Chad, Cradle of Humanity ?

With Australopithecus bahrelghazali (Abel) and Sahelanthropus tchadensis (Toumaï), two hominids fossils, Chad is a potential Cradle of Humanity

Toumaï Shows the Teeth

More Details in the French Version : Tchad berceau de l'humanité ?
see also : Meteorite Impacts in the Borkou and Ennedi (Chad)

Press review in relation with these articles :

Ann Gibbons, Fossil Tooth Make Fur Fly, Science, December 2004.

Charles Bremmer, Dispute over an 'oldest tooth'. The Times, December 02, 2004.

Ann Gibbons, T he First Human. The Race to Discover our earliest ancestors.Doubleday, April 2006, 306 p.

Dalton Rex, Brickbats for fossil hunter who claims skull has fase tooth. Nature, Vol 430, August 26, 2004, pp. 956.

A short article may have unexpected consequences :

1) SAJS 100, March/April 2004.
The collage of a fossil linked to human origins is denounced by a geographer and an orthodontist.

2) SAJS 100, September/October 2004.
What did you think would happen? 28 eminent paleoanthropologists, on the basis of scans, declared in writing that the geographer and the orthodontist did not have the competence to put a tooth back either on the right or on the left of the mandible, and that the tooth in question is from the right side.
Note that petitions and votes of this kind are rarely used to advance science and that, in several celebrated cases, have boomeranged against the authors.

3) SAJS 101, March / April 2005.
28 paleoanthropologists did not notice that the digital crown images were in complete contradiction with the description in the same paper, while the scale of the documents was only one tenth of reality (SAJS 100, September-October 2004, 'Sahelanthropus tchadensis: the Facts' and 'International Letter of Support' by Michel Brunet and MPFT).

As is usual for a third molar, the mesial root is vertical and the distal root is inclined towards the rear of the mandible. Also the description made by the MPFT of section A of the CT scans indicates the front of what is actually the rear. The description of section B indicates the lingual surface to the right which corresponds to a left tooth, which was not the aim of the demonstration.

The confusion is total and is very unfortunate for those who confidently signed an 'International Letter of Support' of the description of a right tooth on the basis of a sheet of CT scans whose description was reversed, as was subsequently recognized by the MPFT in 'Note by M. Brunet et al.’ (SAJS 101, March-April 2005).

South African Journal of Science

Vol 100, March/April 2004

'Further details concerning Sahelanthropus tchadensis.' Alain Beauvilain and Yves Leguellec

Vol 100, September/October 2004

'Sahelanthropus tchadensis: the facts Michel Brunet and MPFT†.'

'Beauvilain and Le Guellec reply'

Vol 101, March/April 2005

'Further correspondence concerning Sahelanthropus tchadensis.' Alain Beauvilain, Yves Leguellec

'Note from M. Brunet et al.'

***South African Journal of Science*, Vol 100, March/April 2004**

Further details concerning fossils Sahelanthropus tchadensis (Toumaï)

Alain Beauvilain and Yves Leguellec

The fossils attributed to Sahelanthropus tchadensis from the Late Miocene of Chad are of great scientific and cultural importance, and are the subject of debate as to whether they represent a hominid or an ape. It is therefore necessary to provide full and accurate details of the fossil collection. We demonstrate that the third lower molar of the right mandible illustrated in the Nature article (1) that announced the discovery is in reality a left m/3, and an incisor collected at the site did not feature in the paper. These particulars affect the interpretation of the fossils, including estimates of the minimum number of individuals involved and the associations between the various specimens.

Introduction
The discovery and annoncement of fossils attributed to Sahelanthropus tchadensis (1) has confirmed the importance of the Late Miocene of Africa as a period of major interest for the study of hominid origins. The interpretation of Sahelanthropus as an early hominid (possibly even a bipedal one) is the subject of debate, with some authors claiming that it was probably a quadrupedal ape (2). Brunet (3) has maintained his view that Sahelanthropus was a hominid. Because of the nature of this debate, and its implications for palaeoecology and human origins, it is necessary to provide further particulars regarding the hypodigm and anatomy of Sahelanthropus. The aim of this contribution therefore is to record geomorphological details about site TM 266 (and its environs), which yielded the Sahelanthropus fossils, and also to describe the events surrounding the discoveries themselves, as it seems there is some confusion about the fossil associations, and even about the total number of hominoid specimens recovered from the site. Given the great value of the discovery in terms of human heritage and its scientific importance, we deem it necessary to provide information about the context of the findings, as well as particulars of the fossils themselves.
The first author led the field trip that resulted in the original discovery of the skull popularly known as Toumaï (4) as well as the later expeditions that recovered the mandible fragments and the isolated teeth attributed to Sahelanthropus. He was thus intimately involved with the events as they unfolded.

Geological context
Toros-Menalla is the name given by palaeontologists to an extensive region about 150 km long in the west of the Djourab Erg, Chad. A notable feature in the centre of this zone is a low scarp generally about 10 m high and about 40 km long (Fig. 1). This escarpment, facing east, is clearly indicated on the topographic map IGN NE-33-VI at a scale of 1:200 000 over a distance of more than 25 km (from 16°20’N, 17°15’E to 16°15’N, 17°28’E). This relief possibly represents an ancient fault which became reactivated as a result of tectonic forces related to Tibesti volcanism. Emi Koussi (a large volcanic cone 80 km in diameter, culminating at an altitude of 3415 m, that was built up from the end of the Tertiary, Late Miocene to sub-Recent times) is only 400 km away. Today, evidence of tectonic activity is closer to the Djourab, such as the northern margin of the Chad basin where the Angamma cliffs attest to an underlying flexure.
It is unfortunate that the area has been described as ‘a desolate plain, the monotony of which is interrupted only by the modern dunes’(4), which gives a distorted impression of the potential of the area both for stratigraphy and palaeontology. A similar restricted impression is given by the stratigraphic section published by Vignaud et al. (5), which shows 6-m-thick sediments comprised of only two stratigraphic units. It should be pointed out that this section was measured at TM 266, but that, along the 40-km length of the scarp, the succession is really thicker and more complex than this (Fig. 1), with lacustrine, peri-lacustrine, subaerial, fluvial and aeolian deposits.
The scarp forms either a unique escarpment, or a succession of lower terrace-like steps (Fig. 1), sometimes separated from each other by several hundred metres. Even though it is levelled off at its two extremities, in the central part a combination of corrasion and deflation by the wind has in places uncovered a well-defined scarp or series of structural flats and bluffs (Fig. 1). There, the uppermost part is rich in geomorphological and archaeological evidence indicating the shores of the last phase of Lake Mega-Chad. However, the sediments in the lower parts of the scarp are considerably older, dating from the Late Miocene on the basis of mammalian biochronology, and are not related to the Pleistocene to Recent history of Lake Chad, but rather to a much older precursor.
To give an idea of the palaeontological richness of the area, it suffices to point out that from the first reconnaissance of the scarp from July 2001 to March 2002, 59 fossiliferous localities were mapped, which together yielded 1627 fossils. These represent 20% of all the fossils that have been stocklisted since 1994 by the scientific project. Among these fossils, new taxa of Giraffidae, Hippopotamidae and Anthracotheriidae as well as Sahelanthropus were discovered. The collection also contains an abundance of microfauna and other vertebrates, both aquatic and terrestrial, such as Deinotherium, Anancus and fish. Of these localities, it is TM 266 with its 691 fossils (in March 2002) that accounts for the richest assemblage. The fossiliferous potential of this area is still important, notably for finding small fossils and in relation to the entire surface that is covered with a thin layer of sand or dunes that the wind may sweep away.

