is probably derived from guano. A significant

feature is that at Varswater the phosphate has remained unaltered as lime

phosphate (Singer & Warren, in preparation).

The discovery of shark teeth and other marine forms (vide infra) at the

same level as and with the land mammals indicates earlier invasion of the

area by the sea. Studies on borehole cores and the presence of shelly sands and

gravels suggest the formation of sand bars, especially at Varswater, while the

clay and clayey sands at Baard’s indicates a possible lagoonal stasis. Further

studies are being carried out in an attempt to clarify the geological genesis and

morphology of the region.

The major portion of the phosphatic deposits is located at both sites at

up to 100 feet above sea-level. The archaic fossils are related to or found just

above these deposits. The ‘rolled’ nature of much of the fossil material suggests

some transport prior to final deposition. The upper phoscrete and calcrete

layers, as well as the clayey deposits, contain the faunal forms represented

also at ‘Elandsfontein’ (Ceratotherium, Equus, Sivatherium, Homoioceras and various

antelopes), typical of the late Middle Pleistocene—early Upper Pleistocene

(‘Hopefield Period’).

280 ANNALS OF THE SOUTH AFRICAN MUSEUM

On a comparative basis with other sites, it seems likely that the stegolo-

phodont—stegodont—archidiskodont complex and the hipparionids derive

from a period extending from the Lower Pleistocene to the early Middle

Pleistocene. The stegolophodont and stegodont specimens together indicate a

very early phase of the Lower Pleistocene. Elsewhere (e.g. at Olduvai)

archidiskodont material overlaps this phase and extends up to the Middle

Pleistocene. It will be demonstrated below that the hipparionids from Lange-

baanweg belong to the same species as that from the lower layers of South

Serengeti, so that their presen e is suggestive of the earliest phases of the

Pleistocene.

FAUNAL ASSEMBLAGE FROM LANGEBAANWEG

The following identifications have been made thus far:

MOLLUSCA ARTIODACTYLA

Trigonephrus sp. Hippopotamus sp.

Stvathertum(*) olduvaiense

CARNIVORA cf. Oreotragus sp.

cf. Arctocephalus cf. Damaliscus sp.

Canis sp. cf. Redunca sp.

Hyaena brunnea Homoioceras sp.

REPTILIA

PROBOSCIDEA ae Teesineh

Stegolophodon sp.

AVES

Stegodon sp. 4

Archidiskodon sp. cf. Struthio

Palaeoloxodon sp. SELACHII(?)

Isurus cf. glaucus

PERISSODACTYLA Lamna nasus

cf. Ceratotherium Carcharias sp.

Hipparion (Hipparion) albertense baardi C. ferox

subsp. nov. Rhinoptera cf. dubia

Equus helmet Glopias vulpes

(1) According to Arambourg’s recent re-description (1960) of Pomel’s type specimen from

St. Charles (Oran), Libytherium maurusium has precedence (see also Singer and Boné, 1960).

(?) Identified by Dr. B. Schaeffer, American Museum of Natural History, New York.

METHODS AND NOMENCLATURE

For the description of the occlusal surfaces of the teeth, the nomenclatures

used by Arambourg (1947, 1959), Cooke (1950), and Sondaar (1961) have

been modified (see pp. 325-8) and synthesized (figs. 10, 11).

The measurements of the teeth have been taken according to the method

described by Sondaar (1961).

e s

{Lance BAAN

i} OA

;

i>)

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 281

The indices devised by the authors are described at appropriate places

in the text.

The following abbreviations refer to the collections and sites of the source

material:

Archaeol. Surv. = Archaeological Survey of South Africa, now in the Univer-

sity of the Witwatersrand, Johannesburg, South Africa.

B.M.N.H. = British Museum (Natural History)

C = Cornelia

L = Langebaanweg

M.M.K. = McGregor Memorial Museum, Kimberley

Nas. Mus. = Nasionale Museum, Bloemfontein

Old. = Olduvai

S.A.M. = South African Museum, Cape Town

Uc. = University of California, Berkeley.

GEOGRAPHICAL DISTRIBUTION OF AFRICAN SITES WHERE HIPPARIONIDS

HAVE BEEN RECORDED (figs. 3-7)

Site Specimen Reference

Morocco

Camp Berteaux (Gara Ziad) (a) Molar teeth

Femur fragment with proximal

epiphyses Bourcart, 1937

(6) Left M? Ennouchi & Jeannette,

1954

Oued el Akrech (near Argoub

el Hafid) Ennouchi, 1951

Fouarat (south of Kenitra Choubert, Ennouchi &

(Rharb)) Marcais, 1948

ALGERIA

Tafna (Guiard, Prov. Oran) Dalloni, 1915

Oued el Hammam Adult skull no. 141 Arambourg, 1959

Adult skull no. 125

Adult skull with dentition no. 116

Juvenile palate with dentition

no. 122

Adult mandible no. 143

Adult half mandible no. 8g

2 juvenile mandibles with milk

teeth nos. 95-96

Several teeth series, juvenile and

adult

Isolated teeth

Humerus, distal fragment no. 159

Radio-ulna no. 123, fragments

no. 13, no. 22 and no. 27

Femur: fragment without epiphyses

and distal fragment

Tibia: distal fragment

Pelvic fragment

282 ANNALS OF THE SOUTH AFRICAN MUSEUM

Site

Oued el Hammam (cont.)

Marceau

Mascara

St. Donat

St. Arnaud (Cemetery)

Ain el Hadj Baba

Ain el Bey

Mansourah

Setif

Beni Foudda (Ain Boucherit)

Ain Hanech

Oran (Puits Kharouby)

Ain Jourdel

TUNISIA

Garet Ichkeul

Tozeur Beds

Utique

Djebel M’dilla

FAYUM

Wadi Natrun, Gart el Moluk

Hill

ETHIOPIA

Omo Valley

CENTRAL AFRICA

Tchad, Koro Toro region

Goz-Kerki

Koula

Bochianga

Quadi Derdeney

Koula Ri Katir

(It is not clear at which of

these five sites Stylo-

hipparion has been found)

Specimen

Astralagus, 2

Calcaneum, I

Numerous complete metapodials

(5 Mc III, 5 Mt III), many (6)

of them with lateral digits

P2 and Ma, P,

Skull fragments of one individual:

palate with nearly complete left

dental series, and half right den-

tal series; symphysis with incisors

“Teeth of two jaws’

(a) upper molars

(6) lower molars

Metapodials with developed lateral

digits

Upper teeth: right P*-M?

left M?

Lower teeth

M2, left P; or P,, Ma

M;, P

Left upper P (P* ?)

(a) 1 incisor, 5 M,, 1 M@ (brought

back by Brumpt from the du

Bourg de Bozas Mission, 1903)

(6) right P? no. 396

left M3 no. 395

left Ma fragment no. 370

right M, or M, no. 3 & no. 4

right humerus fragment

Reference

Arambourg, 1959

Arambourg, 1956

Joleaud, 1936

Arambourg, 1956

Pomel, 1897

Arambourg, 1956

Thomas, 1884

Pomel, 1897

Arambourg, 1949

Pomel, 1897

Thomas, 1884

Arambourg, 1949

Roman & _ Solignac,

1931

Solignac, 1927

Andrews, 1902

Haug, 1911; Joleaud,

1933

Arambourg, 1947

Coppens, 1960

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

Site

EAST AFRICA

Uganda: East shore of Lake

Albert in Kaiso layers

Kenya: Kanam lower beds.

Kanjera

Tanganyika: North-east shore

of Lake Eyasi (W. and N.

of Mumba Hill)

South Serengeti (various loca-

lities)

Olduvai

SOUTH AFRICA

Cape Province:

(i) Namaqualand (40 mi.

east of Springbok)

a) Sydney-on-Vaal

(6) Pniel, near Barkly

West

Vaal River at Chris-

tiana

Orange Free State:

Cornelia (Uitzoek)

Transvaal:

Kromdraai Cave

Bolt’s Farm

? Makapansgat

Specimen

I incomplete upper molar

BM Mri2615

Teeth

Loose teeth, fragmented mandibles

Left P,-M, and M,

Right P,, M,, M, and part of

(above specimens are S.A.M.

9982)

Left M3 (MMK 431)

Left P+ or M4 (MMK 5225)

Left M? or M$ (Arch. Surv. 113)

(a) right M? (or M?) (Nas. Mus.

C558)

left M? (Nas. Mus. C555)

left M, + M, (Nas. Mus.

C556)

(6) symphysial fragment and in-

(¢)

cisors

right M, (Nas. Mus. C797)

left P, (Nas. Mus. C796)

(d) left P;-M, (Nas. Mus. C795)

ower teeth (discovered by Broom)

NOTE: M2? indicates upper molar; M, lower molar, etc.

283

Reference

Hopwood, 1926

Reck & Kohl-Larsen,

1936

Dietrich, 1942

Hopwood, 1929, 1937

Haughton, 1932

Cooke, 1950

Van Hoepen, 1930

Van Hoepen, 1932

Cooke, 1950

Cooke, 1963

SUMMARY OF PUBLISHED GEOLOGICAL EVIDENCE FROM

AFRICAN HIPPARIONID SITES

The present state of our knowledge of the geological evidence from sites

in Africa where hipparionids have been discovered indicates that they first

appeared in Upper Miocene deposits. A summary of the data is considered

useful for the assessment of the status of the Langebaanweg material.

MIOCENE

The continental layers of vertebrate fossils at OUED EL HAMMAM

(between Mascara and Mercier Lacombe, upstream from Bou Hanifia,

284 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fic. 3. Map of Africa indicating major areas of discovery of hipparionids. Details of areas

A-C are shown in figs. 4-6, while area D is enlarged in fig. 1.

Algeria—figs. 3, 4) occur between two marine horizons, namely, (1) a Burdi-

galian one, characterized by its mollusc and echinid fauna and (2) an Upper

Miocene Lithotamnium limestone. On this stratigraphical basis, the Hipparien

horizon has been dated as Tortonian, i.e. Upper Vindobonian (Late Miocene)

immediately antedating the Tortono-Sahelian transgression which occurs at

the end of the Miocene marine period in this area (Arambourg, 1959).

The MARCEAU deposit, and especially its lignite and lacustrine clay

and limestone which yield the Hipparion specimens, are referred to a period

anterior to the Pliocene transgression, i.e. they belong to the Upper Miocene

and may be equated with Oued el Hammam (Flandrin, 1942).

285

CAPE PROVINCE

HIPPARION FROM LANGEBAANWEG,

‘anbny *0% f~nayxyoy yoresyy “61 feTTIG IW [eqelq “gi fanezoy, £1

: VISINO J,

-eqeg (pep a uty ‘91 {Aog Ja ury pue papmnof ury “SI fneaoieyy “PI f3yeu0g ‘3g “1 Spneury ‘1g “SI SyoouRPy uly “11 Seppnog

Tuag puv yWayonog ury ‘or {yngg “6 SYyemnosuey] “g Svrvoseyy “ZL feuyey, g SwewWUEH je pono *S {(ueiQ) Aqnoreyy sting ‘PF {(ueid) sapreuy 3g °&

> VINao

*xnvoylog durery *z fyeqery "I a

: ODDONOP,

‘Ajoureu ‘paquiosap useq sey uorndgizy YTyM Woy sotyTed0] 9y} SuNvorpur (G “Sy uly) qorysepy oy} Jo deypy “PY ‘org

286 ANNALS OF THE SOUTH AFRICAN MUSEUM

There is good evidence that CAMP BERTEAUX (near Taourirt, Eastern

Morocco) belongs to this period. The Hipparion layers, from which Mastodon

cf. angustidens has been recovered (Bourcart, 1937), form the base of a lacustrine

series of ‘argiles smectiques’ which are lying on marine horizons with Tortonian

fauna and are covered with Pontian deposits. Choubert and Ennouchi (1946)

have dated the Hipparion layers as Upper Tortonian, being a sub-continental

phase at the end of the Miocene period. Thus, they are somewhat younger

than those from Oued el Hammam (Arambourg, 1959).

The brack water deposits of TAFNA (near Guiard, Province of Oran)

show a ‘Sarmatian facies’ of the Upper Vindobonian. On faunal and litho-

logical grounds, they were first referred to the ‘classical Pontian’ (Dalloni,

1914), but recently they were re-studied and correlated with the Oued el

Hammam horizon (Perrodon and Tempere, 1953). However, the geological

evidence is not clear.

PLIOCENE

Pliocene palaeontological sites are rare in the Maghreb. Some fossiliferous

lacustrine formations in the SETIF—CONSTANTINE area have been referred

to this period (Middle Pliocene). Included are also the limestone deposits of

the ST. ARNAUD Cemetery and of the MASCARA region, AIN EL BEY,

AIN EL HADJ BABA, ST. DONAT, and possibly UTIQUE in Tunisia.

Most of these exposures correspond to the fluviolacustrine filling of a

vast depression between the Tellian and Sahara—Atlas. According to Arambourg

(1956) the Azpparion sitifense layers of St. Arnaud Cemetery, Ain el Hadj Baba

and Mascara are lithologically different (sands and lacustrine limestones)

from the Stylohipparion and Equus numidicus deposits of Ain Boucherit and Ain

Jourdel (gray marls and red conglomerate). While these represent the base

of the ‘Villafranchian’ (similar to Lac Ichkeul and Fouarat), they are referred

to the Middle Pliocene. No satisfactory geological evidence is available for

TOZEUR (Tunisia): it was referred by Boule (1910) and Solignac (1931) to

the ‘Pontian’ stage because of the presence of Merycopotamus and Hipparion.

Now that the occurrence of these forms is known in Upper Miocene deposits

in India and North Africa, there is a special need for obtaining better

stratigraphical evidence for dating this site.

At WADI NATRUN (Gart el Moluk Hill) in the Nile Valley, fluvio-

marine gypsiferous clays (containing, among other rare vertebrates, Hipparion

albertense) are interbedded with limestones and shales, and deposited in an

arm of the Mediterranean in the Nile Valley. On a stratigraphical basis they

are dated as Middle Pliocene.

PLEISTOCENE

North Africa

The Apparion fossiliferous horizons in the Maghreb correspond strati-

graphically to the marine regression following the major Pliocene extension.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 287

The most significant exposures are St. Arnaud and Oran in Algeria, and Lac

Ichkeul, near Bizerta, in Tunisia (figs. 3, 4). The sequence has been extensively

described and discussed (Howell, 1959). The general feature is that of a

300-foot series of marls and lacustrine clays and silts, with interbedded

fossiliferous sandstones, gravels and calcareous conglomerates.

At LAC ICHKEUL the base is formed by deep-water Plaisancian and

lagoonal Astian marine deposits capped with fresh-water beds from the early

Pleistocene, with intercalation of a one-foot thick conglomerate.

At BEL HACEL eroded dune sandstones overlying transgressive Astian

marls and sands are bearing the ‘Villafranchian’ fauna: they are further overlain

by the 100-metre high Sicilian beach.

In the ST. ARNAUD area, two distinct fossiliferous horizons are visible.

The lower one (Ain Boucherit, i.e. Beni Foudda) is a coarse calcareous conglo-

merate; the upper one (Ain Hanech) is a cracked, rather sandy or gravelly

clay, intercalated in the thick marl filling the old marshy or lacustrine depres-

sion. Faunal and stratigraphical considerations make the St. Arnaud upper

deposit somewhat younger than Lac Ichkeul.

At ORAN (ST. CHARLES), where Libyhipparion (?) libycum was first

recovered and named by Pomel (1897), the consolidated dune sandstones and

clays are concordantly overlying the calabrian sandstones, of which they

represent the regression facies. The calabrian itself is in direct contact with the

Flabellipecten flabelliformis sands and sandstones of the marine Pliocene; and at

the base, the sequence reaches the Tortono-Sahelian marks and red stone of

the Upper Miocene (Arambourg, 1960).

Ligniferous clays of the PUITS KHAROUBY near Oran from which

H. massoesylium Pomel was recovered, has usually been placed—on the basis

of its mollusc fauna—in the Upper Pliocene, but there is little geological

evidence for it being older than Lower Pleistocene.

East Africa

Most of the fossiliferous horizons of East Africa (figs. 3, 5) are linked with

tuffs, ashes or lapilli projected and deposited by volcanic eruptions. This

volcanic activity spans over a long period of time, probably from Lower

Miocene up to Upper Pleistocene and even to historical times. Therefore it is

difficult, on purely geological grounds, to date the different fossil deposits.

Usually it is only by convergence of tectonic, stratigraphical, lithological and

chemical methods on the one hand, and by faunal studies on the other that it

is possible to propose a satisfactory date.

At KANAM, on the slope of the dissected volcanic centre of Homa

Mountain, the lacustrine brown calcareous fossiliferous clays, with their

interbedded fine tuffs from intermittent volcanic explosions, are commonly

referred to the Lower Pleistocene. |

The KAISO series is complex. It has been repeatedly studied (Wayland,

1926; Fuchs, 1934), and recently Lepersonne (1949) subdivided it into three

288 ANNALS OF THE SOUTH AFRICAN MUSEUM

‘SP

Lake Edward

ee te ae a

aaa’ BS a

Lake Kivu = oS

ee 7

RUANDA

= Lake

Eyasi

Ti ANG ANY, |

Fic. 5. Map of Central-East Africa (B in fig. 3) indicating the localities from which Hipparion

has been described, namely, 1. Lake Albert; 2. Kanjera; 3. Mumba Hills, Lake Eyasi; 4. Olduvai

Gorge; 5. South Serengeti, Laetolil beds.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 289

main stages of sedimentation, the middle one being highly fossiliferous. It is

a clayey deposit, alternating with thin beds of fine sands and sandstones, and

discontinuous ironstone horizons and limonite lenses (Kaiso bone beds),

containing silicified wood and a high proportion of vertebrates. It overlies a

silty lower deposit with some minor gravel horizons, resting on a basal ironstone

layer (laterite) capped by unstratified sands above the down-faulted peneplain

surface. This lower part of the Kaiso series, including the fossil horizon, is

usually placed in the Lower Pleistocene (Howell, 1959; Bishop, 1963).

The OMO beds are lithographically comparable to the Kaiso deposits.

The ‘Villafranchian’ fauna, and the Hzpparion material in particular, occur in

sandstone horizons between lacustrine volcanic clayey tuffs. In the evolution of

the Rudolph depression in its Nilotic or open phase, Arambourg (1943) has

distinguished two major extensions of the lake during the Pleistocene: (1) the

vertebrate beds occur in a sand and sandstone horizon of the lower Omo

Valley, representing a first phase of lacustrine deposition prior to the general

tectonic uplift of the end of the Lower Pleistocene. (2) Posterior to the tectonic

dislocation, the second phase corresponds to the series of horizontal terraces

on the slope of the Rudolph Basin—they are not fossiliferous. This distinction,

confirmed by the study of the fauna, permits the fossil beds to be dated as

Lower Pleistocene, probably younger than Kaiso and Kanam.

It is more difficult to specify the geological location of the SERENGETI

Hipparion material. The Laetolil beds in the Vogel River area show a sequence

of upfaulted subaerially deposited yellow and gray tuffs (Kent, 1941). Unfor-

tunately the fauna collected by Kohl-Larsen ‘in den Schluchten und Wasser-

rissen’ of the South Serengeti was probably recovered from various horizons

which were altered and mixed-up by subsequent erosion (Dietrich, 1942). The

gray tuffs, about 20 m. thick, which appear to contain the ‘older fauna’ (Kohl-

Larsen, 1939), are dated as Lower Pleistocene, more or less contemporary

with Omo and Olduvai I. Efforts have been made to identify this stratigraphic-

ally older fauna by such means as the type of mineralization and colour.

Conclusions based on this method can only be accepted with reserve because

of the fact that possible weathering may have altered the original appearance

of the fossils. Nevertheless, Dietrich (1942) believed that all his Hypsohipparion

material derived from the lower gray tuffs.

The vast sequence of stratified beds exposed in the OLDUVAI GORGE

(northern Tanganyika) has been extensively discussed in recent years, and

lately by Pickering (1960) and Leakey (1963). Independently of the chemical

dating of the beds, which provides evidence of a great duration and antiquity

of Bed I, dating has also been attempted on studies of the fauna and on a

climatological basis. Little has been added from the strictly geological point

of view. It is important to note that the major climatic, geological and appar-

ently faunal ‘break’ occurs near the base of Bed II and not, as was previously

thought, between Beds I and II. Most workers agree that Olduvai I belongs

to the Lower Pleistocene, being more or less contemporary with (and probably

290 ANNALS OF THE SOUTH AFRICAN MUSEUM

prior to) Omo and Laetolil, while Beds II, III, IV are considered to belong to

the Middle Pleistocene.

The KANJERA exposures, on the low cliffs of the Homa Mountain close

to Lake Victoria, consist of a tripartite series of basal greenish ash and tufts,

succeeded by clays with limestone and by upper transgressive beds of brown-

greenish clay. Most of the fossils are obtained from and in close association

with the lower tuff and ash layer, but it is not possible to be certain whether

or not all the material came from this source (Kent, 1942). However, there

is sufficient evidence that the Hipparion may be considered to be more ancient

than the human fossil remains which were recovered from the middle beds.

The Kanjera faunal beds are referred to the Middle Pleistocene, roughly

contemporary with Olduvai IV.

The LAKE EYASI beds are usually referred to an early Upper Pleistocene

period, more on a palaeontological than on a geological basis.

According to Reck and Kohl-Larsen (1936), it is probable that the

Hipparion teeth recovered west and north of Mumba Hill, in the north-east

section of the Eyasi basin (see p. 283), do not belong to the original beds. Indeed

they are very much rolled, which is not the case for the presumably contem-

poraneous material, and constitute a secondary deposition. If this interpretation

is correct, these teeth are probably more ancient than the typical Eyasi fauna,

and should be dated to a period prior to the Upper Pleistocene.

South Africa

The only relatively satisfactory stratigraphical information concerning

Hipparion in South Africa is obtained at the YOUNGER GRAVELS OF THE

CHRISTIANA

BARKLY WEST

Sca/e in miles

2KIMBERLEY

Fic. 6. Map of northern Cape Province, South Africa (C in fig. 3) indicating localities from

which Hipparion has been described, namely, Sydney-on-Vaal, Pniel and Christiana.

2Q1

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

‘sinqsouuvyof = [{ fueqing = q ‘ulaju0;WOO[G = g : UMOYS 91% SOTTO 9014}

jo suonmtsod oy, “sy jo q pue + svoiv 0} UONYoI ut ‘spruorreddiy Jo AraAOOSIp Jo SaTITeIO] UIEIIIN SuNeoIpul votpyY YINog jo depy ‘LZ ‘917

OO OS2 002 OSI OO! OS

uty

yoqbul.ds «

B1]aUJODe

WJe4 S/O.

yebsuedeyey «

292 ANNALS OF THE SOUTH AFRICAN MUSEUM

VAAL RIVER (Christiana; Pniel; Sydney-on-Vaal: fig. 6). In this horizon,

which actually presents the first fossil occurrence in the Vaal area (as fossils

of the oldest gravels have probably been destroyed), three phases have been

recognized from their elevation and their stone artefact associations. However,

most of the fossils seem to have derived from Phase II and III, although it is

not impossible that the fossils found below Barkly West may have been contained

in Phase I of the Younger Gravels (Cooke, 1963). On the basis of a climatologi-

cal interpretation (Séhnge, Visser and Lowe, 1937; amended by Cooke, 1947,

and Lowe, 1952), the Younger Gravels were correlated with the Middle

Pleistocene Olduvai Beds II and IV.

The CORNELIA beds, an open site on the farm ‘Uitzoek’ in the

Orange Free State, consists of exposures of clays, gravels and sand, which

represent an old ‘pan’ filling. Oakley (1954) emphasizes the Pleistocene nature

of the site by referring to a ‘lower Fauresmith industry’ from the beds. There is

no reliable date based purely on stratigraphy for this site.

At present it is also impossible to make any tentative geological dating

for the Notohipparion material from NAMAQUALAND. It was recovered from

a granite level in a well, 60 feet down in the surface limestone, 40 miles east of

Springbok. It is known that the infilling of the valleys with sand has been

occurring in Namaqualand from Upper Cretaceous times, with probable

breaks in the sedimentation, but there is no information on the actual sequence

of gravels and limestone in the well (Haughton, 1932).

The Aipparion occurrence in cave fillings ata KROMDRAAT, (?) MAKA-

PANSGAT and BOLT’S FARM has been dated by faunal and climatological

methods (Ewer, 1957; Brain, 1958), and nothing further can be added from

a purely stratigraphical point of view.

LISTS OF FAUNAL ASSEMBLAGES AT THE AFRICAN HIPPARIONID SITES

It is considered useful to include lists of all the diagnosed, published fauna

from the various sites where Hipparion has been recovered. Omission of particular

sites is due to lack of information, and Hopefield is included because of the

richness of fauna and for comparison with contemporaneous sites. The faunal

assemblage of Langebaanweg is listed on page 280. The sites are listed according

to the order presented on pages 281-3. Wherever possible, generic and specific

names are retained as they appear in the publications quoted.

INDIVIDUAL HIPPARIONID SITES

Oued el Akrech (Morocco) (After Arambourg and Choubert, 1957)

PROBOSCIDEA PERISSODACTYLA

+ Anancus osiris +Stylohipparion sp.

+Elephas africanavus

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 203

Oued el Hammam (Oran) and Marceau (Algiers) (After Arambourg, 1963)

PRIMATES PERISSODACTYLA

+ Macaca flandrin | Dicerorhinus primaevus

RODENTIA + Hipparion africanum

+Hystrix sp. ARTIODACTYLA

CARNIVORA }Palaeotragus germaini

+Samotherium sp.

Dene ae erenses tDamalavus boroccot

TUBULIDENTATA tGazella praegaudryi

+ Orycteropus mauritanicus + Tragocerus sp.

PROBOSCIDEA {Cephalophus sp.

+ Turicius sp. AVES

+ Mastodon sp. {Struthio sp.

Ain Hanech (Algeria) (After Arambourg, 1947; 1949)

CARNIVORA ARTIODACTYLA

Hyaena sp. Hippopotamus amphibius

Canis sp. + Omochoerus(*) sp.

PROBOSCIDEA Giraffa sp.

Bs {Libythertum maurusium

tAnancus osiris Bie

+Elephas cf. planifrons + Buba ie :

+E. aff. meridionalis or recki P-

Oryx sp.

PERISSODACTYLA Alcelaphus sp.

+ Atelodus aff. simus {Gazella sitifensis

{Stylohipparion libycum

Equus sp.

(4) According to Leakey, 1958= Mesochoerus.

Garet Ichkeul (Tunisia) (After Arambourg, 1949)

PROBOSCIDEA ARTIODACTYLA

+ Anancus osiris Hippopotamus sp.

+Elephas cf. planifrons {Libytherium maurusium

Bos sp.

PERISSODACTYLA Oryx sp.

tAtelodus aff. simus Alcelaphus sp.

+Stylohipparion libycum {Gazella aff. sitifensis

Equus Redunca sp.

Tozeur (Tunisia) (After Roman & Solignac, 1934)

PROBOSCIDEA PERISSODACTYLA

{Mastodon cf. longirostris + Hipparion sp.

294

ARTIODACTYLA

+ Merycopotamus aff. dissimilis

tCapreolus matheront

ANNALS OF THE SOUTH AFRICAN MUSEUM

+ Tragocerus amaltheus

| Hemitragus perimensis

Antilopinae

Wadi Natrun (After Studer, 1898; Andrews, 1902; Arambourg, 1947 and 1963)

MAMMALIA

PRIMATES

+Libypithecus markgraft

CARNIVORA

+ Machairodontidae gen. sp. indet.

Phocidae gen. sp. indet.

Lutra sp.

LAGOMORPHA

? Leporidae

PROBOSCIDEA

+ Mastodon sp.

PERISSODACTYLA

+Hipparion sp.

Omo (After Arambourg, 1947)

PRIMATES

+ Dinopithecus brumpti

CARNIVORA

| Homotherium (?) ethiopicum

PROBOSCIDEA

{Deinotherium bozasi

tElephas (Archidiskodon) recki

+Elephas (Archidiskodon) cf. planifrons

PERISSODACTYLA

+Atelodus cf. germano-africanus

+ Stylohipparion albertense

Equus cf. zebra

ARTIODACTYLA

tHippopotamus (Tetraprotodon) prot-

amphibius

+ Omochoerus heseloni(*)

+ Metridiochoerus andrewsi(*)

(*)= Mesochoerus heseloni (Leakey, 1958).

($) = Notochoerus euilus (Leakey, 1958).

ARTIODACTYLA

{Hippopotamus protamphibius

{Stvachoerus giganteus

Camelidae gen. sp. indet.

{Libythertum maurusium

+ Hippotragus (?) cordiert

REPTILIA

Crocodilus

{Euthecodon nitriae

| Sternothaerus dewitzimanus

+ Trionyx pliocaenicus

PISCES

Protopterus

Synodontis

+ Notochoerus capensis (*)

+Phacochoerus africanus fossilis(*)

Giraffa camelopardalts

+Guraffa gracilis

+ Sivatherium olduvaiense(*)

+ Menelikia lyrocera

Kobus (Kobus) sigmoidalis

Kobus (Kobus) sp.

+ Redunca ancystrocera

Alcelaphus sp.

+ Strepsiceros imberbis

Aepyceros melampus

Antidorcas sp.

Oryx cf. gazella

Tragelaphus nakuae

Taurotragus cf. procanna

Gazella praethomson

Syncerus aff. brachyceros

(?) =Pronotochoerus jacksoni (Leakey, 1958).

(*) According to Leakey, 1958= Tapinochoerus meadowsi, but does not belong to the Omo

Beds.

(®) See footnote to Langebaanweg fauna (p. 280).

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 295

Koro Toro (Tchad) (After Abadie, Barbeau and Coppens, 1949; Coppens,

1960)

MAMMALIA PERISSODACTYLA

PRIMATES {Stylohipparion sp.

i Ceratotherium simum

+‘ Australopithecus’ sp. Hane

ARTIODACTYLA

| Hippopotamus cf. protamphibius

}Sivatherium(*) sp.

ES GEOSeDE* Giraffa camelopardalis

tArchidiskodon africanavus (+) Alcelaphus sp.

+A. recki(?)

tAnancus sp. REPTILIA

+Stegodon sp. Crocodilus niloticus

CARNIVORA

HAyaena cf. striata

(+) According to Cooke, 1960= Elephas africanavus.

(?) According to Cooke, 1960=Palaeoloxodon recki.

(?) See footnote to Langebaanweg fauna (p. 280).

Kaiso?

CARNIVORA + Metaschizotherium hennigi

+? Homotherium ethiopicum Ceratotherium simum

PROBOSCIDEA

ARTIODACTYLA

+ Hippopotamus imaguncula

+H. protamphibius

+Stegodon kaisensis

tArchidiskodon exoptatus

PERISSODACTYLA + Notochoerus euilus

tStylohipparion albertense | Mesochoerus limnetes

Kanam!

CARNIVORA PERISSODACTYLA

Crocuta crocuta +Stylohipparion albertense

+ Equus oldowayensis

Ceratotherium simum

Diceros bicornis

PROBOSCIDEA ARTIODACTYLA

tAnancus kenyensis {Hippopotamus imaguncula

+Stegodon katsensis + Metridiochoerus pygmaeus

tArchidiskodon subplanifrons +Nyanzachoerus kanamensis

+A. exoptatus Giraffa camelopardalis

+ Deinotherium bozasi + Libytherium olduvatense

1 Compiled from Hopwood, Leakey & McInnes in Leakey, 1951; Dietrich, 1950; McInnes,

1953; Cooke, 1963; Leakey, 1958.

296 ANNALS OF THE SOUTH AFRICAN MUSEUM

Eyasi!

PRIMATES

Homo sp.

RODENTIA

+ Pedetes surdaster

Thryonomys swinderianus

CARNIVORA

Crocuta crocuta

Caracal caracal

Panthera leo

Panthera pardus

TUBULIDENTATA

+ Orycteropus aethiopicus

PERISSODACTYLA

Equus burchelli

+Stylohipparion sp.

Diceros bicornis

Laetolil (South Serengeti)?

PRIMATES

+ Sumopithecus oswaldi

RODENTIA

}Pedetes surdaster

+ Hystrix galeata

CARNIVORA

Canis mesomelas

C. mesomelas latirostris

tCanis africanus

Crocuta crocuta

Caracal caracal

? Panthera pardus

TUBULIDENTATA

t Orycteropus aethiopicus

PROBOSCIDEA

tAnancus kenyensis

tArchidiskodon subplanifrons

TA. exoptatus

tA. recki

+ Deinotherium bozasi

1 See footnote p. 295.

ARTIODACTYLA

Hippopotamus amphibius

Potamochoerus koiropotamus

Giraffa camelopardalis

Strepsiceros strepsiceros

Taurotragus oryx

Syncerus caffer

+ Homotoceras nilsoni

Kobus ellipsirymnus

tAdenota kob

{Redunca redunca

Pelea sp.

+ Oryx beisa

Aepyceros melampus

+Gazella granti

PERISSODACTYLA

+ Stylohipparion albertense

Equus burchelli

+ Metaschizotherium hennigi

Ceratothertum simum

+ Serengeticeros efficax

ARTIODACTYLA

Hippopotamus amphibius

+ Notochoerus euilus

+ Okapia stiller

Giraffa camelopardalis

+Libytherium olduvaiense

+ Tragelaphus buxtom

Taurotragus oryx

Syncerus caffer

? Kobus ellipsirymnus

}Redunca redunca

+ Hippotragus equinus

+Damaliscus angusticornis

Aepyceros melampus

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 207

Olduvai (After Leakey, 1958, and Cooke, 1963)

Numbers I, II, III and IV refer to the various beds

PRIMATES

+Simopithecus oswaldi (1, 11, III, IV)

+S. jonathani (II, ? IV)

+Paranthropus boiser (1, 11)

+Homo sp. (1, IT)

CARNIVORA

Canis mesomelas (II, IV)

C. mesomelas latirostris (11, IV)

+Canis africanus (II, IV)

Aonyx capensis (II)

Crocuta crocuta (1)

Panthera leo (11, I11)

P. pardus (? 1)

PROBOSCIDEA

tAnancus kenyensis (1, I1)

+ Archidiskodon exoptatus (1)

jae reckt) (Ll ELT EV)

+ Deinotherium bozasi (I, 11)

PERISSODACTYLA

+Stylohipparion albertense (1, II, III,

IV)

Equus burchelli (11, 111, IV)

+E. aff. grevyz (1, II, III, IV)

+E. oldowayensis (1, II, III, IV)

+Metaschizotherium hennigi (I, 11)

Ceratotherium simum (1, II, III, IV)

+Serengeticeros efficax (11)

Diceros bicornis (1, 11, III, IV)

ARTIODACTYLA

+Hippopotamus gorgops (I, II, III,

IV)

+Notochoerus hopwoodi (111, 1V)

+N. compactus (11)

+ Mesochoerus olduvaiensis (1, II, III,

IV)

+Potamochoerus majus (I, II, III, TV)

Phacochoerus africanus (II, III, IV)

+P. altidens altidens (II, III, IV)

+P. altidens robustus (I, 11)

+ Tapinochoerus minutus (1V)

+ T. meadows: (I, 11, III, IV)

tAfrochoerus nicolt (II, III, IV)

| Metridiochoerus andrewsi (1)

Orthostyonyx brachiops (11)

+Okapia stiller (? 1)

Giraffa camelopardalis (11, IV)

+G. gracilis (11)

tLibythertum olduvaiense (1, II, III,

IV)

Strepsiceros strepsiceros (? I, II, IV)

1 Strepsiceros imberbis (? I, II, IV)

} Tragelaphus buxtoni (11)

Taurotragus oryx (I, II, IV)

+ Homoioceras nilsson: (IV)

+Bularchus arok (II, III, IV)

+ Adenota kob (1, II, III)

tHippotragus equinus (III, IV)

1H. niro (II, IV)

t Oryx beisa (1)

+Damaliscus angusticornis (II, IV)

tD. teste (1, II, IV)

+Alcelaphus kattwinkeli (11, III, IV)

}Beatragus hunteri (1, 11, IV)

{Gorgon taurinus semiticus (I, II, IIT,

IV)

{Gazella gazella praecursor (I, II, IV)

TG. granti (I, II, IV)

{ Phenacotragus recki (1V)

+Pultiphagonides africanus (1, 11)

{Pelorovis oldowayensis (II, IV)

Vaal River Younger Gravels (After Cooke and Wells, 1946; Cooke, 1949, 1963;

Wells, 1964)

CARNIVORA

cf. Crocuta crocuta

PROBOSCIDEA

{Gomphotherium sp.

