The most striking feature of the dentition of Zinjanthropus is the enormous size of the premolars and molars. All of the cheek-teeth are significantly greater even than those of the crassident South African australopithecine, Paranthropus, and, in some dimensions, a few of the teeth approach and even surpass some of the Gigantopithecus teeth of China. In Table 29, the dimensions of each of the sixteen teeth of Zinjanthropus are given.

Notes on methodology and terminology

In this account, we have used the term mesiodistal crown diameter (or simply m.d.) in preference to the term ‘length’, for some workers (e.g. Olivier, 1960) have designated this dimension ‘maximum breadth’. In fact, as Moorrees (1957, pp. 79–80) has pointed out, almost as many workers have called this measure crown breadth as have called it crown length!

Similarly, we have used the term labiolingual crown diameter (l.l.) for anterior teeth and buccolingual crown diameter (b.l.) for cheek-teeth, instead of ‘breadth’ or ‘maximum thickness’. Where both anterior and cheek-teeth are reflected in a single table (e.g. Table 29), the term buccolingual is preferred.

To measure these two dimensions, a Helios Vernier sliding caliper was used, with points especially sharpened to a fine taper, in order to enable them to be inserted between teeth in a jaw. The instrument was calibrated in millimetres and tenths of millimetres, as well as tenths and hundredths of an inch.

Although over fifteen partial or complete australopithecine maxillae have thus far been described from Africa, Zinjanthropus is only the second australopithecine specimen to have presented the world with a complete adult maxillary dentition. The first was a fine Australopithecus maxilla from Sterkfontein, Sts 52a. The presence of all sixteen teeth in the alveolar process of the Olduvai specimen, as well as of almost the entire palate, has made possible a very detailed study of the teeth and supporting structures. In the present chapter, the following aspects will be treated: the shape of the dental arcade and related features of the palate; the arrangement of the teeth in the arcade.

The shape of the dental arcade, alveolar process and palate

The maxillary dental arcade of Zinjanthropus presents an evenly arched, parabolic curve (pls. 20 and 29). The incisors and medially-placed canines are arranged in a low arch, extending between the right and left anteriorpremolars. From immediately behind the canines, the premolar–molar series continues with no trace of a diastema. The premolarmolar teeth lie in two slightly diverging rows, which are nearly straight save that the distal ends of the series are somewhat inturned. In other words, the arcade is very similar in shape to those of other australopithecines and of the Homininae (Le Gros Clark, 1955; Robinson, 1956). It differs markedly from the typical pongid pattern, in which the canine and post-canine teeth form approximately straight rows, parallel or even slightly divergent anteriorly, and separated from the anterior teeth by a distinct diastemic interval in the overwhelming majority of instances (Le Gros Clark, 1952b).

Cultural status

It has long been known that pebble-tools were made in Africa during the period of the australopithecines. But there have been two schools of thought on the relationship between the fossils and the implements. According to one view, Australopithecus was the maker of these stone implements. The notion has been forthrightly expressed by J. D. Clark as recently as 1963: ‘To take the example of the australopithecines:–Few would now doubt that they were representative of a hominid form responsible for the Pre-Chelles-Acheul Culture’ (p. 356). This view has been espoused by Arambourg (1958), by Oakley (1961, 1964) although he earlier (1954, 1956) took the opposing view, by Washburn (1959), Washburn and Howell (1960), Washburn and De Vore (1961), and others.

Exponents of the second view have doubted whether the australopithecines were stone toolmakers (Robinson and Mason, 1957, 1962; Inskeep, 1959; Mason, 1961; von Koenigswald, 1961, etc.).

In favour of the first view are the contemporaneity of the australopithecines and the pebbletools, and the discovery of a fragmentary, apparently australopithecine maxilla and a few possible pebble-tools in the same layer near the top of the Makapansgat deposit (Brain, Van Riet Lowe and Dart, 1955; Dart, 1955b). This discovery Dart (1955b) described as ‘providing the first concrete evidence that an australopithecine type…was contemporaneous with, and may have been responsible for, the concomitant pebble culture found in this sealed Central Transvaal cavern deposit’.