TM 39 talus 2

Fig. 1. View of the east-facing scarp in the Toros-Menalla (site TM 039) region of the Djourab Desert, Chad, showing the Miocene deposits cropping out as small bluffs (usually composed of sandstones or other resistant layers such as diatomites), overlain by modern sand dunes. The erosion in this area is largely due to wind deflation, but it is thought that the scarp owes its origin to the presence of a fault line, which is obscured under sands on the downslope side of the scarp. Note that the total thickness of Miocene strata in the region greatly exceeds 10 metres. Despite the possibility of reinterpreting the fossils, the mammalian assemblages collected appear to comprise exclusively Late Miocene taxa.

Palaeoanthropology
Hypodigm of Sahelanthropus tchadensis.
There are some differences between the list of fossils attributed to Sahelanthropus tchadensis in the article in Nature (1) and the one that was published on the Internet (6) based on specimens that were collected in the field by the first author and by his Chadian colleagues and Laurent Viriot, itself based on the field catalogue made at the time of the discoveries (7). The following account deals with the fossils in the chronological order of their discovery. The catalogue numbers comprise four parts : the first, TM 266, refers to the locality ; the second indicates the year of discovery ; the third is the catalogue number of the specimen in the order of discovery ; and the fourth (if present) denotes different elements of a single fossil.
TM266-01-060-1 (the skull) was simultaneously discovered next to TM 266-01-060-2 (a symphysis fragment with incisor and canine alveoli) on 19 July 2001, by Ahounta Djimdoumalbaye. From the same place, an unpublished incisor was found on 21 December 2001 at N’Djamena by Laurent Viriot, while picking through screening residues. The latter specimen comprises half an incisor, split vertically from the apex of the crown to the tip of the root [Fig. 2(4)]. This tooth differs from the upper incisor (TM 266- 01-448) that was illustrated in Nature (ref. 1, fig. 2a).
It should be pointed out that it has been the practice of the Mission Paléoanthropologique Franco-Tchadienne not to write numbers on fossil hominid specimens, a policy that has been enforced since the discovery of the mandible of Australopithecus bahrelghazali (8). The catalogue numbers of specimens TM 266-01-447 (molar in ref. 1) and 448 appeared for the first time in the CD database of fossils established at the University of Poitiers and for the second time in the Nature article in July 2002. We are unaware of specimen TM 266-01-447 and the exact sites and circumstances of the discovery of fossils TM 266-01-447 and 448.
On 1 November 2001, the complete right upper canine belonging to TM 266-01-060-1 was found, exactly where the skull was located, as a result of screening by Fanoné Gongdibé. The tooth consisted of two fragments which fitted perfectly onto the right canine root.
TM 266-02-154-1 (right hemi-mandible with P4 M2 and roots of p/3 and m/3) and TM 266-02-154-3 (crown of left M3) were recovered on 20 January 2002, about 105 m to the north-northeast of the Toumaï skull site. This lower jaw was the subject of two photographs in the Nature paper (ref. 1, fig. 2b,c) one in occlusal view, the other in the form of an axial CT scan.
TM 266-02-154-2 (lower canine; ref. 1, fig. 2d,e) was first recognized on 30 January 2002, at N’Djamena, in screening residues from the spot where the mandible and m/3 were found. It was completely encrusted with sand and could not be photographed at the time of discovery. Finally, TM266-02-203 (mandibular symphysis with premolar, canine roots and incisive alveoli) was discovered on 12 March 2002, a dozen metres south of the mandible and m/3.

planche.mars.04.96pts site

Fig. 2. Fossils attributed to Sahelanthropus tchadensis, from site TM 266, Toros-Menalla, Chad. Scale : 10 mm.
1) TM 266-02-154-1, oblique lingual view of right hemi-mandible showing the specimen in the condition that it was found, lacking the m/3 crown, and exposing the roots of the p/3.
2) TM 266-02-154-1, occlusal view of right hemi-mandible made soon after discovery, showing the roots of m/3.
3) TM266-02-154-3, occlusal view of left m/3 made soon after discovery, and before the specimen was cleaned (mesial at the top). (4) TM 266-01-448 ?, unpublished damaged incisor found near the holotype skull of Sahelanthropus tchadensis.

Anatomy
The photograph in Nature of the right hemi-mandible, TM 266-02-154-1 (ref. 1, fig. 2b,c), shows the complete tooth row from p/4 to m/3. In occlusal view, the m/3 appears as a well-preserved tooth, whereas the three other teeth are headline worn and damaged. In contrast, in the tomographic section, the roots of the m/3 appear to be damaged, the two roots being broken. In fact, the m/3 molar illustrated, of which only the crown is preserved, was found isolated several decimetres from the hemi-mandible. It is here catalogued by the first author as TM 266-02-154-3. It was given the suffix 3, following 1 (mandible) and 2 (canine) of the other pieces attributed to TM 266-02-154, the only indication in the field catalogue. This m/3 was later fixed onto the lower jaw in the laboratory at the University of Poitiers.
Figure 2(12) shows the mandible as it was on the day of its discovery. Note that the right m/3 crown is missing. The roots of the p/3 are clearly visible in anterior oblique view [Fig. 2(1)], extending deeply into the corpus of the mandible and curving distally as they do so. The extremely buccal position of the mesial root (on the left in the figure) indicates that the canine root was probably large and that its apex was positioned appreciably further distally than its cervix. The inferred large size of the canine root indicates that this mandible probably belonged to a male individual.

dent M3 96 pts site

TM 266 M3 2

Fig. 3. Interpretation of TM 266-02-154-3, left m/3. The disposition of the protoconid, hypoconid, hypoconulid, metaconid and entoconid reveals unequivocally that this lower molar is from the left side. Scale:10 mm.

Additional photography.