298 ANNALS OF THE SOUTH AFRICAN MUSEUM

+‘Archidiskodon’ subplanifrons +Stylochoerus compactus

t‘Archidiskodon’ brooma }Phacochoerus aethiopicus

+‘ Archidiskodon’ transvaalensis Phacochoerus africanus

+*Palaeoloxodon archidiskodontoides’ Hippopotamus cf. amphibius

(SS ga: 2) }Sivatherium (?) cingulatum(*)

t‘Loxodonta’ cf. atlantica + Stvatherium olduvaiense haughtoni(*)

cf. Loxodonta africana cf. Alcelaphus caama

PERISSODACTYLA +‘ Alcelaphus robustus’

SD onppRTiOn sleytlers i Megalotragus eucornutus

é onnochaetes cf. gnou

gues Pelee cf. Connochaetes sp

+ Equus plicatus : i

cf. Damaliscus sp.

cf. Sylvicapra grimmia

cf. Aepyceros melampus

cf. Antidorcas marsupialis

t‘Equus sandwith’ (=E. plicatus ?)

Equus cf. burchelli

Equus cf. quagga

cf. Diceros bicornis 1 etal Tes

ARTIODACTYLA cf. Hippotragus sp.

+ Mesochoerus paiceae cf. Strepsiceros strepsiceros

+ Notochoerus capensis Taurotragus cf. oryx

+ Tapinochoerus modestus Syncerus cf. caffer

+* Tapinochoerus’ meadowst - t‘Homotoceras’ cf. bani

(+) See foomote to Langebaanweg fauna (p. 280).

Cornelia (Uitzoek) (After Cooke, 1963)

PERISSODACTYLA T aurotragus Oryx

+ Stylohipparion steytlert +‘Homoioceras’ baini

Equus burchelli + Kobus venterae

+E. plicatus Damaliscus cf. albifrons

+ Eurygnathohippus cornelianus +Damaliscus sp.

ARTIODACTYLA Alcelaphus caama

+ Hippopotamus gorgops arto

: # ea sorgops +Connochaetes laticornutus

Notochoerus compactus

+ Megalotragus eucornutus

Phacochoerus africanus SY aapilin ies

+ Orthostonyx sp. i

Giraffa camelopardalis LG elnoee

tLibytherium olduvaiense

Strepsiceros strepsiceros

+Gazella sp.

+ Antidorcas marsupialis

Kromdraai (After Cooke, 1963)

INSECTIVORA Crocidura cf. bicolor

+Proamblysomus antiquus Suncus cf. etruscus

+ Elephantulus langi +? Myosorex robinsoni

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

PRIMATES

+ Parapapio jonesi

+Papio robinsont

+Paranthropus robustus

LAGOMORPHA

Lepus capensis

RODENTIA

+ Mystromys antiquus

Tatera cf. brantsi

?Desmodillus auricularis

Grammomys cf. dolichurus

+Rhabdomys cf. pumilio

Mastomys cf. natalensis

Malacothrix cf. typica

Steatomys cf. pratensis

+Palaeotomys gracilis

tCryptomys robertst

CARNIVORA

Canis mesomelas pappos

+Canis atrox

+Canis terblanchet

Vulpes pulcher

Bolt’s Farm (After Cooke, 1963)

INSECTIVORA

+Proamblysomus antiquus

+Atelerix major

+ Elephantulus langi

+Elephantulus antiquus

E. cf. brachyrhynchus

Suncus cf. etruscus

| Myosorex robinsont

CHIROPTERA

Rhinolophus cf. capensis

tcf. Myotis sp.

PRIMATES

+ ?Parapapio broomi

+? P. whiter

tCercopithecoides williamsi

RODENTIA

Pedetes cf. caffer

| Herpestes mesotes

{Crossarchus transvaalensis

tCrocuta spelaea

tC. ultra

+ Hyaena bellax

{Felix crassidens

| Therailurus piveteaur

{Panthera aff. leo

t? P. white

TP. shaw

+ Megantereon eurynodon

HYDRACOIDEA

{Procavia antiqua

+P. transvaalensis

PERISSODACTYLA

tStylohipparion steytlert

? Equus burchelli

TE. plicatus

TE. helmet

ARTIODACTYLA

+Potamochoerops antiquus

+Mystromys hausleitnert

Tatera cf. brantst

tDasymys bolti

+Rhabdomys cf. pumilio

+ Thallomys debruyni

Leggada cf. minutoides

L. cf. major

Malacothrix cf. typica

tPalaeotomys gracilis

HAystrix africae-australis

tCryptomys robertst

CARNIVORA

Canis mesomelas

C’. mesomelas pappos

Aonyx cf. capensis

Suricata suricatta

+Crossarchus transvaalensis

+Hyaena bellax

+Leptailurus spelaeus

299

300 ANNALS OF THE SOUTH AFRICAN MUSEUM

| Therailurus barlowi

+Panthera aff. leo

+ Machairodus transvaalensis

PROBOSCIDEA

+Loxodonta atlantica

HYDRACOIDEA

Procavia capensis

PERISSODACTYLA

+Stylohipparion steytlert

Equus burchelli

{E. plicatus

ARTIODACTYLA

+ Tapinochoerus meadowsi

Makapansgat (After Cooke, 1963)

INSECTIVORA

tChrysotricha hamiltoni

{Elephantulus langi

Suncus cf. etruscus

| Myosorex robinsont

PRIMATES

+Simopithecus darti

{Parapapio jonest

+P. broomi

{P. whiter

{Papo robinsoni

tAustralopithecus africanus

+Cercopithecoides williamsi

LAGOMORPHA

Pronolagus randensis

RODENTIA

+ Mystromys hausleitneri

tM. darti

? Tatera cf. brantsi

Grammomys cf. dolichurus

Pelomys cf. fallax

+Rhabdomys cf. pumilio

Aethomys cf. namaquensis

Mastomys cf. natalensis

Leggada cf. minutoides

+Potamochoeroides shawi

+P. antiquus

Tragelaphus scriptus

Taurotragus oryx

Syncerus caffer

Damaliscus cf. pyrgatus

? D. cf. lunatus

+ Alcelaphus robustus

tA. helmen

Connochaetes taurinus

+ Makapania broomi

+? Raphicerus campestris

+Gazella wellsi

1 ? Phenacotragus vanhoepent

} ?Antidorcas marsupialis

Dendromus cf. mesomelas

+ ?Malacothrix makapani

Steatomys cf. pratensis

+Palaeotomys gracilis

+ Aystrix major

H. africae-australis

+ Xenohystrix crassidens

+ Gypsorhychus makapani

tCryptomys robertst

CARNIVORA

? Canis mesomelas pappos

+Cynictis penicillata brachyodon

+Crocuta cf. brevirostris

+ Hyaena makapani

+ Therailurus barlowi

+ Megantereon sp. nov.

HYRACOIDEA

+Procavia antiqua

+P. transvaalensis

+Procavia sp.

PERISSODACTYLA

+? Stylohipparion steytlert

tEquus helmet

+ Metaschizotherium (?) transvaalensis

Ceratothertum simum

Diceros bicornis

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 301

ARTIODACTYLA

Hippopotamus amphibius

+ Notochoerus euilus

+Potamochoeroides hypsodon

+P. shawit

Giraffa camelopardalis

+Libytherium cf. olduvaiense

Strepsiceros strepsiceros

Tragelaphus angasi

Taurotragus oryx

Syncerus caffer

+S. cf. makapani

+Cephalophus pricet

C.. caerulus

tRedunca darti

Redunca arundinum

R. fulvorufula

Oryx gazella

+ Alcelaphus robustus

tA. helmet

Connochaetes taurinus

{ Oreotragus major

+ Makapania broomi

Aepyceros melampus

{Gazella gracilior

+ Phenacotragus vanhoepeni

‘ELANDSFONTEIN’, HOPEFIELD (SOUTH AFRICA)

(Modified from Singer, 1957)

PHOLIDOTA

Manis sp.

PRIMATES

+Simopithecus oswaldi hopefieldensis

+Homo sapiens rhodesiensis (“Saldanha

Man’)

LAGOMORPHA

Lepus sp. (cf. capensis)

RODENTIA

Bathyergus sp. (cf. suzllus)

Georychus sp. (cf. capensis)

HAystrix sp. (cf. africae-australis)

Otomys sp. (cf. saundersiae)

Parotomys sp. (cf. brantsz)

CARNIVORA

Canis mesomelas

Canis adustus

tLycaon pictus magnus

Mellivora capensis

Herpestes sp. (cf. ichneumon)

Herpestes eogale

Hyaena brunnea

+Crocuta spelaea

Lynx caracal

Leptailurus serval

+Panthera leo spelaea

| Megantereon gracile

PROBOSCIDEA

tLoxodonta (Palaeoloxodon) cf. anti-

quus reckt

+? Archidiskodon sp.

PERISSODACTYLA

+Equus (Hippotigris) plicatus

+E. helmei

+E. cf. sandwithi

Ceratotherium simum

Diceros bicornis

ARTIODACTYLA

+ Mesochoerus lategani

1M. paiceae

+ Tapinochoerus meadowsi

Hippopotamus amphibius

+Sivatherium olduvaiense(*)

+Giraffa cf. gracilis

| Homoioceras sp.

Taurotragus oryx

Redunca arundinum

Raphicerus campestris

Antidorcas marsupialis

tAntidorcas sp.

Tragelaphus(*) sp.

302 ANNALS OF THE SOUTH AFRICAN MUSEUM

cf. Hippotragus niger Connochaetes sp.

tcf. H. leucophaeus { Lunatoceras sp.

| Hippotragus sp. - }Pelea sp.

+Damaliscus cf. dorcas tcf. Gazella wellst

{Damaliscus sp. +Gazella sp. nov.

(1) See footnote to Langebaanweg fauna (p. 280).

(2) In the other faunal lists Tragelaphus has not been substituted for the earlier labelling

Strepsiceros.

Refer now to table 1 on pages 303 to 313.

COMMENTARY ON THE PUBLISHED FAUNAL ASSOCIATIONS AT

HIPPARIONID SITES IN AFRICA

MIOCENE

Oued el Hammam and Marceau

Arambourg (1959) has clearly demonstrated that these two sites contain

fundamentally the same fauna: both include Hipparion africanum, Samotherium

sp. and Ayaena algeriensis, and cannot but be contemporaneous, as is confirmed

by their stratigraphy.

It seems clear that these assemblages correspond to a very special and,

so far, little-known stage of development of the fauna of Africa, posterior to

the Burdigalian and anterior to the ‘classical Pontian’. The fauna from these

sites are very different from the well-represented assemblage of the Lower

Miocene in East Africa (Losodok, west of Lake Rudolph; Rusinga and Moboko

Islands in the Kavirondo Gulf of Lake Victoria) and South West Africa. The

typical Burdigalian assemblages contain, inter alia, Mastodon cf. longirostris,

Deinotherium hobleyi, Aceratherium, Teleoceras, and Anthracotherudae, Propalaeo-

choerus, Bunolistriodon, Dorcatherium, Creodonta, none of which are found at

Oued el Hammam, nor are the anthropomorpha which constitute the typical

African ‘touch’ of this Burdigalian fauna. The only similarity between the sites

and a Burdigalian fauna is the presence at Oued el Hammam of an orycterope,

namely, Orycteropus mauritanicus which is comparable with Myorycteropus McInnes

of East Africa. The separation between Oued el Hammam and the Lower

Miocene fauna of East and South West Africa is emphasized by the presence

(at Oued el Hammam) of Giraffidae (Samotherium and Palaeotragus), the

development of Bovidae (Damalavus, Gazella, Tragocerus, Cephalophus) and the

appearance of Hyaena and especially of the equids with Hipparion africanum.

On the other hand, the assemblages from the two North African localities

are at first sight characteristic of the so-called ‘classical Pontian’ of Eurasia

(Eppelsheim, Pikermi, Mont Luberon). However, on detailed study and

comparison of the several groups involved, Arambourg has shown that Eurasian

‘Pontain’ assemblages and the material from Oued el Hammam may belong

to the same ensemble, but they are not identical: specific and generic differences

exist and there is not even a single species in common (table 2).

393

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

uRyques,

iseAW

aTTIVsasI0TO

viofueyy

Mey

NIALL ETO

TEBE EAN} KO)

II teanplo mrs

I eanplO

poe]

owO

UIVUL YT

OSIvY]

eMquInyyy

[IMSL BERR Ee

roy

DUMNVVEIEONV'I

YIIMIIPUO AA,

[RAZ BIEIVA\

syWivoaH jo Dp

I aWIavy,

Peqst10p.y

ppyedoy

BI[aUIO')

JOATY [ee

wey sajog

TevIPULOIS]

”~

SUeIxIIeMS

yessuedeyeyy

*(Zg61) ta8utg os[e 999 z

“ds snaayjiqouis' y.

EHO Y NSE:

elDP S'

elUDYIDUOL “65

Rip]omso snsaypiqowns

SAaLVWNITdd

Issagialy9s “JO sisaagorurpy

ee ‘ds suodqyy "Joy

vs wosffoas "JO “y

ees zs sasuaqva “yo snydojouryy

VuUALdOUIHO

X WUOSULGOL XALOSOKTY y

x snasnija "JO snouny

Xx 59 a0 56 4010919 *}9 *4)

Nel pies see SISUASTUND] DANPIION) ».

BY a|PaNGlas is “+ wsuards apps Cuojq7y

SNaulpecd: as * ssnymudysdyIv1g “JO

snnbyup snynquoy dap T +

xX ae = "+ asun) snynquoydapy

2 By sey MODUL X19] AIF ye

XY rus oe "+ pajads 0q70}0.40) 1") x.

is WUOZUDY DYILAMGOSMLUY’) »

SNNBYUD SNUWOSC]QWVOLT ye

VUOALLOUSNI

UlI3}UOFAIIIS

sune yj,

ANNALS OF THE SOUTH AFRICAN MUSEUM

304

uelquies,

isvAq

aT[TesasSI0[O

eiofueyy

IME yy

AI feAnplo

IDE LANA)

IT, AN 2 ©)

Th BEANE) (O)

TOE]

oulg

weUuey

OsTe sy]

eMquInyAy

[DOHEL WSR ORS

rawysyy

DAEUMNVVEAONVT

YIIMIOPUO AA

syyieay] jo "5

[eeIILLA

peqsHopy

PreyedoH

eI[IUIO')

JOATY [ERA

wade y Sjjog

TevIpuloIy

SUBIYIIVMS

bebe

yessuedeyeyy

UIOJUOFYIIIG

sunv J,

‘ds sunpy

VLOATTOHd

syyoxvs sngay

sisuadva sndaT

** sisuapups snoipjouolg

VHdYOWOOVT

wsupyyjiad sapr0zayj1goI4a") »

sis “ds owopy »

suarqvs "JO "HH

SISUASAPOYL OULOFT y

sisuadns sndoayquv)a 7, »

SNUDIA[D SNIBYIUJO]DAISNY ye

oie SNISNGOL “Tx

125109 SNGOMYJUDLV

SnUuIsn *q

"* 1UOSULGOL “Ty

sqaousnsun "7x

OI ca

tpozt din

quupunau o1dog

CEE ale:

UMOOLY “Te

wsauol * Tx

snnbyuv ovdodving »

suadur snaayjiqourgy »

"+ solo snaayjuqor.10) »

35)

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

rn %

calla

wn %

”~

o- Bd

ee LOLDUL XISOFT xe

"+ 177aQGwUpd sMUojojOLlg »

“+ SUlIDLF SCU0I0ADIDE

ISJUDLG “JO SNUOJOLDT

SnyDIINSIUN “JD “PAT

“+ 14aUdn} "JO snUuojod TAT

aDISLapUNDS "JD “Cm

SNIDLOALL “JO SAUO0JE

sisuagog "JO sduojzvays

qupgoynwu “We é x

pad] "JO xiLyjoovIvINY

“* spjaulosaul “JO °G

snnbyup snuoipuagy »

siugsaquns snuojsor905"

oe oe solout ‘yo OFF

saploynuuu "JO DpvssaT

SISUI]DIDU "JO SMUOISDTY

snanpang “tO *[,

wudnigap Sio7]DY T, x

snprygosdiys “Jo stuoyjap

sisuanboupnu "jo scuoyjap

ound "JO scuopqvyyY x.

xpyof ‘Jo suojag

xe 19109 SCLUASD

smNy IOP “FI SCUOWUDIE

SUUDINIUND SNPIpoOUsaq

IJUDAQ “JO VALI}D J,

HLUP “W x

1IUpLaISNDY “TAT x

snnbiyun sduogsy

oe oe alfa ‘yo Saf

aS SE) [2 Oa 6

ey + LaISDPANS “Ty

"+ apppsuasDy $322P2dq x

2Y susuadva sniay”

VILNUAOU

ANNALS OF THE SOUTH AFRICAN MUSEUM

306

uelyquies

seh

dTIesos10[O,

eiafueyy

Mey

ANI MESO)

INE BANAT)

iEesUPT@

PLENTY)

poe T

owlO

UIVUL YT

OsTeyy

emquinyy

”~

TEH woxorg

Jowyayy

DAUMNVVEFTONV'T

”~

YIOMIOPUO AA

syyeoH jo "5

TeRIAALLA

Peqsioy y

preyedoy

eI[auIOT)

JOATY [PVA

wey Sa]Og

~ MS eM

IeVIPULOIY]

SUBIYIIEMS

yessuedeyeyy

UI9JUOFAINIS

—

sunvy,

sisuadns pioarnjayy

SYOJDGAUL U0kI01C—

snuspu snd *T

snag uovsd7T

DULDYD * A

4ayang saqin4

SNISNpD *F)

me ML ON

SUSOMQILY *) x

XOMD “Ty

(uovad'7 ¢) snuvoidfo “Fy

soddvd svpjawuosaut +)

SISOLNID] SDJAULOSAUL *F)

Sppauosaut stud‘)

VAud ISSTA

“*snqins "JO snduatyjog

Isplaqod SMUMOZGCL »

sasuadva "Jo snysd10ax)

dUdqvynu *)

AOUNL *)

Yaop snyatyLosqae »

LOUIUL SNUWLOM IT ye

snupiapuins sdiuouodsy J,

SUAPISSDLI XIUPSAYOUIN

oe DIVIIDG “ET

SUDAISND-ADNA[D * FT

DIDISILI “JO XISAET

397

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

~ A

~ my yO

CORY Sordoe

fa esl

paingut snpLvog *d x

snp.vog “gq

CHESS Ges

0a) DiayqUDg

* imvajand *T, %

UNOPADG SNANIWWAIY T,

oa1as "T

Tle

snavjags snjniiv}qaT x.

JDIDLDI ]DIVDAD!)

“+ qwIvLDI KUCT

SUAPISSDAI SUIOT ye

*01199 “Hx

qwpqoyvul “HT »

wd “Hx

pauunig Duavdyy

SUJSOMINILG “JD "Dye

LESIONS

DINISNUIA “4

pavpags *') x

D]JIGUD VINIONI “Tx

DINIOAI VINIOL)

VRE TES

wsuagsag]s DUIVAIAT x

xafiof vuavdo2aT

uopodyavig vpvIINUaG *)

pyoppruad s19vUd?)

SISUI]DDASUD.A] SNYIADSSOA) ye

snsourpnpvg XDIUP x

uowmauyo “FT

La 2

SAZOSAUW SIISIGLAFT ye

DIPDILANS DIVING

sisuadno ‘yo "Pr

sisuadvo xcuopr

ANNALS OF THE SOUTH AFRICAN MUSEUM

308

ueyquies,

tseAy

dT[TesoS10[O

eiofuey

IMey

INI WET)

UTE BN KO)

II reAnplo

I reanplo

Hohe

[HOweT

oulg

WeUeyy

OsIVyy

eMquInyA

[gee Sal

Jowyayy

OUMNVVEEO)

YIOMI3PUO AA,

syqieay jo “D

[ePAIELA

peqsi10] 1

PreyedoH

eIyauIO')

IOATY [BBA

UIey SIjOg

IevIPULOIy

SUBIYIIEMS

yessuedeyeyy

U9} UOFYIIIS

suney,

IsMZ0g WiNLIYJOULAT »

UOPOXO]OID IVT 5

oe ‘ds a ae

SISUI]VDASUDL] "FY ye

BUN |i

a 024 ym

SNJDIGOXA “Py

suosfiunpdqns UuopoysipyrApy

sue puvnsfo “7

DIYUD]ID DJUOPOXOT »

‘ds uopoydojosas'y.

SISUISIDY UOPOTAIS'

‘ds uoposa}s'y.

SISUAMUBY SNIUDUP

‘ds wnrwayjoyquor »

VuAdIDsoOdoudd

‘ds "OQ

snargorzan snqo1sajrcuQ)

VIVINGOITOEOL

snpoydas0pa4y7 “Jo

VICUdINNId

WiNrGOUyJa UNLLIYJOULOFT

w. oa ‘ds ‘Wx

SISUA]DVASUDL] SNPOMDYIDIA, ye

fe ‘Aou ‘ds "fy

UOPOUINA “JA

"* aqlandd UOaLaqUdaayY

309

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

MOS

Serr

Sar

Serer

atta!

baba.

bbe

“ds sn1a0ys0j0NV

SNJIDGULOD SNLBOYIOION, 4p

SN]INA SNLBOYIO{ON y.

"* Ipoomdoy ‘Nx

SISUAGDI SNABOYIOJON. ye

aVvdins

VIALOVCOILLYV

“ssnaguyquin “Fy

snagryquivjoLg *FT x

“* DINSUNSDUL * FT

sqos10s snumpododqry ».

VLNOGOONV

siusong Sosa

"+ XDIJA $0199] ATUILIG ye

“+ wuts wnrayj0yD1a")

sisuajpvasuviy (i) ‘Wx

UUUIY WNLMIYJOZIYISDIAJAY ye

BSNUDIJIULOI SNFGLYOYJVUTCANT 4

MOT ELAS SG (eS

posonbd “FT

CAEN Cle

sngooud “7 x

SISUBMVMNOP]O “T ve

1CQ9LT “FJV “Dy

wpayoing snnbay

5.0 So eliapays Shs

pISUILIg]D U0LUDGGLYOIAS x

os 2 ds avurwvddizy

VIALOVGOSSTUAd

‘ds “dx

SUSUI]DVASUDAD “Tx

DNDYUD "Tx

sasuagv? D1AvIOLT

VaACIOOVUAH

ANNALS OF THE SOUTH AFRICAN MUSEUM

310

uryquies,

seh

aT[I¥sas10[O

eiofueyy

ERS

be bebe by

AI teAnPplO

eA PTO

bbq Bah

Sbaby babe

II Feanplo

I RAO)

ToweT

weury

OSV yy

eMqunyjy

HUTT WAP fon REL

Jowyoyyy

DUMNVVEHLONV'I

YIOMIIPUO AA

syyeop jo 9

IE TSTETUN

PeqsHIO]

PryedoH

eI[QUIO')

IOATY [eV

Wide SIjog

IevIPUlOIy

SUBIYIICMS

yessuedeyeyy

UT9}UOFAINIG

sune J,

a 291]098 D1GOYO

AVGIMAVULD

snnbyup sGo1a0yr0Uunj0F »

Pa OARYScliae

UOposqCY SaPLOLIOYIOULDIOT

o o. ‘ds . Ox

sqorysv1g XAUOJSOYILE

SISUMUDUDY SNLBOYIVZUDAAL ye

snnbyuv sns90yr0j CF]

WMasDYZIABULAUL SNLIOYIOI CF]

; wosyIDl * J»

IDZUDMU SNLIOYIOJOUOLT yp

“+ snanwUsdd "WW

USMALPUD SNLIOYIONP JIA

YONU SNLIOYIOL/ VY »

IsMOpDaUl * T+

"* sngnurut snsaoyr0urqD J,

SNISNGOL SUAPYID *T x

SuapIy]D SUaPYID “Tx

snuvai4fp snsvoya0ovY J

P snlDut * Tx

snuppogos0y snsaoyr0unjog

On SOOUULY] “TAT xp

** 1UDGa4D] "WW

WUo0pasay "WV

SISUALDANP]O “T %

"+ anand SNLBOYIOSIIN

311

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

—

were

Sey WES!

~ A

n~%

nme

Kn %

nA K

xX | X

AK APM

Lagu “FT

COO AOE Tk

OL "HT

snuinba snsv.jodqizy

ee oe ‘ds oF

snpoasqvo vajag

oe ‘ds oN

pynfnsoanf “yy

wnupuniw “yy

GGL Cole

pounpas DIUNPIY »

94a] SNSDLJOUQ

ee ME PMLA:

aDIJUIQ “Ye

snuudasquja snqoy

pruuius vadvo1apdy

snjjnsand “Fy

0d “Dy

ae ae "+ snasodg snygojvy da’)

YOLD SDYILDING we

“ds sp1av010Wl0z]

WUW| ,SVAIIOOUOFT , x

** 1UOSS/IU SDLIIOOUOFT ».

WUDGDYDU "JD “Sy.

daffoa snaaoudy

xA40 sndD.LJOAND TJ.

X a oe oa oe ispouv * 7,

we oe ee sngquuas oy

"+ quojxng snydvjasv. T, x

re UES

SOLINSJI.YS SOLINSJOLIGS

AVAIAO"d

asUuatvanp}0 “JO “T

asuarwanpjo winiayzCQrT x

SULIDAT “Dy

syppivgojaups vip.

ANNALS OF THE SOUTH AFRICAN MUSEUM

312

uryquies,

iseAg

dT[IVsaS1O[O

eiofueyy

Mey

ANIL HEAAU AKO)

IIT t@ANPIO

EAP TO

IL eA YO)

[YOR e T

oulgC,

wWeUuL yy

OsIeyy

emquinyyy

[HEL VSP (Fe

Jouwypoy")

‘Al

SUMNVVEEO)

YIIMIIPUO AA,

were

syyeay jo"

ETT

PeqsHIoy.y

PreyedoH

eI[aUIO')

TOATY [RRA

ULI SAOg

TeVIPWOIS]

SUPIYIIEMS

jessuedeyeyy

UIQUOFFIOIS

Sune],

1UL004g DIUDGDYDINY »

“. ‘ds ‘O

** sJanduo) “CO »

4O(DUL SNFV.QOILC)

sndv.yoaso snxv.oas(Q

SNIYUMAS SNULIND) UOTLOF »

: SIPUDIT +) 5

SNINULOIYD) °F) ye

nous *)

SNNbYUd *1)

SnULinD} Sajavyz0uUod!)

(¢) Lajuny sndn4vag

qurajsuaqyou, “Py

19UL]IY "PW x

SNJSNGOL “PF

Ts DUDDI “WW

YIYULNIDY SNY JOOP 4

+ oe ‘ds Vage

‘ds ‘7

oe ‘ds ‘dx

suodfiqqn “$9 “q

snyoun) "2"

ETT @

snainadd "Jo "ql

SIULODNISNGUD SNISUYODUD CT »

Aa)

psiaq XCLC »

‘ds "11x

ous

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

“WLUW [9 pPoNnO Wo yey} puv viseiny jo

eUuNey UeUOg [BOISSe[D, 9Y} UIaMj}Oq UOsTIedWIOT *3 ATAV],

qwipulias *q qanoL SnSvAJOaD] DT

sisuarsas]D "ET ann1uojps DuaDvd ET

SNNUDILUNDUL °C wapnos sngosajati—

WNAUDLLIIPAUl * FT

wnuvn.fo * FT annus Uorlvgquyy

sypjuarso snuryso1ani¢g

snaapug *q UaYIDUAIA}YIS SNULYLOLBIGT

WVWWYH Td ddN0 VISVENA

[JouTIxs = y !your}xe = x]

*(gS61 ‘Aayvary J03J@ pojooI109 o1¥v sepING oy], “wep [euossad “‘1a8utg Aq

pesvyda. st playadoxy 105 3st] ay} nq ‘Z pur g ‘F saqqvy “EQ61 ‘axooD 0} Surps099x) vorpy YINOG pur Jsvy UT sozs Jv soSvTquuasse [eUNL] 9U990}sI9Tq 9} JO sisdoudg “1 AIAV],

xX aX is sa) SIsUaCDMNOP]O S220L0]2g

X |X see ge snuvnafp sapruosvyquynd x

Xe ee ee ee ee ‘ds (é) Vx

XG |IPXG Xai Xa eXaile Xe exe lac: a ae ot SUIDUGNSADUL SDILOPUIUP’ ye

é Xe pes Se a quagaoyuve * dx

Xx x oS sie "+ 2y9ad SNIDQOIDUGY T ».

X w. we ee So ‘ds O59)

XY Ss es on -* gpa0utpay *D)y

XG XE DE XG ee ee ee ws 1pU0g “De

NE Sal ie Y 20 So 3.8 SU 28]]902 *4) x

XG ee oe ee LOULIDAT ‘Dx

xX Xe XaleX ve 5 ee yumd Dy

AG X |X * 4osunaavad vIJaZ Vd V))IZV4) »

x 2 He "+ 24a77DAI snrUdLIONT

x xX | X x xX x< sre ses "+ snduppau so.aat apr

X ae a SNINULONG SNFV.AJOPVTIVAT ye

Gh || 6 os oe "+ WINLIUL $D.4990] DUNT

Xi X X ai sae BO SLUPSAGUDI SNABNYIOY ye

314 ANNALS OF THE SOUTH AFRICAN MUSEUM

These differences indicate that the fauna of Oued el Hammam is more

primitive than that of the Eurasian ‘Pontian’, thereby suggesting an ancient

biogeographical autonomy of Africa. The presence of a fauna more primitive

than the ‘classical Pontian’ does not necessarily mean that it antedates the

Pontian. However, as is indicated in the section on the geological aspects of

the sites (vide supra p. 283), the Oued el Hammam deposit lies under lacustrine

or marine deposits. These contain clearly recognized Upper Miocene fauna of

the Sarmatian or of the Tortono-Sahelian age, which were formed by the

marine transgression at the end of the Miocene. As indicated by Arambourg,

there is no possible correlation between Oued el Hammam and the ‘classical

Pontian’ which is formed posteriorly to the Sarmatian.

PLIOCENE

The Hipparion-bearing sites in North Africa and in the Nile Valley,

which are usually considered to be Pliocene, cannot be satisfactorily dated on

the basis of the fauna alone. The faunal assemblages at Tozeur, Ain el Bey

and Mascara are rather poor. Their antiquity is suggested by the presence of

Merycopotamus, Helladotherium and Rhinoceros pachygnathus. The Pliocene nature

of these sites can only be determined from their geology.

At Wadi Natrun, none of the rare fossils recovered is characteristic of a

particular period. Determination of stratigraphical relationships is necessary

for the assessment of the Pliocene date of the fossiliferous beds.

PLEISTOCENE

The simple comparison of lists of fauna from numerous sites often results

in misleading or incorrect conclusions for the following reasons:

1. The assemblages may represent a sampling of fauna from different

biotopes, and furthermore, the sampling (i.e. collecting) may have been made

by means of different methods and for differing requirements. For example,

in the case of small rodents, they may be an incidental part of a general collec-

tion recovered during a thorough investigation, or they may have been sought

for exclusively from, say, breccia by a specialist. However, they may have

been overlooked in the field when the investigator was merely collecting the

larger bones. Thus they would be absent from the collected assemblage, but

this would not reflect the fact that they may have been present in situ. Thus

someone studying or re-studying a particular collection, and being unaware

of the conditions of recovery of the material, may accurately record frequencies

but these may yet be misleading.

2. Different palaeontologists may use criteria and terminology at variance

with others, often because of a familiarity with material from particular areas

or countries. In addition the fragmentary nature of certain specimens may be

responsible for differing interpretations.

3. Just as different biotopes may express slight differences in similar

forms, so climatic differences may influence similar forms even in adjacent

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 315

regions. These differences may be interpreted on a specific rather than on a

subspecific level, thus influencing statistical analyses.

Consequently, a practical method of comparing sites within a geographical

region is by studying the results obtained by either a single palaeontologist

or by a particular institute. This requirement is partially met for the ‘Villa-

franchian’ sites of North Africa, for the australopithecine breccias in South

Africa, for Olduvai Gorge, and possibly for Laetolil and Rawi in East Africa.

It is to be remembered that the success of the investigation within each of these

three geographical units varies.

The North African ‘Villafranchian’ fauna is characterized by the associa-

tion of the latest mastodonts with primitive elephants, of the Sivatheriinae

with Giraffa, and of the hipparionids with the modern Equidae.

However, there is sufficient evidence available to suggest that such associa-

tions, found in several places in Africa, are not necessarily contemporaneous.

Unfortunately, the term ‘Villafranchian fauna’ is loosely used with a different

interpretation in different areas, but it may only be applied, sensu stricto, in

a chronological sense to the Mediterranean basin.

Arambourg (1947) proposed a relationship between the fossil-bearing

sites of Africa according to the presence or absence of archaic elements, such

as Archidiskodon planifrons, Anancus, Stegodon and Chalicotheriidae. During the

Lower Pleistocene some of these forms seem to have disappeared, being replaced

by A. recki, while other primitive groups with tertiary affinities like Stylohipparion,

Deinotherium and Libytherium persisted.

By means of the Proboscidea sequence, a subdivision of the ‘Villafranchian’

can be formulated:

A. Lower ‘Villafranchian’

Characterized by the PRESENCE of Anancus, Stegodon, Stegolophodon (?),

and the appearance (in a stratified deposit) of Archidiskodon (or Elephas sub-

planifrons, planifrons, africanavus, and exoptatus): Kanam, Kaiso, Garet Ichkeul.

B. Middle ‘Villafranchian’

Characterized by the extinction of Stegodon, the persistence of Anancus,

Mastodon, Archidiskodon (or Elephas) planifrons or subplanifrons and exoptatus and

the first appearance of Palaeoloxodon recki (or E. meridionalis): Koro ‘Toro,

Laetolil, Ain Hanech, Olduvai I.

C. Upper ‘Villafranchian’

Characterized by the presence of Anancus, Stegodon, Mastodon, and the

continued presence of A. exoptatus, E. africanavus and even Deinotherium: Omo.

The post-‘Villafranchian’ horizon commences when Anancus, Stegodon, and

Mastodon as well as primitive forms of Archidiskodon (exoptatus) and Elephas

(planifrons, africanavus) become totally extinct, coincidental with the appearance

of Loxodonta africana and atlantica as well as the further development of P. reckt.