General considerations

‘Splitting’ and ‘lumping’: an historical perspective

The well-known trends of ‘splitting’ and ‘lumping’ are very much evident in the history of australopithecine taxonomy and nomenclature. Thus, at one time or another, the following generic labels have been applied to supposed members of this group: Australopithecus, Plesianthropus, Paranthropus, Telanthropus, Meganthropus, Praeanthropus, Zinjanthropus, and Hemanthropus (originally called Hemianthropus). We know today that many of these nomina did not represent valid taxa or good biological units.

By and large, splitting seems inevitable at an early stage in the sequence of discoveries of a group; lumping is equally inevitable at a later stage when many specimens have accumulated, when ranges of variation for the group in question and for related groups have been minutely explored, when the time dimension is better known, and when a deeper understanding has been gained of the functional and cultural capacities associated with a group of fossils showing a particular morphological pattern.

Because we now possess much of the detailed knowledge necessary to detect where affinities lie, the time is ripe for a reconsideration of hominid fossil designations and the combining of previously separated groups of fossils into broader and more valid taxa.

The fact that the author has been led by his studies of Zinjanthropus and other australopithecines to regard them all as members of a single genus is based not upon any enhanced biological insight, but upon the irresistible body of evidence that has accumulated over the four decades since Australopithecus africanus was named and launched by Dart (1925).

The external surface of the basis cranii of Zinjanthropus is beautifully preserved (pls. 6 and 14). With the exception of a few parts of the left temporal bone and the apex of the right petrous pyramid, the entire basis is preserved, down to the finest anatomical detail, as far forward as the body and greater wings of the sphenoid bone. Furthermore, the parts which are missing or damaged on one side are intact on the opposite side.

The occipital bone

The planum nuchale

The character of the nuchal crest has been dealt with in the preceding chapter. Immediately below the nuchal crest, on either side of the external occipital protuberance and of the posterosuperior part of the external occipital crest, is a clear oval impression, about 27 by 16 mm., in the area between the superior and inferior nuchal lines. In modern man, the semispinalis capitis muscle is attached to a corresponding impression. Anterolateral to the last impression is another somewhat irregular impression, corresponding to the attachment of m. obliquus capitis superior. Posterolaterally, the lower lateral limb of the nuchal crest is rugged, as it flows towards the adjacent pars mastoidea and mastoid process: this part probably provided attachment to the splenius capitis muscle.

The inferior nuchal line is not very clearly defined. It runs below and in front of the impression for semispinalis capitis; then continues laterally for a short distance, turning sharply forwards opposite the impression for m. obliquus capitis superior, finally flowing over the lateral part of the exoccipital and terminating in the jugular process.

This is the second volume of the new series of publications dealing with the discoveries made at Olduvai Gorge since 1951. The title of the first volume included the dates ‘1951–1961’. These have been dropped from the present volume, and will also not occur in the remainder of the series, since, inevitably, data obtained since 1961 will need to be included. The next two volumes in this series will deal with the cultural sequence and excavations of the ‘living floors’ in Beds I and II, at Olduvai, by Mary D. Leakey, and the detailed geology of the deposits exposed in the Gorge by Richard L. Hay. It is not yet certain which of the two will be ready for publication first.

A fifth volume is in preparation by Phillip Tobias, the author of this volume on ‘Zinjanthropus’. In it he will describe, in detail, the fossil hominid remains from Olduvai that are attributed to Homo habilis, and to Homo erectus, as well as the new mandible of an australopithecine of ‘Zinjanthropus’ type from Peninj, north of Olduvai. The fifth volume will also include chapters dealing with the postcranial hominid material from Olduvai by Dr John Napier, Dr Peter Davis and Dr Michael Day.

In the present volume, my friend and colleague, Phillip Tobias, has produced a truly magnificent piece of research—a major contribution to the study of the anatomy of the australopithecines, or ‘near men’ as I call them.

The cranial vault

The curvature and components of the vault

(1) The parieto-occipital plane of Zinjanthropus rises steeply from the external occipital protuberance. The vault of Australopithecus is more evenly curved in this area.

(2) The parietotemporal side-walls of the vault also rise steeply, contributing to the full, rounded character of the vault. This feature finds a parallel in the latest cranium of Australopithecus from Makapansgat, MLD 37/38.