The occlusal outline of the tooth TM 266-02-154-3 immediately evokes a hominoid lower molar [Fig. 2(3)], being pentacuspidate with the mesio-distal diameter of the occlusal surface greater than its bucco-lingual one, and the mesial margin [at the top in Fig. 2(3)] broader than the distal one, which is also more convex indicating that it is most likely an m/3. The protoconid is larger than the hypoconid and the hypoconulid but is smaller in area than the entoconid (Fig. 3). The metaconid is slightly mesio-distally shorter than the protoconid but is bucco-lingually slightly broader. There is a prominent mesio-distal groove separating the buccal cusps from the lingual ones, located slightly buccally to the midline of the crown; despite the wear on the tooth, the transverse grooves between the protoconid and the hypoconid on the one hand and the hypoconid and the hypoconulid on the other are clearly visible. The transverse groove separating the metaconid and entoconid is worn away except at its lingual and buccal extremities. In Fig. 2(3) the mesial fovea is obscured under a sandy concretion, but the distal fovea is discernible. In the Nature article (ref. 1, fig. 2b), the concretion has been removed, but the mesial fovea is difficult to discern, probably because it may have been worn away. A remarkable feature of this tooth is the presence of a large metaconulid positioned disto-lingually to the metaconid (sometimes known as the ‘sixth’ cuspid). Such cuspids are rare in extant humans, occurring in 7% of m/1’s (tooth 36 in dentist’s terminology) and about 2% of m/3’s (tooth 38). Sixth cuspids also occur in samples of Pan and Gorilla lower molars. Given that the protoconid, hypoconid and hypoconulid are on the left of this tooth, this indicates that the specimen is a left lower molar, meaning that a left m/3 has been mistakenly glued into a right mandible.
We provisonally estimate that there is a minimum of three individuals in the TM 266 hominoid hypodigm: one represented by the type specimen, skull TM 266-01-060-1; a second by the hemimandible TM 266-02-154-1, the teeth of which are more worn than those in the type specimen; and a third by the mandibular symphysis, TM 266-02-203, which shares parts in common with the right lower jaw. It remains to be determined whether the third symphyseal fragment with incisor and canine alveoli (TM 266-01-060-2 in ref. 1) and the isolated teeth from the site represent the same individuals as these three.
Two more hominoid specimens were recovered in the neighbourhood, an edentulous hemi-mandible (TM 247-01-02), found by Mahamat Adoum on 15 July 2001, and a symphysis broken into four fragments (TM 292-02-01 with left C, p/4m/1, right p/3, m/1m/2) collected by the first author and two of his colleagues on 15, 16 and 17 January 2002. These fossils probably represent two additional individuals of Sahelanthropus.

Palaeoecological implications
After the discovery of Australopithecus bahrelghazali (‘Abel’) on 23 January 1995 in Chad Sahara, the unearthing of Sahelanthropus (‘Toumaï’) on 19 July 2001 confirms the existence of a Lake Chad Story, countering the prediction made by Coppens of an East Side Story (9). And yet, if Toumaï is not a hominid (2), its discovery could be used to support the hypothesis of the East Side Story.
Other mammals found in the same levels, such as anthracotheres, which are abundant in Chad but have not been reported from the richly fossiliferous deposits of the same age in East Africa (10), reveal that there was indeed marked provincialism in the Late Miocene faunas of Africa and show also that the known environmental conditions of the Chad Basin could have supported the emergence of human species.
The present analysis, based on some Sahelanthropus paratypes, is only one element of the discussion about this genus. It does not modify the basis of the debate concerning the systematic position and palaeoecology of Sahelanthropus.

We thank the Centre National d’Appui à la Recherche, a Chadian organization dependent on the Ministère de l’Enseignement Supérieur and its successive heads, Abakar Adoum Haggar and Baba El Hadj Mallah, the French Ministry of Foreign Affairs, who employed A.B., and the SCAC (Cooperation Section of the French Embassy, N’Djamena) for research support, and the University of Paris X Nanterre for approving the long secondment of A.B. to Africa. We are particularly anxious to thank our Chadian colleagues, Mahamat Adoum, Likius Andossa, Ahounta Djimdoumalbaye, Fanoné Gongdibé and Mackaye Hassan Taïsso, for their hard work, excellent companionship in the field and willingness to participate in scientific endeavour despite the hardships of desert conditions.

1. Brunet M., Guy F., Pilbeam D., Mackaye H.T., Likius A., Ahounta D., Beauvilain A., Blondel C., Bocherens H., Boisserie J-R.,De Bonis L., Coppens Y., Dejax J., Denys C., Duringer P., Eisenmann V., Fanone G., Fronty P., Geraads D., Lehmann T., Lihoreau F., Louchart A., Mahamat A., Merceron G., Mouchelin G., Otero O., Pelaez Campomanes P., Ponce De Leon M., Rage J-C., Sapanet M., Schuster M., Sudre J., Tassy P., Valentin X., Vignaud P., Viriot L., Zazzo A. and Zollikofer C. (2002). A new hominid from the Upper Miocene of Chad, Central Africa. Nature 418, 145-151.
2. Wolpoff M., Senut B., Pickford M. and Hawks J. (2002). (Communications arising): Sahelanthropus or ‘Sahelpithecus’? Nature 419, 581-582.
3. Brunet M. (2002). Reply to Sahelanthropus or ‘Sahelpithecus’? Nature 419, 582.
4. Press release, July 2002, University of Poitiers, 22 pages, widely disseminated by the press, in which the popular name for Sahelanthropus tchadensis, Toumaï, was coined.
5. Vignaud P., Duringer P., Mackaye H.T., Likius A., Blondel C., Boisserie J-R., de Bonis L., Eisenmann V., Etienne M.E., Geraads D., Guy F., Lehmann T., Lihoreau F., Lopez-Martinez N., Mourer-Chauviré C., Otero O., Rage J-C., Schuster M., Viriot L., Zazzo A. and Brunet M. (2002). Geology and palaeontology of the Upper Miocene Toros-Menalla hominid locality, Chad. Nature 418, 152-155.
6. Internet site http://chadcradlehumanity.monsite-orange.fr/

7. Beauvilain A. (2003).Toumaï, l’aventure humaine. La Table Ronde, Paris.
8. Brunet M., Beauvilain A., Coppens Y., Heinz E., Moutaye A.H.E. and Pilbeam D. (1996). Australopithecus bahrelghazali, une nouvelle espèce d’Hominidé ancien de la région de Koro Toro (Tchad). C. R. Acad. Sci. Paris 322, 907-913.
9. Coppens Y. (1994). East Side Story: the origin of humankind. Sci. Am. (May 1994), 8895.
10. Lihoreau F. (2003). Systématique et Paléoécologie des Anthracotheriidae (Artiodactyla: Suiformes) du Mio-Pliocène de l’Ancien Monde: implications Paléobiogéographiques. Thesis, University of Poitiers.

Gibbons Science 2004

Ann Gibbons, Fossil Tooth Make Fur Fly, Science, December 2004.

South African Journal of Science, Vol 100, September/October 2004

Sahelanthropus tchadensis: the facts

Michel Brunet and MPFT†

Sir, In a recent article in this journal, Beauvilain and Le Guellec suggest that our initial description of Sahelanthropus tchadensis (2) was flawed by the inaccurate identification and association of specimens. Their claims are without foundation. Beauvilain and Le Guellec (1) offered supplementary information on the hypodigm and geological context of the oldest known hominid, Sahelanthropus tchadensis, discovered at TM 266, in the Djurab Erg, northern Chad, by the Mission Paléoanthropologique Franco-Tchadienne (2, 3) (MPFT). Their stated intention was to describe ‘the events surrounding the discoveries themselves’ (1) (p. 142), although Beauvilain has already done so at length (4). Their contribution in the South African Journal of Science alleges that : 1) Vignaud et al. (3) failed to present correctly the stratigraphic nature of the hominid site TM266 ; 2) Brunet et al. (2) neglected collected hominid specimens ; 3) the mandibular specimen TM266-02-154-1 described by Brunet et al. (2) is in fact a fossil hominid chimera manufactured by a left m/3 incorrectly glued to a right hemimandible ; 4) together, these supposed errors affect the Minimum Number of Individuals (MNI) count for the site. Beauvilain and Le Guellec (1) do not question the attribution of the fossil to the hominid clade (2) rather than to an African ape, but attempt to show that the MPFT members reached inaccurate conclusions. The team firmly stands by its original statements.