316 ANNALS OF THE SOUTH AFRICAN MUSEUM

North Africa

In a comparison of Garet Ichkeul with Ain Hanech, the fauna of which

are substantially the same, Arambourg (1949) concluded that Garet Ichkeul

was the more ancient (Lower ‘Villafranchian’) because of the more abundant

E. planifrons and Stylohipparion at the former site, while the latter had remains of

P.. reckt.

There is good evidence that there is close identity between Ain Hanech

(and Bel Hacel) on the one hand, and Garet Ichkeul, Ain Boucherit (i.e. Beni

Foudda of Pomel) and Oued el Akrech (Arambourg and Choubert, 1957),

on the other. On the basis of the fauna alone, it does not seem possible to

compare these in any great detail with the other North African Pleistocene

Hipparion-bearing sites.

Total number of — Extinct Extant Extinct Extant

identified species Species Species Total Extinct

%o %

SOUTH AFRICA

shaving. ss Bee 34. 25 9 73°5 36-0

Sterkfontein a ae 37 26 II 70°4 42°3

Makapansgat .. iy 70 43 27 61°4 62°7

Swartkrans =i ae 39 32 5] 82 21°9g

Kromdraai se ae 4I 29 12 70°7 41°4

Bolt’s Farm Sy 40 56 35 21 62°6 60-0

Vaal River a a 35 20 15 57 - 75°71

Cornelia .. ate ae 23 17 6 74. 35°3

Hopefield ae ae 45 23 22 51-2 95°6

Florisbad Se a 32 15 17 46-9 113°4

Viakkraal. . a Be 18 9 9 50 100

Cave of Hearths te 45 10 35 22°2 350°0

Wonderwerk As al 25 & 17 32 213°0O

RHODESIA

Chelmer .. oe ao II 6 5 — —

Broken Hill as if 31 7 24 22°6 342

Mumbwa 36 4c 18 3 15 16-6 500

EAST AFRICA

Kaiso aie a ap 10 8 2 80 25

Kanam .. a: Se 16 12 4 75°0 33°3

Omo a ee sis QI 15 6 71°4 40

Laetolil .. ers Se 33 20 13 60-6 65-0

Olduvai I of aye 34 28 6 82-2 Q21°4

Olduvai II x: Re 49 38 II 77°4 29°0

Olduvai III he 8 2 18 5 78-4 27°8

Olduvai IV ¥ sie 39 30 9 76-9 30°0

Rawi ae ae sic 7 4 3 = =

Kanjera a Be 17 14 3 82°3 21°4

Olorgesailie ne Hes 9 7 2 — =

Eyasi Ne es BE 24 9 15 37°5 166-0

‘Gamblian’ BY aa 12 7 5 58:4 71°4

TABLE 3. The relationships between identified extinct and extant fauna from African Pleistocene

sites.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 317

East and Central Africa

The OLDUVAI GORGE has been extensively studied. From the four

Beds, 55 species have been identified (table 1), Bed II and IV containing

respectively 49 and 39 of them (table 3). However, recent discoveries and

more detailed analyses will considerably alter these figures. Dietrich (1942),

Arambourg (1947), Hopwood (1951) stressed the general Middle Pleistocene

character of the Olduvai faunal assemblage. It has been pointed out that no

available faunal evidence suggests that the lower part of Olduvai is of Lower

Pleistocene age. This view will probably be considerably modified when

descriptions and analyses of recent discoveries are published shortly. In this

paper, only the available published data are assessed. Leakey (1963) emphasizes

the point that ‘Bed I covers a very long period of time with a gradual change

from a very wet climate at the beginning to savanna conditions and then

subdesert ones’. It is no longer possible to speak simply of ‘the fauna of Bed

I’. He also points out that the fauna from Bed I, although ‘Villafranchian’

and older than Omo, does not conform with the Lower ‘Villafanchian’ fauna

as represented in East Africa at Kanam East and West. Furthermore, it is no

longer certain whether Bed I contains Hippopotamus gorgops, Taurotragus oryx,

Tragelaphus strepsiceros, and even Palaeoloxodon recki which is super-abundant in

Bed II and Bed IV.

Olduvai I a 56 BOM, Cave of Hearths .. 56 BROW

Swartkrans 97 So) DUG) Broken Hill. . a 22k)

Olduvai III 6 50 PH7/OKs} Wonderwerk se B19 GRP)

Olduvai II .. w3 2020) Eyasi Bi A TS

Olduvai IV se -- 30°0 Florisbad_ .. A AO)

Cornelia .. 55 50 | BAROR Hopefield .. WN BI Mug) Gey

Taung a 50 50 BHO) Vaal River ahs bo GPO)

Omo ir se a 4020 Laetolil ie ae a5) ex)e(s)

Kromdraai ae eA ThA) Makapansgat ch Bi OTA:

Sterkfontein ee a9.) GAG Bolt’s Farm. . oc ea O 25.6

Bolt’s Farm.. a a 60.0 Sterkfontein Me a] OA

Makapansgat Ae a O27 Kromdraai a 7 Ova]

Laetolil ue a0 O50 Omo be 3 Ast 71°4

Vaal River of dol GR Taung ae ft SN ROR

Hopefield .. “i od) ORS Cornelia... We BN BAO)

Florisbad_ .. ae 56 MUGOZL Olduvai IV Ey fay 7O%9

Eyasi MY is ee) L660 Olduvai II a Bia Nt afgp oes

Wonderwerk At 5.6 PUG}O() Olduvai III its ae 78°4

Broken Hill. . a 342.40 Swartkrans ae 54) BLO

Cave of Hearths .. 22) 350° 0 Olduvai I Be SG ho 200)

TABLE 5. List of sites in the order of

TaBLe 4. List of sites in the order

of increasing proportions of extant

species. Ratios (extant/extinct) are

determined only for those sites where

more than 20 species have been

identified.

increasing proportions of extinct

species. Ratios (extinct/total) are

determined only for those sites

where more than 20 species have

been identified.

9ZIDI[2}I-UoU

| aca ‘a't ‘satgads younxe JUoJayIP JO Jaquinu [20] 9y} Jo adejUIoI0d & sv passaidx9 ov S9}IS OM] 0} UOUULUOD satdads ouN)xXa Jo JaquInu ay} aur eNbyqo ayy Mojag

peztoryeyt

‘says om} Auv 3% paynuapt soloads yourjxa JUSIAYIp JO Joaquin [%j}0} IY} 9Iv soINSY paziorypeit-uou IY} a[IYM ‘sajIs OM} OF uoWNOI Satdads yOUNXa JO

Joquinu ajnjosqe 9y} 93NjsUOD sandy pozioryeyt sy sur] sNbiyqo ay} saoqy ‘sais 9UI00)sIo[q OM) Auv 3v p23zuasoidos BUN] JOUNXS VY) JO UONJaIdI9JUT [LONSHRIG “g AIAV],

6.1 | g-F | 0.G1 | z.€ | 0.Ga | 1.G1 | €.6€6 | b.g | g-b | z.z | gE | G.z nt peqs1oy,q

£.3 | 0-g | g-9 | 1-%1 | 0-91 £.9 oN ‘sedge

cP gf G.1h | 0.gS | 1-g6 | €-o1 | F-gt | g-a | L.z F.g | 1-8 | bb | Got g-1 Q-1 “+ AT 1eANplO

§ ‘

€& ca iv ¢.Ev | b.6€ | 1-G1 | 3.33 | 3-b | of 9-g | g-o1 | 1-9 | 6.1 0.3 £.z ** TIT reanplo

a ZI

re} ty €P 6& v.€G | v.62 | F.0% | €.z Bal | 6.9) gS | het Pu 9-1 O10 II teanplo

I & GC ZI

&b £& ab €€ &P 1.6% | &.0& | 6.% lex) I Sof) |] @ai2 |} @)ot Let 6.1 oe I reAnplo

4 IT &1 £6

cE Ga €P 6¢ Lb 66 g-02 | 6.6 | L.€ 6.G¢ |G.4 | o.€ g-1 on [OIe’'T

b Z G IT 6

of ba ge La a4 €& 6z g-F 6.% | g-11 | a-& che ‘+ oWIO

Z Y) 6 Or 9

1G gi gf bs tb VE gz 1G 1. | €9 | 0-g | I-11 | 1.6 "+ TUE, uayoug

I I I I I lf I |

0z gl LE Gz gh gf La Gz F1 9-gz | o.%1 | Z.o1 | 9-€1 | G.L 0-% | 0-6 "+ YIOMIOpUo AA

& I onell =o p I I

61 61 oF Bz gh ge of Gs gi bi Pastel ORO level tele Gite Ole G | ¥-o1 | 1.6 | 6.3% ** suyjaeapy jo "5

2) I b

gz 6z Lb GE bS GY ge bE Go Go ga F.61 | g9-4z | o-41 | GS.g b |} g-9 | 0.4 ‘* JOATY [eed

Z & & b & G I G & P

€& a& 6h LE 9S LY ob iad La Gz of gf 1-11] %.6 | &.g | 2.6 | G.t | G.g sie pleyodoxy

G b b G b § F & & § Z

vi gz ch €& aS th of 1& oS rete £z 6z gf g-o1 | S.F 1. |v.6 | 6.P De BIpausory

8 6 C & 6 I I 6 & b 8 B

LY bb G9 oS ol zg GG oS raid ob 66 LY €& Lb ¥.Gz | g.1% | g-Lz | 9-SE | ¥.S1 | ** wae safog

€ I I I I & 9 8 G g

ob gé 6S LY Lo LS 6h th 9& LE gé gb gh bb 1G g-a& 0% | a-3o | G.s1 | ** reeipwro1ryy

6 I § 4 G §1

gt Iv 19 6h 69 6S aS LY 66 oF ov oS aS gh GG oh b.G1 | g-1€ | g-1n | ** = sueszyqaemg

I I I [ I G G § [ al SI

9S aS €L 19 1g iL z9 gS oS oS gh 6S G9 gs 19 09 G9 g-4a | a-L1 | ** yedsuedeyepy

C lt I G B I é LI él OL

ob GE GG &F £9 €¢ ob 1p 6& €& 6& &P gb 1b GY ch 44 L4e) b.bo | ** urayuozyi9ajig

I I I lf I I & § § é OT; Or PI ST

ov bE GG &b £9 €S GP oF aE €& FE Gt gt oh aS gh 1G gS 1b DiC *+ Zuney,

It 8 9 9 Or Or

tr is3) iS e~] ee) n a n

Be ae eee ee eee Ve ee er

ee ee eee pee fe cae fe se

» E. BE. E. E. = a) a @, eh a 5 q 3D 5

= BS lee A fete eae Bie a Se ee a

< fey = a, 8 4 5 me 8 5

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 319

OMO has classically been equated with Olduvai I and II but recently a

proposal has been made that it is more similar to the base of Olduvai IT (tables

4, 5). The extinct fauna common to Omo and Beds I and II are 30:3% and

20°4% of the total respectively (table 6). There is no substantial difference

between the faunal assemblages except for the presence of Anancus and hominids

at Olduvai. The absence of <injanthropus and ‘Homo habilis’ at Omo, does not

per se imply that Omo is younger than Bed I, because the Olduvai Lake shores,

permitting the possible development of living sites, favoured a concentration

of hominids.

The SERENGETI (Laetolil) fauna is closely related to beth Olduvai and

Omo. The extinct species at Serengeti in common with Olduvai I are 23:1%

with Olduvai II are 23:4%; and with Omo are 20:6% (table 6). The essential

difference seems to be the presence at Laetolil of Anancus and Archidiskodon

subplanifrons, both of which are absent at Omo, and also the presence of Meta-

schizotherium hennigi (which is also present at Kaiso). Leakey (1958) believes

that the differences between the Laetolil and Bed I faunas do not imply a

temporal separation but rather illustrate a difference in ecological conditions

existing contemporaneously. He indicates that it has not been proved that

Laetolil is older than Olduvai I. Furthermore, it should be pointed out that

Laetolil probably contains a mixed assemblage of different faunal stages, as

was already suggested by Dietrich.

KAISO and KANAM are badly represented from a faunal viewpoint:

10 (8 of which are extinct) and 16 (12 of which are extinct) species respectively

have been identified. In spite of the fact that information about their recovery

is unsatisfactory, there is no doubt that a real archaic character is attached

to these sites, from which Chalicotherium, Stegodon, Anancus, Stegolophodon and

Hippopotamus imaguncula have been recovered. These forms are considered by

most authorities to represent the most ancient East African Pleistocene fauna

known at present.

KORO TORO consists of five ‘apparently equally old’ deposits which

contain about 30 identified species of mammals. Abadie, Barbeau and Coppens

(1959) and Coppens (1960), studying the fauna of the lowest level, conclude

that it belongs to the Lower ‘Villafranchian’ because of the contemporaneity of

Mastodon, Stegodon and Elephas. However, the rarity of both stegodont and

mastodont (more typical of the base of the Lower Villafranchian and the

Pliocene) and of P. recki (more typical of the Kamasian), and at the same time

the abundance of E. africanavus suggest that we are dealing with an intermediate

stage between the very base (e.g. as at Kaiso where africanavus is present without

recki) and the top (e.g. as at Omo where recki and africanavus are abundant

but stegodont and mastodont are absent).

Kent (1942) and Arambourg (1943, 1947) emphasized the great similarity

between the Olduvai and the Omo faunal assemblages, and indicated that

they considered the Omo fauna to be slightly older. Arambourg suggested

that the various East African deposits may be considered to correspond to

320 ANNALS OF THE SOUTH AFRICAN MUSEUM

different phases of the transition between the Upper Pliocene and the Lower

Pleistocene: the Serengeti, Kaiso and Kanam tuffs being the most ancient

and being the equivalent of the ‘Villafranchian’ horizons in North Africa

and Europe, while Omo and Olduvai I should be considered as being younger

and corresponding to the first true Pleistocene deposits. This view has been

partially modified because of the recent discoveries at Olduvai. Leakey (1963)

describes a major climatic, faunal and geological ‘break’ near the base of

Bed II, the lowest part of which (overlying the ‘marker bed’ at the top of Bed

I) he considers to be of Upper ‘Villafranchian’ age and comparing very

closely with that of Omo. According to Howell (1959), ‘the faunas (from the

Villafranchian sites of Central and East Africa) differ somewhat in composi-

tion, that from Laetolil beds being probably the youngest, overlapping basal

Olduvai and that from Kaiso being perhaps the oldest. The Omo fatina overlaps

both Laetolil and Kaiso and that from Kanam is probably broadly equivalent.’

South Africa

The fauna of the TRANSVAAL cave breccias (Taung, Sterkfontein,

Makapansgat, Kromdraai and Swartkrans) has been extensively studied.

The fauna of these sites show a very high degree of similarity: 11-°8—32-6%

of the total number of extinct species are common to the different sites (table 6).

A similar range (15:4—35°6%) has been calculated for the BOLT’S FARM

faunal assemblage found in the vicinity of Sterkfontein. The six sites show an

extinct/total ratio of species of 61-74% (table 4) and an extant/extinct ratio

lower than 63°% (table 5), which indicates a considerable antiquity. The

conditions of accumulation at and the geology of these sites are also similar.

On the basis of the fauna (Ewer, 1957), as well as of mineralogical and climato-

logical studies of the breccia (Brain, 1958; Robinson, 1961), a relative age

sequence has been derived, namely, Sterkfontem—Taung, Makapansgat,

Swartkrans and Kromdraai, extending from the Lower to the early Middle

Pleistocene.

HOPEFIELD and FLORISBAD faunas seem to have rather similar

frequencies: not only are 15:1°% of the total of extinct species common to

both sites, but their close faunal relationship is expressed by a similar ratio of

extinct/total species (51°% and 47% respectively). However, the extant/extinct

ratio (viz. 96% and 113°%%, respectively) suggests a greater antiquity for Hope-

field. Furthermore, there are considerable differences in the types of hominids

and artefacts recovered from these two sites. Although Hipparion has not been

recovered from these two sites, they are included as important Middle—Upper

Pleistocene sites.

From purely a consideration of the fauna, the YOUNGER VAAL RIVER

GRAVELS and CORNELIA could fit satisfactorily in the chronological

sequence between the Transvaal cave breccias and Hopefield—Florisbad with

the following intermediate ratios (tables 4, 5): 57-74% of extinct species;

35-75% of extant/extinct species; and 27-6% of all the extinct species being

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 321

common to both sites (table 6). According to Cooke and Wells (1946) ‘while

it is possible that some of the living species may post-date the Younger Gravels

phase of deposition, it appears most probable that the bulk of their material

forms a fairly coherent whole, representative of the fauna of the later part of

the first wet phase’. The fauna from the Younger Vaal River Gravels and

Cornelia deposits could be broadly considered as a Middle Pleistocene fauna,

especially in the light of the recent reinvestigation by Cooke and Wells who

indicated that the Cornelia fauna is comparable with the Younger Vaal

River Gravel material (unpublished; quoted by Cooke, 1963, page 96; see

also Wells, 1964).

Although the Lower Pleistocene sites of East and of South Africa have

ratios of extinct species higher than 60% and an extant/extinct ratio lower

than 70%, direct comparisons between East and South Africa have severe

limitations. From an analysis of the fauna (table 1), based on the lists drawn

up by Cooke (1963), it appears that not a single species as recorded is common

to both East and South African Lower Pleistocene sites. We believe that this

lack of relationship is really less marked than indicated because different

names have been given to the same species in East and in South Africa. This

is mainly because the taxonomy in the two regions has been developed by

independent investigations, e.g. the species of Simopithecus (Freedman, 1957;

Leakey and Whitworth, 1958; Singer, 1962). To some extent this shows the

poor state of our knowledge of the African fauna as a whole. Furthermore,

Pickering (1960) warned that a critical approach will always be required

when comparing a plains fauna, such as that found at Olduvai, with an assem-

blage obtained from a cave deposit which is usually the case in the Lower and

early Middle Pleistocene of South Africa.

There is some affinity between the Younger Vaal River Gravels, Cornelia

and Hopefield assemblages on the one hand, and the East African early Middle

Pleistocene collections on the other, ranging from 2:9-11:8% of the total

number of extinct species for any two particular sites among these groups

(table 6).

CONCLUSIONS BASED UPON THE FAUNAL AND GEOLOGICAL EVIDENCE IN THE

LITERATURE

The large amount of data discussed in previous chapters may be utilized

for the construction of a tentative illustration of the relationships between the

sites at which Hipparion has been discovered in Africa (fig. 8). By and large this

supports the views expressed by Cooke (1963) and Bishop (1963).

CHRONOLOGICAL RANGE OF Hipparion

Apart from the actual dating of the geological deposits from which Hipparion

has been recovered in Africa, there is the general question of its first appearance

and its ultimate disappearance on the continent.

It seems that Africa was not the area of origin of hipparionids, as the

322 ANNALS OF THE SOUTH AFRICAN MUSEUM

NORTH AFRICA EGYPT oe arate | eee AFRICA SOUTH AFRICA

Se & RHODESIA

eA Vlakkraal

Wonderwerk

Chelmer Florisbad

Olorgesailie B.Hill Hopefield

Kanjera

4 i Cornelia

Ternifine : Vaal

Kromdraai

Swartkrans

PLEISTOCENE

Ain Hanech

Bel Hacel Sterkfontein ext.

Makapansgat

Sterkfontein

Taung

Koro Toro

—-'-0-00r-

Ain Boucherit| Garet

Fouarat| O.e. Akrech |Ichkeul

Langebaanweg

PLIOCENE

Wadi

Natrun

O.e.Hammam

Marceau

Fic. 8. Tentative correlation between Hipparion.sites in Africa. A few other sites have been

incorporated for reference. B. Hill = Broken Hill; O.e. = Oued el; A.e. = Ain el; Gamblian =

Gamble’s Cave; Bolt’s = Bolt’ s Farm.

evidence points to probable migrations from Eurasia. Nevertheless, it is some-

times claimed that Africa contains the evidence of the early, if not the earliest,

existence of Hipparion in the Old World.

At the other end of the chronological range, the hipparionids survived

much longer in Africa than in any other area known at present. With very

few exceptions, most of the sites where African Hipparion have been located

are of Pleistocene origin, i.e. a period when they no longer existed in Eurasia

and in America.

Consequently it is considered necessary to compare the geological range

of occurrence inside with that outside Africa.

UPPER LIMIT

According to available information, the upper limit of Hipparion outside

Africa seems to correspond roughly to the Upper Pliocene. It is generally

accepted that the Villafranchian is defined by the appearance of some modern

genera, among which is Equus. However, there is no reason to deny, a priori,

a coexistence of Hipparion with modern Equus which developed on a parallel

line from Pliohippus. Such co-existence has actually been demonstrated for

Africa. Outside Africa, Equus not only heralds the Pleistocene, but it practically

replaces the more primitive Hipparion in its habitats.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 323

AMERICA| CHINA INDIA MIDDLE- S.E. CENTRAL & AFRICA

EAST EUROPE W. EUROPE

Azov Sea Val d’Arno, Sénéze

(Khopry)

Roussillon

Gédallé, Berechti

Malouchteni

Maragha

Pikermi | Polgardi, Vélles

Taraclia Samos Mont Lubéron

Fic. 9. A diagrammatic representation of the probable chronological range of Hipparion in

different continents.

Kanjera

Cornelia

z Vaal 1°)

w |

re) d

oO Koro Toro u

= v

7) a

a Ain Hanech Omo i

= Laetolil

Ain

Fouarat,Garet Ichkeul, Boucherit

Kanam, Kaiso

Pinjor

Patrot

Wadi Natrun

A.e. Hadj Baba

Hemphill

PLIOCENE

Mascara

Upper- Dhok Patan

Clarendon

Vallés - Penédeés

Rhéne-Valley

Teruel

Sebastopol

Odessa

Pao-Te (Red Clays)

Qued el Harnmam

Marceau

Eppelsheim

UPPER-MIOCENE

The youngest formation bearing Hipparion remains in America is the

Blanco formation (Upper Pliocene), where its presence is even questioned by

some palacontologists. Even if accepted, remains are very scarce.

In China, the latest occurrence is in the Nihowan deposits of the Sang

Kan Ho Valley, east of Pekin (fig. 9); in India, in the Tatrot and the Pinjor

zones of the Siwaliks, while the records are still very insufficient for Mongolia.

In the U.S.S.R. (including its Asian portion), the most recent occurrence is to

be found in the Azov Sea shore deposit at Khopry; in South and Central

Europe at Berechti and Malouchteni in Rumania, and Goddll6 in Hungary.

In western Europe, Hipfparion has not been found in the Villafranchian deposits

of the Val d’Arno and Senéze; and it is rather uncommon in the Roussillon,

although it is still found in Perrier where it is rather exceptionally associated

with Equus.

Without going into details which are irrelevant in the framework of this

paper, it may be stated that outside Africa, Hipparion does not seem to have

extended into the Pleistocene. In many instances, it was already becoming

rather scarce during the Middle Pliocene and definitely more rare in the

Upper Pliocene.

LOWER LIMIT

The lower limit of Hipparion, or the time of its first appearance, presents a

more difficult problem, as it is not at all easy to correlate the chronological

324 ANNALS OF THE SOUTH AFRICAN MUSEUM

interpretations of all the localities. In addition, a different faunal basis (marine

or continental) is applied to various areas, and a direct comparison of the

fossil associations is not always possible. Furthermore, some geological terms,

like ‘Pontian’, have been used in a different context and with a different

meaning by various authors, e.g. as a facies, or a stratigraphical or a faunal

horizon. Consequently it is especially difficult to appreciate the meaning of a

particular statement, factual though it may be, without danger of misinterpreta-

tion. Therefore, the soundest approach may be, first, to locate the earliest

occurrence of Hipparion, within a region, and then, to correlate as far as possible

the interregional data.

In America, Hipparion is not found in the Barstov formation, appearing

first in the Clarendon formation (fig. 9). It is also found right at the base of

the Mint Canyon formation, and in the Hemphill formation. The upper

Clarendon and the Hemphill formations are unanimously considered as ‘Lower

Pliocene’ and correlated with the European ‘Pontian’. But for more than thirty

years, the Mint Canyon has been a major topic of discussion and argumenta-

tion. While most scholars follow Stirton’s opinion (1939) and locate the conti-

nental basal horizon of Mint Canyon in the Pliocene, Maxson (1930), among

others, basing his opinion on the debatable malacological fauna of the overlying

marine Cierbo beds, considers it to be middle Upper Miocene age.

In China, Hipparion appear in the red clays of Chan-si, Chen-si and

Kansou. These Pao-Te formations have not been properly subdivided. At

different times, investigators have pushed them back to the Upper Miocene

(Teilhard and Young, 1931), or restricted them to the Lower Pliocene (Teilhard

and Leroy, 1942).

In India, the first Hipparion are recovered in the Chinji zone of the Siwaliks,

which is referred either to the Middle (Pilgrim, 1938) or to the Upper (Lewis,

1937) Miocene, or even to the Lower Pliocene (Colbert, 1935).

In the U.S.S.R., the earliest occurrences are those of Moldavia and Odessa

and the Sebastopol fauna (Borissiak, 1914). Russian geologists refer the former

deposit to the Middle Sarmatian, and the latter two to the Upper Sarmatian,

i.e. the Upper Miocene. Similar information was recently obtained from the

Upper Sarmatian in the Istanbul vicinity (Chaput and Gillet, 1938; Yalcinlar,

1952).

In central and southern Europe, because of the exceptional associations of

H. primigenium with the typical Miocene Anchitherium, Eppelsheim might be

considered as one of the first known areas of occurrence of the genus. But in the

Rhone Valley (Denizot, 1939), at Vallés-Pénédés in Catalogne (Villalta and

Crusafont-Pairo, 1946, 1947, 1948) and in the Teruel Basin (Sondaar, 1961),

abundant remains of Hipparion have been recovered in deposits which are

dated as Tortonian. Maragha, Pikermi, Samos, Polgardi, Baltavar, Vélés

and Mont Luberon, referred to as typical ‘Pontian’ sites, are probably somewhat

younger, and should be placed in the Lower Pliocene.

In spite of many unsolved problems in correlating these sites, it seems

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 325

difficult to deny that there is in North America, in Asia and in Europe constant

and repeated indications of the appearance of Hipparion in the Upper Miocene

(fig. 9).

GENERAL DESCRIPTION OF Hipparion TEETH

UPPER DENTITION

The pattern of horse teeth has been described many times and detailed

structures have been extensively discussed in previous publications. The

distinctive characters of horse teeth that have been described are based on

the original description proposed by Osborn (1907). In 1918, he successfully

applied this description in his study of the North American Equidae. Subsequent

UPPER TOOTH

Fic. 10. Features of the occlusal surface of an upper molar tooth of hipparionids. Abbreviations:

B. prf—bouclé prefossette; Eclo—ectoloph; Hy—hypocone; Hyg—hypoglyph or hypoconal

groove; Hys—hypostyle; Me—metacone; Mel—metaconule; Melo—metaloph; Mes—meso-

style; Mets—metastyle; Pa—paracone; Pas—parastyle; p. c—pli caballin; p. hys—pli hypo-

style; p. plo—pli protoloph; p. prf—pli prefossette; p. prl—pli protoconule; p. ptf—pli post-

possette; Pr—protocone; Prf—prefossette; Prg—preprotoconal groove; Prl—protoconule;

Prlo—protoloph; Ptf—postfossette.

326 ANNALS OF THE SOUTH AFRICAN MUSEUM

alterations and additions have been proposed by, among others, Stirton (1941),

Arambourg (1947, 1959), Gromova (1952), Hopwood (1937) and Cooke

(1950). It is not proposed to make any further modifications here, but it is

necessary to comment on the dental nomenclature, without any phyletic

implications, so as to outline the basis adopted in this monograph.

Typically three crests are recognized. A mesiodistal ectoloph, joining

paracone and metacone on the buccal surface, an anterior protoloph and a

posterior metaloph. The latter two are more or less transverse in the primitive

condition but in advanced Equids they are half-moon-shaped. Being lopho-

dontic specializations of the protoconule and the metaconule, they build the

protoselene and the metaselene, and they meet the ectoloph at parastyle and

mesostyle, respectively (fig. 10).

Lingually deflected from the main selenic lophs, but more or less attached

to them there is a protocone and a hypocone. Their rather deep bordering

grooves, filled with cement, tend to isolate them from the crests. These grooves

are the pre- and post-protoconal grooves (valleys, sinuses) which lie anteriorly

and posteriorly to the protocone, respectively, and the hypoconal groove

(sinus) or hypoglyph which is related to the hypocone (the posterior being

usually very well marked, and the only one noticeable). Sometimes the protocone

shows a ‘spur’: the ‘protoconal spur’.

When the pre- and post-protoconal grooves are maximally deepened,

they become confluent in a medivallum or internal depression. Thereby they

produce complete isolation of the protocone. This is one of the major and

characteristic features of the upper molars of Hipparion.

Elevations of the cingulum have also been described: parastyle, mesostyle

and metastyle along the ectoloph; hypostyle on the posterior surface which

ultimately develops a cusp, seemingly independent of the cingulum.

The protoloph and the metaloph crests enclose, more or less completely,

the pre- and postfossettes. The more constant and deeper plications in the

enamel wall of the fossettes (‘marks’) and also isolated ‘horns’ have been

assigned special names: anteriorly, the pli protoloph and the pli postfossette,

respectively; posteriorly in the anterior mark, the pli prefossette and the pli

protoconule which isolates a ‘boucle préfossette’ or ‘prefossette loop’ (Stirton,

1955), while posteriorly in the posterior mark is the pli hypostyle (fig. 10).

Furthermore, a pli caballin appears constantly in the post-protoconal

groove or in the internal depression. It is a lingual extension from the outer

border of the selene, and eventually, after wear, it displays a two- or threefold

division.

LOWER DENTITION

A distinction is commonly drawn between a mesial trigonid, with buccal

protoconid and lingual metaconid, and a distal talonid, with hypoconid,

entoconid and hypoconulid, separated on the buccal aspect at the level of a

fairly constant external depression (fig. 11).

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 327

LOWER TOOTH

K & as ae

Fic. 11. Features of the occlusal surface of a lower molar tooth of hipparionids. Abbreviations:

Ecsd—ectostylid; End—entoconid; Ensd—entostylid; Enfd (Pid)—entoflexid (Post. int.

depression); Exd—external depression; Hyd—hypoconid; Hyld—hypoconulid; Med—meta-

conid; Mefd (Aid) —metaflexid (Anter. int. depression) ; Mesd—mesostylid; Pasd— parastylid;

Prd—protoconid; Prsd—protostylid; Ptysd—ptychostylid.

The stylids on the buccal side are protostylid, an inconstant ectostylid

and a ptychostylid, while on the lingual aspect there are the metastylid and

the entostylid.

On unworn teeth, a crest, the paralophid, can be distinguished extending

from the protoconid to the protostylid in the anterolingual corner.

On the lingual aspect there are two prominent internal depressions or

invaginations, the metaflexid anteriorly and the entoflexid posteriorly.

The metastylid and the metaconid are two rounded formations joined by

a narrow isthmus, giving the appearance of a bow tie and forming the so-called

‘double knot’.

SUMMARY OF THE CHARACTERISTICS OF HIPPARION TEETH

Based on Gromova’s detailed description (1952, pp. 70-6), the following

characteristic features are noted.

328 ANNALS OF THE SOUTH AFRICAN MUSEUM

Upper dentition

1. Less hypsodont. But note that the more recent (African and American)

Hipparion are hypsodont.

2. Pli protocone is present.

3. Enamel plications are usually more developed.

4. There is a rather high percentage of open marks.

Lower dentition

1. Relatively less hypsodont.

2. Double knot:

(a) ‘caballus’ type in African Hipparion.

(b) ‘stenonis’ type in American Hipparion, and also in some Chinese.

External depression: this is rather profound in the Miocene forms.

4. Posterior internal depression: this is elongated, curved anteriorly and

lingually. Hence its oblique and ‘broken’ appearance.

5. Anterior internal depression: not only does this have an antero-external

angulation (like in Equus), but it also has a postero-external angulation,

with long sharp ‘horns’ directed towards the buccal aspect.

At the posterior extremity there is a deep invaginated plication.

6. ‘Talonid is bifid on Mg.

7. Anterior depression often shows plications at both extremities.

8. There is a tendency to build stylids (proto-, ecto-, hypostylids) in milk

and permanent molars.

REVIEW OF ADDITIONAL ENAMEL ELEMENTS (STYLIDS)

OF THE LOWER TEETH

DESCRIPTION

Ectostylid=Ectostylid of Gromova (1952), Arambourg (1959), Hopwood

(1937), Sondaar (1961).

=Protostylid of Osborn (1907), Stirton (1941).

—=Buitestyltjie of Van Hoepen (1930).

# Ectostylid of Osborn (1918), Sefve (1912, 1927).

This stylid arises from the basal cingulum. It is an accessory external

column or pillar that is always independent and located on the buccal side,

close to the mesio-vestibular border of the hypoconid and in the external

groove between hypoconid and protoconid, i.e. between talonid and trigonid.

In early wear it is seldom apparent on the occlusal surface because it rarely

reaches more than a short distance above the crown-root junction and it is

usually embedded in very thick cement. It can usually be observed better

on the buccal aspect where the cement is thinner.

Ptychostylid =Ptychostylid of Arambourg (1947).

=Pli caballinid of Stirton (1941).

=Ectostylid of Osborn (1918).

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 329

This is not an independent pillar, but a fold on the mesial border of the

hypoconid, in the interlobar angle on the buccal side. It may be present when

the ectostylid is present, but there does not seem to be any relationship between

these two formations.

Protostylid = Parastylid of Gromova (1952), Van Hoepen (1932), Stirton (1941).

=Protoconid fold of Cooke (1950).

#Protostylid of Osborn (1907), Stirton (1941).

This is a broad laminated pillar, compressed mesiodistally, found close

to the protocone where it is located on its antero-buccal side. With increasing

wear it fuses fairly rapidly with the protocone. This stylid also arises from the

cingulum and it may develop into an independent element on the antero-labial

angle of a tooth.

Aypostylid

This is a postero-external plication, appearing as a raised element on the

talonid. Sometimes it may develop into an independent pillar facing the

talonid.

Entostylid

This is a rare and inconstant enamel formation in the entoflexid. It may

be completely isolated in the cement (pl. 8, A) or connected to the mesio-lingual

corner of the entoconid.

FREQUENCY

Independent of the degree of attrition, which may tend to hide the presence

or the development of these formations, the stylids are very differently developed

in various genera, species and even in individuals. The stylids may only be

observed on some of the teeth of a particular jaw, and occasionally they are

found to be present only on one tooth in a complete dentition. Observations

have been made by various authors since the time when these stylids were

first described by Gaudry (1862) and Weithofer (1888) on the milk and

permanent teeth of Hipparion mediterraneum from Pikermi.

According to Gromova (1952), these additional elements are usually

poorly developed in the non-hipparionid equids. The stylids are best developed

in the Hipparion group in which they are always present on the milk teeth,

but to a varying degree on the permanent teeth. In this respect the African

Hipparion are the most progressive.