(3) The vault of Zinjanthropus, as seen in norma verticalis, is long-spheroid, whereas that of the reconstructed Paranthropus is short-spheroid and that of Australopithecus long-ovoid. Zinjanthropus is intermediate in this respect between the two South African forms, though approaching more closely to the second reconstruction of Paranthropus.

(4) Most dimensions of the parietal bones in Zinjanthropus fall within the range of variation of Australopithecus, while a few dimensions exceed the range. In anteroposterior extension, the australopithecine parietal is slightly smaller than that of Homo erectus, whereas, in mediolateral extension, the australopithecine parietal is far smaller than that of H. erectus. In contradistinction, the anteroposterior extent of the parietal bone in pongids is far smaller than in australopithecines.

(5) The chord-arc indices of the edges of the parietal bones show that in Zinjanthropus, the parietal bone is slightly less curved along the coronal margin than in Australopithecus. The anteroposterior curvature of the parietal bones of Australopithecus tends to be greater, and the mediolateral curvature less, than in H. erectus.

The fossil remains, to which the name ‘Zinjanthropus boisel’ anthropus boisei' has been given, comprise an almost complete cranium with the entire maxillary dentition. In the same level of Bed I were found some fragments of a second cranium and parts of the shafts of a gracile hominid tibia and fibula, while from the slope just below were recovered an upper lateral incisor, an upper molar and a lower premolar of a very gracile dentition. According to Leakey, ‘the hominid skull was found as a single unit within the space of approximately one square foot by about six inches deep…The expansion and contraction of the bentonitic clay, upon which the skull rested and in which it was partly embedded, had resulted, over the years, in its breaking up into small fragments…’ (1959a, p. 491). However, despite its fragmentary condition, the individual pieces of bone are in a beautiful state of preservation: there is no sign of warping or squashing or other pressure effects, nor are the surfaces defaced by weathering. As a result, a wealth of anatomical detail is preserved in the cranium, which compares most favourably with the very best of the australopithecine remains hitherto described from South Africa. In completeness and state of preservation, the Zinjanthropus cranium must take its place alongside the child skull from Taung and the adult cranium (Sts 5) from Sterkfontein as the most perfect of the early hominid crania thus far found in Africa.

The Olduvai Gorge in the Republic of Tanzania came to the attention of the world shortly after my mother Mary discovered the ‘Zinjanthropus boisei’ skull on July 17th 1959. The field of African prehistory, and in particular the study of human evolution, has changed and developed dramatically over the past 50 years. I am particularly pleased that Cambridge University Press have decided to republish the 5 monographs that comprehensively cover the many scientific studies that have been undertaken on the Olduvai material collected by my parents, Louis and Mary, working with a number of colleagues. As the Golden Anniversary of the discovery approaches, it is timely to reflect on the importance of that find.

I was lucky to arrive at Olduvai two days after the discovery and I well recall the excitement of the occasion. My parents were operating on a very tight budget and the field season was short. Fortunately, on hand was world-renowned photographer Des Bartlett who, aided by his wife Jen, fully recorded on film the first few days of excavations and reassembly of bone fragments back in camp. As pieces were glued back together, and the shape of the skull and its morphology became clear, my parents showed uncharacteristic and unrestrained emotion! At the time, ages for fossils were wild guesses and radiometric dating had not been done anywhere in Africa.

In Table 18 are recorded a number of measurements of the Zinjanthropus cranium, especially metrical characters which have not been dealt with in the preceding sections. Most of these measurements relate to the calvaria alone, but several—such as nasion–basion length and basion–prosthion length—are, strictly speaking, cranial rather than calvarial measurements. To compare with the Zinjanthropus data, I have assembled as many measurements on the South African australopithecines as could be culled from Broom et al. (1950), Broom and Robinson (1952), Dart (1962b) or obtained by personal measurement; as well as some comparable measurements on Homo erectus.

Considerable difficulties arise when one attempts to apply measurements devised and defined for the description of modern human crania to ancient hominids, as well as to non-hominid Primates. These problems have been fully discussed by Davidson Black (1931), Washburn (1942) and Weidenreich (1943) among others. Among the most difficult termini of measurements are inion and euryon.

Cranial length and the toro-occipital index

In modern hominine crania, the inion and the opisthocranion (furthest occipital point) are different points, the former low and the latter high on the cranium. On the other hand, in some Neandertal crania, in H. erectus, in the australopithecines and in many pongid crania, the two points coincide to mark the posterior terminus of the maximum cranial length (glabella–opisthocranion).