Geological issues
Given the thorough geological survey performed by the MPFT in the Djurab since 1994, including georadar investigations (3, 5, 10), it is clear that the whole Toros-Menalla (TM) fossiliferous area shows no evidence whatsoever of a faulting during at least the last 7 Myr. Except for the modern dunes, the entire landscape is very flat. In the TM area, all of the small scarps pointed out by Beauvilain and Le Guellec (1) are the consequences of aeolian ‘over-digging’ at the foot of the dunes, as are very common across the Sahara and in many other sandy deserts.
These geomorphological features do not in our view represent any ‘reactivated ancient faulting’ as they allege (p. 142). The section given in Vignaud et al. (3) is synthetic, showing the different encountered facies related to the alternation of humid and dry periods. The section at TM 266 is accurate as published, and all fossil specimens reported in references (2) and (3) were derived from this local section as published therein. There is no doubt about their provenance.

Inventory issues
The MPFT practices of inventorying and publishing fossils does not differ from those normally practised in palaeontology. All collected specimens, including hominids, are registered under an inventory number comprising the name of the site, the year, a specimen number, and, in the case of several pieces of the same individual, a part number (e.g. the cranium nicknamed ‘Toumaï’ was numbered TM266-01-060-1). This attribution of an inventory number occurs at different stages when processing the discoveries : 1) for most of them, directly in the field after a precise identification; 2) in the laboratory (at N’djamena or Poitiers) for all the specimens recovered after sieving, or after cleaning for some specimens completely embedded in matrix. Beauvilain, a geographer in charge of MPFT logistics, was unfamiliar with this process and the way in which specimens were assembled into the published hypodigm of S. tchadensis (2).
The first paper describing the new taxon S. tchadensis (2) included only the specimens which were definitively identified as hominid by their anatomical characters. These specimens belong to several individuals, as we reported in Nature (2) : the holotype cranium is one of these specimens. Beauvilain and Le Guellec (1) wonder why a very worn incisor (TM266-01-460) and a damaged partial mandibular symphysis (TM266-02-203) were not included in the paratype series. They were not included because their exact affinities are yet to be fully developed and determined. Further studies of more fragmentary remains have identified additional individuals, and the MNI count will expand as both excavations and preparation work continue.

Issues of restoration and interpretation
In the Djurab desert, the discovery of fossils is facilitated by an intense erosion of the sediments in which they are embedded through the action of the sand blown by winds across the sedimentary units. When the fossils are unearthed, the same aeolian erosion also affects them. The damage mostly consists of polishing, cracking, breaking, dispersion of the different parts, and, finally, total destruction of the specimens if they are not collected almost immediately after their first exposure. For example, the cranium TM266-01-060-1 of Sahelanthropus tchadensis was partially unearthed when found and had suffered from such a weathering by sandblasting. During its exposure, it lost most of its front teeth. As reported in Beauvilain and Le Guellec, the broken right canine belonging to this skull was found separately. There is no doubt that this canine belonged to this cranium, because, as correctly noted by Beauvilain and Le Guellec (1) (p. 143): ‘The tooth consisted of two fragments which fitted perfectly onto the right canine root’. However, they erroneously reported it as a ‘complete’ canine whereas in fact it is the distal half of the canine. This canine was published in its natural position on the cranium (2).
A parallel case occurred for the mandible in question. The right third molar had been displaced from the tooth row by erosion and transported by wind to where it was recovered, some decimetres distant from the mandible with the remaining teeth. After recovery, we established that it belonged to the mandible itself, and we attached it where it belonged, as a right m/3 of the hemi-mandible TM266-02-154-1. Beauvilain and Le Guellec now assert that this tooth has been attributed to and mounted on the incorrect side. This moderately worn tooth bears substantial occlusal anatomy which unambiguously identifies it as a third molar. The identification of the side was based on two decisive independent criteria, one set physical and the other set biological. First, there is an unambiguous match between the lower surface of the tooth and its roots, which remained in the mandible. There is no doubt about the integrity of this join (Fig. 1A, B, C, F, G). This is further confirmed by the matching interproximal facet preserved on the mesial surface of the tooth in question and the second molar retained by the mandible. Second, the anatomy of the third molar allows unambiguous siding. As in all hominoid teeth, the buccal cusps are the more worn, with a larger, most heavily worn cusp (the protoconid), marked here by heaviest occlusion and placed mesiobuccally (Fig. 1D, E). The occlusal rims of the lingual cusps stand out slightly but distinctly, from less wear due to well-known masticatory mechanisms common to modern humans, fossil hominids, and fossil and modern apes.

Plate.CTScan.96pts site

Fig. 1. Right hemi-mandible TM266-02-154-1 of Sahelanthropus tchadensis. A, B, and C: CT scans (courtesy: University Museum, University of Tokyo) at the level of the m/3. The mandibular corpus and the retained roots of the m/3 are in light red. The crown of the third lower molar found separately and claimed to be a left side is in blue. A precise matching between the m/3 crown and the corresponding roots in the hemi-mandible can be observed.
The interstitial space between the m/3 and its roots corresponds to thickness of the glue used to affix the tooth to its roots.
A, Sagittal sections with mesial side at right from right to left, CT scans are respectively shot at 3.33 mm, 4.41 mm, 7.83 mm, and 8.70mm from the buccal edge of the tooth ;
B, transversal sections with lingual side at right from right to left, CTscans are respectively shot at 2.67 mm, 3.69 mm, 4.11 mm, and 9.36mm from the mesial edge of the tooth ;
C, sections parallel to the occlusal surface, at the cervix level and below, with mesial side at top from right to left, CT scans are respectively shot at 6.93 mm, 7.14 mm, 7.44 mm, and 7.80mm from the occlusal edge of the tooth. Mesially, the mesio-buccal and mesio-lingual roots remaining in the corpus (see F) appear in light red and show exact matching with the m/3 crown (in blue) ;
D, occlusal view of the complete specimen with its m/3 ;
E: occlusal view of the m/3 ;
F: occlusal view of the m/3 roots ;
G, disto-lingual view of the join (white arrow) between the m/3 and its distal root.
All scale bars are 0.5 mm.

Conclusion
The logistical contributions of Beauvilain to the fieldwork in Chad are gratefully acknowledged, but the claims and assertions by him and Le Guellec (1) have no bearing on either the interpretation of the geology of TM 266 or of the associations, taxonomy, or phylogeny of Sahelanthropus tchadensis.