Ectostylid

Statistical analysis of Hipparion elegans provides a frequency of 12%

ectostylids on the premolars and 1° on the molars, and in H. moldavicum

frequencies of 7:°5°% and 6-7% respectively. Sondaar (1961) presents the

following data on the presence of ectostylids:

330 ANNALS OF THE SOUTH AFRICAN MUSEUM

H. periafricanum: o

H. concudense aguirrei: 0

H. gromovae: 1 in 200 specimens

H. concudense: 5 in 84 specimens

H. koenigswaldi: 20 in 150 specimens

H. primigenium: 23 in 80 specimens

Therefore it can be concluded that the ectostylid is not quite a rarity

in Eurasiatic Hipparion permanent teeth, especially on the premolars. Neverthe-

less, the ectostylids appear to be much more frequent in the later forms of

Hipparion, i.e. in Africa, where it has often been stated to be ‘a constant feature’

(Arambourg, 1956, 1959). Comments on this will be given below. Furthermore,

it can be stated without dubiety that the ectostylids have been commonly

observed in most of the specimens from Pleistocene deposits: not only are

these stylids very frequent, but they are occasionally strongly developed when

they reach along the whole length of the crowns of the teeth.

Ptychostylid

This is the external hypolophid fold which is quite common among all

hipparionids, reaching a maximum development in Neohipparion eurystyle Cope

found in America, in which as many as four may occur. These stylids do not

arise from the cingulum and cannot be considered as true stylid cusps. There

does not seem to be any direct association between these stylids and ectostylids.

Protostylid

According to Sondaar (1961), this stylid is almost a diagnostic feature of

the genus Hipparion even though it varies quite considerably in its extent. It is

often a fold attached to the protoconid, but it may develop into an isolated

pillar. Here again, the African Hipparion seem to present the most progressive

features: it is found in them as early as the Upper Miocene (i.e. H. africanum)

and as late as Notohipparian namaquense, in which it is a very tall and isolated

pillar.

A high frequency is also found in some Eurasiatic forms, namely, in H.

elegans 75°5°% is noted in P,—P, and 80% in M,—M,; in H. moldavicum 88% is

noted in P,—P, and 96% in M,—-M,

Sondaar (1961) states that it is a common, although not constant, feature

in all the Spanish species of Hipparion, except H. truyolst. Sometimes a double

protostylid has been observed, and this has been especialy noted in H.

koenigswaldi, in which there is a high frequency of ectostylids (vide supra).

In connection with the American material, Stirton (1942), who used

‘parastylids’ for “protostylids’, points out that ‘the statement that the parastylid

does not appear in the American Hipparion is not supported by the evidence.

Though the isolation is not as persistent nor as complete in the New World

forms as in some Eurasiatic Hipparion (H. platyodus Sefve), it does appear in

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 331

early stages of wear in some teeth. This is true not only in Neohipparion, but in

Nannippus and Pliohippus, though extremely rare in the latter.’

HAypostylid

This is very frequently observed in Equus stenonis and in zebra but it is

very rarely seen in Hipparion. However, it is sometimes found in the African

forms, the most typical example of which is a M, from Omo (Joleaud, 1933).

EVOLUTIONARY HISTORY OF THE STYLIDS

The development of these stylids seems to be influenced by a common

factor: a higher frequency of ectostylids in a particular group is correlated

with a more precocious appearance of the protostylid (i.e. a higher protostylid)

in the same population, and vice versa.

In a particular group it is noted that the protostylid seems to be more

frequently, more strongly and more permanently developed than the ectostylid.

There seems to be an evolutionary pattern in the tendency to build stylids,

from Merychippus (possibly from Parahippus which already shows a weak proto-

stylid) through Miocene times up to Hipparion. Arambourg (1959) notices

that this ‘tendency to develop cingular formations’ is already present in the

Upper Miocene H. africanum from Oued el Hammam. The tendency seems to

have developed further during Pliocene and early Pleistocene times, being

most obvious in the African Hipparion, in which the most recent forms show the

strongest development.

Protostylids and ectostylids appear on the milk teeth of Merychippus in

which they show slight development. They are constantly present and show

a fair degree of development on the milk teeth of all Hipparion, and they appear

on the permanent dentition of the African Pleistocene Hipparion.

Along another evolutionary line of the Equidae, the Pliohippus—Equus

sequence, it has been shown that protostylids and ectostylids remain weakly

developed, the hypostylid being temporarily more developed, as, for example,

in Equus stenonis. With the appearance of E. caballus, all the stylids have

practically disappeared.

EVOLUTIONARY SIGNIFICANCE OF THE CONES AND STYLIDS

In a phyletic perspective, it has been established that the additional

stylids have originated from enamel buds on the basal cingulum of brachyodont

teeth. They developed first on the milk dentition and only later in the evolu-

tionary sequence, and in particular groups, did they become permanent

features of the adult dentition, appearing initially on M, and M, and eventually

on the premolars. For example, it is known that the parastylid was present

in Parahippus, but only on the milk teeth as an ill-developed feature. Later it

became a constant and developed characteristic, even of permanent teeth, in

Merychippus and later equids of the same phylum. The ectostylid is present in

Merychippus, but only on the milk teeth, and then only occasionally. It becomes

332 ANNALS OF THE SOUTH AFRICAN MUSEUM

an element of the permanent dentition in Hipparion, where it appears to be

constant in some of the latest representatives of the group, i.e. in the Pleistocene

forms of Africa.

The tendency to raise enamel buds and isolate them is thus an early

ontogenetic trend, which in evolution has progressively influenced the later

(adult) stages of individual development.

Later in this paper it is suggested that this trend was necessitated by the

special architecture of the lower dentition with its characteristic development

of highly individualized conids.

The same trend, both ontogenetic and phylogenetic, seems to be recogniz-

able in the formation and the isolation of the major cones in the upper dentition.

It is really remarkable that the separation of the protocone and the hypocone

from the main lophs is a gradual process, both ontogenetic and phylogenetic.

The isolation of the protocone is rather ancient, and in the permanent teeth

of Hipparion it has extended right down to the base of the crown. Contrariwise,

the isolation of the hypocone is hardly noticeable on the permanent teeth,

but in very early stages of wear it has been observed (vide infra p. 373). However

in milk teeth the isolation of the hypocone reaches a much more characteristic

degree in similar hipparionid groups, e.g. in the South Serengeti Hypsohipparion

and in the Langebaanweg specimens (see p. 368).

The difference between the expression of cones and stylids in milk and

permanent dentitions on the one hand, and the more explicit individualization

of these features among later representatives in some phyletic lines, on the

other hand, seem to be linked with the increasing hypsodonty of the Equidae

teeth in Upper Cenozoic times.

ECOLOGICAL CONSIDERATIONS

Gromova (1952) outlined the ecological significance of the development

of the styles (stylids). She promoted the idea that they played a role in strengthen-

ing the tooth so as to meet the heavier requirements of coarser food in a dry

country where the grass was becoming very tough. The presence of additional

pillars and folds probably increases the trituration power of the teeth and

their resistance to pressure forces. Thus their appearance on the milk teeth

would prevent or decrease rapid wear.

Furthermore, Gromova developed a very suggestive correlation between

elongation of the protocone, development of the stylids, thickening of the

enamel, reduction of the external groove and higher hypsodonty. These

functional features are different methods of improving the efficiency of a tooth

with increased power of trituration. Consequently it was suggested that the

styles were developed as an adaptation for improving the grinding of coarser

food in drier climates.

Sondaar (1961), in turn, described a similar correlation among the

Spanish Hipparion which have the highest frequency of stylids. In 4.

koenigswaldi the maximal development of plications in the upper molars increased

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 333

the enamel surface with a double hypoconal groove, while the lower teeth

display wavy internal depressions—entoflexid and metaflexid. However, H.

elegans, with the same functional features, only displayed a very high frequency

of protostylids. Most of the African Hipparion developed the stylids and a

maximal hypsodonty, but not the extreme plications. There seems to be a

balanced compensatory effect between stylids, plications and other elements

so that, dependent on the specific region, one or other of these elements would

develop to a varying degree. In America, Hipparion adapted to a hard and

tough grass by an increased hypsodonty and a thickening of the enamel. There

can be no doubt that numerous plications and deep invaginations of the

enamel pattern, as well as elevations of folds and stylids, provide the teeth with

a more efficient trituration surface, and are very suitable for herbivorous

animals in an area of increasing aridity (Stirton, 1941).

TAXONOMY BASED ON STYLIDS

The observation of the presence of stylids, both proto- and ectostylids,

provided a new basis for generically differentiating the various African

hipparionids. Haughton (1932) decided to create the genus WNotohipparion

because of the appearance of these stylids. Van Hoepen (1932) followed the

same line of reasoning and erected the genus Stylohipparion, a distinction

supported by Joleaud’s publications of S. libycum from Omo (1933).

To some extent Dietrich (1942) followed the same tendency and, on the

basis of the presence or absence of ectostylids, he divided the Serengeti LOWER

cheek teeth into two groups. The group without stylids he termed Hypso-

hipparion and he referred those with stylids to Stylohipparion. However, Dietrich

did not attach a true taxonomic value to this separation because he found it

impossible to distinguish between the UPPER teeth of the two groups. He

indicated that the frequency and development of styles (-ids) can be influenced

by environment or selective pressure. Consequently he suggested that Stylo-

hipparion might not represent a true genus, but only a variety formed under a

strong developmental pressure (‘Entwicklungswucht’).

However, outside Africa, the variation in the development of stylids did

not lead to such taxonomic differentiation. Arambourg (1947) questioned the

validity of a generic difference (Stylohipparion) based upon the presence of a

feature which did not seem to separate clearly the Eurasiatic Pliocene forms.

Nevertheless, he opined that the very high frequency and the strong develop-

ment of the ectostylid in African forms supported the validity of the generic

differentiation.

Gromova also discussed this matter. She did not consider a true generic

differentiation of African hipparionids valid. Ectostylids are not an entirely

new characteristic in them exclusively but only represent a further development

(either in frequency or structure) of a feature not at all exceptional (albeit

irregular and even rare) in Eurasiatic Hipparion. Gromova proposed a sub-

generic rank Hipparion (Stylohipparion) for the African group on the basis of

334. ANNALS OF THE SOUTH AFRICAN MUSEUM

obvious peculiarities of this geographical unit. However, she writes of these

ectostylids: ‘the importance as a diagnostic feature and generic character has

been somewhat exaggerated’.

This criticism can be expanded (see pp. 387—92) and even the sub-generic

status of Stylohipparion can be questioned. In spite of the ‘air de famille’ (Aram-

bourg, 1959) of all African hipparionids, the major reason for placing the.

Pleistocene African hipparionids in a special genus or sub-genus is precisely

the presence and constancy of a very high frequency (which the present authors

doubt) of the ectostylids. Although this is generally accepted, it is a highly

debatable basis of separation. As mentioned above, Dietrich (1942) described

a very large collection of lower teeth without ectostylids and placed them in

the ‘genus’ Hypsohipparion. However, Arambourg (1947) suggested that these

teeth had ‘erroneously’ been attributed to a hipparionid and that, in fact, they

belong to E. zebra. This suggestion tends to remove Hypsohipparion from the

hipparionid scene so that the ‘frequency’ of the ectostylids among African

hipparionids still remains artificially high. Arambourg’s suggestion and his

deletion of this group seem to have been largely accepted. Gromova’s mono-

graph does not include the Serengeti material and does not refer to this paper

of Dietrich (1942). Arambourg (1956) has maintained his viewpoint and

speaks of ‘the typically African genus Stylohipparion characterized by the

presence of a broad ectostylid’. Naturally this does not positively exclude

Hypsohipparion, but in 1959, in a revision of the ‘few’ fossil African hipparionids,

Arambourg explicitly limits the Pleistocene material to ~Stylohipparion,

‘characterized by a permanent ectostylid’.

However, it can now be stated that the Serengeti Hypsohipparion material

can no longer be excluded because of the assumption that it belongs to zebra.

More than 90% of the small series of Langebaanweg equid upper teeth

unquestionably belong to Hipparion so that it cannot be asserted that all the

lower teeth belong to zebra merely because they do not possess an ectostylid.

Not only is this statistically highly unlikely but also the morphological features,

for example, the division of the talonid in Mg (on L938, the only Mg available

at Langebaanweg) would definitely exclude this group from belonging to zebra.

Consequently, as the Langebaanweg material, in which the ectostylid is

constantly absent, belongs to Hipparion, there are no longer sufficient grounds to

exclude a priori the Serengeti Hypsohipparion teeth from the hipparionid group

only on the basis that they do not possess an ectostylid.

Furthermore, it can now also be stated that the concept that African

Hipparion possess a very high frequency or permanent presence of ectostylid is

no longer applicable. Consequently the validity of the generic status of Stylo-

hipparion becomes highly questionable. Later in this paper this will be discussed

more fully in the light of the evidence presented. At present it suffices to state

that in Africa the Pleistocene Hipparion may be found with and without

ectostylids.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 335

REVIEW OF PUBLISHED AFRICAN Hipparion CRANIAL MATERIAL

MIOCENE

Oued el Hammam and Marceau

The presence of Hipparion in the Miocene of the Maghreb has been known

since 1932 when Suess discovered a mammalian fossiliferous deposit upstream

from Bou Hanifia (Oran) during the construction of the Oued el Hammam

dam. On several occasions brief comments have been made about this material

by, among others, Arambourg (1947, 1951, 1954). A rather rich collection

of cranial and postcranial material, with fairly complete milk and permanent

dentitions, as well as isolated teeth (table 7), have now been recovered from

two Upper Miocene sites in Algeria (figs. 3 & 4): Oued el Hammam (Oran)

and Marceau (Algiers). A detailed description was given by Arambourg

(1959): he discussed their relationships to and differences from Eurasiatic

forms of similar age, and other African forms of the Plio-Pleistocene period.

He suggested that this material should constitute a new specific form, namely

Hipparion africanum for which he proposed the following diagnosis:

Fipparion with a large skull, medium-sized limbs and heavily built extremi-

ties. Face and snout are elongated; nasal aparture is long and broad;

orbits are situated far back; pre-orbital fossae are long, simple and distant

from the orbit. Dentition of medium size: P?-M? = 141—154 mm.

Upper molars with strongly plicated enamel; compressed elliptical or

lenticular protocone. Cingular formations developed on lower milk

teeth but a laminated protostylid sometimes persists on permanent molars.

The limbs are rather short, with robust metapodials that have

well-developed lateral digits.

Arambourg emphasizes the ‘African character’ (vide infra) of the dentition

of this new species: enamel plications and narrow elongated protocone, as

well as a tendency to develop cingular formations. Because of these features

which distinguish it from all Eurasiatic forms, and because of its stratigraphical

location in the Upper Miocene, he suggested that this species should belong

to an independent African stock, isolated on the continent of Africa since the

end of the Miocene.

Camp Berteaux

Bourcart (1937) mentioned the discovery of Hipparion at Camp Berteaux

(Taourirt, Eastern Morocco). This material and additional specimens, described

by Ennouchi and Jeannette (1954), are now considered to belong to Hipparion

africanum.

Other sites

It is not possible to comment on the Hipparion material that was merely

mentioned by Dalloni (1915) and Solignac (1927), which was derived from

336 ANNALS OF THE SOUTH AFRICAN MUSEUM

UPPER Specimen ip2 Pp pA Mi M2 M3

A-P length a os 141 33°0 26-0 25°0 23°0 23°0 20°0

Mar.* 27°0 24°0

Transv. breadth .. A 122 22°0 24°0 23°0 16-0

141 ‘22°5 24°5 25°3 23°0 20°5 18°5

Mar.* 26°0 25°7

Eicighitaeser a Be 122 48-0 58-0 60-0 47°0

Protocone length is 14! 7°8 9:2 10°2 FO@ 7°5 7°6

Mar.* 8-4 8

Protocone breadth ie 14! 5°70 4°0 3°5 Qo] 3°5 2°5

Mar.* 5:0 4°0

LOWER Specimen 2, Ie 1B, M, M, M, DM, DM, DM,

A-P Length Be 143 30°0 23°6 25-0 25°3 26:0 23-4

89 PLAC) OO) BIRO) “PBOG) “KOO. SRO)

3 29°O 24°0 23°5 20°7 21:0 27°0

Mar.* 31°75

6 30°75 24°2 26-0

105 31°6 29°0 30°4

Transv. breadth .. Mar.* 18:0

6 14°0 14°4 13°5

105 12:0 12°0 9

* — Marceau: no number given.

TaBLE 7. Measurements (mm.) of the teeth of Hipparion africanum (from Arambourg, 1959).

the Upper Miocene layers of the lower Tafna Valley (near Guiard, Oran),

and the Djebel M’Dilla ‘Pontian’ horizons (of Tunisia) respectively.

Discussion on the ‘African character’ of Hipparion africanum Arambourg, 1959

The relationships of H. africanum, of the Upper Miocene of the Maghreb,

to other African hipparionids have been discussed by Arambourg. He dis-

tinguishes three species, stratigraphically separated: H. africanum of the Upper

Miocene, H. sitifense of the Pliocene of the Constantine and Oran area, and

Stylohipparion libycum of the ‘Villafranchian’ of North, East and South Africa.

Despite the great temporal span, Arambourg believes that there is a ‘family

air’ or unity common to these three forms, which distinguishes them from

the Eurasiatic material, and which stresses the common origin of the African

hipparionids, isolated since the end of the Miocene.

From the description given of the skeletal material of H. africanum, it

seems obvious that there is a marked difference between it and the Eurasiatic

forms. Until the present, only teeth have been available for comparison, so that

the lack of adequate comparative postcranial material has made it impossible

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 337

to determine to what extent H. africanum shares common skeletal features

with the African hipparionids of the Pliocene and Pleistocene. Dr. Leakey

indicates (personal communication) that considerable postcranial material

has now been recovered from Olduvai. These are to be described by Dr. R. H.

Stirton of Berkeley, California, who will elucidate the skeletal relationships.*

On the basis of the teeth alone, it is difficult to acknowledge a close

resemblance between the three African species. As there is a long span of time

separating H. africanum from Stylohipparion libycum, one expects to find some

differences, but it is first necessary to assess and appreciate the importance of

the resemblances stressed by Arambourg before it is possible to evaluate the

phyletic implications.

Of the ‘family air’ features linking H. africanum and the later African

species, Arambourg considers two as most typical: (i) In the upper cheek

teeth, the enamel plications and the shape of the protocone of the molars;

(ii) In the lower molars, there is a tendency to develop additional cingular

stylids. Two other classical features which will be discussed are the hypsodonty

of the cheek teeth and the shape of the double knot in the lower molars.

Enamel Plications: It is commonly believed that the enamel plications have

little phylogenetic significance. The plications are very variable, being much

influenced by functional and ecological conditions. In any event, because of

the vast temporal and spatial differences, and especially climatic and biotopical

environments, this feature cannot be used as a basis of comparison. This is

just an a priori consideration.

However, Arambourg writes (p. 93) that H. africanum is characterized by

the complex pattern of enamel plications, which are more numerous than on

_ most of the European forms. Although there is quite considerable variation in

the complexity of the enamel at a particular period and in a particular area

(compare H. crassum and concudense; Greek and Samos Pliocene forms), it

cannot be denied that the Eurasiatic forms normally possess a rather plicated

enamel mark-wall pattern. Furthermore it is certainly far more developed

than that of the African Pliocene and Pleistocene species, which are actually

characterized by their small number of plications.

The multiple pli caballin seen in many teeth does not seem to be typically

African either: it is observed in many other groups on both sides of the Mediter-

ranean Sea, and also elsewhere—compare H. sitifense (Mascara), H. gracile

(Pikermi), H. koenigswaldi, H. concudense (South Aragon), Hypsohipparion albertense

(Serengeti), and even in America in Neohipparion eurystyle Cope, in which as

many as four may occur on the anterolabial border of the hypoloph.

Protocone: The elongated and narrow protocone is undoubtedly a typical

and constant feature of all the African hipparionids so far known from the

Pliocene and Pleistocene deposits. It also seems to be a distinct, strongly

expressed evolutionary trend in the whole group (see pp. 354-5; table 14;

* See Oldwia Gorge 1951-1961, 1 by L. S. B. Leakey (1965) which was published while

this paper was in press.

338 ANNALS OF THE SOUTH AFRICAN MUSEUM

figs. 12, 13). A satisfactory comparison of individual teeth at different stages

of wear is difficult, and published sketches of this feature may be deceptive.

Nevertheless, it is obvious that a comparison of the indices of the length and

shape of those protocones available for study definitely confirms this “African

feature’ of H. africanum. For all teeth, it shows a constant greater ‘African’

index than any of the European species considered (except for the breadth/

length index of P?, which in any event, is a rather atypical tooth). It is important

to note that such a progressive feature, characteristic of the African phyletic

line, can be traced back as far as the Upper Miocene.

The double knot: Gromova (1952), in a recent review of all the available

evidence, indicates that the original type of knot is the ‘stenomis’ one: 1t presents

a very different metaconid and metastylid, the latter being strongly angulated,

and separated from the metaconid by a narrow deep valley. This original

type, found in the Merychippus ancestor, has been retained mostly in the

American Hipparion. However, there seems to be a constant modification

amongst Old World Hipparion: towards a ‘Hipparion’ type of knot in Eurasiatic

forms, characterized by a symmetrical and rounded metastylid and metaconid

separated by a wide and shallow valley; and towards a more ‘caballus’ type

in Africa, where the metastylid is sub-triangular. The description and drawings

of H. africanum indicate clearly that it has a ‘Hipparion’ type of knot, more

typical of the Eurasiatic form and bearing no resemblance to the African

forms. As H. africanum is the oldest species of Hipparion known in Africa, it

needs to be decided whether it has acquired this ‘progressive’ feature very

rapidly or whether the ‘Hzpparion’ type of knot is not as progressive as was

previously considered, having already been present in some Merychippine

ancestor, and developed in parallel with the ‘stenonis’ type.

Stylids : The tendency to develop stylids is quite marked in many, but not

all, African forms. It is said to be expressed in H. africanum by a small laminated

or rounded, rather isolated protostylid on the permanent P,-M, of specimen

no. 2, and by a small ectostylid on the vestibular aspect of most of the milk

teeth. The tendency is found to be constant among all Hipparion. The stated

degree of development of these structures does not seem to be particularly

‘African’: it is certainly surpassed by some Eurasiatic species, for example,

H. koenigswaldi from Teruel. Of course, it should not be overlooked that the

feature seems to have been expressed rather weakly in Merychippus times. The

minor development that it shows in the Upper Miocene of Algeria may hardly

be considered ‘African’ when compared with the Stylohipparion of the ‘Villa-

franchian’. However, the feature suggests a rather strong and precise

evolutionary trend at work.

Hypsodonty: H. africanum is stated to be poorly hypsodont. The indices are

by no means comparable to the ‘African’ specimens of the Pliocene and

Pleistocene times, and they are of the same magnitude as the other Hipparion

groups of the Mediterranean basin. H. africanum is even less hypsodont than

H., sitifense which is stated to have low teeth.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 339

PLIOCENE

Mascara

A palate with an almost complete dentition and a symphysial region with

incisors, probably belonging to the same skull, have recently been described

by Arambourg (1956).

This Pliocene form is characterized by very small dimensions of the teeth

(table 8), which also show few and simple plications and little hypsodonty. The

protocone, completely independent of the protoloph right to the base of the

crown, is relatively compressed.

The specimens have been referred to H. sztifense Pomel with which they are

stated to share ‘almost identical morphological features and size’.

Lee Bg Te M}! M? Me P2-M?

1. MASCARA

A-P length ae ale 560, BXa) 20°5 20 19°5 19 18°8 129

Transy. breadth .. ae so 1((8)) LO) 20 19°5 18 18°6

Crown height at 56. fd OHO OSG TwO%s we 13

Protocone length .. a og HOG F gB}OG}) OI 2) Goa 8-8

Protocone breadth .. “if 5) GINS ALOE AOL OG) G06) 3

Protocone shape .. oe! WOH 60 55 55 46 34

Protocone length/Tooth isoretin Sol) PR BY 41 33 38 46

2. ST. ARNAUD CEMETERY

H.. sitifense (type specimen)

A-P length site 7 ass 23

Transy. breadth .. a ae 22°5

3. AIN EL HADJ BABA

R 1

A-P length a is 50 222 O2) 120 20°2

Transv. breadth .. ae a QI 20 QI 20°3

Protocone length .. ae a 6727716 (HOM Gog

Protocone breadth .. ie ae GO A) Bor GB

Protocone shape .. “3 51°6 66°6 52°2 47°5

Protocone length/Tooth feneth sa: BY) PFO) | GRO, -GNiop)

TABLE 8. Summary of measurements (mm.) of teeth of North African Pliocene Hipparion.

St. Arnaud Cemetery

The type specimens of H. sitifensis (sic) Pomel 1897 were recovered from

the St. Arnaud Cemetery. These two upper teeth, figured by Pomel (see also

Arambourg, 1956) are little hypsodont and rather small (although not quite

as small as the Mascara teeth). A broken calcaneum is also mentioned.

Lower teeth of a correspondingly small size and medium hypsodonty,

collected by Arambourg but as yet undescribed, have no ectostylid.

Ain el Hadj Baba

Upper isolated, but partially serial right (?) P4-M?, a left M?, lower

teeth, and limb bones were described by Thomas (1884) who referred them

to H. gracile (Kaup).

340 ANNALS OF THE SOUTH AFRICAN MUSEUM

However, according to Arambourg (1956), the narrow and small number

of plications of the enamel, the double pli caballin, features of the protocone,

dimensions of the upper teeth and absence of the ectostylid on the lower teeth,

make the Ain el Hadj Baba specimens very similar to H. sitifense types and

co-types, as well as to the Mascara specimens. Therefore he considered them to

be H. sitifense.

From the metapodial, it may be observed that the lateral digits were

still very strongly developed.

Wie (Northern Tunisia)

This area also includes the continental sands of the Ferryville region near

Bizerta (fig. 4).

Solignac (1927) indicated that the fragmented remains had been assigned

by Depéret to H. crassum Gervais, but Arambourg (1956) is inclined to refer

them to Stylohipparion.

St. Donat (Algeria)

Teeth of two maxillae, which Arambourg (1956) interprets as being

possibly referable to H. sitifense, were found at this site. Joly (1909) recognized

further teeth from this site and hastily referred them to H. gracile, but they

should probably be compared with Arambourg’s 1956 material and pooled

with H. sitifense.

PLEISTOCENE

North Africa

Oran: The teeth recovered by Pomel in the Oran area are among the

first discoveries of Hipparion in Africa and formed the basis of the first description.

Two sites were mentioned:

(i) ‘St. Pierre sandstone quarry, on the property of Mr. Brunie, in the

St. Charles district, in the east quarter of Oran, which is also the

type locality of Libytherium maurusium.’

Two lower teeth were described by Pomel (1897) as the type specimens of

Hipparion (?) libycum. The one, a left P; or Py, has an ectostylid while the other

is without an ectostylid, but the antero-external region is broken. The specimens

were successively referred, first by Van Hoepen (1932) to Stylohipparion steytleri

because of its similarity to P; (Nas. Mus. C795) from Cornelia (vide infra, p. 347) ;

then by Joleaud (1933) to Libyhipparion libycum because of its fundamental

similarity to the du Bourg de Bozas Mission material from Omo, although it is

more elongated and somewhat larger; and later by Arambourg (1947, 1956) to

Stylohipparion libycum (or albertense).

A distal portion of a third metatarsal was also recovered from this deposit.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 841

(ii) Puits Kharouby.

Five upper molars were described by Pomel (1897) and referred to H.

massoesylium. However, Pomel himself was not too sure whether or not these

upper teeth belonged to those lower teeth which he had described under the

name Hipparion (?) labycum. Their reciprocal kinship was ascertained by Joleaud

(1933) who consequently referred H. massoesylium to his Libyhipparion libycum,

and by Arambourg (1947, 1956), who for the same reason, placed them in

Stylohipparion libycum (or albertense).

Ain fourdel (Constantine): The material consists of a M, and a lower

premolar, the additional stylids of which seem to have been overlooked by

Thomas (1884) who referred them to Hipparion gracile.

Joleaud (1933) emphasized their similarity to Pomel’s Oran specimens and

to the material discovered in the Orange Free State of South Africa. He proposed

that the Ain Jourdel material should be placed in a new species Stylohipparion

(2) thomasz.

Arambourg (1947) can see no reason for not pooling Ain Jourdel (one

of the two teeth at least) and ‘H. massoesylium’? (= Joleaud’s Libyhipparion

libycum) with Stylohipparion libycum.

Beni Foudda (Constantine): This site has been referred to as Ain Boucherit

by Arambourg.

A molar recovered here was attributed by Pomel (1897) to Hipparion

ambiguum but was referred by Arambourg (1947, 1956) to Stylohipparion libycum

(or albertense).

Wadi Natrun (Gart el Moluk Hill): Andrews (1902) described a left upper

premolar (?P#) of large dimensions (A—P 29 mm.; transv. breadth 28 mm.),

but apparently it is little hypsodont (height 38 mm.). However, it is a rather

worn specimen.

When the specimen was discovered, very few African Hipparion had been

recovered, and the typical features of the few that had been recovered were

still largely unknown or as such unrecognized. The best comparative material

at that time was that from Oran: the enamel pattern of the specimen from

Wadi Natrun appeared definitely more complex and it differed in the absence

of isolation of its hypostyle. Therefore, in spite of its transversely more com-

pressed protocone, which explicitly suggested some close resemblance to

Hipparion theobaldi (from the Siwaliks), the specimen was referred to the more

widespread H. gracile.

After the discovery of further material from Omo and Kaiso, Joleaud

(1933) emphasized the difference in the parastyle and mesostyle features from

Libyhipparion ethiopicum, and he stressed the resemblance with the Kaiso speci-

men, referring to it the Wadi Natrun premolar, which he called Hipparion

cf. albertensis (sic).

342 ANNALS OF THE SOUTH AFRICAN MUSEUM

Central Africa

Koro Toro: The Hipparion material recovered from the ‘Villafranchian’

horizon of the Tchad has been provisionally referred by Coppens (1960) to

Stylohipparion. He is preparing a detailed description of these specimens, together

with the associated fauna. Therefore, it is not possible to comment on this

material at this stage.

East Africa

Omo: The first specimens were recovered by the du Bourg de Bozas

Mission (1903) and were described by Joleaud (1933). They consist of a few

isolated teeth, namely:

I incisor;

5 lower cheek teeth (1 right premolar, one left Ms, two left molars

and one right molar) ;

1 fragmented upper right molar.

To this group Arambourg (1947) added a left M! or M?, first interpreted

by Joleaud as Hippotigris.

A second series of specimens was recovered by Arambourg (1947) during

his 1932-33 expedition, and comprises 1 right P?; 1 left M*%; 1 fragmented

right upper molar and 1 right M, or Mg.

The first of these groups was described by Joleaud (1933) who erected

the new genus and species Libyhipparion ethiopicum,! generically distinct from

the South African specimens from Namaqualand and Cornelia on the one

hand, as well as from the Constantine material of Thomas, on the other, and

specifically separated from Pomel’s St. Charles teeth.

In a later reconsideration of these Omo specimens, to which the second

series of teeth from the same area was added, Arambourg (1947) rather empha-

sized (a) the close relationships between all the Ethiopian hipparionids; and

(6) their profound similarity to the East African (Kaiso, Serengeti pro parte,

and probably Olduvai) material and the Orange Free State discoveries.

Furthermore, Arambourg considered it reasonable to refer most of Pomel’s

and Thomas’ North African (St. Charles, Puits Kharouby and the Constantine)

material to S. albertense or possibly to S. libycum.

The characteristic features common to these specimens are:

(1) pronounced hypsodonty reaching 80 mm. in unworn teeth;

(ii) upper molars with very complicated enamel pattern and laterally

compressed and strongly elliptical protocone;

(iii) lower teeth with strongly developed additional stylids, rather large

oval or flattened and pointed ectostylid;

(iv) the presence of a ptychostylid; and

(v) ridged or pillar-like protostylid and well-built entostylid.

(4) With exception of left upper molar, considered by Joleaud as belonging to Equus

(Hippotigris).

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 343

Therefore, there are no reasons for proposing any generic or specific

distinctions between these Pleistocene African Hipparion, which according to

the priority rules have been referred to Stylohipparion albertense.

Olduvai: Some of the first fossils identified in the Olduvai Gorge, when it

was first discovered by Kattwinkel in 1911, were the distal ends of 3 metatarsals

of hipparionids, which Hopwood (1937) referred to Stylohipparion cf. albertense

(Hopwood).

The dimensions of the lower articulations of metatarsal III were 40 « 30

mm. The broken metatarsal II and metatarsal IV are stated to have had

‘originally’ almost the same lengths as metatarsal III, and to have had the

appearance of broadening distally.

Further specimens of teeth, recovered by Kattwinkel and Reck in 1913,

as well as an upper dental series collected by the British expedition, were

referred to the same genus and species. The diagnosis proposed by Hopwood

(1937) for this Stylohipparion cf. albertense is:

Three-toed equid, with high-crowned molars of the Hipparion group.

Lower teeth are laterally compressed and display a strong, often laminated

ectostylid. Genotype: Stylohipparion hipkint Van Hoepen (1932).

Unpublished data

In recent years a large collection of equid material has been recovered

from the Olduvai Gorge by Dr. L. S. B. Leakey. All the specimens from Bed II

(and a few from Marsabit Road, Olorgesailie and Omo) are now being studied

by Dr. R. A. Stirton in California. Dr. Leakey kindly permitted the authors

to make a brief survey of this material. In order not to encroach on the final

description, only a few features relevant to this paper are mentioned:

(i) In this mixed sample there are 471 Equus teeth, of which 224 are

from the upper dentition. Of the 186 teeth of hipparionids, 100 are

upper. The vast majority of these are from Bed II, only a few deriving

from other sites.

Mee

This is then a reasonable sampling to ascertain the expected propor-

tions of upper and lower teeth. It will be indicated below that at

Langebaanweg there are approximately equivalent proportions of

uppers and lowers, but at the latter site there are almost no Equus

specimens.

(iii) The data is also useful to assess the situation at South Serengeti

(vide infra). In this respect it is interesting to note that in the Olduvai

collection, among the lower teeth, those of hipparionids have elong-

gated, strongly marked ectostylids, typical of Stylohipparion, while

those without ectostylids lack the typical features of Hipparion as well

as the protostylids of the Langebaanweg specimens. They are typically

Equus in type.

344 ANNALS OF THE SOUTH AFRICAN MUSEUM

(iv) There is a great range of variation in the size of these Hipparion

teeth, and also in the length and shape of the protocone. On the

whole, the range seems smaller than that of the Langebaanweg

teeth, but the protocone is more elongated. In the Hipparion lower

(as well as in these Equus) teeth, there is a complete absence

of protostylids and hypostylids.

South Serengeti: A large collection of Hipparion has been made from various

localities in the South Serengeti, and was described by Dietrich (1942). A

complete list has never been published. However, from the available data, it

is possible to ascertain that the collection consists mostly of isolated and frag-

mented specimens, a few maxillary and mandibular fragments (among which

are symphysial portions with incisors), many loose upper and lower teeth,

both milk and permanent. There are also some rare postcranial specimens,

namely, a proximal fragment of a femur, a distal fragment of a tibia, two

third metatarsals, one third metacarpal and tarsals. Dietrich artificially

reassembled the isolated teeth into series, and most of the dental series figured

in his publication are reconstructed. From the scattered data and illustrations,

it is possible to establish an incomplete list of the South Serengeti dental

material (table 9).