The planum nuchale: tilt and height

The tilt of the planum nuchale

The planum nuchale of the occipital squama is tilted at only a slight angle to the plane of the foramen magnum which, in turn, is tilted upwards at an angle of 7° to the F.H. (see section on ‘The Plane of the Foramen Magnum’). Thus the planum nuchale of Zinjanthropus is much more horizontal than is that of the Pongidae, in which the nuchal surface rises steeply upwards to the nuchal crest. This low, nearly horizontal planum nuchale is characteristic, too, of Australopithecus (e.g. MLD 37/38 and Sts 5), though in the very unusual and apparently deformed cranium VII from Sterkfontein, the planum is much steeper and more pongid in appearance. Broom and Robinson (1950, p. 25) suggested that this unique feature in Sts VII, which they likened to the effects of artificial deformation on Amerindian crania, was due to ‘slow postmortem crushing without very manifest breaking of the bones’. In Sts VIII, as well, the planum nuchale ‘makes an angle with the main part of the base of the basioccipital of about 49°’ (Broom and Robinson, 1950, p. 28). The latter plane is thus not nearly as horizontal as in Zinjanthropus. Of the Paranthropus crania published, the only ones in which this area is satisfactorily preserved are those of two juveniles from Swartkrans, in which the plane of the nuchal surface is very similar to that in Zinjanthropus (Broom and Robinson, 1952, pp. 26–30).

Attrition of individual teeth

Although the M3's of Zinjanthropus are not fully erupted, the other teeth are already markedly worn with one or more areas of exposed dentine on every one. Such early wear is characteristic of the Australopithecinae. The degree of wear may be compared with that in South African australopithecine maxillae, of which two specimens of Australopithecus (Sts 52 a and Sts 37) and four specimens of Paranthropus (the Kromdraai specimen, SK 13, SK 52 and SK 49) have the M3's not fully erupted, some being a little more advanced and some at a slightly earlier stage of eruption than Zinjanthropus: in all these maxillae, the amount of attrition is far less than in Zinjanthropus. The degree of attrition in the latter has already levelled the cusps and left numerous areas of dentine exposure (pl. 29). The exposed dentine is black in colour, in contrast with the yellowish-grey enamel. Areas of exposure are thus particularly clearly delimited.

The incisors have a substantial strip of exposed dentine, extending for the full mesiodistal diameter of the teeth, save for the mesial and distal enamel walls (pl. 32). Wear is slightly greater on the right I than on the left, the incisal surface being hollowed on the former, but straight to convex on the latter. The canines show the typically hominid form of wear from the tip (cf. Le Gros Clark, 1950a), which has resulted in a nearly flat occlusal surface with an irregularly biconvex area of dentine exposure, somewhat larger on the right canine than on the left.

The supra-orbital torus (Pls. 1 and 15)

Dominating the face is a powerful supra-orbital torus which is far more robust than in any of the australopithecines hitherto discovered. To illustrate this point, the vertical thickness of the torus was measured at the highest point on the orbital margin: the value in Zinjanthropus (13·6 mm. left, 13·7 right) is nearly twice as great as in most specimens of Paranthropus (7·4 in SK 46; 7·2 in SK 847; 7·8 and 5·6 in SK 48), with the exception of SK 52 which approaches the value in Zinjanthropus (12·4). The values in Sts 5 (Australopithecus) are 7·8 and 9·6 mm. The values quoted by Weidenreich (1943, p. 30) for the toral thickness in Homo erectus pekinensis reveal a range in the middle of the orbital margin of 11·5–17·4 mm. (n = 5 crania, 8 sides). If our technique of measuring this vertical thickness of the torus is comparable with his, then three out of five Choukoutien crania have thicker tori even than Zinjanthropus. This is in keeping with the overall robusticity of the cranium of H. erectus, whereas the australopithecine cranium is an essentially gracile one, though with heavy ectocranial embellishments and tori. Again, if the anteroposterior width of the torus be measured from the mid-point of the anterior surface of the upper margin of the orbit (just lateral to the trochlea) to the temporal crest or line behind, the values in millimetres are as shown in the table in the next column.