We thank the Chadian authorities (Ministère de l’Education Nationale, de l’Enseignement Supérieur et de la Recherche, Université de N’djamena, CNAR).
We extend gratitude for their support to the French ministries, Ministère français de l’Education Nationale (Faculté des Sciences, Université de Poitiers), Ministère de la Recherche (CNRS : SDV & ECLIPSE), Ministère des Affaires Etrangères (DCSUR, Paris and SCAC N’djamena), to the Région Poitou-Charentes, to the RHOI (F.C. Howell and T.D. White) funded by the NSF and also to the Armée Française, MAM and Epervier, for logistic support.We specially thank G. Suwa (University Museum, University of Tokyo) for the CT scans, and C.O. Lovejoy and T.D. White for stimulating discussions. We especially thank also all the other MPFT members who joined us for the field missions, G. Florent, C. Noël and Sabine Riffaut for administrative and technical supports.

Bibliography
1. Beauvilain A. and Le Guellec Y. (2004). Further details concerning fossils attributed to Sahelanthropus tchadensis (Toumaï). S. Afr. J. Sci. 100, 142-144.
2. Brunet M., Guy F., Pilbeam D., Mackaye H.T., Likius A., Ahounta D., Beauvilain A., Blondel C., BocherensH., Boisserie J-R., Bonis L. de, Coppens Y., Dejax J., Denys C., Duringer P., Eisenmann V., Fanoné G., Fronty P., Geraads D., Lehmann T., Lihoreau F., Louchart A., Mahamat A., Merceron G., Mouchelin G., Otero O., Pelaez Campomanes P., Ponce De Leon M., Rage J-C., Sapanet M., Schuster M., Sudre J., Tassy P., Valentin X., Vignaud P., Viriot L., Zazzo A. and Zollikofer C.(2002). A new hominid from the Upper Miocene of Chad, Central Africa. Nature 418, 145-151.
3. Vignaud P., Duringer P., Mackaye H.T., Likius A., Blondel C., Boisserie J-R., Bonis L. de, Eisenmann V., Etienne M-E., Geraads D., Guy F., Lehmann T., Lihoreau F., Lopez-Martinez N., Mourer- Chauviré C., Otero O., Rage J-C., Schuster M., Viriot L., Zazzo A. and Brunet M. (2002). Geology and palaeontology of the Upper Miocene Toros-Menalla hominid locality, Chad. Nature 418, 152-155.
4. Beauvilain A. (2003). Toumaï, l’aventure humaine. La Table Ronde, Paris.
5. Brunet M., Beauvilain A., Coppens Y., Heintz E., Moutaye A.H.E. and Pilbeam D. (1995). The first australopithecine 2500 kilometres west of the Rift Valley (Chad). Nature 378, 273274.
6. Brunet M., Beauvilain A., Geraads D., Guy F., Kasser M., Mackaye H.T., Maclatchy L.M., Mouchelin G., Sudre J. and Vignaud P. (1998). Tchad: découverte d’une faune de Mammifères du pliocène inférieur. C. R. Acad. Sci. Paris 326, 153-158.
7. Brunet M. and M.P.F.T. (2000). Chad: discovery of a vertebrate fauna close to the Mio-Pliocene boundary. J. Vert. Paleont. 20(1), 205-209.
8. Schuster M. (2002). Sédimentologie et paléoécologie des séries à vertébrés du paléolac Tchad depuis le Miocène supérieur. Ph.D. thesis, Université Louis Pasteur, Strasbourg.
9. Ghienne J.M., Schuster M., Bernard A., Duringer Ph. and Brunet M. (2001). The Holocene giant Lake Chad revealed by digital elevation models. Quat. Int. 87, 81-85
10. Schuster M., Duringer Ph., Ghienne J.F., Beauvilain A., Mackaye H.T., Vignaud P. and Brunet M. (2003). Discovery of coastal conglomerates around the Hadjer El Khamis Inselbergs (Western Chad, Central Africa): a new evidence for lake Mega-Chad episodes. The Costal conglomerates of Lake Mega-Chad. Earth Surf. Proc. Landf. 28(10), 1059-1069.

†Jean-Renaud Boisserie a,b, Djimdoumalbaye Ahounta c, Cécile Blondel a, Louis de Bonis a, Yves Coppens d, Christiane Denys e, Philippe Duringer f, Véra Eisenmann e, Pierre Fronty a, Denis Geraads g, Gongdibé Fanoné c, Franck Guy a,k, Thomas Lehmann a, Fabrice Lihoreau a,h, Andossa Likius h, Antoine Louchart i, Hassane Taisso Mackaye h, Olga Otero a, Pablo Pelaez Campomanes j, David Pilbeam k, Marcia Ponce De Leon l, Jean-Claude Rage e, Mathieu Schuster f, Pascal Tassy e, Xavier Valentin a, Patrick Vignaud a, Laurent Viriot a and Christoph Zollikofer l.

a CNRS UMR 6046, Université de Poitiers, 40 Avenue du Recteur Pineau, 86022 Poitiers Cedex, France;
b Laboratory for Human Evolutionary Studies, University of California, 3060 Valley Life Sciences Building, Berkeley, CA 94720,U.S.A.;
c Centre National d’Appui la Recherche, BP 1228, N’Djaména, Tchad;
d Collège de France, place Marcellin Berthelot, 75005 Paris, France;
e Muséum National d’Histoire Naturelle et CNRS UMR 8569, rue Cuvier, 75005 Paris, France;
f CNRSUMR7517, Université Louis Pasteur, 1 rue Blessig, 67084 Strasbourg, France;
g CNRS UPR 2147, 44 rue de l’Amiral Mouchez, 75014 Paris, France;
h Université de N’Djaména, BP 1117, N’Djaména, Tchad;
i CNRS UMR 5125, Université Claude Bernard, 2743 Bd du 11 novembre 1918, 69622 Villeurbanne, France;
j Museo de Ciencias Naturales, C/Guttierez Abascal 2, 28006 Madrid, Spain;
kPeabody Museum, Harvard University, 11 Divinity Avenue, Cambridge, MA 02138, U.S.A.;
lAnthropologisches Institut/Multimedia Laboratorium, Universität Zürich-Irchel, Winterthurer Str. 190, 8057 Zürich, Switzerland.

('LETTRE INTERNATIONALE DE SOUTIEN')
(in italics Press release diffused in November 2004)

July 1, 2004
Dr. Graham Baker
Editor
South African Journal of Science

Dear Graham,
A recent paper published in the South African Journal of Science has put into question the identity of a unique and important Chadian hominid fossil and the competence of the team that described it in Nature.
Attached you will find a comment that we and a set of professional paleoanthropologists wish to publish in your journal to set the record straight about the correct identification of the specimen in question. All of our coauthors were sent the Beauvilain and Le Guellec SAJS article via pdf, along with j-pegs of CT images and detailed photographs of the tooth in question (artificially isolated from the mandible so that its siding would be judged exclusively on the basis of crown morphology). The e-mail that we sent to these colleagues on June 22 is reproduced below.
As you will see, to date, we have 27 coauthors from 13 countries, all offering unequivocal support for the identification of the tooth as from the right side. Many of these coauthors provided anatomical justification for their identifications, and several expressed surprise about how this article passed through the review process into print without being questioned.