Plate Number in

Description Origin and Coll. Number Dietrich (1942)

Milk Dentition

Upper r dP2-dP4 so. WO SEXO) XIII 96

r dP2-dP3 .. Gadj 10 XIII 97

1 dP2-dP4 Gadj 2.39 XX 162

Lower r dP2-dP4 Vo 313 XIII 93b

r dP3-dP4 Marambu XIII 93a

r dP4 : Gadj 10-13.3.39 XIII 94

Permanent

Upper 1 P2—-M3 Vo, Gar. XIII 87

1] Ma-Ms .. Vo 670 XIV 102

1] Po—-M3 Vo 330, Vo 670 XV 107, 108

1 Poa—P4. Vo 670 XX 160

1p ey a Vo, Gar. XIII 88

r Mo2-Ms .. Vo 670 and Marumba XIV 103

Lower 1 M3-—Pe2 Vo 313, 670 XV 106

] P2a—-M3 Der., Gadj., Gar., Vo XV 105

1 P2a—-M3 Vo, Gar. XIV 101

] Pa—-M3 Vo 330, Vo 670 XIII go

1 Mr Garussi 2.39 XIV 99

1M Gar. River 200 XIII 95

r M3-Pe2 Vo 670, Gar., Olduvai Hill XVI 109

r Po-M3 Vogel River XIII 89

r Poa-Mg3 .. Vo 670, Gar., Vo XIII o1

r P3-M3 .. .. Gadj to XIII 92

Symphysial region with incisors Vogel River 9-10-38 XVI 112

TABLE 9. Hipparion material mentioned by Dietrich (1942) and derived from

South Serengeti region

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 345

In contrast to the relatively unchanged nature of the postcranial skeleton

(vide infra) since Pliocene times, the advanced and progressive features of the

dentition of the South Serengeti hipparionids have been emphasized by Dietrich.

The teeth, both upper and lower, are strongly hypsodont (crown length:

g cm.). The upper molars show numerous enamel plications, and an elon-

gated, ungrooved protocone, completely isolated right down to the base of

the crown, and embedded in a thick cement layer. The incisors (see p. 350) are

longitudinally grooved on their anterior surface (in one case I, is missing).

It does not seem possible to distinguish different forms of hipparionids at

Serengeti on the basis of the upper teeth. On the contrary the lower teeth show

some constant features whereby it seems possible to distinguish two groups.!

A large group of teeth, all apparently recovered from the gray tuffs, does not

show any additional stylids while another group of some 50 (40 isolated teeth

and 2 incomplete dentitions) displays the ectostylids characteristic of most of

the African hipparionids then described. In these teeth the ectostylids are

sometimes very strongly developed, pillar-shaped or flattened and ridged.

Protostylids and hypostylids are also regularly present, and a very peculiar

entostylid (see p. 360) may be observed on one tooth (Garussi River 200, fig. 95

of Dietrich, 1942). These teeth derive from deposits that seem to span over a

longer period of time than do those from which the teeth without stylids were

recovered: some of them probably occur in the gray tuffs, contemporaneously

with the first group, while others belong to a probably younger horizon.

However, as has been stated previously, the stratigraphy of the South Serengeti

is very poorly documented. Other than the presence of the stylids and the

average shorter total length of the dental series at a comparable stage of

wear (156 mm. v. 170 mm.), there seem to be no appreciable differences

between the two groups. However, the value that can be attached to the

assessment of the total length of the dental series in these specimens is doubtful

because (a) of the ‘reconstruction’ of most of Dietrich’s series, and (6) of the

variation of the total length due to wear.

In spite of the above distinction, Dietrich does not consider it necessary

to separate generically the South Serengeti material. ‘This attitude is reasonable

because

(1) lack of knowledge of a stratigraphical sequence rules out a clear

chronological sequence;

(2) too little was known of the skeleton, both cranial and postcranial ;

(3) the upper teeth showed no features whereby they could be separated

into two groups, and yet it is a statistical probability that, if there were two

species at the site, all the uppers could not have belonged only to one of these

species; and furthermore,

(4) the development of the ectostylids, ranging from absence through

mild formation to strong development, is an expression of dental adaptation

to environmental factors (vide infra, p. 352).

1 See also discussion on Arambourg’s concept of the non-validity of Hypsohipparion, p. 333.

346 ANNALS OF THE SOUTH AFRICAN MUSEUM

On the basis of the data available then,! Dietrich considered that within

the Serengeti material there was an original and conservative African stock of

hipparionids for which he erected a new genus Hypsohipparion Dietrich, 1942,

with Hipparion albertense (Hopwood) from the Lake Albert Kaiso bone beds

as the type specimen. Into this genus and species he placed the Serengeti

material without stylids, as well as all the upper teeth and the postcranial

remains. The diagnosis that he proposed was:

An advanced three-toed, short-snouted Hipparion of the dimensions

of a small caballus, the M? of which reaches a crown height of 9 cm. The

protocone shows a tendency to become laminated, non-grooved and com-

pletely isolated from the protoconule. Teeth show no reduction of the main

cusps and styles. Plications are more marked than in all Pliocene species.

Lower teeth (deciduous premolars, permanent premolars and molars)

are (?)? all without ectostylids.

Dietrich believed that the group manifesting the stylids was displaying

the developmental trend of a selective pattern. He provisionally placed this

group in Stylohipparion as a practical measure, partially comparable to Stylo-

hipparion libycum, Stylohipparion ethiopicum, and Stylohipparion steytleri. However,

he maintained that the generic labelling had no strict taxonomic value because

of the scarcity of the material on which this was based. He did not see the

value of assigning his material to a new species.

Arambourg (1947) questioned the validity of the generic status of Hypso-

hipparion on the basis of a possible incorrect determination of what he considered

to be lower zebrine teeth (see p. 333). In later publications (especially 1956 and

1959) he seems to have definitely adopted this view permanently. In his opinion

the Serengeti material belongs partly to Stylohipparion and partly to E. zebra.

This opinion is discussed and considered unacceptable elsewhere in this paper.

Lake Eyasi: The few isolated teeth recovered from the Lake Eyasi shore

(west and north of Mumba Hill) are very much rolled and unsuitable for any

detailed study (vide supra, p. 287). They were referred to Hipparion sp. (Reck and

Kohl-Larsen, 1936).

Lake Albert (Uganda): From the east shore of the Lake, one incomplete

upper molar (B.M.N.H. M12615) was described by Hopwood (1926) as the

holotype of Hipparion albertensis (sic). The tooth is rather hypsodont, with a

complicated enamel pattern. Its antero-posterior diameter is 24-5 mm.

Dietrich (1942) used. it as the genotype of his new genus Hypsohipparion

albertense, the proposed diagnosis of which is noted above.

South Africa

Namaqualand (Cape Province): Nine isolated mandibular teeth of the

same jaw, left P,-M, and M, and right P,-M, and M, (fragmentary) were

1 Through the kind collaboration of Dr. K. H. Fischer, Berlin, measurements of some of

the ‘Hypsohipparion’ teeth were made available to the authors: see table 20.

2 Dietrich includes this ‘query’ in his original diagnosis in German (p. 97).

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 347

described by Haughton (1932), who erected a new genus and species Noto-

hipparion namaquense. The type specimen is S.A.M. 9982. Unfortunately a

diagnosis was not provided.

A small ectostylid and a protostylid are occasionally present. It is difficult

to comment on the height and relative hypsodonty of these teeth because the

crown is rather low, as the specimens are in a very advanced stage of wear

(table 10). The teeth possess a thick cement.

Cooke (1950) proposed a diagnosis for the genus Notohipparion Haughton:

‘rather low-crowned, heavily cemented hypsodont lower cheek teeth with an

extra antero-external cingulum fold or column, either isolated or fused with

the parastylid, present in all the permanent cheek teeth except the second

pre-molar, and a deep groove separating the strongly developed metaconid

and metastylid. The upper dentition is unknown’.

Because of the moderate hypsodonty and the weak development of

additional stylids, Dietrich (1942) considers Notohipparion to be more primitive

than Stylohipparion. ‘This opinion had also been held by Van Hoepen (1932).

Arambourg (1947) compared these lower molars with the ‘single’ African

Pleistocene genus Stylohipparion but, probably because of the poor stratigraphy,

he did not revise its taxonomic status. In later comments on the African

Hipparion, he (1959) makes no further mention of these specimens.

P; IEA P, M, M, M,

A-P Length .. oe oe D205 27°0 26-0 23°0 24.°0 29°0

Transv. breadth et es a: 16-0 17°0 17°0 16-0 14°5 12°5

Height .. Si ae ne Soy, a AOS 17°5 25°0 23°0 30°0 34°0

Transv. breadth/A—P length no) GEO 63 65°4 69:6 60:4 43

Height/A—P length .. a: 6) 22 64:8 96-1 100 125 117

Hypsodonty Index .. Bi 55/0 LO 97 67°9 69°5 48-2 36-8

TABLE 10. Measurements (mm.) of teeth of ‘Notohipparion namaquense’ Haughton

Christiana (Cape Province): From this site an upper second (? third) left

molar (Archaeol. Sury. 113), considerably worn, undescribed and not illus-

trated, has been referred to Stylohipparion steytleri by Cooke (1950).

Transvaal australopithecine cave breccias: ‘Two lower teeth, recovered by

Broom in the breccia filling the Kromdraai Cave, have been referred to

Stylohipparion steytleri (Cooke, 1950). Hipparion teeth recovered from Bolt’s

Farm and Makapansgat Limeworks have been similarly referred (Cooke, 1963).

These specimens have not been described.

Cornelia (Uitzoek): Eleven isolated upper and lower teeth include a

group of five and another group of two teeth belonging to single individuals

(table 11):

(i) one right M1 or M? (Nas. Mus. C558), one left M? (C555), one left

M,—M, (C556),! described by Van Hoepen (1930) as a new species

1 In the original publication (1930), C556 was stated to be composed of lower milk teeth

(DM, and DM,), and only in 1932 were they correctly acknowledged as permanent teeth.

348 ANNALS OF THE SOUTH AFRICAN MUSEUM

Hipparion steytleri, the type being C558 and the paratypes being

555-559.

(ii) A series of five left lower teeth P,-M; (C795), recovered from the

type locality, were referred by Van Hoepen (1932) to the same

species.

(ii) A right M, (C797), in very early wear and therefore displaying

a completely uncharacteristic pattern, was used by Van Hoepen

(1932) to erect a new genus, namely, Stylohipparion, and a new species

hipkini, to which a left P. (G796), recovered from the same locality

and formation, was provisionally referred.

Van Hoepen (1932) subsequently referred steytlert to the new genus Stylo-

hipparion. He considered Stylohipparion to be ‘obviously the terminal stage of

the evolutionary line which originated from Hipparion with Notohipparion.’

Joleaud (1933), who compared steytleri, with Libyhipparion ethiopicum, and

hipkint with the Ain Jourdel specimen of Thomas, accepted the specific distinc-

tion between steytler: and hipkini. Cooke (1950) considered this distinction to

be invalid. This opinion is strongly supported here, for both a priori and a

posteriori reasons: first of all, no upper teeth of hipkini have been recovered,

and Van Hoepen’s species is based on one single lower tooth (C797), which

constitutes a very inadequate basis for creating a new taxonomic unit.

Furthermore, the tooth is in a very early stage of wear, which is the case of

the obviously atypical, peculiar pattern of its occlusal surface. Thus the

‘difference’ between it and the Stylohipparion steytleri is not at a species level.

Van Hoepen provisionally referred specimen C796 to the new species,

but it is even less different from the known steyélert specimen. In any event,

being a P,, it does not show the typical features of a species.

On the other hand, and quite apart from this question of a specific distinc-

tion, the genus Stylohipparion Van Hoepen has been generally accepted by all

subsequent authors. Hopwood (1932), Dietrich (1942), Arambourg (1947,

1956, 1959), Cooke (1950) and Gromova (1952), among others, have considered

Stylohipparion as a valid unit (at least at a subgenus level), which expresses the

‘African’ trend of Pleistocene hipparionids. The following diagnoses have

been proposed:

Arambourg (1947): “Tridactyl equid, with very hypsodont dentition.

Upper molars with compressed protocone of very elliptical section, wavy

parastyle; especially complicated enamel pattern, particularly around

the prefossette. Lower molars with broadly developed ectostylid, laterally

compressed, and close to the antero-external hypoconid pillar.’ (Direct

translation from the French.)

Cooke (1950): ‘High-crowned rather hypsodont lower cheek teeth

with a strongly developed isolated pillar external to the ectostylid, possess-

ing no external groove between hypoconid and hypoconulid, having no

protoconid and hypoconid and small rather widely separated metaconid

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 349

and metastylid. High-crowned upper cheek teeth with isolated oval

protocone and possessing a small flange on the antero-internal side of the

parastyle.’

We wish to point out that the ‘ectostylid’ referred to by Cooke is actually

the ptychostylid, while the ‘strongly developed isolated pillar’ is the ectostylid.

UPPER TEETH

St. steytlert (van Hoepen)

M? or M? M$

(C558) (C555)

A-P length we 22°0 21-0

Transy. breadth .. 22°0 18-0

Height .. a 54°0 66-0

St. steytlert St. steytleri St. hipkini

LOWER TEETH ley PR, M, M, M, M, M, P,

(C795) (C556) (C797) (C796)

A-Plength .. 25:0 28-0 22°0 24°5 — 21°70 28°5 31°0

Transv. breadth 15°0 12°O 12°0 II°5 12°0 I2°5 15°0 14°5

Height .. ee O40 74°0 69-0 73°0 32°0 41-0 81-0 46-0

TABLE 11. Dimensions (mm.) of teeth of ‘Stylohipparion’ from the Orange Free State, South Africa.

From Uitzoek also comes an anterior portion of a lower jaw (Nas. Mus.

C679), containing four large first and second incisors, which are flattened

anteriorly and arranged almost in a straight line. Each incisor shows the

‘mark’ or the cement-filled depression in the enamel. The third incisor is small

and lies behind and in contact with the second incisor.

The above description (Cooke, 1950) is a slightly modified interpretation

(also suggested by L. H. Wells) of the original description by Van Hoepen

(1930), made, according to Cooke, so as to avoid the ‘startling supposition’ of

the complete absence of the third incisor. However, it should be pointed out

that specimen GADJ. 10 from South Serengeti shows complete absence of

not only I;, but also of the canines (Dietrich, 1942).

One of the special features of this specimen C679, besides the reduction

(? complete, i.e. absence) of I,, is that the lingual aspect of I, and of I, has

two surfaces, meeting at an angulated thick ridge. Each surface also has

longitudinal parallel ‘costae’.

A new genus and species Eurygnathohippus cornelianus van Hoepen (1930)

was created to include the above material. No measurements were published,

but from the illustrations and the cast it is clear that the jaw fragment belongs

to a very large skull.

Dietrich (1942)(1) made the suggestion that this mandibular symphysial

? Dietrich incorrectly describes the specimen as belonging to a milk dentition (‘Milch-

vordergebiss’, p. 97).

30°

ANNALS OF THE SOUTH AFRICAN MUSEUM

region (for which no cheek teeth are known), should belong to Stylohipparion,

or at least to the advanced Hipparion of the Cornelia layers.

Ewer (1963) refers to Eurygnathohippus as a chalicothere. The basis for this

opinion is not clear to the authors.

NOTE ON Hipparion INCISORS

Hipparion incisors have rarely been discovered in Africa. In a recent,

comprehensive study of a large assemblage of Hipparion material from South

Aragon, Sondaar (1961) indicates a similar shortage there. The small number

of recorded specimens from Africa are:

(1) Upper incisors: the H. africanum type specimen (no. 141) from Oued el

Hammam presents a complete snout, and, from the same site, there are

also three maxillae with incisors. They have been described by Arambourg

(1959) without comment, because there is no comparative material.

(ii) Lower incisors :

(a)

(4)

(c)

The Libyhipparion ethiopicum (Joleaud, 1933) collection from the

Omo Valley contained one incisor (vide supra, p. 342).

Dietrich (1942) mentions some incisors from the South Serengeti

without going into any detailed description (vide supra, p. 344).

Eurygnathohippus cornelianus (Van Hoepen, 1930): an anterior portion

of a lower jaw. This may belong to an individual of the genus Hipparion

or Stylohipparion, as has been suggested already by Dietrich (1942)

and in a personal communication (1964) from Dr. L. S. B. Leakey

(who bases his view on material recently discovered at Olduvai).?

Each of these specimens raises difficult problems, which cannot be solved

until more complete material becomes available. A few of these problems

are:

(1)

(3)

The ‘Libyhipparion’ incisor does not show the sub-elliptical section of

Hipparion. It is completely subdivided into two clearly separated

‘marks’, no other example of which is known among Hipparion in the

literature. It has been suggested that this feature is an extreme mani-

festation of the enamel plications, typical of the African hipparionids.

It is impossible to state whether or not the feature is really exceptional.

It is not demonstrable in the other few available incisors.

Eurygnathohippus: The ‘costae’ and the blunt ridge on each specimen

are unique in this small series. A possible explanation for these

features is that they are adaptations to browsing and strengthen the

teeth, set practically parallel to the horizontal symphysis.

The occasional reduction (? absence) of the third incisor is referred

to above.

1 Confirmed while in press. See Leakey, L.S.B.: Olduvai Gorge 1951-1961, 1. p. 26.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 351

A SUMMARY OF PUBLISHED POSTCRANIAL REMAINS OF AFRICAN Hipparion

The postcranial skeleton of African hipparionids is very poorly known.

The only described specimens are:

Camp Berteaux: A femur fragment with the proximal epiphysis.

Oued el Hammam : Distal fragment of humerus (no. 159) ; radio-ulna (no. 123)

and fragments 13, 22 and 27; femur without epiphysis and a distal

fragment; distal fragment of tibia; pelvic fragment; two astragali;

one calcaneum; five metacarpals and five metatarsals, many with

lateral digits.

Ain el Hadj Baba: Metapodials with lateral digits.

Oran: One fragmentary calcaneum and one third metatarsal.

Omo: Right humerus.

Olduvai: Distal end of three metatarsals. There are also quite a number of

undescribed specimens. The authors examined a number of such

specimens recently collected at Olduvai Gorge. Data on the

metapodials is provided below.

South Serengeti: Proximal fragment of a femur; distal fragment of tibia;

two metatarsals III; one metacarpal III; tarsals.

Arambourg (1949) has probably described the largest collection known

from one site: it consists mainly of numerous complete metapodials (many of

them with lateral digits) which have been referred to Hipparion africanum.

From the description and the measurements, this North African Upper Miocene

Hipparion seems to have been about the same size as the smallest H. mediterraneum

of Pikermi. The extremities indicate that africanum was rather short and

heavily built. The lateral digits were more robust and strongly developed than

those of most other hipparionids, and they certainly were still functional.

From the Pliocene, only a few metapodials with very well developed

lateral digits are known from Ain el Hadj Baba. They have not been adequately

studied. It has been assumed that they belong to H. sitifense.

From the Pleistocene of East Africa a right humerus is known from Omo,

and a few fragmentary long bones and metapodials have been recovered at

Olduvai and South Serengeti. The material has not been described. Size and

morphology of the metatarsals do not seem to differ noticeably from H. gracile.

Dietrich (1942, p. 101) states that the South Serengeti metatarsal III ‘must

Breadth, distal end A-—P diameter,

distal end

Olduvai (Hopwood 1937) .. ue 40 30

South Serengeti (Dietrich 1942) .. 37 35

H. elegans... te ate ae 29°6 2571

H. moldavicum ae ue, oe 33°9 26°8

H1. mediterraneum ne Oe oy. 32°58

TABLE 12. Dimensions (mm.) of metatarsal III of Hipparion

352 ANNALS OF THE SOUTH AFRICAN MUSEUM

have a length similar to H. gracile, i.e. 23-24 cm.’ The lower articulating

surface is 37 mm. broad and 35 mm. A-P, so that it is only possible to state

that the distal end is broader and thicker than in most of the European Hipparion

(table 12). The lateral metapodials reach almost to the distal articulating

facet and are rather strongly developed, but it is not possible to state firmly

whether or not they possessed three phalanges. Nothing is known about the

front extremities at present.

The only species with similar dimensions of the lower end of metatarsal

III is A. longipes with 40 and 33 mm. for the breadth and A—P dimensions

respectively. H. africanum has a distal breadth of 34-7 mm. It is difficult to

assess whether these greater dimensions of the East African Pleistocene forms

are indicative of larger overall dimensions of the animal, or whether they only

refer to more massive metapodials. Although there is no direct correllation

between the different structures, it may be mentioned that the larger size of

some of the South Serengeti molars may suggest that the first possibility is

not excluded. In general, too, Pleistocene mammals tend toward giantism.

On the basis of published material, it was not known whether or not the

African Pleistocene hipparionids were three-toed horses. There is some informa-

tion concerning the existence of metatarsal II and metatarsal IV at Olduvai,

of which it is stated that ‘they originally had practically the same length as

metatarsal III and that they are broadening distally’ (Hopwood, 1937).

However, no phalanges are mentioned.

Unpublished data

Examination of the Hipparion third metapodials from Olduvai Bed II,

now in Berkeley, California, reveals conclusively that the lateral digits were

robust, reached the distal ends of the metapodials, and were as well developed

as those of Hipparion from Pikermi. The metapodials II and IV are not repre-

sented by complete specimens or by the distal articular end. Measurements of

the fragments are considered to be of no value. Despite the lack of complete

lateral digits, the evidence of the articular grooves formed by them on the

third metapodials convincingly indicates that the East African Hipparion were

three-toed. Furthermore the data suggests that they possessed slightly longer,

but more massive, limbs than the Pikermi Hipparion (see also p. 363).

Table 13 indicates the dimensions of the available third metapodials

from Bed II, compared with two specimens from Pikermi.

It appears that the African hipparionids have hardly modified their

locomotor apparatus since the Pliocene.

THE ECOLOGICAL ADAPTATIONS OF Hipparion

Considerable plasticity exists in biological organisms. Factors such as

climate (with such variations as in temperature and humidity) considerably

affect water supply, food, animal and plant associations and the actual nature

of the environment, e.g. savanna, sand dunes, forests. The range of an intra-

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 353

Proximal Distal

Metacarpal II Total Proximal breadth Distal breadth

length A-P (max.) A-P (max.)

EQ5 7s LLS Dash =: Q17 36 46 33 42

F. 3457 .: 206 37 44 34 °45

1955. BK II 45 ae 223 37 >45 36 48

MEAN .: 215 37 45 34 45

Pikermi UC634.02** 206 31 40 29 38

Metatarsal II

1957- SHK II 557* 253 38 48 35 45

1957. SHK IT 729/730 260 37 47 36 45

1955. BK II 68 ee 257 40 46 35 45

1953. BK II . 242 37 45 34 44

HOR eo Sal Mya) oe 254. — — 33 >42

EQ575) BERGHE 663) - —- 41 46 — —

UORGY IN IND gi ye — — — 38 48

1941. SIF 797 ys = = = 32 43

MEAN ae 253 38-6 46°4 36 44°7

Pikermi UC63415** .. 241 37 45 33 40

* These appear to belong to one individual.

** University of California, Berkeley (Dept. of Paleontology).

+ There is a unilateral pathological growth at the distal end.

TABLE 13. Dimensions (mm.) of third metapodials of Hipparion from Olduvai Bed II.

All the specimens, except the two labelled ‘Pikermi’, are from Olduvai.

species variation may be considerable and will allow extensive pliability in

adaptation to changing environmental conditions. Such modifications may not

change the genetic background or consequently, the taxonomic status. In

comparing different anatomical patterns, this concept must be borne in mind.

On the other hand, it is also obvious that the environment may favour,

either the greater adaptive potential of some individuals in a particular popula-

tion or mutations, so that selective forces may operate to promote different

races and eventually a new species. Therefore, in a survey of a biological

group which extends over a vast continent, it is essential to appreciate the

plasticity of the organism.

Such a study has been successfully developed for the family Equidae, and

the hipparionids in particular, because of the vast amount of available fossil

material, both in America and in Eurasia. Following other scholars, Gromova

(1952), in her revision of the genus Hipparion, has emphasized its variability

and evolutionary trends in relation to ecological factors.

The aims of this section are

(a) to summarize the main conclusions about the ecological and functional

significance of Hipparion characteristics;

(b) to summarize the available information concerning the African biotopes

occupied by Hipparion from Upper Miocene to Pleistocene times; and

African Hipparion under the influence of these bionomical circumstances,

354 ANNALS OF THE SOUTH AFRICAN MUSEUM

and thereby to contribute to the appreciation of the taxonomic status of

the African groups.

ECOLOGICAL AND FUNCTIONAL SIGNIFICANCE OF HIPPARION FEATURES

The overall dimensions of the Hipparion skeleton do not allow definite conclu-

sions about the biotope. It is known that the smallest races of extant Equidae

are to be found in the driest areas characterized by drier food and, in the

northern part of their dispersion, in areas of poor economic conditions. It is

probable that similar factors have influenced the Hipparion-associated fauna,

and that their general dimensions illustrate in some way the nature of the

biotope. However, larger and smaller forms (? species) have repeatedly been

recovered from the same deposit so that it appears that they co-existed in the

same climate, but they may have occupied, within the same area, slightly

different ecological niches. On the other hand, this is not a general rule because

exceptions have been documented. For example, at Pavlodar (on the right

bank of the Yrtych, Moldavia, U.S.S.R.) where it was possible to make inde-

pendent studies of the nature of the biotope at different ecological stations, a

species characterized by longer and more slender extremities and reduced

lateral digits, viz. H. longipes, lived in a drier habitat than the shorter-limbed

H. elegans did. Therefore, it is not possible to make any direct inferences about

the biotope from a mere consideration of the overall dimensions of the animal.

Various factors are probably involved which cannot be adequately isolated.

Relative dimensions of cheek teeth and incisors seem to be largely influenced by

the quality of the food: a drier, more steppe-type grass usually develops larger

teeth. However, the incisors and the cheek teeth may show differing reactions

and adaptability to this ecological feature. This has been demonstrated by

comparing species presenting large cheek teeth with species having small

cheek teeth, e.g. H. gratum and H. longipes, which show comparable incisors.

The relative size of the teeth in proportion to the overall dimensions of the

animal must also be considered, a smaller animal normally having relatively

larger teeth.

The development of the preorbital fossa, which has become an important

taxonomic consideration, is in the present state of our knowledge not related

to any ecological feature. Animals with large and with small preorbital fossae

are found together in the same xerophytic (H. proboscideum and H. matthewi, at

Samos) and moist habitats (H. moldavicum at Taraklia, or H. theobaldi in the

Siwaliks). Preorbital fossae may even be missing (H. platygenys at Taraklia)

without any apparent ecological reason.

The isolated protocone; completely detached from the protoloph and the

protoconule right down to the base of the crown, is a constant feature of the

hipparionids. It obviously weakens the structure of the tooth. This has to be

compensated for, especially in the case of a hard and dry grass diet, by a

greater development of the cement in the anterior and the posterior valleys so

as to bind and strengthen these individual elements (Stirton, 1931).

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 355

Two elements are involved in shaping the protocone: antero-posterior

length and transverse breadth. It is known that, relative to the total dimension

of the tooth, the A—P length of the protocone is more constant than its transverse

breadth. Therefore, a protocone may look elongated, but this is only a reflection

of its narrowness. Thus it is advisable to express the functioning structure of

the protocone by means of two indices:

(2) the protocone length is expressed relative to the general mesiodistal

measurement of the tooth, i.e. the ‘length index’.

(6) The so-called ‘shape index’ expresses the length of the protocone relative

to its breadth (table 14; figs. 12, 193).

These indices highlight the differences in adaptation between the African

and Eurasiatic Hipparion.

The two indices for the African forms are compared to those for 7 Eurasiatic

Hipparion species. Available measurements of the total dental series have been

utilized. It may be noted from figure 12 that the length index of the African

Hipparion is constantly higher for each tooth than that of the corresponding

tooth of any of the Eurasiatic species. However, in both groups the length

index tends to increase in a mesiodistal direction. On the contrary, the shape

index (fig. 13) is constantly lower for the African forms, decreasing for both

African and Eurasiatic groups in a mesiodistal direction. Thus it may be

concluded from the two indices that the protocone of the African Hipparion is

both relatively longer (elongated) and narrower than in the Eurasiatic forms.

The shape of the protocone varies with progressive wear, the section

tending to become more oval and less elongated. But independently of these

modifications, for particular locations in the dental series at comparable

levels of attrition, there seems to be also an adaptive elongation or broadening

of the protocone which is linked with the type of food it has to deal with. It

is suggested that the efficiency of the protocone’s function in trituration lies

in the action of its buccal and lingual enamel crests, developed perpendicularly

to the lateral chewing movements of the mandible. It is obvious that the

efficiency is maximal for a narrow and elongated protocone, for here the

crests are at one and the same time the longest and the most perpendicular.

Thus they would favour a dry and harder bunch grass. This suggestion is

strengthened by the contemporaneity of the lengthening of the protocone and

the development of a typical xerophytic vegetation in America.

Enamel plications

It has been successfully demonstrated that the complexity of the enamel

pattern on the occlusal surface of Hipparion molars has developed synchronously

with the drier environment in Upper Miocene and Pliocene times. Like the

elongated protocone, the numerous enamel plications are oriented perpendicu-

larly to the movement of the jaws, and are increasing the triturating power of

the teeth for coping with harder grass. Although there seems to be a parallel

increase in the degree of plications with increasing toughness of food, there

ANNALS OF THE SOUTH AFRICAN MUSEUM

2 Xopuy odvyg pure xopuy yisue7T suoc0j}01g ‘PI alav],

*.yq99} UIOM AloVeIOPOW IO 9]II], 1OJ ‘1Q61 ‘teepuog WOT poyepNd[Vd IO 60} SurpAlovy 9

‘oGS61 “eAOWOAT) WOT, P9ye[NoTed IO 60} Surpr0s0y p

‘9S61 ‘Sinoquivay Wo Ay pozE[NI][Rd 10 £0} SurpsOIIV »

‘6S61 ‘Smoquivry Wo poyepNoyeo 10 £0} Surpszovoy q

‘yysusy 9U0D0}0Id 0} DATVL[II YYPRoIq 9UOI0}0Ad = (*]'G) xopuy odeyg

“YISUIZ] YIOO} 07 SALIVA YASUI, 9UOD0}0I1d = (*['"J) XOpUy YISUS'T »

356

6.9P 1.G§ G.6r7 | £.06 | 9.09 | €.4a | €.19 | 6.9% | g-99 | 3-9% | 6.zg | 1-9% | S-09 | 6-L4z | 1.9G | §.08 PS | S.9€ | beh | 1.GE |

:ueoUl

erouon) 9

9-9P o.GE :Yyq907 UIOM

AJo}esO pou IO ap] AOJ UBOUT *q

6.9S-P.96 | 9-Gh-S.9z :UOIVRLIVA JO 98uRI “Dd

‘aN 10 LN

vd ‘ed

6.40 | o.a | ¥.zS | 1.6% | 9.PS 1 | G.E9 | £.-4% | g.&S | z-0& | 1-LP | g.bE | 9.06 GE bE gh | 6.26 gb aN

6.LY g& 6.47 | 1.18 | a-LG | 9.92 gS | £.Lz | L.bg | 3.9% 1-09 of LG | €.2§ ZW 101A

6°19 | 9-16 gh g& | 9-9F | 9-28 ZW

GG AS MW Mracohs) We /Z, chile, WN

6.6F L.VE 1.1G | 3-0& | S.tg gz | 9-9 | a-gz | 6.g9 | a-ba g-09 | £.Gz | &.65 | F.gz =f 1) ey

ere) Qo GG ib | &.E | g.oP rd

g-b9 | 9-Fz | 09g bE | o.Sb | &-SE ed

olKG || Garces OLA OSa| GOO Geos.| Oreo) | hee \nGeeG) (NOatcllec.GQ | 12 |k6.6S) | Q.ce LG zz | bg | 9.Sz al

TS sled “EtS) TT US tat aS lea ES “eal eS ILA Ts alee las: EAL Is =) 08 | JESS Ud

alpppmssiuaoy glunuv4forsad pluna1app oul

SamuvvqgasuvT ‘H gasuapnauos "FT | ,avaowuoss * FY ‘H pscuasqn "EY ‘H psupiaya "FT asuafins “FT | quinuvaisfo “Py

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 357

@ H. africanum

@ H. sitifense

Hypsohipparion

Langebaanweg

. elegans

H. moldavicum

. platygenys

. periafricanum

. gromovae

. concudense

. koenigswaldi

P/EN

Fic. 12. Protocone length index in the various teeth of African and non-African Hipparion species.

3 58 ANNALS OF THE SOUTH AFRICAN MUSEUM

e@ H. africanum

m H. sitifense

Hypsohipparion

Langebaanweg

. elegans

. moldavicum

. platygenys

. periafricanum

. gromovae

. concudense

- koenigswaldi

dpoepocono px

Fic. 13. Protocone shape index in the various teeth of African and non-African Hipparion species.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 359

must be a limit to the former. The quantity of available enamel is fairly constant

and probably limited, so that an increase in the complex pattern of the enamel

will result in a thinning out of the plications. Eventually this would produce

too weak a support for the triturating pressure required for hard grass. When

this point is reached, the plications again reduce in number and the protocone

lengthens as a compensatory mechanism.

Therefore, it is not always possible to ascertain from the small number of

plications whether they represent a primitive condition corresponding to a

humid environment, or whether they are secondary, resulting from a reduction

under the influence of a hard grass food. Then other factors must be considered,

e.g. the shape of the protocone, to solve the problem.

A directly comparable situation is encountered in the development of

proboscidean teeth. Here an increase in the available grinding surface of the

enamel ridges (lamellae) could represent the increasing plications in the teeth

of the Equidae, especially as they are also oriented perpendicular to the grinding

movement. This parallelism is also demonstrated in the hypsodonty of these

two groups.

Hypsodonty

In the evolution of the Equidae most lines show a constant increase in

the crown height of the cheek teeth. However, from Merychippus, in Miocene

times, the progressive hypsodonty has advanced along two phyletic lines

(Hipparion and Pliohippus) at a faster tempo, again under the influence of a

general increase in aridity. The same observation has been made among

Bovidae. It seems obvious that it is once more an adaptive response of the

organism to the xerophytization of the diet.

The dietary changes involved in the evolution of Cenozoic hypsodont

horses probably corresponded to a shift from browsing on softer herbaceous

plants to grazing on harsh, siliceous grasses, which acted as abrasive agents on

the teeth. These changes are, in turn, a consequence of an increasing aridity

as has been clearly demonstrated for the Great Plains of America (Stirton, 1947).

The association of widespread sandy deposits and of bunch grass of the St¢pzdium

type, with the increasing aridity of this area is actually known to be contempo-

rary with the development of hypsodonty from Lower Miocene times onward.

This increased height of the crown has played an important role as a protective

modification in the dentition. It is supposed, although impossible to prove

from the fossil record, that a mutation or other factors, favoured an increase

in activity of the odontoblasts and ameloblasts to lay down more dentine and

enamel when the tooth was still in the formative stage. The root closure (fusion)

may have been considerably retarded as well. Furthermore, it may be suggested

that dry climate has influenced the teeth by necessitating some protective

mechanism not only because of the harsh conditions of the grass, but also

because of the sand mixed with it as a foreign substance. Stirton (1947) states

that in the Great Plains area, a close inspection reveals much sand adhering

360 ANNALS OF THE SOUTH AFRICAN MUSEUM

to grass or lodged in between blades and stems, where it is blown by wind or

splashed after thunder showers.

The double knot in the lower cheek teeth of the hipparionids displays three

typical features:

(a) a primitive ‘stenonis’ type, inherited from Merychippus and generally

retained throughout the evolutionary sequence among American species;

(b) a more progressive, typical “Hipparion’ type has been developed by most

European forms; and

representatives.

It has been suggested that the double knot of the lower teeth was shaped

and typified under the influence of the protocone structure in the upper teeth,

with which it is functionally linked in occlusion. There is good reason for

developing a long, narrow metastylid parallel to an elongated, elliptical

protocone, as is found in most African Hipparion. Here, the shape of the metasty-

lid in particular, and consequently of the whole knot, conveys useful information

about diet and climate. However, the correlation is probably not a simple

one, because American Hipparion display an elongated protocone and a rounded

‘stenonis’ type of metastylid simultaneously.

Total tooth structure

Upper teeth are rather compact. The outer enamel forms a fairly con-

tinuous and parallelipipedic structure, only grooved slightly by a small hypo-

glyph. The only weak structure which can be broken away from the tooth

under excessive strain is the protocone, separated from the rest of the tooth by

the internal depression. In the Langebaanweg material, this plane of fracture

is noted in L935, Lo42, Lo45, Lg56, S.A.M.11722 and S.A.M.11724 (pls. 1, 7,

8, 9). It has been noted above that, with increasing dryness of the food, this

weakness is largely corrected by the development of a narrow protocone closely

adherent to the protoloph.

Lower teeth, on the other hand, are built on a very different pattern.

The many conids and stylids are much more individualized: they are contig-

uous but are separated from each other by rather deep depressions right down

to the base of the crown. These intervals are filled with cement, but it is softer

than the enamel and does not always supply sufficient compensatory protection,

as can be seen in many fossil teeth where the cement has disappeared. In drier

environments, the hard food influences this structure of the lower teeth,

demanding additional strengthening of the enamel-dentine pillars which are

embedded in their cement coating to prevent both fracturing and excessive

wear of the tooth. Support can be provided in two ways, either

(i) by reducing the depth of the internal and external depressions separating

protoconid from hypoconid, metaconid from metastylid, and metastylid

from entoconid, or

(ii) by additional stylids.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 361

Reduction of the depth of the depressions is commonly observed. As has

been stated previously, this occurs on the lingual surface by the development

of a ‘caballus’ type of double knot, while on the buccal surface it is secured by

a broader connection between the two halves of the ectolophid.

The alternate solution lies in the construction of additional pillars:

ectostylid and ptychostylid in particular, just in front of the external depression,

and protostylid and hypostylid on the antero- and postero-external corners.

Entostylids are occasionally developed in the postero-internal depression. In

one particular case (a ‘Stylohipparion’ Gar. Fluss no. 200, figured by Dietrich,

1942, pl. XIII, fig. 95), a strong additional pillar is seen between metastylid

and metaconid.

The higher frequency of these additional pillars in milk teeth seems to have

been necessitated by the lowness of the crown of these teeth and their softer

construction. This is the obvious way to prevent too rapid attrition which

cannot be compensated as there is no continuous growth of the tooth.

As suggested previously, enamel plications, shape and size of protocone,

structure of the double knot, increased hypsodonty and additional pillars are

many different methods of meeting the mechanical requirements of a tougher

diet. They do not all necessarily develop together at a parallel tempo. Some

of them are functionally linked but other elements develop compensatory roles

so that all need not be present at one time. Therefore, no single element can

give reliable information about climate, biotope and the quality of food, but an

assessment of the whole picture and of the reciprocal values of the different

features must be considered.

The extremities

Duerst (1926) has successfully demonstrated how a humid climate and a

soft grass develop a heavier type of equid with broader extremities (eurysome

type), while a leptosome type of a smaller, more slender animal with more

gracile extremities tends to be built in drier biotopes characterized by a

xerophyte vegetation. This seems to apply especially to metapodials which

become relatively longer.

The progressive reduction in size and the disappearance of the lateral

digits in equids, with the concentration of the whole body weight on the third

metapodial, has always been interpreted as an adaptation to increasing speed

in steppe surroundings. It has been repeatedly suggested that the high tempo of

evolution of Pliohippus towards monodactyly is an expression of its drier habi-

tats, while the lateral digits of African Hipparion are kept more or less functional

right into the Pleistocene, probably because of a more humid environment.

THE UPPER MIOCENE AND PLEISTOCENE BIOTOPES OCCUPIED BY AFRICAN

HIPPARIONIDS

Our knowledge concerning past climates and biotopical conditions in

those areas where African hipparionids have been recovered is still inadequate.

362 ANNALS OF THE SOUTH AFRICAN MUSEUM

The fragmentary information spans over a very long period of time (probably

10-12 million years) and is derived from a large continent where the climates

have certainly been variable. Consequently it is not always possible to draw

accurate conclusions from the collected data, nor is it possible often to correlate

data from one site with those of another.

Furthermore, this information is deduced from fossil assemblages of which

some are listed on the basis of debatable fragments, while others are wrongly

identified. For example, at Olduvai Gorge, recent and more extensive collections

will permit a better analysis.

Even where there is sufficient and accurately identified material, it

should be remembered that a fossil specimen is not necessarily found in the

typical habitat of the living form. It is well known that in periods of drought

or volcanic eruptions (which have played such an important role in the Ceno-

zoic sedimentation of East Africa, from Omo to Eyasi), animals have concen-

trated around swamps, temporary pools and tuffs, possibly fairly distant from

their normal biotope.

This biotope is often less restricted than one imagines, and the adaptation

of some organisms is quite fantastic. So, it seems futile to draw definite conclu-

sions from the presence of one or even a few animals. Real associations must

be considered to avoid misinterpretation of the presence of a species at a

particular site.

(a) In the Upper Miocene, the Oued el Hammam faunal assemblage

contained Palaeotragus, Samotherium, Damalavus, Gazella, Dicerorhinus, Hippopota-

mus and Hyaena. With slight generic differences, it corresponds very closely to

the fauna of the East African plateau of today, and indicates the typical

savanna with mimosa biotope of the tropics. According to Arambourg (1959),

the presence of Macaca and Cephalophus in the contemporaneous deposit at

Marceau suggests the proximity of forests. This is supported by the location

of the deposit in a higher and already tilted portion of the Tellian Atlas, and

by the lignite abundant in the deposit.

(b) The Lower and Middle Pleistocene faunal assemblages of the East

African Hipparion-bearing sites clearly indicate the same type of steppe or

plain and highland (mimosa) savanna environment, very similar to that of

the present day, ‘with a marked bias toward somewhat moister conditions in

most cases’ (Cooke, 1963). The main feature at Omo, Kaiso, Kanam, Olduvai

I-IV and Eyasi is undoubtedly that of an open savanna. No true forest associa-

tion is to be found in these areas, although forest strips and moist woodland

occurred occasionally, as is indicated, for example, by the presence of Nesotragus

moschatus. The cyclical and seasonal variation of the rainfall has provided

more humid conditions at times, and consequently the development of swamps

and (or) lakes in the lowlands, with very luxurious vegetation, are responsible

for the abundant Suidae, Bovidae, Giraffidae, Elephantidae and Equidae. A

swampy environment is explicitly suggested at Olduvai by Aonyx, and at

Omo by Kobus. A lake shore or grassland adjacent to water is evident both at

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 363

Omo and Olduvai (e.g. Hippopotamus, Omochoerus) where this is also reflected

by the nature of the deposit.

The presence of Oryx and Beatragus hunteri would suggest arid conditions

prevailing in the locality.

Although most of the fauna are suggestive of plains or a plateau, Tragelaphus

nakuae and Dinopithecus brumpti indicate that a mountainous biotope was

probably part of the picture at Omo.

and the prevailing feature remains that of a steppe or a savanna fauna. The

Vaal River Gravels suggest a fauna of thornbush and grassland. The cave

breccia—probably less representative of a true biocoenose—contains both

open plains and dense bush-dwelling forms. It is believed that the rainfall

varied considerably during the Pleistocene, with oscillations from 20% to 40%

annually (Brain, 1958). The fossil forms represented there would have tolerated

such changes. In any case, even in the driest conditions (Taung) well-watered

bush valleys could be located at not great distances.

In conclusion, it may be stated in broad terms, that African Upper

Miocene and Pleistocene Hipparion sites seem to have been areas of (dry to)

moist savanna and grass steppe, or of open woodland with mixed savanna. It

is not possible to define with any more precision the environment proper to

Hipparion in this general biotope which extended over half of Africa for several

million years. It is probable that local circumstances and features have directed

the relationships and balance of the various species in particular areas: but as

a whole, and in spite of replacement of species, the African fauna has maintained

an essentially constant composition throughout late Cenozoic times, which

indicates that the broad ecological pattern did not vary greatly from the

Upper Miocene onwards. The main ecological feature, viz., the savanna, may

at times have been extensively altered by moister climates producing swampy

conditions and even patches of forest. Then later, it would be altered again

by a rather low rainfall. Nevertheless, drastic changes probably never occurred.

There seems little evidence of a true forest fauna or of conditions too dry to

sustain a normal steppe biocoenose (Cooke, 1963).

EFFECTS OF BIONOMICAL CONDITIONS ON THE EVOLUTION AND THE MIGRATION OF

AFRICAN HIPPARIONIDS

It has been generally accepted that the three-toed hypsodont Hipparion

were typical steppe forms. The savanna biotope is commonly emphasized as an

important equid environment of the late Tertiary in the Great Basin of the

United States, and the vast reduction of this habitat is strongly suggested as a

suitable explanation for the extinction of Hipparion by the end of the Hemp-

hillian (Shotwell, 1961). The life span of Hipparion in North America, namely,

Clarendonian-Hemphillian, corresponds rather precisely to the extension of

the savanna or steppe biotopes in this area. The recent observations of Sondaar

(1961) strengthen this view: at Nombrevilla, in Spain, where Decennatherium,

364 ANNALS OF THE SOUTH AFRICAN MUSEUM

Aceratherium and Mastodon were recovered together with Hipparion, there is

good evidence for a steppe environment. However, it has become clear that

HMipparion is not restricted to such narrow ecological conditions. Thenius (1950)

interprets H. gracile as a forest-dwelling form, and Kurtén (1952), on the

basis of the Chinese ‘Pontian’ fauna, enforced the necessity of a reconsideration

of this hypothesis. Because of the almost identical frequency (13°) of Hipparion

in both the gaudryi and dorcadoides faunas, characteristic, respectively, of the

‘forested’ or ‘southern’, and of the ‘steppe’ or ‘northern’ provinces (the classical

distinction of Schlosser, 1903), it was suggested that the Chinese Hipparion

thrived in various surroundings, or at least that the genus as such was not

exclusively a steppe form. Furthermore, it is not possible to make a direct

correlation between the geographical distribution of a particular species and

its biotope. Of the eleven different species recorded by Sefve (1927), some are

confined to one locality and others turn up in widely different areas; in some

instances, up to four different species have been recorded from the same fossil

pocket.

The adaptability of hipparionids explains their extraordinary diffusion

both in North America and in the ancient world, from Upper Miocene times

onwards. Ecological conditions influenced structural features in different and

varying ways, often difficult to explain. A similar geographical distribution

and even the simultaneous presence of different fossil forms in one particular

pocket are compatible with slightly different ecological niches or habitats in

a small vicinity.

Under favourable conditions, observations have often been made of the

association of more massive animals (with larger skull, shorter limbs, longer

lateral digits and strongly plicated enamel) with open forested areas, character-

ized by soft and wet ground and tender vegetation. Lighter forms (with slender

and elongated limbs, reduced lateral digits and simpler enamel pattern,

elongated and narrow protocone and increasing hypsodonty) are found in a

more xerophytic environment. Outside Africa, H. primigenium, theobaldi, crassum

are representatives of the former group; H. elegans, longipes, moldavicum,

proboscideum, matthewi are typical of the latter.

In general, the various characteristics observed in the African Hipparion

are Clearly indicative of animals adapted to a rather xerophytic steppe environ-

ment, although this never seems to have been very severe, and was certainly

not constant everywhere. It has been stated above that nowhere in Africa

are Hipparion-bearing sites located in true desert areas, and none of the forms

recovered shows the extreme leptosome type displayed by some North American

forms, e.g. H. whitney.

Hipparion africanum, with its large skull, short and massive limbs with

lateral digits well developed, and rather less hypsodont teeth, is probably,

from the inadequate data available, the least adapted to a steppe environment.

However, the enamel plications and the narrow protocone show an orientation

in that direction. It is difficult to assess whether this moderate expression of

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 365

the steppe features corresponds to the precocity (in time) of H. africanum or

reflects a humid and even partially forested environment.

The moderate transverse flattening of the protocone, the few and simple

plications, the absence of additional stylids, the reduced hypsodonty and the

small dimensions of the teeth of Hipparion sitifense, as well as the development

of the lateral digits, are to be interpreted in the same way, although the absence

of skeletal remains does not permit evaluation of the degree of massiveness.

The postcranial skeleton of the African Pleistocene forms is practically

unknown. However, quite a few metapodials have been recovered in the

Olduvai Gorge (Tanganyika). They are not particularly slender, and the

lateral digits are still well developed, but the absolute length of the third

metapodial is quite remarkable (metacarpal: 217-223 mm.; metatarsal:

242-260 mm.), being 5-9% longer than the corresponding elements from

Pikermi (table 13).

The typical features of the dentition are strongly marked hypsodonty, the

fairly complicated enamel pattern of the ‘mark’ walls, and the elongation of

the protocone, which may become very flattened. The extreme flattening

corresponds to a spectacular development of the additional stylids. It seems

clear that these features must be interpreted as a positive pattern of adaptation

to a steppe environment throughout Africa. However, the absence of any

extreme slenderness of the metapodials and the complicated enamel pattern

seem to strengthen the conclusion drawn from the faunal assemblages, viz.

that the xerophytic environments of the African Hipparion were never very

severe.

In the absence of sufficient cranial and postcranial remains, and because

of the vicarious role played by the different tooth structures in their adaptation,

it is unnecessary to comment any further on the special features of every single

group so as to define more precisely the particular environment of each.

Discussion is further limited by the fact that climate and biotope seem to have

been rather uniform over a great part of Africa, and have certainly not under-

gone drastic changes. Throughout the continent, African Hipparion reflect in

their known morphology the vastly extended steppe or savanna environment

with an absence of much aridity.

UNPUBLISHED MATERIAL FROM THE VAAL RIVER DEPOSITS, SOUTH AFRICA

Sydney-on-Vaal and Pniel

The exact location of these two specimens is not known (fig. 6). Cooke

(1949) describes other material from these sites which are in the northern

Cape Province.

Material:

MMK 431 —left M3

MMK 5225—left M, or M,

These specimens were diagnosed by the authors as belonging to Hipparion

366 ANNALS OF THE SOUTH AFRICAN MUSEUM

in the collection of equid material which was kindly sent on loan by the Director

of the McGregor Memorial Museum, Kimberley, South Africa.

MME 431 (pl. 1)

This is a left M? with fairly high crown, in an early stage of wear.

Half of the ectoloph: fairly deep, flattened arc, being angulated at the

parastyle and slightly angulated at the mesostyle.

The parastyle: the lateral angle is chipped away at the occlusal surface

but near the base it is seen to be prominent. It is rounded and projecting, and

is separated by a fairly deep groove from the ridged anterior angle. The surface

is very oblique, almost in the same plane as the anterior surface.

The mesostyle is partly broken. It is fairly large and rounded, and it has

a deep groove.

The metastyle tends to be broad with a ridged lateral angle, a smaller

ridged posterior angle and a slightly concave surface, which is oblique.

The protocone is isolated. It is very long, being about half the length of

the tooth. Anteriorly and posteriorly, it is sharply angulated. Its inner surface

is flattened, and the lateral, outer surface is wavy, presenting an elongated

cigar shape.

The protoconule is short and has a flattened arc. Posteriorly it becomes

markedly angulated and turns laterally to join the base of the pli caballin.

The hAypocone is elongated and flattened. Its two sides are almost touching.

The hypoglyph is deep and circular, demarcating an angulated hypostyle.

The hypostyle is joined to the metastyle by a concave arc of enamel.

The pli caballin is long, with unequal sides almost touching the protocone.

The prefossette: the anterior wall has a small plication and a very deep,

narrow pli protoloph which tends to separate the medial part of the prefossette.

The posterior wall has shallow plications and a deep narrow pli protoconule

which tends to separate the medial part of the fossette.

The postfossette: the anterior wall has an irregular shape, with one plication

and a very deep and narrow pli postfossette, which tends to separate off a

flattened portion of the medial part of the postfossette. The pli hypostyle is

duplicated. It is fairly deep, tending to isolate the medial part of the postfossette.

The anterolateral angle of the postfossette tends to form a rounded, almost

isolated pillar.

MME 5225 (pl. 1)

A left M, or M,, probably M,.

The tooth is partly fragmented, especially on its inner aspect.

The protoconid is partly broken mesially. It appears that the other wall is

flattened. The protoconid is shorter than the hypoconid. The metaflexid is

elongated, flattened and biconcave with rounded anterior and posterior ends.

The metaconid is broken away. Its outer wall appears to be a flattened arc.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 367

The parastylid is partly broken but its posterior wall turns sharply inwards

and almost touches the outer wall of the metaconid.

The metastylid: the outer wall is flattened in a mesiodistal direction. The

entoflexid is elongated, flattened and anvil-shaped.

The entoconid is rounded and pillar-like, being almost completely separated.

The hypoprotonocid groove is deep and wide, being rounded internally. It

has a narrow mouth laterally and a bulge posteriorly.

The piychostylid is finger-like, projecting anterolaterally. The sides are

almost parallel and it has a narrow base.

The ectostylid groove is a marked indentation, broad and shallow.

The hypoconid has a flattened, elongated outer surface with posterior parts

tapering sharply inwards.

The ectostylid is an isolated triangular pillar, the outer wall being chipped

at the occlusal surface, but it can be seen to be flattened near the root. The

inner wall is flattened and meets the outer wall at a sharper angle. The two

walls are separated anteriorly by a somewhat irregularly concave base. The

inner wall of the enamel joins the metaconid.

The metastylid forms a distinctive broad arc near the base of the tooth where

it is intact.

MMK 5225 MMkK 431

A-P length .. Bie - 64 23°5 A-P length PHOG

Breadth — Breadth .. QI°l

Max. A-—P fenete MEG OcOnidil 8-1 Breadth/A—P 94.°7

Max. A-P length hypoconid. . 12 Protocone length A—P 12°5

Max. A-P length metastylid.. 6-2 Prefossette A—P length 10°8

Max. A-P length entoconid .. 4°6 Prefossette breadth ats 8°3

Max. A-P length entoflexid.. 11-8 Postfossette A—P length .. gio) u@ses

Max. A-P length metaflexid .. 6-8 Postfossette breadth Wo 7-0

Breadth hypoconid 5°6 Protocone A—P/total A—P length 56-2

Breadth entoconid 471 Crown height... : .. €a'70

Crown height .. a .. Ca 40

TABLE 15. Dimensions (mm.) of MMK 431 and MMK 5225 from the Vaal River Deposits.

Taxonomic status

The above two teeth are referred to Hipparion (Stylohipparion) libycum (see

p- 387-92 for diagnosis).

MATERIAL FROM LANGEBAANWEG, CAPE PROVINCE, SOUTH AFRICA

DESCRIPTION OF HIPPARION TEETH

Milk dentition

Material:

S.A.M. 11717—right DM?

S.A.M. 11718—right DM?

368 ANNALS OF THE SOUTH AFRICAN MUSEUM

S.A.M.11717 (pl. 2)

A right DM? in medium stage of wear with a thick layer of cement

particularly on the inner side.

Half of the ectoloph: it is shallow and has a slight arc. It is sharply angulated

at the mesostyle and slopes up into the parastyle.

Parastyle: the lateral angle is chipped but it is seen to project markedly

in a lateral direction near the base of the tooth. The surface is oblique and not

quite in the same plane as the anterior surface of the tooth. There is a marked

groove. The anterior angle is rounded and forms a ridge.

The mesostyle is stumpy, projecting, and flanging at the surface. A fairly

marked overlap exists anteriorly. The surface is rounded, not grooved. The

sides tend to be parallel.

The metastyle is slight and rounded.

The posterior ectoloph has a vertical bulge at the centre of its surface.

The protocone is oval in a mesiodistal direction, with the medial side slightly

more flattened than the other side. It is distinctly separated from the other

cones by a thick layer of cement.

The protoconule is an elongated, flattened arc with a fringed effect.

The protoglyph is deep and wide with an irregular base.

The hypocone is elongated, with a rounded medial side and a flattened

lateral side, tending to come to a point posteriorly. Anteriorly it is almost

completely isolated by a very deep indentation of the hypoglyph. Opposite,

there is an indentation tending to separate it from the metaconule. The

metaconule also has a fringed appearance.

The hypoglyph has a broad mouth. Lateral to the mouth of the hypoglyph,

the enamel shows a V-shaped indentation.

The pli caballin is long, narrow and arched, tending to touch the protocone.

At the base of the pli anteriorly, there is a trace of a duplication of the pli.

The prefossette: the anterior wall shows two deep plications and a deeper,

narrow pli protoloph which tends to isolate the medial part of the prefossette.

The posterior wall of the prefossette shows numerous plications which are

markedly complex. One of the plications adjacent to the pli protoconule is

large and has become isolated, forming the typical island lateral to the base of

the pli caballin, and wedged between the medial portion of the pre- and

postfossette. The pli protoconule is deep and narrow, almost touching the

medial wall of the paracone. The pli prefossette is deep and complicated with

a bifid base, tending to isolate the posterolateral angle of the prefossette.

The postfossette: the anterior wall has numerous plications that are not as

complicated as the posterior wall of the prefossette. The pli postfossette is

deep and narrow, tending to isolate the medial part of the postfossette. The

posterior wall of the postfossette shows numerous plications, the most lateral

one being rather deep. The pli hypostyle is deep and angulated, tending to

isolate the medial wall of the postfossette.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 369

S.A.M.11718 (pl. 2)

This is a right DM? of the same individual as 11717.

The anterior portion of the tooth is broken away but the remainder of the

tooth shows the same features as 11717 with the following differences:

(a) Protocone is more elongated and more oval.

(6) Hypocone is more oval.

(d) Pli protoconule is more complicated, and is duplicated.

Dimensions S.A.M.11717 Hypsohipparion

(DM®) (DP?-DP* __ )

A-P ue ar 31°9 28-34

Transv. breadth .. 26:2 23-26

Heights es. ae QI 28-34

(medium wear) (early wear)

TABLE 16. Dimensions (mm.) of Langebaanweg milk

molar compared with ‘Hypsohipparion’.

The features which are observable on the two Langebaanweg milk molars

are identical to the description given by Dietrich (1942) for the Hypsohipparion

material (19 individual teeth and 2 maxillary fragments). The isolation of the

hypocone and the shape of the protocone of Dietrich’s (1942) specimen illus-

trated in his figure 162 is typical of the Langebaanweg milk teeth. The

dimensions of the latter fall into the range of Dietrich’s specimens (table 15).

Lower permanent dentition

Material:

Pe Left: L947, L1465p (pl. 6).

Right: Lg3q9 (pl. 5).

POF Ey: Lett: 1937, 1.943 (pl. 6).

Right: Lg41, L946, Lo56, L1451a, L1465a (pls. 5, 6,

8.1):

M, or M,: Left: Lg44, Lo52, Lo50, L1448 (pls. 6, 8).

Right: L948, Lg54 (pl. 7).

M;: Left: L938, Lo49, L1465B (pls. 8, 9).

Right: L1465F (pl. 2).

One specimen, L1465c, a M, or Mg, is excluded because of its size and

its fragmented appearance which does not permit recognition of any typical

hipparionid features.

The following specimens seem to belong to single individuals:

1. L937, Lg39.

2. Lo41, L943, Lo44, Lo48.

3. L954, L957.

4. L938, L959.

370 ANNALS OF THE SOUTH AFRICAN MUSEUM

Classification of specimens according to wear:

A. Unworn or slight wear: Lo49.

B. Early to medium wear: L947, L1465p, Lo39, L937, L943, Lo41, L946,

Li451A, L1465a, Lo44, Lo52, Lo59, L1448, Lo48, Lo38, L1465z,

L1465F.

C. Late wear: L954, L957.

General description

All the Langebaanweg specimens obviously belong to one group, viz.

Hipparion (Hipparion) albertense baardi subsp. nov. The dimensions are presented

in table 17. The following features are distinctive:

1. Entoconid: Usually in P, it is more or less flattened in a bucco-lingual

direction. It tends to be more rounded or quadrangular in shape in Ps, P,,

M, and M,. However, in early wear a slight indentation of the enamel produces

a ridged effect on its lingual surface mesially.

2. Metaconid: It has a slight tendency to a bilobed formation due to a

fairly marked indentation on the mesiolingual aspect, which is only recognizable

in a very early stage of wear.

3. Double knot: This is of the ‘caballus’ type which is usually found in

African Hipparion.

4. Metastylid: It has a triangular shape with the right ante found distally

on the lingual side of the entoflexid.

5. Metajlexid: It constantly shows unique invagination on both its

anterior and its posterior walls.

6. Entoflexid: The buccal wall shows varying degrees of waviness

(plications).

7. Internal depression: This valley, lying between the metaconid and

metastylid, is very broad, shallow and irregular in P,, while in P,—P, it is

broad but slightly deeper giving it a U-shaped appearance.

8. Ectostylid: This is constantly absent, but on Lg52 there is a small

elongated thickening of the enamel at the crown-root junction. However this

‘bud’ does not arise from the cingulum but extends from the posterior root onto

the crown just above the junction (pl. 8).

9. Ptychostylid: Present on P,, and occasionally it is very well-developed.

In P,—P, there are irregular indentations on the bucco-distal aspect of the

protoconid, i.e. in the valley (external depression) opposite the ptychostylid.

Inconstantly, slight indentations are seen on the wall of the hypoconid in the

depression.

In M,—-M, the ptychostylid and the indentations are not constantly

present. In M, these features are both absent.

10. Protostylid: It may appear as a laminated ridge or as an isolated

pillar. Usually, when present, it extends along the whole height of the crown.

371

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

uoNNUap JUsUeUIOd JaMO] SamuveqesuUr’'T jo (‘uru) suoisaowid “£1 aTdv J,

13S F.Gz 6.3S 3-6 1.G 1-bP 0-18 g-ZI I-O1 z.9G 9-91 g-6z L.G9 NET

:1@IM “paul 0} IYSIS

UL Y}00} JO uUvo/[

Coz Z-QS gS ¢.6 z.G 6.19 0-gL L.i 1-6 6.95 1-41 6.gz L.09 7d

“yj99} “AIPUT Jo Uva]y

g&z 8 +B g- 0S 8:6 z.G 0.1P 1.94 €.a1 6 F.o96 P.gi Q-Q9z g-LS uvow ‘sqy

602 C.z reyes 8-9 9 g-17 Lol g-O1 LeL L.aS G.1 €.Gz 6S aSghi J

6S aGoVi JT

gla €.61 P.EG G1 €.bz Lo : 6F6 J

V.Lz 9-41 v 6.62 6.6 6.96 G.11 o-16 SCEOEO Te STA)

giz L.Gz F.oF P.11 9-¥ 1.96 g-GL T.11 9:8 L.GG 1-QI 6.93 9-29 ‘+ gbbi J

o61 6.62 z-6G 6-9 6.F 6.65 6.89 g-O1 &.L 6.96 1.GI G.gz G.oS ** 666 J

Cgi 6.06 89 T.9 L.G ¢.6& iL €.11 8 L.¥9 ¢.g1 Q-9z a. LY “> $56 7

6g1 c.g 6 .bS z.6 G 9-96 6.08 L.6 gL 1.69 L.gi C.gz C.9h zG6 "J

Fiz Go 6.00 3-11 c.P z-oP tg g-1 6.6 L.&G L.G1 € .6% L.%g gr6 TJ

Qza 6.6% z-66 L.01 ZF 1-66 1-16 oii I-01 Q-3aS GI P.Qz b.t9 Fh6 J

2*yAq 10 TAL

Q1z 6.9% ¢.&9 ¢.9 D.G P.EP T.&9 6.§1 g-II z.QS 9-gI oS 69 viGhr J

L.gi Lo6 7

gl P.E& L.L9 6 1-9 €.SP F.Lo o-SI 6.9 3-19 g-Li 1.6% €+&G eeeQvOngT

goz gz z-89 L.6 9-9 1.69 G.gl 6.31 1-01 a. PS z-QI 6.6% 6.39 Aa

BS Orgran ehitOndh

F.96 G.oG I-O1 1-G g-G1 3 -QI oS aSofr ‘JT

8-91 6.01 686 ‘JT

6a1 6.1P .E¢ 1-01 v.G F.9P 9-9 o-QI 3.6 £.0G Li ¢.§¢ 1.1P S Lb6 7 ®q

(uo SUI] | yorssoadop

yysu9] (pruojey | -oafoad) 0OL | 4sody8ua7T

S15 Ug | eee eT OLE) piuosojua | uorssordop — uorssoadop | uorssoadop

poe EL ea aA SOE ENO ee om (gree ON ESC Ec SC] (CRESS eee (Oe nese ee eel rer n A

372

ANNALS OF THE SOUTH AFRICAN MUSEUM

It can always be observed as a prominent ridge along the mesio-buccal border

of the tooth. It is constantly present on P,;—P,, and always present on M,—Mg.

11. Entostylid: One entostylid has been observed (L946) in the entoflexid

on the bucco-mesial aspect of the entoconid. It presents as a distinct flattened

enamel island (pl. 8).

12. Talonid: Bilobed.

Upper permanent dentition

Material:

Psion P24:

M! or M?:

M3:

? premolar:

Cheek teeth:

Left: L934, L936, Lo42, L1463, L1467z, S.A.M.11716

(DES hy ey Gy Fe

Right: Lg55, L1467c, S.A.M.11722, S.A.M.11724

(pls. 1, 3, 7, 9).

Left: L945, L953, Lo58, L1467B, L1467p, L1467H,

SeAuMeni7 19) (plsi3574.) 0):

Right: L940/950, Lg51, L956, L1459 (pls. 1, 2, 4).

Right: L935 (pl. 9).

Left: S.A.M.11723 (embedded in plaster for sectioning).

?: L1785 (fragment).

The following specimens probably belong to single individuals:

no fB OO ND

. L935, 8.A.M.11722, S.A.M.11724.

. L1473, S.A.M.117109.

. L942, L945, Lo56.

. Lo40/Lo950, Lg51.

- L934, L953, L955, L1459.

. L14678, L1467c, L1467p, L1467E.

completely unworn tooth, L1467A, recovered from the same site as

L1467B—-E and H (viz. Baard’s Quarry), does not present the typical features

of hipparionids. It shows the same type of fossilization as the other L1467

specimens, but, because of its exceptional dimensions (ca. 32 x 28 mm.), it is

provisionally excluded.

The teeth are classified according to their degree of wear as follows:

completely unworn or just erupting: S.A.M.11716.

early wear: L1463, S.A.M.11719.

moderate wear: L935, L940/950, L942, L945, Lo51, L956, L1467B-£,

§.A.M.11722, S.A.M.11724.

advanced wear: L934, L936, L953, L955, Lg58, L1459, L1467H.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE

General dimensions (mm.) of P?-M?:

1. A-P length: range of variation: 21-8-30°5

Mean: 26:6

2. Transverse breadth: range of variation: 23:5-30°3

Mean: 26-9

3. Breadth/length index: range of variation: 90:7—111°4

4. Hypsodonty:

Means

All P?-M?

Unworn teeth

Early wear

Moderate wear

Mean: 99:1

Absolute

crown height

(mm)

* The ratio between crown height and breadth.

Hypsodonty

Index*

(%)

49°6

36-4

42°3

51-6

373

Height/length

Index

(%)

193

249

233

201

To indicate the relationships between the Langebaanweg dimensions and

those of other African Hipparion, the following data has been extracted from

table 18 and summarized (table 19):

A-P length

Transv. breadth ..

Breadth/length

Height (unworn and early

wear) .. a :

Hypsodonty index (un-

worn and early wear)

Height/length index (un-

worn and early wear) ..

Hi. africanum — H, sitifense

23-27

20°5-26

97°2

19-23

18-22°5

48-60

38° 3-41°3

223-240

“Hypsohipparion’

albertense

(South Serengeti)

25-31

24°5-29°5

95°1

249

Langebaanweg

26-6

26°9

991

67°4

39°3

238

TABLE 19. Comparison of dimensions (mm.) of Langebaanweg upper teeth with summary of

data of other African Hipparion.

From these figures, it is clear that the Langebaanweg Hipparion has

hypsodont upper teeth, but this hypsodonty has not reached the same degree

as it has in the South Serengeti specimens.

Description of selected individual teeth:

Two specimens, an unworn left premolar (S.A.M.11716) and a left M?*

or M2?

(S.A.M.11719) in early wear, have been selected for detailed

ANNALS OF THE SOUTH AFRICAN MUSEUM

paren

Viz

£61

L&s

60a

1G1

ogi

bow

o61

Qos

6&%

gol

98!

OGG

*yj001 JUoUeUIIEd 19ddn Samueveqosury jo (“WIUL) sUOISUDTTIG] “QI ATAV.],

6.L¢ P.GE 1.67 v.P z-6 Z-QOl 1.QG V.Gs 8-gh FA 07 WL ‘YI99,],

[eR PAR OS Cea

0-96 g- PE 6.64 L.V 9-6 6.96 6.Lz L.gz g+1S 33 Uap Gap Sep ee yyy,

(UCR pO TS oe

9-67 1.S6 6.9F 9-0 6.6 1-66 6.9% Q-9z L.6b sie uvoy [P1ouoy

1-93 FS "+ salir Ww'y's

6.aP 9-65 PT .o& v II L.S6 9 -9z Q.Lz £9 Be ** 61411 7

v.96 L.ov L.1¥ if) cmt L.o6 ¢.Gz 1 -Qz ol gIZ11 ‘W'y's

G.Q% L.1G 9-F €.6 C.0€ 6.SP HLOovI "Ty

z.9G G.a& L.bV a. T.6 6.Po1 §.0€ 6.9% ra alohi

1-2G 6.16 gS g-P 6-9 F.go1 5 -+Qs ¢ .gz z.PG ‘+ alohi J

g-0S of 9g-SS § 6 1-96 8+ QZ of 8-95 oLobi

€.9P G.LE 6.9P v.v 6 6.901 1-9z va a.4S oe ak aLori

L.ga G.oS z.P £.9 6.9% z.69 sis oe fohi TT

oS C.1E oS g-& g-L 101 6 .vz 1-z 9 - QP 6GSh1 7

ZOl g-Lor ¢.Ez Q-1z 6% gS6 7

o-VG GP 966 7

6.89 9& oS G oI 7-06 €.Lz Q- Lz oF CG6 J

£.z9 g-Sb €.LY a.G II V.1it Q-9z 1-z th £56 J

9-9P €.G¢ €.E¢ 9-0 6 Loi z-Lo ¢.Gz 9S 196 J

L.Go €S Gh6 TJ

3.6% vS "* 3b6 7

L.LY b F.g 1G 0S/ov6 |

6.96 9-F gL Lo 7 966r7

9.6% 1.0F Cf6 J

1.99 6.66 z-QP z.G g-O1 zOl Qz G.Lz ab bE6 TJ

xopuy (xoput y)8uor] =) (xopur odeyg=) | YyIproiq yisusy] yp8u9]

AyuoposdAfy Sus] W100], yysus/Yypesiq | 9U0I0}0I1g | 9U0I0}01g |—————_|_ ypeoug yysua'T WYSIOYY

2Uu090}01g yproig

YISuUsT 9U090}01g

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 375

description. They display the characteristic features of the other Langebaan-

weg specimens. The typical features of the whole group will be outlined

below (see p. 376).