We understand that Professor Brunet is providing an independent response to Beauvilain and Le Guellec, and we hope that our contribution might be published alongside this and the apology that Dr. Martin Pickford has sent you by e-mail.
With all best wishes,
F. Clark Howell Tim D. White

Sir, – We, the undersigned, have carefully examined the photographs and digital crown images of a fossilized third molar from the upper Miocene of Chad. This tooth was originally identified by the discoverers (Brunet et al. 2002, Nature 418, 145151) as a right third mandibular molar. Arecent paper by Beauvilain and Le Guellec (2004, S. Afr. J. Sci. 100, 142144) claimed that this tooth had been misidentified and was in fact a left lower third molar. Based on crown anatomy evident in the images examined by us, we confirm the identity of this tooth as a right molar, as originally published by Brunet et al. (2002).

F. Clark Howell, Laboratory for Human Evolutionary Studies, Museum of Vertebrate Zoology, The University of California at Berkeley, Berkeley, CA 94720, U.S.A. E-mail: fchlhes@socrates.berkeley.edu
Tim D. White, Laboratory for Human Evolutionary Studies, Museum of Vertebrate Zoology and Department of Integrative Biology, The University of California at Berkeley, Berkeley, CA 94720, U.S.A. E-mail: timwhite@socrates.berkeley.edu
David R. Begun, Department of Anthropology, University of Toronto, Toronto, ON M5S 3G3, Canada. E-mail: begun@chass.utoronto.ca
Yaowalak Chaimanee, Paleontology Section, Bureau of Geological Survey, Department of Mineral Resources, Rama VI Road, Bangkok 10400, Thailand. E-mail: yaowalak@dmr.go.th
Marie-Antoinette de Lumley, Institut de Paléontologie Humaine, 75013 Paris, France. E-mail: prehist@mnhn.fr
Philip D. Gingerich, Museum of Paleontology, University of Michigan, Ann Arbor, MI 48109-1079, U.S.A. E-mail: gingeric@umich.edu
Colin Groves, School of Archaeology and Anthropology, Australian National University, Canberra, A.C.T. 0200, Australia. E-mail: colin.groves@anu.edu.au
Erksin Güleç, Department of Physical Anthropology and Paleoanthropology, University of Ankara, 06100, Sihhiye, Ankara, Turkey. E-mail: erksin.gulec@humanity.ankara.edu.tr
Yohannes Haile-Selassie, Cleveland Museum of Natural History, 1 Wade Oval Drive, Cleveland, OH 44106, U.S.A. E-mail: yhailese@cmnh.org
Leslea Hlusko, Department of Integrative Biology, The University of California at Berkeley, Berkeley, CA 94720, U.S.A. E-mail: hlusko@socrates.berkeley.edu
Jean-Jacques Jaeger, Equipe Phylogénie, Paléobiologie & Paléontologie, I.S.E.M., CNRS-Université Montpellier II, UMR 5554, Cc 064 Place Eugéne Bataillon, 34095 Montpellier Cedex 5, France. E-mail: jaeger@isem.univ-montp2.fr
Jay Kelley, Department of Oral Biology, College of Dentistry, University of Illinois at Chicago, Chicago, IL 60612, U.S.A. E-mail: jkelley@uic.edu
Meike Köhler, Institute de Paleontologia M. Crusafont, c/ Escola Industrial 23, 08201 Sabadell, Barcelona, Spain. E-mail: kohlerlm@diba.es
Wu Liu, Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing 100044, China. E-mail: liuwu@ivpp.ac.cn
David Lordkipanidze, Georgian State Museum, 0105 Tbilisi, Georgia. E-mail: geonathist@ip.osgf.ge
C. Owen Lovejoy, Matthew Ferrini Institute for Human Evolutionary Research, Department of Anthropology, Kent State University, Kent, OH 44242, U.S.A. E-mail: olovejoy@aol.com
Lawrence B. Martin Department of Anthropology and of Anatomical Sciences, The Graduate School, Stony Brook University, NY 11794-4433, U.S.A. E-mail: lawrence.martin@stonybrook.edu
Monte L. McCrossin, Department of Sociology and Anthropology, MSC 3BV, New Mexico State University, P.O. Box 30001, Las Cruces, NM 88003-8001, U.S.A. E-mail: mmccross@nmsu.edu
Jacopo Moggi-Cecchi, Laboratori di Antropologia, Dipartimento di Biologia Animale e Genetica, Universitá di Firenze, via del Proconsolo 12, 50122 Firenze, Italy. E-mail: jacopo@unifi.it
Salvador Moyà-Solà, Institute de Paleontologia M. Crusafont, c/ Escola Industrial 23, 08201 Sabadell, Barcelona, Spain. E-mail:moyass@diba.es
Lorenzo Rook, Dipartimento di Scienze della Terra, Università di Firenze, via G. La Pira, 4, 50121 Firenze, Italy. E-mail: Lrook@geo.unifi.it
Pat Smith, Department of Anatomy, The Hebrew University Hadassah Medical School, Jerusalem, Israel. E-mail: pat@cc.huji.ac.il
Gen Suwa, The University Museum, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan. E-mail: suwa@um.u-tokyo.ac.jp
Mark Teaford, Center for Functional Anatomy and Evolution, Johns Hopkins University School of Medicine, Baltimore, MD 21205, U.S.A. E-mail: mteaford@jhmi.edu
Phillip V. Tobias, School of Anatomical Sciences, University of the Witwatersrand, Johannesburg, South Africa. E-mail: tobiaspv@anatomy.wits.ac.za
Alan Walker, Departments of Anthropology and Biology, The Pennsylvania State University, University Park, PA 16802, U.S.A. E-mail: axw8@psu.edu
Phil Walker, Department of Anthropology, The University of California at Santa Barbara, Santa Barbara, CA 93106, U.S.A. E-mail: pwalker@anth.ucsb.edu
Steven Ward, Department of Anatomy, NEOUCOM, Rootstown, OH 44272- 0095, U.S.A. E-mail: scw@neoucom.edu

South African Journal of Science, Vol 100, September/October 2004

Beauvilain and Le Guellec reply

Brunet et al. (1) criticize three main pointsin our article (2) : a) the tectonic origin of the Toros-Menalla scarp, b) the determination of side of an isolated molar that was stuck onto a right mandible attributed to Sahelanthropus tchadensis from siteTM266, and c) the inventory of fossils. We responsif to each of these criticisms in turn and conclude that our original hypothesis and interpretation are likely to be correct.

Geological issues in the Djurab
Brunet et al. (1) report that georadar did not yield any evidence of faulting in the Toros-Menalla region, and they ascribe the formation of the scarp to ‘overdigging’ by wind. Whilst wind deflation does indeed cause depressions in deserts, such basins are seldom linear, usually being undulating in outline or spoon-shaped.
The fact that the scarp in question is rectilinear over a distance of more than 40 km suggests an underlying tectonic origin. Elsewhere in the Chad Basin (3) there is evidence of neotectonic activity including the western shore of Lake Chad, which runs straight NNW-SSE for nearly 250 km, a trend that continues along the Dilia Valley (Niger) over an even greater distance. The Bahr el Ghazal Valley is almost straight NNE-SSW for more than 450 km, with a cliff on the left bank in its lower reaches. This neotectonic activity frequently reactivates ancient fractures (4).
Georadar was indeed used in January 1999 but only at sites east of the Bahr el Ghazal (the Toros-Menalla region is on the west side). On fossil and recent dunes it revealed reflectors only 1 to 2 metres thick (5), and in sandstone it ‘did not yield an image at depths greater than 1 metre.’(5)
This georadar equipment is not designed to locate evidence of faulting, and it was not used to search for faults in the Djourab. Thus failure of the equipment to reveal faults in the region does not provide a sound argument against our view that the linear stuctures that occur there are likely to be of tectonic origin.
We thus maintain our original view that the Toros-Menalla scarp, which yielded all the fossils of Sahelanthropus tchadensis, owes its origins to tectonic activity. It has subsequently been deflated by the wind, but this has only modified its form rather than being the original cause of it.