S.A.M.11716 (pl. 5)

A left P*, unworn.

Occlusal surface: Paracone and metacone are V-shaped and they are

curved towards the medial side. The protoconule and metaconule are less

angulated than the lateral cone and tend to curve less towards the medial

side. The paracone and the metacone tend to overhang the fossettes.

Protocone: The cement is split off near the apex of the tooth. It is lozenge-

shaped, completely isolated and the apex tends to curve in a buccal direction.

Shape: The tooth has medium hypsodonty. It has a gentle convexity from

root to apex in a lingual direction. The styles stand out as clear costae, forming

deep hollows between them. One-third of the distance from the occlusal surface,

the hypoglyph becomes increasingly deep so that in the upper half of the tooth

the hypocone becomes well demarcated.

Appearance of the tooth, sectioned just above the middle :

The halves of the ectoloph are deep and arc-shaped with a slight angularity

at the mesostyle.

The parastyle is prominent. Most of the prominence is taken up by the

lateral angle which is rounded. The surface, flattened but with a slight groove,

is markedly oblique and almost in the same plane as the anterior (mesial)

surface of the tooth.

The mesostyle is prominent, and, because its rounded surface is flanged, it

overlaps the buccal depressions on both sides.

The metastyle is very slight (almost negligible), its rounded ridge decreasing

in prominence from the occlusal surface to the root.

The protocone is slightly elongated and oval. Completely separated from the

protoconule.

The frotoconule is a flattened arc, angulating fairly sharply toward the pli

caballin.

The hypocone is elongated posteriorly, rather flattened. There is no tendency

to isolation. There is a deep V-shaped hypoglyph, and the hypostyle is

angulated.

The pli caballin is bifid, the posterior process just touching the protocone.

The prefossette: There are plications which are so deep and complex that

the ccntre of the prefossette is very narrowed and reduced. The anterior wall

is plicated and the antero-lateral angle is knob-shaped, tending to be separated

from the prefossette. The pli protoloph is deep and narrow, touching the very

deep and narrow pli protoconule so that the medial portion of the prefossette

is an almost isolated triangle. The plications on the posterior part of the pre-

376 ANNALS OF THE SOUTH AFRICAN MUSEUM

fossette are very deep and complex. One of the more medial ones tends to

flange out to form an almost isolated irregular pillar just lateral to the base of

the pli caballin and wedged between the most medial part of the prefossette

and postfossette.

The postfossette: The plications on the anterior wall are also deep and

complicated but not as numerous as on the posterior wall of the prefossette.

The pli postfossette is deep and irregular-shaped, almost touching the base of

the pli hypostyle, thus tending to isolate an oval portion of the medial part of

the postfossette. Just lateral to the pli postfossette, the plications tend to isolate

a small rounded pillar.

S.A.M.11719 (pl. 3)

A left M? or M?.

It is slightly higher crowned than $.A.M.11716, and slightly less curved.

Half of the ectoloph: The paracone is deep and quadrangular-shaped,

angulated at the mesostyle and parastyle.

The parastyle is prominent and narrow. The sides tend to be parallel,

slightly widening at the base.

The metastyle is partly broken. It forms a slight ridge.

The frotocone is elongated, oval and angulated anteriorly and posteriorly.

It is completely isolated.

The protoconule forms a broad flattened arc, slightly angulated toward the

pli caballin.

The plz caballin tends to be duplicated with the anterior tongue longer and

almost touching the protocone.

The hypocone is rather short, tending to be separated from the metaconule.

There is a very deep hypoglyph that is very broad-mouthed.

The prefossette: There are a few anterior plications, continuous with a

deep narrow pli protoloph which tends to isolate the medial wall of the fossette.

The pli protoconule is shallow. The plications tend to be absent. However,

the original plications have become isolated to form an irregular circular

island (prefossette loop) just lateral to the pli caballin and wedged between

medial parts of the pre- and postfossette.

The postfossette: The plications tend to be worn away but it can be seen

that they were complex anteriorly. The pli postfossette is very deep and touching

the medial wall of the metacone. The pli hypostyle is deep, angulated and

wide, and it has almost isolated the medial part of the postfossette.

Typical features of the upper teeth:

1. Protocone: The characteristic isolated protocone of Hipparion is observed

in all the complete molars mentioned above. On L935, L942, L945, Lg56,

S.A.M.11722 and S.A.M.11724, as well as on L1785, the lingual portion of

the tooth has been broken away along the pre- and postprotoconal grooves and

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 377

does not allow any observations of the protocone. However, there is no reason

to suppose that the protocone in these teeth displayed different features, and

it is as umed that the description proposed for the other teeth applies to these

specimens as well.

The protocone is elongated in a mesiodistal direction, and clearly shows

angulation in those teeth not worn or in early stages of wear. It becomes more

elliptical or oval in the later stages of attrition. Thus the shape index (breadth/

length) varies from 36-4 to 58-9, with an absolute mean of 48-9 for all the

teeth and an average of 48-3 for those teeth which have not reached an advanced

stage of wear (table 14).

The length index ranges from 42-7 to 28-7, with a mean of 35:1 for all

the teeth, of 34-7 for teeth in early wear, and of 35-7 for those in advanced

wear. This smaller range of variation indicates that the actual length of the

protocone does not change as much with wear as does the breadth, and that

the apparent elongation of the protocone of the unworn teeth is more due to

its narrowness than to its actual length.

Thus it is clear that the Langebaanweg specimens display this typical

feature of African hipparionids (see p. 355). In so far as the few available speci-

mens in varying stages of attrition from other African sites permit comparison,

there seems to be a great similarity in the shape and dimensions of the protocone

between H. africanum and the Langebaanweg Hipparion. The ‘African’ character

is even more strongly expressed in ‘Hypsohipparion’ albertense (tables 14, 20;

figse 52.19).

2. The hypocone is commonly angulated, tending to isolate in early wear.

Later it becomes more rounded, and the isolation from the metaloph is no

longer discernible, the anterior groove tending to disappear. Parallel variations

in the shape of the hypoglyph are observed with progressive attrition.

3. The marks are constantly closed.

4. The plications are numerous, profound and often bifurcated, especially

in early or moderate stages of wear (e.g. S.A.M.11716, Lg40/950, Lg51).

The enamel pattern is most complex on the posterior wall of the prefossette

and on the anterior wall of the postfossette. The plications have been counted

according to the method suggested by Gromova (1952), and the average

plication formula for the Langebaanweg specimens has been established:

(Bess) Coa ey

(1-2)

There is a constant isolation of a prefossette loop in the lingual-distal angle

of the prefossette, being bilobed and large in early stages of wear, and rather

triangular and arrow-shaped in advanced wear (e.g. specimens L934, Lo51,

S.A.M.11716; best example is the latter on pl. 5).

The pli caballin is constant, often double, in early and moderate stages

of wear. It is not observed in very advanced wear.

In all these features there is not a clear basis of distinction between the

ANNALS OF THE SOUTH AFRICAN MUSEUM

378

‘bo61 ‘r9yosty “YS “Aq Aq poejeorunururos syusWaIMseay ‘"a8usIIG YINOG

woul (69-98 ‘ss ‘TIT 214% ‘2v61) YyoreIq asuazsaqgyo .uoiuvddiyosd(f], JO {99} WIOpuRI Jo (“UIUT) SUOISUDHIC] ‘0% ATV],

9-9S GI C.9z WwW

z-6S gi La a.

9-89 $.gI Le W

€.GL G.1z C.9% "7

€.6¢ G.L1 C.6z oT

LY ¢.G1 && eal

HLAGL WAMOT

6.0P v.LP L.v G.11 66 bz a.¥z eW

L.1v v.GP L.v €.11 1-96 C.bz Gz ZN

cv 6.697 P.G rept 66 La €.Lz IW

6.0P g-&P g-S L.G1 C6 ¢.6z 1 “al

a-oF ¢.&P 1-9 a-S1 9-88 G.Lz 1§ ed

¢.cP L.VE z-9 9-1 V.PL 6% 66 2d

(xoput odeys =) (xopur y}8u9z] = ) yysuo'T

yisus] 9U000}01g YSuI] YOO J, yIpesiq yisuo] Yypro1g yysuo'T HLAAL WAddA

9u090}01g | au0D0}01g | WPPOTE

YipeoiIq IJUOI0}OIg

YySus] 9U0I0}01g

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 379

Langebaanweg material and the ‘Hypsohipparion’ upper teeth from Serengeti,

the plication formula of which is found to be very similar:

(Ce) = (25) (Ore) = (C2)

(0-3)

POSTCRANIAL EQUID REMAINS

At Langebaanweg seven phalanges have been recovered. Because of

their equid structure it could not be excluded a priori that they belonged to

Hipparion represented by the dental remains at the sites. It must be pointed

out that a few teeth of Equus (to be described in another publication) have

been recovered at these sites. However, they only constitute 10° of the total

number of equid remains collected thus far. Consequently, because of the

lack of comparable Hipparion remains from Africa, the Langebaanweg specimens

are hesitantly included in this paper, more as a documentation for comparison

by others than a positive contribution to the reconstruction of the Hipparion

limb.

Material:

First phalanges: L1456, L1462a, L1462B, L1462c (pls. 10, 11).

Second phalanges: L1449, L1462p (pl. 12).

Third phalanges: 11444 (pl. 13).

All the absolute dimensions namely, length, diameter of the proximal

and distal ends and of the median shaft are usually considerably larger than

those (after Gromova, 1952) of the three species of Hipparion compared, i.e.

elegans, moldavicum and longipes (table 21). There is virtually no overlap between

the two series, although the highest figures for H. longipes (the largest of the

three species) are sometimes very close to the lowest figures for Langebaanweg

(table 22). Sondaar (1961) mentions one first phalanx of the forelimb third

digit with an exceptional length of 69:8 mm. In Europe the highest mean is

known from Eppelsheim, H. primigenium, with a length of 64-3 mm. and a

proximal extremity breadth of 41-3 mm.

A comparison of the dimensions of the proximal end of the first phalanges

from Langebaanweg with the breadth of the distal extremity of metatarsal III

from Olduvai and South Serengeti (tables 12, 13) makes it obvious that the

Langebaanweg specimens require a much broader metatarsal for articulation.

For H. longipes the respective breadths of the distal end of metatarsal III and

the proximal end of the first phalanx are 40 mm. and 42 mm. The mean

distal breadth of the known specimens of metatarsal III from East Africa is

44°7, the maximum being 48 mm., the widths of the proximal end of the first

phalanx at Langebaanweg are 50-56 mm. Therefore, either the Langebaanweg

specimens belong to a slightly more massive limb of an Hipparion or else they

belong to Equus.

Compared with Equus zebra and burchelli, the Langebaanweg specimens are

larger in a number of dimensions, but there is a greater amount of overlapping

*(6F61) eAoWOID WIZ pozEpNdTLd 10 0} Surpr0ddV gq

(S61) eAoWOID WO I pozE[No[Vd 10 60} Surpz090V »

*(yjayoing pue vigaz) snnby ueolyy utapour pue (suvow) uownddyy oneiseing yM poreduioo ssuoq quiy Somuveqosuey oy} jo (wur) suorsusuNg ‘1% wavy,

¢.L6 For z-66 oo1 | g-9f a-LL V.LL L.zg G.&L G.LL 6.18 6.S9 x73 oe xopuy y/5

66 C.1P ¢.FE LE 9-92 L.gz G.aE 6.28 g- bz a.Gz F.g6 oh ris yeys Wpeoiq “uIpy “DH

G.ch z-S9 ol Ale z-oL 6.29 9-95 | 1.9 b.PS 19 g-&¢ g-LS i: 3 xopuy W/a

6z Bz Le Lo 6.L1 L.Li &% G.1z L.gt €.91 1.42 P.Gz ‘* puo [e\sIp ‘weIp g-y “Wy

Oo Lor | 9-%6 tor | ¥.%g T.L9 6.68 G.96 G.oL z.Le £.98 as “ xopuy v/q

S ov &P GE C.96 8 -Qz Z-1E GE F.6§ ole F .62 [Levay oA pus [eIsIp Yapeorg “"]

5 Gil Gir 111 | Z£.¥11 ; 9-96 o-o6 0-6 g6 €.68 £.68 1o1 | g-Sor ae xopuy y/9

2 ot ob Iv C.aP 1.66 B- aS G.66 1.66 g-0& z.06 g Lt g-1S Bs pue ‘xoid yypearg, “Dy

a VE G.a& of ¢.6z L.6z g-6z ¢.E a.&E g-L4z g-4z 1.06 1.96 (uvIpow) ‘jans‘yue yysueT “g

S) ov ov LE LE V.bE L.GE oP 6.66 V.bE L.&& a-LP 6.9P es yjsus] winwixepy, “VW

& 6hhiq | azghiy

EA LIOId GUIHL ‘it XNVIVHd

a = zg

5 Lich | L2v Ocmiiecalen recs | z.ge | 967s | one 1.66 | 9-S& | S.gh | &.+F L.66 |-+ +: — xapuy y/9

9 a& a& G.Le G.Le va B+ os 6.06 L.Qz 5.3 €.o1 9-98 1.06 6.10 "* uvIpsul Qyeys ypeog “9

3] GS cot 13 Go 6.41 Li 1G G.1% gi ¢.G1 1.Gz o-ES 6.S% F.Lz ‘* pus [eIsIp ‘wep q-y “Wy

a LS v.EG g-oS 1.69 €.0G T.oP 9-6 ¢.ES V.LY t.vP G.PS L.&S Lea es |e ‘s xopuy y/q

ty ov ov LE LE 6.gz L.Lo G.& g- bE 5 -Qs 6.Sz o-1P v.1v oie |} ee pus [eisip yIpesig “Wy

ie) g- Sz Q-3% Z -Qz L.1% Q-0% Lo La €.Lz a& ao oO eer Ons

4 ‘oie ‘xoid ‘wep g-v ‘qd

s! oL L.Lg looeal G.ol F.19 9-gl 6.09 gL ¢.Gg L.gl 0g 99 +.L9 OVafeys OI 2 og xopuy g/9

Z oS P.oP L.GP 6.17 L.6P ¢.&P 6.9P 1-.LP LY z.1P 6.2S L.LY V.LP L.z9 es ws xopuy y/9

“f GE G.c& a& 1 Qo 6.Gz L.1& Z-Qz 1.0% ov 8-96 V.LE bY ** puoe *xoid ‘wep g-y “yD

6.1L to G.09 09 1.09 F.cG F.09 z.&9 P.GG €.2G z-99 z-oL F.oL g-6L ae! 3 xopuy y/q

oS gh ad bY g- FE 6.26 oP L.1v g-o0& g-o0& oS 0.bS ¢.GG 9S 3 pus ‘xoid yypeoig “gq

of GL o£ tL 6.66 9-66 ¢.69 G.L9 ¢.GG z-QG ¢.cL LL 0.6L z-oL oe ‘+ -ysusy (e107, “Vv

puryy 210,J puryy I10,J pulpy 210] pury 210,J pulpy 210,J gShiy | ozghiy | azgbry | veghryT

(ayeur9y) q01qaz “7 plUngtappjOoUL * FT esagisuo] “Ey eSUD9]9 *H] somuveqosur'y LISId GUuIHL ‘I XNVIVHd

8 (11%g1 NIZ)

0 COME ELITE l

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 381

and the dimensions are closer together than in the comparisons with the

Hipparion above (figs. 14, 15; tables 22, 23, 24). The difference between the

Langebaanweg and the Equus proximal phalanges average 6%.

The situation is less clear for the relative dimensions. The lowest index for

Langebaanweg is constantly smaller than the highest for burchelli, and very

often even overlaps H. longipes. The relatively smaller indices indicate that in

the Langebaanweg specimens the length versus the breadth and the breadth

versus the A—P diameter are somewhat greater than in Hipparion and burchellz.

In these limited comparisons it seems that, to some extent, the Langebaanweg

architecture differs from these groups.

It cannot be denied that the Langebaanweg measurements fit more

satisfactorily within the range of Equus, but more material and comparisons

are needed before this discussion can be taken further.

First phalanx, III digit

Cc OCOD WA Axx x x

C7 eeoxdG x

D CKO xx@ x

8 tee H. moldavicum

@ (F :

3 ees H. longipes

F CX AA xx x (

A (Fore) :

® (Hind) E. burchelli

G COO We O/Ax x V (Fore) ¢

¥Y (Hind) zebra

COKO GLbxA x x Langebaanweg sp.

E/A 0 60 vey ha A

10 20 30 40 50 60 70 80 90

Abs. dimensions and Indices scale

Fic. 14. Absolute dimensions and indices scale for the first phalanx of the third digit in various

species of Hipparion and Equus compared with the Langebaanweg specimens.

382 ANNALS OF THE SOUTH AFRICAN MUSEUM

Second phalanx, III digit

(Fore) H ee

B (Hind) - moldavicum CRE DoxeA x

@ (Fore) tnlona:

5 . longipes

Cans ibn) com> te x

A (Fore) Indices

ore é

A (Hind) E. burchelli

xs tere, E. zebra @ O% Oe x ana

DI—_ x Langebaanweg sp. CROW LAxx

E OO 67 PAxA

2 E/A

F KH 0 OxxyQZA

o }x 0064 vA ¥ F/E

G CO oy VAL

ORI Xe G/A

Abs. dimensions 10 20 30 40 50 60

Indices scale 50 60 70 80 90 100 110

Fic. 15. Absolute dimensions and indices scale for the second phalanx of the third digit in various

species of Hipparion and Equus compared with the Langebaanweg specimens.

ORIGIN AND DIFFERENTIATION OF AFRICAN HIPPARIONIDS

In recent years, it has become evident that Africa was more a true and

independent centre of evolution and dispersion than a refuge for ‘Pontian’

Eurasiatic animal forms. Discoveries at different levels of the history of the

mammals and in different regions have supported this concept proposed

originally by Pilgrim (1941), and later vindicated by, inter alios, Arambourg

(1959) and Cooke (1963).

Several independent observations have definitely established the original

role played by Africa in the development of late Tertiary mammals in general,

and have illustrated its contribution to the establishment of the ‘Pontian’

fauna in particular:

(1) The endemic evolutionary differentiation of the Proboscidea and

Hyracoidea from the Oligocene onwards;

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 383

A. ABSOLUTE DIMENSIONS

H. longipes Equus burchelli Langebaanweg

(largest) (largest) (smallest)

PHALANX I

Total length ae se hs 69°5 75°0 70°2

Breadth prox. end BC or 42 50 50

Diameter prox. end Se ne 31°7 35 36-8

Diameter prox. artic. surf. ie 28-2 27°3

Breadth distal end we oh 34°8 40 41°2

Diameter distal end 66 Bi Q1°5 22 PIG)

Breadth median shaft .. hh 30°3 32 34.°1

PHALANX II

Maximum length ie ee 42 40 47°2

Length median ant. surface .. 34-5 34 34°1

Breadth prox. end ste NG 39°5 46 47°8

Diameter prox. end Je she 28°3 29 31

Breadth distal end ae aK0 39°4 43 41

Diameter distal end ae ee 23 29 24°

Minim. breadth shaft .. oe 32°9 41-5 38° 4

PHALANX III H. moldavicum

Anterior length .. ahs = 54°1 47°5 61

Maximum breadth ae 3 54 59 72

B. INDICES

PHALANX I H. longipes

Prox. breadth/length .. of 63:2 Tlier3, 66-2

Prox. diam./length se ae 48-9 50 47° 4.

Prox. diam./breadth .. ie 80-9 70 68

Distal breadth/length .. se 53°5 57 53°7

Breadth med. shaft/length a: 43°6 45-7 44°3

PHALANX II

Prox. breadth/length .. oe 98 HS IOI

Distal breadth/length .. wi 98-5 107 86°3

Distal diam./breadth .. ius 65:7 1 2er5) 57°8

Minim. breadth shaft/length .. 82-7 104 81°3

TABLE 22. Comparison of dimensions (mm.) and indices of Langebaanweg smallest phalanges

with the largest of H. longipes and E. burchelli. The italicized figures indicate those dimensions in

which the Langebaanweg specimens are smaller.

H. medit.

PHALANX III Langebaanweg (Pikermi) JH. elegans H. moldav. _ E. burchelli

ant — post ant post ant = post

Length ant. surface .. 61 50 A5wO) A772) elo Gach AO Aes

Maximal breadth ae 72 55 46°3 44°2 54 A3°4 59) 1585

Index B/L ae 118 110 LOT 93" O 103 78-9 128 123

TABLE 23. Dimensions (mm.) of phalanx III of Langebaanweg and of some Eurasiatic species

and E. burchelli.

384 ANNALS OF THE SOUTH AFRICAN MUSEUM

Hipparion Equus Langebaanweg

elegans | moldav.| cabs.l. | zebra

Index of length

Phal. III, ant. 135°9} 145°1I | 116-4 | I10 W277 123

gd digit/Phal. II | post.140°1) 157°3 | 118-8 | 116-9

Index of length ant. 78°7| 86-9 | 64:2 | 57°6|]Li44q4 | Li444 | Li444 | Li444

III/I

post. 85 g1°-2 | 66-4 | 61-3 | Lrg62a | L14628 | Lr462k | L1456

85°3 75'°8 Taceh 79°3

TABLE 24. Comparison of relative lengths of phalanges.

(In the case of Langebaanweg indices have been calculated from the possible associations.)

(ii) the presence in the Lower Miocene of East Africa and in the Upper

Miocene deposits of North Africa of a complete specialized faunal assemblage

with typical African characteristics before parallel diversification took place

in Europe;

(iii) the constant specific differences between comparable Pontian

Eurasiatic and African forms; and

(iv) the African pre-Pleistocene differentiation of Bovidae with the evi-

dence of the local origin of Cephalophini, Neotragini and probably the

Alcelaphini (Wells, 1957).

It is suggested that, because of the stability of its climate, its geography

and its vegetation, Africa has provided a suitable habitat throughout the

Tertiary. Furthermore, groups have evolved and later have developed parallel

forms. Elsewhere, these have become extinct, but survival was favoured in

Africa. Even in modern times this seems to be the case, e.g. the Proboscidea,

Giraffidae, Artiodactyla, and Anthropomorpha. The surviving species, far

from being newcomers and refugees from outside, are local representatives

of a fauna which has disappeared elsewhere.

Arambourg (1952) and Cooke (1960), inter alios, demonstrated that the

Plio-Pleistocene forms of elephants may have differentiated within the continent

of Africa, paralleling the steps of progress observed elsewhere, but not necessarily

linked to Eurasia by periodic invasions as was previously thought.

Analogous circumstances may have influenced the origin and the diffe-

rentiation of the hipparionids. There is abundant evidence that Hipparion

evolved in North America from some Merychippus stock in Miocene times.

Migration probably took place at a rather rapid pace, Hipparion being found

on both sides of the Mediterranean basin before the end of the Miocene.

Hipparion has been recovered from brackish water deposits dating from the

Sarmatian in the area of Sebastopol (Borissiak, 1914) and Istanbul (Chaput

and Nafiz, 1934; Chaput and Gillet, 1938). It is also known from the Upper

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 385

Tortonian Beds in the Rhone Valley (Denizot, 1939). Villalta and Crusafont-

Pairo (1946a, b, c; 1947; 1948) have described Aipparion from the Vallés—

Penedés area in Catalonia, in continental and brackish water deposits inter-

calated in sediments ranging from the Burdigalian to the Tortonian, with

Hipparion appearing in the Upper horizon (Vallesian, parallel to the Meotian

of Eastern Europe, i.e. Upper Sarmatian). In the basin of Teruel, Hipparion

was recently found to appear in Los Algezares in some gypsum intercalation

of the middle light-red clayish deposit overlying the Los Monotos series,

dated as the Vindobonian (Sondaar, 1961).

Parallel to these occurrences, and probably very close to the last one in

the chronological sequence, hipparionids have been recorded in the Upper

Miocene of the Maghreb, i.e. H. africanum (Arambourg, 1959) from Oued el

Hammam and Marceau.

It is still not known which Atlantic or Pacific bridge the hipparionids used

to invade the Old World. The only reasonable assumption which has been

proposed is that the migration route probably lay very much to the north.

The absence of any important migration of Camelidae and Antilocapridae

from America, or of hyaenas, antelopes and giraffes to America, provide good

evidence in support of this high latitude of the migratory passage (Simpson,

1947). On the other hand, and as a confirmation, the most ancient Old World

Hipparion seem to have been much better adapted to a humid and forested

biotope of temperate, or even cold, climates than most of the American forms,

which evidence a drier and steppe environment. Rather than indulging in

some fantasy that a direct connection between America and Africa existed by

way of a hypothetical chain of islands across the North Atlantic (Joleaud,

1g1ga, b), it seems more reasonable to accept the normal Bering route which

certainly meets the climatic requirements of the migrants.

Furthermore, in the present state of our knowledge, it is neither possible

to formulate any conclusion about those species of Hipparion responsible for

the origin of the African group, nor to reconstruct the phyletic sequence of

African Hipparion. The fact that H. africanum is found contemporaneous with, or

possibly prior to, every other European form and that it already shows typical

African differentiation, refutes a European origin of the group from any of

the known ‘Pontian’ species. On the other hand, and for the same reasons,

there seems at present to be no possible way of establishing an African origin

for the European hipparionids. However, the metapodials of H. africanum are

rather strong and the lateral digits are well-developed, which must be inter-

preted as ‘primitive’ features, less specialized than in most of the early Eurasiatic

forms. But other characters, e.g. the articulation of the external cuneiform

with the second metatarsal only, a typical mechanical adaptation to tridactyly,

show more primitiveness in the relatively contemporaneous H. catalaunicum

(Pirlot, 1956). It seems that here there is evident some expression of an inde-

pendent mosiac pattern of evolution in different (Eurasiatic and African)

groups. Although they obviously descend from a common ancestral stock, they

386 ; ANNALS OF THE SOUTH AFRICAN MUSEUM

have been precociously isolated on different continents and have developed

their independent stages of evolution.

For several reasons it is not possible to depict accurately the actual evolu-

tionary development of the African Hipparion. As has been stated previously,

the fossil record is rather poor. At more than thirty sites where Pleistocene

hipparionids have been recovered, the available material consists almost

exclusively of isolated teeth, and virtually nothing is known of the skeleton.

Furthermore, because of the plasticity of the group, and of the fluctuating

character of many features, which seem to be highly adaptive to ecological

requirements, it would appear hazardous to reconstruct artificially an ortho-

genetic series. Gromova (1952) states that the diversity among Old World

hipparionids usually remains at the level of a particular genus and does not

exceed specific differences. On the basis of our present knowledge, this typifies

the African representatives of the group: overall dimensions, proportions,

size of the teeth, plications of the enamel, development of stylids, elongation

of the protocone, preorbital fossae and development of lateral digits. Most

of these features are readily adaptive and susceptible to functional interpretation

under climatic and ecological circumstances.

On the other hand, it cannot be denied that parallel to the ecological

adaptation which has been emphasized previously, some general and fairly

continuous trends have characterized the evolution of the hipparionids through-

out the world. From Merychippus onwards, in late Cenozoic times, the various

groups have shown a permanent tendency to

(a) increased hypsodont dentition;

(b) lengthening and flattening of the protocone;

(d) development of additional stylids; and

(e) evolution of the double knot in three different but constant directions,

i.e. towards a ‘stenonis’ type in America, towards a ‘“Hipparion’ type

in Eurasia, and towards a ‘caballus’ type in Pleistocene Africa.

There is insufficient evidence to state that the reduction of lateral digits

and the elongation of the third metapodials have obeyed a clearly directed

evolutionary trend. The enamel pattern has not been controlled by any

constant progressive influence, but it has only responded to the changing

ecological conditions.

There is little doubt that these trends have been developing on more or

less parallel and independent pathways and at different and varying speeds

along the several lines, with the resulting mosaic type of differential radiating

evolution.

For the reasons stated earlier in this paper, the degree of relationship

between Hipparion africanum and the Pleistocene stock of African Hipparion

is not obvious. There is a great gap in our record, and this is not sufficiently

bridged by the fragmentary remains of H. sitifense. Therefore it is difficult to

appreciate to what extent the distance between H. africanum, H. sitifense and

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 387

the Pleistocene forms illustrates the chronological or the phyletic separation.

All Pleistocene African Hipparion show a marked progression in the

expression of trends (a), (b) and (d) above; and together with the special

orientation of trend (e), this vindicates their common origin and their genetic

homogeneity.

It is suggested that the ‘African features’ and the more stable biotope

are responsible for the prolonged presence of Hipparion in Africa during the

Pleistocene. However, the differential expression of these trends along various

lines of evolution allowed for some differentiation during this period.

A basic group, represented by ‘Hypsohipparion’ and the Langebaanweg

specimens, is characterized by a greater conservatism in the building of addi-

tional stylids: the ectostylid is constantly absent. The flattening of the protocone

and the hypsodonty are strongly manifested in ‘Hypsohipparion’, but are less

marked at Langebaanweg. This basic group seems to be restricted to the

Lower Pleistocene (the basal grey tuffs of Serengeti and the archaic forms at

Langebaanweg).

‘Stylohipparion’ is more advanced in the expression of additional stylids:

the constancy of the ectostylid indicates a group subjected to a particularly

clear trend of evolution (‘Entwicklungswucht’ of Dietrich, 1942) in the process of

building a new HAipparion form. This successful group has rapidly extended

throughout the continent, being found in Lower and Middle Pleistocene

deposits from the Maghreb to the northern Cape Province, and _ possibly

differentiated on a geographical basis. The oldest record derives from North

Africa. It is not impossible that the group migrated eastward and then south-

ward, and, after being widespread throughout Africa, became extinct first in

North Africa during the Lower Pleistocene, then in East Africa where it is

still found in the Middle Pleistocene deposits.

TAXONOMIC STATUS OF THE AFRICAN HIPPARIONIDS

Family Equidae Gray 1821

Sub-family Equinae Steinmann & Déderlein 1890

Elongated face with completely closed orbits. Molar teeth, at first mode-

rately, then later strongly hypsodont, with closed valleys, the external portions

of which become isolated in the shape of pits, called fossettes or marks. Abundant

cement layer on the outer surface and in all the hollows of the tooth. Complete

homeodonty, the largest tooth being P, of a triangular prismatic shape,

pointed forwards. Ulna is first fused with the radius; however, later in the

development, it is reduced to its proximal portion (olecranon process), which

has the appearance of a mere apophysis of the radius. Limbs are tridactyl in

primitive forms, with complete lateral digits although they no longer reach

the ground, except during galloping, when they become functional. In advanced

forms, however, limbs are monodactyl, and the lateral metapodials are reduced

388 ANNALS OF THE SOUTH AFRICAN MUSEUM

to small splints, eventually only persisting in their proximal portion.

Unguligrade.

HIPPARION de Christol 1832

Face is relatively short, with profound lacrimal fossa. Prismatic hypsodont

molars, of moderate height, slightly curved, with more strongly plicated

enamel pattern than in Equus. The protocone pedicle has narrowed, isolating

the protocone, giving the appearance of a column which is attached to the

protoconule only at its base. On its occlusal surface, the protocone shows an

island of dentine, circled with enamel, of a rounded or oval shape. The enamel

of the fossettes is rather strongly plicated. Upper and lower incisors show a

typical mark. Metapodial moderate to long. Persistent tridactyly; the lateral

digits are strongly developed and more closely related to the third metapodial

than in Merychippus.

Pomel (1897) included in his definition that the lateral metapodials are

not very different from those of Equus, but they broaden distally into an articular

head; there are three phalanges, the most distal not reaching the ground.

Hipparion africanum Arambourg 1959

Hipparion sp. Arambourg, 1951: 2464, Arambourg, 1954: 295.

Hipparion africanum Arambourg, 1959: 75, Pl. 10, 11, 12, 13, 16, fig. 7.

‘Hipparion with skull of great dimensions, but with limbs of moderate

size and heavy extremities. Face and snout elongated; nasal aperture long

and broad; orbits far back; preorbital fossae long, simple, distant from orbit.

Dental series of moderate size: P?-M%=141 to 154 mm. Upper cheek teeth

with strongly plicated enamel; compressed protocone, elliptical or lenticular.

Cingular formations developed on lower milk teeth. Limbs are relatively

short, with strong metapodials, in which the lateral digits are still

well-developed.’ (Translated from Arambourg, 1959, p. 95.)

Hipparion sitifense Pomel 1897

Hipparion sitifensis (sice) Pomel, 1897: 14, Pl. 1, figs. 11-15, pl. 2, figs. g-10.

Hipparion sitifense Pomel, Arambourg, 1956: 817, pl. 26, figs. 1-5a.

Hipparion gracile [Non] Kaup, Thomas, 1884: 10, pl. 2, figs. 1-3.

Hipparion crassum [Non] Gervais, Thomas, 1884: 10.

Hipparion characterized by its small teeth (A—P and transverse dimensions

of P°-M?, ca. 19 mm.), little hypsodont and markedly curved in the median

plane. The upper teeth have a flattened protocone, completely independent

from the protoloph right down to the base of the tooth. The enamel plications

are few and simple. The lower teeth possess no ectostylid. The lateral digits

are well-developed.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 389

Aipparion (Stylohipparion) libycum Pomel 1897

Hipparion ? libycum Pomel, 1897: 8, pl. I, figs. 1-7, pl. 2, figs. 11-12.

Hipparion massoesylium Pomel, 1897: 11, pl. 1, figs. 8-9.

Hipparium' (sic) ambiguum Pomel, 1897: 15, pl. 2, figs. 2-4.

Hipparion crassum Gervais, Solignac, 1927: 756.

Hipparion steytleri van Hoepen, 1930: 21, figs. 14-19.

Eurygnathohippus cornelianus van Hoepen,? 1930: 23, figs. 20-22.

Stylohipparion hipkini van Hoepen, 1932: 31, figs. 14-20.

Stylohipparion steytlert (van Hoepen), van Hoepen, 1932: 33, figs. 21-23.

Notohipparion namaquense Haughton,® 1932: 421, text-fig. 5.

Libyhipparion ethiopicum Joleaud, 1933: 7, pl. 1, figs. 1-4, 6-8, 10-12.

Equus (Hippotigris) sp., (pro parte) —Joleaud, 1933: 25, pl. 1, figs. 9 and 13.

Libyhipparion steytleri (van Hoepen), Joleaud, 1933: 12.

Stylohipparion cf. albertense (Hopwood), Hopwood, 1937: 130.

Stylohipparion, Notohipparion, Libyhipparion, Dietrich, 1942: 98, pl. 13, figs. 91-95, pl. 14,

fig. 99 pro parte, pl. 15, fig. 105, pl. 16, fig. rog.

Stylohipparion albertense (Hopwood), Arambourg, 1947: 303, pl. 10, fig. 3, pl. 11, figs. 3-5.

Diagnosis :

Three-toed equid with rather hypsodont cheek teeth with a strongly

developed ectostylid, particularly constant on P, and M,. Typically the ecto-

stylid is broad in the centre and narrowed at each end, the greater length

being mesiodistally and antero-lingually orientated. The upper teeth show the

typical ‘African’ features, namely, complex enamel plications of the ‘mark’

walls, and the narrow elongated protocone.

Hipparion (Hipparion) albertense serengetense, subsp. nov.

Hipparion albertensis (sic) Hopwood, 1926: 17, fig. 4.

Hypsohipparion albertense (Hopwood), Dietrich, 1942: 97, pl. 4, fig. 39, pl. 13, figs. 87-90,

93b, 96, 97; pl. 14, figs. 101-104, pl. 15, figs. 106-108, pl. 16, figs. 110, 112; pl. 20, fig. 160.

Equus (Hippotigris), Arambourg, 1947: 306.

Stylohipparion albertense (Hopwood) (pro parte) Arambourg, 1947: 306.

Diagnosis :

Three-toed equid with very hypsodont teeth (the M? of which reaches a

crown height of 80-90 mm.). The lower teeth are characterized by the constant

absence of the ectostylid. The protocone is very elongated, the cement is very

thick, the enamel plications being complex. The dental series is short, average

length being 156 mm.

Hipparion (Hipparion) albertense baardi, subsp. nov.

Type specimen: L946 (pl. 8) in the S.A. Museum, Cape Town.

Paratypes: Lo51, S.A.M.11717 (pls. 4, 2) in the $.A. Museum, Cape Town.

Type site: Baard’s Quarry, Langebaanweg, C.P.

1Jt must be a misprint. Everywhere else, except in the title, Pomel writes Hipparion

ambiguum.

2 This is tentatively included here on the basis of our belief that this symphysial fragment

is referable to other Stylohipparion specimens from this area.

’ This may prove to be a transitional form between the typical Hipparion and the more

progressive Hipparion (Stylohipparion).

390 ANNALS OF THE SOUTH AFRICAN MUSEUM

Diagnosis :

Three-toed equid with rather hypsodont teeth, about 70 mm. crown height.

The lower teeth are characterized by the constant absence of the ectostylid

and by a tendency to form other additional stylids, especially protostylid

extending along the total height of the crown. The protocone is elliptical but

less elongated than in serengetense, enamel is thin and plications are complex.

DISCUSSION

1. THE AFRICAN NON-PLEISTOCENE HIPPARIONIDS constitute

one genus (Hipparion) and two species (africanum and sitifense). The lack of

marked affinity to any of the Pleistocene forms which constitute a homogeneous

unity precludes any consideration of this group in terms of the subgenera

proposed for the Pleistocene forms.

2. THE AFRICAN PLEISTOCENE HIPPARIONIDS constitute a

certain unity marked by the characteristics of the genus Hipparion. Furthermore

they have in common a number of ‘African features’ which distinguish them

from the non-African forms. Nevertheless these different features do not

necessitate generic distinction.

Within the group there exist variables of differing degrees of importance.

The most obvious differentiating characteristic is the presence or absence

of ectostylids. The evolutionary trend and ecological adaptation of this feature

have been discussed (pp. 331, 332; 333) and it is considered that it constitutes a

differentiation at the subgeneric but not generic level. The statistical constancy

of absence or presence of this feature within groups otherwise identical indicates

that consideration must be accorded on both a non-generic and a non-specific

basis. Consequently the African Pleistocene Hipparion are subdivided into

two subgenera, viz.

Eipparion (Hipparion) de Christol 1832, and

Hipparion (Stylohipparion) van Hoepen 1932.

3. THE CONSIDERATION OF THE SPECIES of Hipparion (Stylo-

hipparion) is beset with the obvious difficulty of the paucity of the material

available. Previous workers, faced with the same problem, have either cautiously

avoided speciation or unjustifiably proposed species that are void, e.g.

Van Hoepen, 1932—steytleri, hipkini. Arambourg (1947, 1956, 1959) refers to

labycum only when he is discussing the North African Pleistocene material ;

otherwise he generalizes to the extent of using the generic (subgeneric) name

Stylohipparion without species determination. We concur with this cautious

attitude. We recognize that material referable to this subgenus has been

recovered from three different geographical areas, namely, the Maghreb,

East Africa (incl. Omo) and South Africa. Joleaud (1933), also aware of the

geographical distribution, recognized in his new genus Libyhipparion three

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 391

LAANGEBAANWEG HYPSOHIPPARION

heey rere FQ RSP ae ee, ES. Sp a ae = A

ZOOS

SLA

Sx SSS

ote

Upper 26.6 x 26.9 Upper 28.5 x 271

Lower 28.9 17.1 Lower 28.0 18.3

ROSA

LPO

aes

ONS.

Fic. 16. Diagrammatic superimposition of an upper and a lower molar tooth of the South

Serengeti “Hypsohipparion’, and the Langebaanweg Hipparion.

species, viz. libycum, ethiopicum and steytleri. However, despite the differences of

the localities, the wide range of variation within the small amount of available

material does not provide the evidence for species separation.

Comparison of the various ‘diagnoses’ provided by Hopwood (1937),

Arambourg (1947), and Cooke (1950), and reference to descriptions by Pomel

(1897) and Joleaud (1933) indicate quite clearly that no true, clear-cut diagnosis

is possible. Therefore, it is proposed to extend Arambourg’s generalized usage

of libycum to cover all the material included in this subgenus, but with the

awareness that subsequent discoveries may bring to light information to

highlight the differences of locality on a subspecies level.

392 ANNALS OF THE SOUTH AFRICAN MUSEUM

4. A better situation exists with reference to Hipparion (Hipparion) which

includes a portion of the material from South Serengeti (i.e. ‘Hypsohipparion’

Dietrich) and the Langebaanweg specimens.

The dimensions and the main features (fig. 16) of the teeth from both areas

are almost identical. However, there are differences in those features which

emphasize the ‘African’ character of the dentition; namely, the hypsodonty,

the elongation and narrowing of the protocone (tables, 16, 19, 20; figs. 12, 13).

For both of these characters the Serengeti material displays a more progressive

stage. Furthermore the latter also exhibits a thicker cement. These differences,

although constant, do not constitute sufficient grounds for species separation.

Therefore the species albertense is retained for both groups. Taking into considera-

tion the ecological adaptation of these features, it is proposed to distinguish

these two groups on a subspecies level. The Serengeti ‘Hypsohipparion’ group,

erected by Dietrich, is now considered to be Hipparion (Hipparion) albertense

serengetense, subsp. nov. The reason for dropping the genus Hypsohipparion has

been discussed above. The Langebaanweg group constitutes the subspecies

Eipparion (Hipparion) albertense baardi, subsp. nov. The name baardi is chosen

because the first specimens were recovered from Baard’s Quarry, owned by

Mr. J. Baard.

It seems that these two subspecies constitute the earliest Pleistocene forms.

The material (to be described by Dr. Stirton) from the base of Olduvai

Bed I should confirm or correct this suggestion.

ACKNOWLEDGEMENTS

This study was supported in part by the U.S. Public Health Service,

National Institutes of Health grant no. GM 10113-02(3, 4) and USPHS

General Research Support Grant 1-So1-FR—05367-o1. In addition, generous

assistance was obtained from

The Wenner-Gren Foundation for Anthropological Research Inc.,

New York;

Fonds National de la Recherche Scientifique (Brussels) ;

Fondation Universitaire de Belgique;

The Boise Fund, University of Oxford; and

The Dr. Wallace C. and Clara A. Abbott Memorial Fund of the

University of Chicago.

We are indebted to the African Metals Corporation (AMCOR) for

permission to work at the Langebaanweg sites, and we are grateful to Messrs.

Glathaar, Krumm, Muller and De Bruyn and Dr. Boardman for their co-opera-

tion. Mr. John Baard, owner of the farm ‘Langberg’, has kindly allowed

excavations. In particular we are thankful to Mr. Robin Warren, a chemist

of AMCOR, for his continuous efforts to recover material and for his enthusias-

tic assistance. Mr. Q. B. Hendey, research assistant to one of us (R.S.) at the

South African Museum, Cape Town, has been extremely. helpful. Some of the

information on the geology of the Langebaanweg sites results from discussions

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 393

with Mr. R. R. Inskeep, Dr. A. Fuller and Mr. D. Needham of the University

of Cape Town, and Mr. Robin Warren.

Dr. R. Bigalke, Director of the McGregor Memorial Museum, Kimberley

has been most co-operativein sending requested material for study. Professor R.A.

Stirton, Museum of Paleontology, University of California at Berkeley, provided

facilities for the study of some of the East African material there, which Dr.

L. S. B. Leakey allowed us to examine. Dr. Stirton kindly permitted us to

publish our observations on the material.

Dr. K. H. Fischer, Institut fiir Palaontologie und Museum der Math.-

Naturwissenschaftlichen Fakultat der Humboldt-Universitat zu _ Berlin

kindly took some measurements (at the authors’ request) on the original

Serengeti material.

Mrs. M. A. Norris patiently typed the manuscript.

SUMMARY

The recent discoveries of hipparionid material at the Langebaanweg

fossil sites (Cape Province, South Africa) necessitated a revision and interpreta-

tion of similar material recovered elsewhere in Africa. At the same time new

data on material from the Vaal River Gravels, Olduvai Gorge and South

Serengeti are added. A review of the geological and faunal data described

from approximately 40 sites in Africa leads to a tentative chronological correla-

tion. Consideration of the anatomical features (mostly of the dentition) provides

opportunity for a discussion of the ecological adaptation and evolutionary

trends. The problems of the origins and migrations of African hipparionids

are dealt with.

It is concluded that the African hipparionids belong to a single genus,

Hipparion. The Miocene forms remain limited to H. africanum. The Pliocene is

represented by H. sitifense, while the Pleistocene forms are referable to two

subgenera, viz. Stylohipparion and Hipparion.

The Serengeti ‘Hypsohipparion’ group, is referred to Hipparion (Hipparion)

albertense serengetense subsp. nov.

The Langebaanweg material forms a unified group and is referred to

Eipparion (Hipparion) albertense baard: subsp. nov.

REFERENCES

ABaprE, J., BARBEAU, J. & Coppens, Y. 1959. Une faune de vertébrés villafranchiens au Tchad.

C.R. Acad. Sci., Paris 248: 3328-3330.

Anprews, C. W. 1902. Note on a Pliocene vertebrate fauna from the Wadi Natrun, Egypt.

Geol. Mag. (n.s., Dec. 4) 9: 433-439.

AramgourG, C. 1943. Contribution a l’étude géologique et paléontologique du bassin du Lac

Rodolphe et de la basse vallée de l’Omo. tre partie: Géologie. Miss. sci. Omo. 1: 157-230.

ArampourG, C. 1947. Contribution a l’étude géologique et paléontologique du bassin du

Lac Rodolphe et de la basse vallée de l’Omo. 2e partie: Paléontologie. Miss. sci. Omo 1:

231-562.

Arampourc, C. 1949. Les gisements de vertébrés villafranchiens de l’Afrique du Nord. Bull.

Soc. géol. Fr. (5) 19: 195-203.

394 ANNALS OF THE SOUTH AFRICAN MUSEUM

ARAMBOURG, C. 1951. Observations sur les couches 4 Hipparion de la vallée de Oued el Ham-

mam (Algérie) et sur l’époque d’ apparition de la faune de vertébrés dite ‘pontienne’.

C.R. Acad. Sci., Paris 232: 2464-2466.

ARAMBOURG, C. 1952a. La paléontologie des vertébrés en Afrique du Nord Frangaise. Int.

geol. Congr. t9: Monogr. région. s. 1, Algérie, hors sér: 1-63.

ARAMBOURG, C. 1952). Note préliminaire sur quelques éléphants fossiles de Berbérie. Bull.

Mus. Hist. nat., Paris (2) 24: 407-418.

ARAMBOURG, C. 1954. La faune a Hipparion de Oued el Hammam (Algérie). Int. geol. Congr.

1g, Alger 1952, Ass. Serv. géol. afr., 2e partie (fasc. 21): 295-302.

ARAMBOURG, C. 1956. Sur des restes d’Hipparion sitifense Pomel, des calcaires lacustres de Mascara

(Oran). Bull. Soc. géol. Fr. (6) 6: 817-827.

ARAMBOURG, C. 1959. Vertébrés continentaux du Miocéne supérieur de |’Afrique du Nord.

Mém. Serv. Carte géol. Algérie 4: 1-161.

ARAMBOURG, C. 1960. Précisions nouvelles sur le Libytherium maurusium Pomel, giraffidé du

villafranchien d’Afrique. Bull. Soc. géol. Fr. (7) 2: 888-894.

ARAMBOURG, C. 1963. Continental vertebrate faunas of the Tertiary of North Africa. Viking

Fd. Publ. Anthrop. 36: 55-64.*

ARAMBOURG, C. & CHOUBERT, G. 1957. Les faunes de mammiféres de l’étage Maghrebien du

Maroc occidental. Conf. int. Ass. quatern. Res. 5. In press.

BisHop, W. W. 1963. The later Tertiary and Pleistocene in Eastern Equatorial Africa. . Viking

Fd. Publ. Anthrop. 36: 246—-275.*

Borisstak, A. 1914. Mammifeéres fossiles de Sebastopol. I. Mém.Com. géol. St-Pétersb. (n.s.) 87:

I-154.

Boute, M. 1910. Sur quelques vertébrés fossiles du sud de la Tunisie. C.R. Acad. Sci., Paris

150: 812-813.

Bourcart, J. 1937. Sur la découverte de molaires d’Hipparion a la base de la série lacustres a

argiles smectiques de Camp-Berteaux (Maroc oriental) C.R. Soc. géol. Fr. 8: 79-80.

Brain, C. K. 1958. The Transvaal ape-man bearing cave deposits. Transv. Mus. Mem. 11:

I-I3I.

Cuaput, E. & GiLtet, S. 1938. Les faunes de mollusques des terrains 4 Hipparion gracile de

Kiuiciik Cekmece pres Istanbul (Turquie). Bull. Soc. géol. Fr. (5) 8: 363-388.

Cuaput, E. & Nariz, H. 1934. Nouvelles observations sur les terrains néogénes de louest

d’Istanbul. Publ. géol. Inst. Univ. Istanbul g: 1-9.

Cuousert, G. & ENNoucut, E. 1946. Premiéres preuves paléontologiques de la présence du

Pontien au Maroc. C.R. Soc. géol. Fr. 11: 207-208.

CotBert, E. H. 1935. The correlation of the Siwaliks of India as inferred by the migrations

of Hipparion and Equus. Amer. Mus. Novit. 797: 1-15.

Cooke, H. B. S. 1949. Fossil mammals of the Vaal River deposits. Mem. geol. Surv. S. Afr. 35:

(3): 1-117.

Cooke, H. B. S. 1950. A critical revision of the Quaternary Perissodactyla of southern Africa.

Ann. S. Afr. Mus. 31: 393-479.

Cooke, H. B. S. 1960. Further revision of the fossil Elephantidae of southern Africa. Palaeont.

Afr. 7: 46-58.

Cooke, H. B. S. 1963. Pleistocene mammal faunas of Africa with particular reference to southern

Africa. Viking Fd. Publ. Anthrop. 36: 65-116.*

Cooke, H. B. S. & Wexts, L. H. 1946. The Power Collection of mammalian remains from the

Vaal River deposits at Pniel. S. Afr. J. Sci. 42: 224-235.

Coppens, Y. 1960. Le Quaternaire fossilifére de Koro-Toro (Tchad). Résultats d’une premiére

mission. C.R. Acad. Sci., Paris 251: 2385-2386.

Dattont, M. 1915. Le Miocéne supérieur dans l’ouest de l’Algerie; couches a Hipparion de la

Tafna. C.R. Acad. Sci., Paris 161: 639-641.

De Curisrot, J. 1832. Comparaison de la population contemporaine des mammiferes de

deux bassins du département de I’Hérault. Ann. Sci. Ind. Midi Fr., Marseille, 2 (5): 1-24.

Denizort, G. 1939. La Crau, la Camargue et l’Etang de Berre. I. Trav. Lab. Géol. Univ. Aix,

Marseille 2: 1-139.

Dretricu, W. O. 1942. Altestquartare Saugetiere aus der siidlichen Serengeti, Deutsch-Ostafrika.

Palaeontographica Abt. A, 94: 43-133.

Dierricu, W. O. 1950. Fossile Antilopen und Rinder dquatorial Afrikas (Material der Kohl-

Larsen’schen Expeditionen). Palaeontographica Abt. A, gg: 1-62.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 395

Du source De Bozas. 1903. D’Addis Abbaba au Nil par le Lac Rodolphe. Géographie 7: 91-112.

Duerst, J. U. 1926. Vergleichende Untersuchungsmethoden am Skelett bei Saiigern. Handb.

biol. ArbMeth., Abt. VII, H. 2, Lief. 200.

Ennoucui, E. 1951. Découverte d’un Hipparion dans les environs de Rabat. Notes Serv. Main.

Maroc 85: 139-144.

Ennoucui, E. & JEANNETTE, A. 1954. L’Hipparion de Camp-Berteaux prés de Taourirt (Maroc

oriental). Notes Serv. Min. Maroc 122: 65-75.

Ewer, R. F. 1957. Faunal evidence on the dating of the Australopithecinae. In Third Pan-African

Congress on Prehist. Livingstone, 1955. London: Chatto & Windus, 135-142.

Ewer, R. F. 1963. The contribution made by studies of the associated mammalian faunas.

S. Afr. F. Sci. 59: 340-346.

FLANDRIN, J. 1942. Le bassin a lignites de Marceau (département d’Alger). C.R. Acad. Sci.,

Paris 225: 370-372.

FREEDMAN, L. 1957. The fossil Cercopithecoidea of South Africa. Ann. Transv. Mus. 23: 121-262.

Fucus, V. E. 1934. The geological work of the Cambridge Expedition to the East African

lakes 1930-31. Geol. Mag. 71: 97-112; 145-166.

Gaupry, A. 1862. Animaux fossiles et géologie de |’ Attique. Paris: Savy. 1862-1867.

Gray, J. E. 1821. On the natural arrangement of vertebrose animals. Med. Reposit., London

15 (1): 296-310.

Gromova, V. 1949. Histoire des chevaux (genre Equus) de ancien monde. Trav. Inst. paléont.

Acad. Sct. URSS 17 (1): 1-373. (In Russian; tr. into French by Satnt-AuBIN, P. de. 1955.

Le genre Equus. Ann. Cent. Etud. Docum. paléont. 13: 1-203.)

Gromova, V. 1952. Les Hipparion (d’aprés les matériaux de Taraklia, Pavlodar et autres).

Trav. Inst. paléont. Acad. Sci. URSS 36: 1-473. (In Russian; tr. into French by Satnt-AUvBIN,

P. de. 1955. Le genre Hipparion. Ann. Cent. Etud. Docum. paléont. 12: 1-473.)

Hauge, E. 1911. Trazté de géologie. Paris: Colin.

Haucuton, S. H. 1932. The fossil Equidae of South Africa. Ann. S. Afr. Mus. 28: 407-427.

Hoepen, E. C. H. van. 1930. Fossiele perde van Cornelia, O.V.S. Paleont. Navors. nas. Mus.

Bloemfontein, 2: 13-24.

Hoepen, E. C. H. van. 1932. Die stamlyn van die sebras. Paleont. Navors. nas. Mus. Bloemfontein

2: 25-37.

Hopwoop, A. T. 1926. The geology and palaeontology of the Kaiso bone-beds. Occ. Papers

geol. Surv. Uganda. 2 (2): 13-36.

Hopwoop, A. T. 1929. A review of the fossil mammals of central Africa. Amer. 7. Sci. (5) 17:

IOI-118.

Hopwoop, A. T. 1937. Die fossilen Pferde von Oldoway. Wiss. Ergebn. Oldoway-Exped. 1913

(N.F.) 4: 111-136.

Hopwoop, A. T. 1951. The Olduvai fauna. In Leakey, L. S. B., ed. Olduvai Gorge: 20-24.

Cambridge: University press.

Howe t, F. C. 1959. The Villafranchian and human origins. Science 130: 831-844.

Joveaup, L. 1g1ga. Relations entre les migrations du genre Hipparion et les connexions conti-

nentales de l’Europe, de l’Afrique et de Amérique au Miocéne supérieur. C.R. Acad.

Sci., Paris 168: 177-179.

Jotzeaup, L. 1919). Sur les migrations 4 l’époque néogéne des Hipparion, des hippotraginés et

des tragelaphinés. C.R. Acad. Sci., Paris 168: 310-311.

Joveaup, L. 1933. Un nouveau genre d’equidé quaternaire de l’?Omo (Abyssinie) : Libyhipparion

ethiopicum. Bull. Soc. géol. Fr. (5) 3: 7-28.

Jory, A. 1909. Les formations continentales néogénes dans les hautes plaines constantinoises.

C.R. Acad. Sci., Paris 149: 323-324.

Kent, P. E. 1941. The recent history and Pleistocene deposits of the plateau north of Lake

Eyasi, Tanganyika. Geol. Mag. 78: 173-184.

Kent, P. E. 1942. The Pleistocene beds of Kanam and Kanjera, Kavirondo, Kenya. Geol. Mag.

79: 117-132.

Kout-Larsen, L. 1939. Vorlaiifiger Bericht iiber meine Afrika-Expedition 1937-39. Forsch.

Fortschr. 15: 339-340.

Kurtén, B. 1952. The Chinese Hipparion fauna. Comment. biol., Helsing f. 13 (4): 1—82.

Leakey, L. S. B. 1958. Some East African Pleistocene Suidae. Fossil Mammals Afr. 14: i-iu1,

1-132.

396 ANNALS OF THE SOUTH AFRICAN MUSEUM

Leakey, L. S. B. 1963. Very early East African Hominidae and their ecological setting. Viking

Fd. Publ. Anthrop. 36: 448-457.*

Leakey, L. S. B. & WuirwortH, T. 1958. Notes on the genus Simopithecus with a description

of a new species from Olduvai. Occ. Pap. Coryndon Mus. 6: 1-26.

LEPERSONNE, J. 1949. Le fossé tectonique Lac Albert-Semliki-Lac Edouard. Résumé des

observations géologiques effectuées en 1938-1939-1940. Ann. Soc. géol. Belg. 72: Mém.

fasc. 1: 3-88.

Lewis, G. E..1937. A new Siwalik correlation. Amer. F. Sci. (5) 33 (195): 191-204.

Lowe, C. van Riet, 1952. The Vaal River chronology: an up-to-date summary. S. Afr. archaeol.

Bull. 7: 135-149.

MaclInngs, D. G. 1953. The Miocene and Pleistocene Lagomorpha of East Africa. Fossil Mam-

mals Afr. 6: 1-30.

Maxson, J. H. 1930. A Tertiary mammalian fauna from the Mint Canyon formation of Southern

California. Publ. Carneg. Instn. 404: 77-112.

OAKLEY, K. P. 1954. Study tour of early hominid sites in southern Africa, 1953. S. Afr. archaeol.

Bull. 9: 75-87.

Oszorn, H. F. 1907. Evolution of mammalian molar teeth to and from the triangular type. New York:

Macmillan.

Osporn, H. F. 1918. Equidae of the Oligocene, Miocene and Pliocene of North America.

Mem. Amer. Mus. nat. Hist. 2: 1-217.

Perropon, A. & TEMPERE, C. 1953. Sur extension du Miocéne dans les Seba Chioukh (Oranie

occid., Algérie). Bull. Soc. géol. Fr. (6) 3: 583-588.

PIckERING, R. 1960. A preliminary note on the Quaternary geology of Tanganyika. Comm.

Coop. Techn. Afr. Sud Sahara (CCTA), Publ. 44: 77-89. (East-Central, West-Central &

South Reg. Comm. Geol., Jt. Mtg., Léopoldville, 1958.)

Piterm, G. 1938. Are the Equidae reliable for the correlation of the Siwaliks with the Cenozoic

stages of North America? Rec. geol. Surv. India 73: 437-482.

Pitcrim, G. 1941. The dispersal of the Artiodactyla. Biol. Rev. 16: 134-163.

Prrtort, P. L. 1956. Les formes européennes du genre Hipparion. Mem. Comun. Inst. geol., Barcelona

14: 1-122.

Pome, A. 1878. Sur un gisement d’Hipparion prés d’Oran. Bull. Soc. géol. Fr. 3: (6): 213-216.

Pome, A. 1897. Les équidés. Carte géol. Algérie, Paléont. Monogr. [12]: 1-44.

Recx, H. & Kout-Larsen, 1936. Erster Uberblick iiber die jungdiluvialen Tier- und Men-

schenfunde Dr. Kohl-Larsen’s im nordéstlichen Teil des Njarasa-Grabens (OstAfrika)

und die geologischen Verhaltnisse des Fundgebietes. Geol. Rdsch. 27 (5): 401-441.

Rosinson, J. T. 1961. The austrolopithecines and their bearing on the origin of man and of

stone tool-making. S. Afr. F. Sci. 57: 3-13.

Roman, F. & Soticnac, M. 1934. Découverte d’un gisement de mammiferes pontiens a

Douaria (Tunisie septentrionale). C.R. Acad. Sci., Paris 199: 1649-1650.

ScHLossER, M. 1903. Die fossilen Saiigethiere Chinas. Abh. bayer. Akad. Wiss., II KI1., 22, Abt.

I; I-22].

SEFVE, I. 1912. Die fossilen Perde Siidamerikas. A. Svenska VetenskAkad. Handl. 48 (6): 1-185.

SEFVE, I. 1927. Die Hipparionen Nord-Chinas. Palaeont. sinica (C), 4 (2): 1-93.

SHOTWELL, J. A. 1961. Late Tertiary biogeography of horses in the Northern Great Basin.

J. Paleont. 35: 203-217.

SINGER, R. 1961. The new fossil sites at Langebaanweg, South Africa. Curr. Anthrop. 2: 385-387.

SINGER, R. 1962. Simopithecus from Hopefield, South Africa. Bibl. Primat. 1: 43-70.

Sincer, R. & Bonet, E. L. 1960. Modern giraffes and the fossil giraffids of Africa. Ann. S. Afr.

Mus. 45: 375-548.

SINGER, R. & Futter, A. O. 1962. The geology and description of a fossiliferous deposit near

Zwartklip in False Bay. Trans. Roy. Soc. S. Afr. 36: 205-211.

Stncer, R. & Hooyer, D. A. 1958. A Stegolophodon from South Africa. Nature (Lond.) 182:

IOI-102.

SounceE, P. G., Visser D. J. L. & Lowe, C. Van Riet. 1937. Geology and archaeology of the

Vaal River basin. Mem. geol. Surv. S. Afr. 35: 1-164.

Sotienac, M. 1927. Etude géologique de la Tunisie septentrionale. Carte géol. Tunisie, Serv.

Mines, Tunis: 1-756.

* African ecology and human evolution. Eds. F. C. Howell and F. Bourliére.

HIPPARION FROM LANGEBAANWEG, CAPE PROVINCE 397

Sorienac, M. 1931. Le Pontien dans le sud Tunisien. Ann. Univ. Lyon (Sci. & Médic.) (n.s.)

1 (48): 1-42.

Sonpaar, P. 1961. Les Hipparion de Aragon meridional. Fstud. geol. Inst. Mallada 17: 209-305.

STEINMANN, G. & DODERLEIN, L. 1890. Elemente der Paldontologie. Leipzig.

Stirtron, R. A. 1939. Significance of Tertiary mammalian faunas in holarctic correlation with

special reference to the Pliocene in California. 7. Paleont. 13: 130-137.

Stirton, R. A. 1941. Development of characters in horse teeth and the dental nomenclature.

FJ. Mammal. 22: 434-446.

Stirton, R. A. 1942. Comments on the origin and generic status of Equus. F. Paleont. 16:

627-637.

Stirton, R. A. 1947. Observations on evolutionary rates in hypsodonty. Evolution 1: 32-42.

SrubDER, T. 1898. Ueber fossile Knochen vom Wadi Natrun Unteregypten. Mitt. naturf. Ges.

Bern 1460: 72-77.

TEILHARD DE CHARDIN, P. & Younc, C. C. 1931. Fossil mammals from the late Cenozoic in

northern China. Paleont. sinica (C), g (1): 1-88.

TEILHARD DE CHARDIN, P. & Leroy, P. 1942. Chinese fossil mammals. Publ. Inst. Géobiol.,

Pékin, 8: 1-142.

Tuentus, E. 1950. Ueber den Nachweis von Anchitherium aurelianense in Pannon des Wiener

Beckens. Anz. 6st. Akad. Wiss. 8: 174-181.

Tuomas, P. 1884. Recherches stratigraphiques et paléontologiques sur quelques formations

d’eau douce de l’Algérie. Mém. Soc. géol. Fr. (3) 3 (2): 1-51.

VILLALTA CoMELLA, J. F. (de) & CrusAront-Pair6, M. 1946-1948. Les gisements de mammi-

féres du Néogeéne espagnol. C.R. Soc. géol. Fr. 1946. 3: 49-513; 7: 133-134; 12: 224-227.

1947, 1: 28-30; 13: 256-259; 14: 278-280. 1948, 9: 167-170; 10: 186-189.

WAYLAND, E. J. 1926. The geology and palaeontology of the Kaiso Bone Beds. Occ. Pap. geol.

Surv. Uganda 2 (1): 5-11.

Werrnorer, A. 1888. Beitrage zur Kenntnis der Fauna von Pikermi bei Athen. Beitr. Paldont.

Geol. Ost. -Ung. 6 (3): 225-292.

WeELts, L. H. 1957. Speculations on the palaeogeographic distribution of antelopes. S. Afr. 7.

Sci. 53: 423-424.

We tts, L. H. 1964. The Vaal River ‘Younger Gravels’ faunal assemblage: a revised list.

S. Afr. J. Sct. 60: 91-93.

We tts, L. H. & Cooke, H. B. S. 1957. Fossil Bovidae from the Limeworks Quarry, Makapans-

gat, Potgietersrus. Palaeont. Afr. 4: 1-55.

YALcINLAR, I. 1952. Les vertébrés fossiles néogénes de la Turquie occidentale. Bull. Mus.

Hist. nat., Paris (2) 24: 423-429.

Ann. S. Afr. Mus., Vol. XLVIII Plate VII

D———__

Occlusal aspect. Scale: 1 cm. A—MMK 431. B—MMK 5225. C—L1467c. D—L1463.

E— L956.

Ann. S. Afr. Mus., Vol. XLVIII Plate VIII

Occlusal aspect. Scale: 1 cm. A—S.A.M. 11717. B—S.A.M. 11718. C—Lo40 / 950.

D—Lo34. E—L1465F.

Ann. S. Afr. Mus., Vol. XLVIII Plate 1X

Occlusal aspect. Scale: 1 cm. A—S.A.M. 11719. B—Lg53. C—Li1467p. D—Lg55.

Ann. S. Afr. Mus., Vol. XLVIII Plate X

Occlusal aspect. Scale: 1 cm. A—L1467H. B—L14678. C—L1459. D—Lg51. E—Lg58.

Ann. S. Afr. Mus., Vol. XLVIII Plate XI

A—D: occlusal aspect. Scale: 1 cm. A—L1467£. B—Lo39. C—L1451a. D—S.A.M. 11716.

E—S.A.M. 11716: section across middle of crown.

Ann. S. Afr. Mus., Vol. XLVIII Plate XII

Occlusal aspect. Scale: 1 cm. A—Lg43. B—Loqgr. C—L1465p. D—Lo47. E—Lg44.

Ann. S. Afr. Mus., Vol. XLVIII Plate XII

Occlusal aspect. Scale: 1 cm. A—Lg48. B—Lg42. C—Lg954. D—Lg36. E—S.A.M. 11724.

Ann. S. Afr. Mus., Vol. XLVIII Plate XIV

Gs F

Occlusal aspect. Scale: 1 cm. A—Lg46. B—Lg52. C—L1465a. D—Lo59. E—L1448.

F—Lo38.

Ann. S. Afr. Mus., Vol. XLVIII Plate XV

Occlusal aspect. Scale: 1 cm. A—Lg45. B—L14653B. C—S.A.M. 11722. D—Lo35.

E—Lg49. F—Lg57.

Ann. S. Afr. Mus., Vol. XLVIII Plate XV1

A, B, C—L1456: anterior, lateral and proximal surfaces respectively.

D, E, L1462<: lateral, proximal and anterior surfaces respectively.

Ann. S. Afr. Mus., Vol. XLVIII Plate XVII

A, B, C—L1462p: lateral, anterior and proximal surfaces respectively.

F—L1462c: anterior, lateral and proximal surfaces respectively.

Ann. S. Afr. Mus., Vol. XLVIII Plate XVIII

A, B, C—L1449: lateral, anterior and proximal surfaces respectively.

Tis

D, E, L1462p: proximal, anterior and lateral surfaces respectively.

Ann. S. Afr. Mus., Vol. XLVIII Plate XIX

A, B, C—L1444: lateral, anterior and inferior surfaces respectively.

VRE ee Fa Cer ee | "tat ee ol aia

INSTRUCTIONS TO AUTHORS

MANUSCRIPTS

In duplicate (one set of illustrations), type-written, double spaced with good margins,

including TaBLe or ConTENTS and Summary. Position of text-figures and tables must be

indicated.

ILLUSTRATIONS

So proportioned that when reduced they will occupy not more than 4? in. x 7 in. (73 in.

including the caption). A scale (metric) must appear with all photographs.

REFERENCES

Authors’ names and dates of publication given in text; full references at end of paper in

alphabetical order of authors’ names (Harvard system). References at end of paper must be

given in this order:

Name of author, in capitals, followed by initials; names of joint authors connected by &,

not ‘and’. Year of publication; several papers by the same author in one year designated by

suffixes a, b, etc. Full title of paper; initial capital letters only for first word and for proper

names (except in German). Title of journal, abbreviated according to World list of scientific

periodicals and underlined (italics). Series number, if any, in parenthesis, e.g. (3), (n.s.), (B.).

Volume number in arabic numerals (without prefix ‘vol.’), with wavy underlining (bold type).

Part number, only if separate parts of one volume are independently numbered. Page numbers,

first and last, preceded by a colon (without prefix ‘p’). Thus:

Smiru, A. B. 1956. New Plonia species from South Africa. Ann. Mag. nat. Hist. (12) 9: 937-945.

When reference is made to a separate book, give in this order: Author’s name; his initials;

date of publication; title, underlined; edition, if any; volume number, if any, in arabic numerals,

with wavy underlining; place of publication; name of publisher. Thus:

Brown, X. Y. 1953. Marine faunas. and ed. 2. London: Green.

When reference is made to a paper forming a distinct part of another book, give: Name of

author of paper, his initials; date of publication; title of paper; ‘In’, underlined; name of

author of book; his initials; title of book, underlined; edition, if any; volume number, if any,

in arabic numerals, with wavy underlining; pagination of paper; place of publication; name

of publisher. Thus:

SmitH, C. D. 1954. South African Plonias. In Brown, X. Y. Marine faunas. 2nd ed. 3: 63-95.

London: Green.

SYNONYMY

Arranged according to chronology of names. Published scientific names by which a species

has been previously designated (subsequent to 1758) are listed in chronological order, with

abbreviated bibliographic references to descriptions or citations following in chronological

order after each name. Full references must be given at the end of the paper. Articles and

recommendations of the International code of zoological nomenclature adopted by the XV International

congress of zoology, London, Fuly 1958, are to be observed (particularly articles 22 and 51).

Examples: Plonia capensis Smith, 1954: 86, pl. 27, fig. 3. Green, 1955: 23, fig. 2.

When transferred to another genus:

Euplonia capensis (Smith) Brown, 1955: 259.

When misidentified as another species:

Plonia natalensis (non West), Jones, 1956: 18.

When another species has been called by the same name:

[non] Plonia capensis: Jones, 1957: 27 (= natalensis West).