Restoration and interpretation of Sahelanthropus fossils
The fossils from the Djurab are well mineralized but are often abraded by wind-blown sand. There are strong daily changes in temperature, which can range from 50°C air temperature at ground level in the day to light frost at night. At midday, pebbles and fossils lying on the surface of the ground can be too hot to touch. These changes lead to fissuring of fossils, which usually have planar fracture surfaces rather than strongly curved ones. Crusts covering the fossils are either in the form of a very adhesive, hard grey, chemically- resistant matrix, or as an iron-rich concretion, coloured by manganese, often separated from the fossil itself by a small space occupied by sand (see Fig. 1).
The skull of Toumaï, which was found upside down on the sand, was protected from erosion by the second type of crust, whereas the right hemi-mandible TM 266-02-154-1 had traces of grey matrix (Fig. 2), which also covered the base of the isolated m/3. For this reason we consider that the cranium on the one hand, and the mandible and the m/3 on the other were not fossilized in the same deposits and are thus probably not contemporary.
The simplest reconstruction of the history of the m/3 and the right hemi-mandible is that they were broken before burial and fossilization some 6 to 7 million years ago, as shown by the matrix covering the broken surface of the distal root of the m/3 (Fig. 2). Later on, they became encrusted under near-surface conditions. Recently exhumed by the wind, abrasion has removed much of their cover of matrix and has polished their surface, while thermal variations have resulted in damage to some of the teeth. The m/3 was presumably the last specimen to erode out of the sediments as its surface is well preserved. Conversely, the other right mandible (TM 266-02-203), discovered in March 2002 without teeth a dozen metres from the mandible and m/3, was the first to erode out.

planche.oct.04 96pts site

Figs 1-3. Photographs of specimens attributed to Sahelanthropus tchadensis (taken of the cranium at 08:00 on 19 July 2001):
1) the M3/ in the cranium ;
2) oblique lingual view of the roots of the m/3 in the mandible (note in particular the relatively planar fracture surface of the distal root, which curves distally and buccally at a constant level (black arrow), and compare it with the antero-posteriorly curved surface marked by an arrow in the image provided by Brunet et al.1) ;
3) root of the right C1/ in the cranium of Toumaï. (Scale bars: 10 mm.)

Biological considerations of the m/3

One of the arguments put forward by Brunet et al. (1) concerning the determination of side of the isolated m/3 is the wear pattern. However, examination of the last upper molar in the skull of Toumaï (Fig. 1) indicates that we need to be cautious about the interpretation of wear on cusps in general. Curiously, in this particular instance wear on the buccal cusps is closely similar to that on the lingual ones.
Because of this, from the beginning of our study (2) we attached more importance to the grooves, which were less affected by wear during life (and more recently by wind abrasion), and are thus better preserved than the tops of the cusps. Firstly, the similarity between the fossil m/3 and a ‘modern’ human molar is striking. The alignment on a regular curve of the three buccal cusps (protoconid, hypoconid, hypoconulid) and their decrease in size mesio-distally constitute a determining criterion. Secondly, the intercuspid column that we named the metaconulid in our article evokes more a morphological variant of the lingual surface of the tooth than one of the buccal surface. Thirdly, the orientation of the main disto-buccal groove accords precisely in its occlusal articular dynamic relationship with the oblique crest (crista obliqua) [‘pont d’émail’] of the non-working maxillary molarwhenthe opposite side of the jaw is engaged in chewing as in other Anthropomorphes (6).
A second line of argument employed by Brunet et al. (1) to demonstrate that the m/3 fits onto the right mandible was a series of scans in which the broken surface of the base of the tooth and the roots in the mandible were shown to be compatible with each other, being separated only by a thin but continuous layer of glue. It is necessary to recognize the detailed work which allowed the fitting of the m/3 crown so precisely onto the roots of the hemimandible. The hard sandy matrix which covered the base of the isolated molar when it was found first had to be removed and then the space now occupied by glue had to be prepared millimetre by millimetre. A similar operation was required for the parts of the root in the mandible, which were similarly covered in matrix (Fig. 2). In general, a section immediately beneath the cervix of left and right mandibular molars reveals radicular surfaces that may be superposed to within about one millimetre. So, it is not surprising in this particular case that a left tooth seemed to correspond to the roots in a right mandible.

The right upper canine of Toumaï.
We apologize for the possibility that our text regarding Toumaï’s right C1/ could be misinterpreted, the canine root being in situ in the skull as shown in Fig. 3. In contrast we could only write that the canine crown found in November 2001 fits onto the root, we being in the Djourab, and the skull at the University of Poitiers.

Curatorial issues
Writing catalogue numbers on fossil specimens is a daily activity in the field. At such an important site as TM 266, all the fossils were collected. That is why, from July to December 2001, 52 postcranial bones, whose zoological group could not be determined in the field, were catalogued, in the expectation that some of them might belong to Sahelanthropus. Among these fragments, 36 are long bones (tibia, femur, humerus and ulna) including intact specimens and broken diaphyses. Considering the excellent preservation of the Toumaï cranium, a careful examination of these bones should yield interesting information, as we consider it likely that postcrandial fossils of a large primate may be present at the site, although nothing has been reported until now. Note that in the catalogue returned to N’Djamena by the University of Poitiers in December 2001, a single fossil had been added to the field catalogue. It consists of specimen TM 266-01-447 (a right M3 according to Brunet et al. (7), whereas the catalogue entry states that it is ‘Classification primate; Description fragments morceaux racines m/1/m/3 ; Dépôt Poitiers (reliquat tamis)’. These specimens were returned to the CNAR, N’Djamena, on 30 January 2002.

Conclusions
We see no compelling reason to modify radically our hypothesis about the geomorphology of the Toros-Menalla region, nor our interpretation of the isolated left m/3 and damaged right mandible from locality TM 266 attributed to Sahelanthropus tchadensis. Finally, given the excellent preservation of fossils at the site, we consider it likely that the collection of postcranial elements collected there may well contain some specimens that belong to Sahelanthropus.

1. Brunet M. et al. (2004). Sahelanthropus tchadensis: the facts. S. Afr. J. Sci. 100, 443-445.
2. Beauvilain A. and Le Guellec Y. (2004). Further details concerning fossils attributed to Sahelanthropus tchadensis (Toumaï). S. Afr. J. Sci. 100, 142-144.
3. Morin S. (2000). Géomorphologie. In Atlas de la province Extrême-Nord Cameroun, pp. 717. IRDLCA/France MINREST-INC/Cameroun.
4. Neev D., Hall J.K. and Saul J.M. (1982). The Pelusium Megashear System across Africa and associated lineament swarms. J. Geophys. Res. 87, B2, 1015-1030.
5. Schuster M. (2002). Sédimentologie et paléoécologie des séries à vertébrés du paléolac Tchad depuis le Miocène supérieur. Thesis, University of Strasbourg.
6. Lautrou A., University of ParisV René Descartes, pers. comm. (author of Anatomie Dentaire, Masson, Paris).
7. Brunet M. et al. (2002). A new hominid from the Upper Miocene of Chad, Central Africa. Nature 418, 145-151.

South African Journal of Science, Vol 101, March/April 2005

Sir – Readers following the debate about Sahelanthropus tchadensis in this journal need to know that the principal argument in the response by M. Brunet et al. to an article of ours is based on a series of CT scans and digital photos which reveal contradictions between the figure and its legend (1). In our reply to this response (2), we did not mention these contradictions, because we considered that it would be unlikely that they would be published as submitted.
However, it is clear that :
1) There is an error in the scales, because hominid teeth a few millimetres long, as suggested by the scale, are unknown. The error is 1 to 10.
2) There is an error in the orientation of the CT scans A, for which the legend reads ‘sagittal sections with mesial side at right’.
The root of m/3, which descends towards the right in the scans, indicates that this is not the mesial surface but the distal one. In lower third molars, the mesial root is vertical and the distal one is inclined towards the rear of the mandible, where there is no other tooth to hinder its development. This point is clearly illustrated by photograph G in the same figure. In addition, in the two central images in line A, the planar interstitial facet, which is on the mesial surface of the third molar, occurs on the left in the image, whereas the rounded surface characteristic of the distal part of the tooth is to the right.
Examination of the sagittal section ‘shot at 3.33 mm from the buccal edge of the tooth’ (the CT scan at the right-hand end) and photographs E and G, and in particular the fractures and wear pattern that the tooth possesses, show that the sagittal sections were taken respectively from the lingual margin of the tooth.
3) There is an error in orientation of the CT scans in line B, which the legend indicates are ‘transversal sections with lingual side at right–from right to left, CT scans are respectively shot at 2.67 mm, 3.69 mm, 4.11 mm, and 9.36 mm from the mesial edge of the tooth’. If the lingual surface is to the right as indicated in the legend, then the photograph is of a left tooth. If this is so, then the text becomes incomprehensible in relation to the CT scans. Apart from that, in hominids, mastication can give rise to aberrant wear patterns. The perceived inclination of the occlusal surface can also depend on the preparation of the remnants of the roots still present just beneath the cervix of the crown and those solidly anchored in the mandible, prior to gluing the pieces together.
White (3) has clearly shown the problems that can be encountered during the process of sticking fragments of fossile together. In general, it is difficult to find perfect contacts between a crown and its roots if these have been separated for millions of years, encrusted separately in a hard matrix and then recently subjected to intense aeolian abrasion and strong temperature changes, which often break the teeth into fragments or polish the exposed surfaces, as has happened in the case of the mandible, or have been cleaned in the laboratory, as happened to the isolated third molar.
Comparison with photograph E reveals that, from right to left, the sections were taken respectively from the distal margin of the tooth.
4) There is an error in orientation of the CT scans C, the legend of which states: ‘Sections parallel to the occlusal surface, at the cervix level and below, with mesial side at top – from right to left, CT scans are respectively shot at 6.93 mm, 7.14 mm, 7.44 mm, and 7.80 mm from the occlusal edge of the tooth. Mesially, the mesiobuccal and mesio-lingual roots remaining in the corpus (see F) appear in light red and show exact matching with the M2 crown (in blue).’
According to the legend, the image on the right was taken closer to the occlusal surface of the tooth and the one on the left of the series closer to the mandibular body. However, the roots and the mandibular body appear in the right-hand image, which is supposed to have been taken closest to the occlusal surface, and practically disappear in the left-hand image, which is labelled as having been taken closest to the mandibular body.
It is a pity that none of the published scans or photographs shows the lateral surface of the m/3 and the small cusplet that we called the metaconulid (1 in the accompanying figure), a peculiar element of this tooth. Incidentally, note that in occlusal view this accessory cusplet appears to have a mark on it (a transverse white streak in the left image) that did not exist at the time of its discovery (right-hand image).The supposed presence of a contact facet between the m/2 and the m/3 cannot be used to support any argument that the two teeth belong to a single individual, for the simple reason that more than two-thirds of the distal surface of m/2, that would normally possess the facet, is missing, having broken off during weathering. The remaining third of the distal surface has been subjected to aeolian abrasion (figure, left image). There thus does not exist any physical evidence of the contact between the two teeth (2 in the photograph). It is therefore not possible to affirm, as is claimed in the MPFT response, that ‘the join ... is further confirmed by the matching interproximal facet preserved on the mesial surface of the tooth in question and the second molar retained by the mandible’ (1).
The preservation of the two fossils is very different, m/1 and m/2 in the mandible having suffered extensive damage by atmospheric agencies and the m/3 being practically intact. Furthermore, if the m/3 belongs to the mandible, then the degreeof wear ought to diminish from the m/1 to the m/3. However, in this case, the m/3 appears to be the most worn of the teeth, showing more extensive areas of dentine than the m/2.

plate.march05.96pts site

The two molars, m/2 and m/3, after preparation (left) and m/3 at the time of its discovery. Key: 1, the small cusplet we call the metaconulid; 2, there does not appear to be any physical evidence of contact between the two teeth.

TM 266 M3 1

TM 266 M3 2

Additional photographs to get his opinion

In the circumstances, it is astonishing that none of the 28 co-signatories grouped under the MPFT observed the numerous problems with their response. Equally, we fail to understand how 28 other authors, including many eminent palaeoanthropologists, after having ‘carefully examined photographs and digital crown images of a fossilized third molar from the upper Miocene of Chad’, could co-sign an international letter of support for the MPFT that is bereft of scientific content (4).
Apart from the fact that the history of science provides no precedent for such a letter, one of the co-signatories recently proposed a list of recommendations of professional ethics (5). As concerns this particular case, we mention only the following instructions from among a rich sample in what is essentially a methodological article: ‘Do not publish without first examining the original’, ‘Do not judge the quality of a paper by its authors’, ‘Read every scientific paper critically’, ‘Do not reject papers or grant applications for personal or political reasons’, as well as the comment that ‘...micro- and macrocomputerized tomography, scanning electron microscopy, laser scanning, ... 3D digitization, and parsimony analysis are only tools. They have no magical qualities.’
We fully endorse this excellent advice and recommend that it be taken seriously so as to reduce the possibility of scholarly misunderstanding in our field.

1. Brunet M. et al. (2004). Sahelanthropus tchadensis: the facts. S. Afr. J. Sci. 100, 443-445.
2. Beauvilain A. and Le Guellec Y. (2004). S. Afr. J. Sci. 100, 445-446.
3. White T.D. (2000). Jaws and teeth of Australopithecus afarensis from Maka, Middle Awash, Ethiopia. Am. J. Phys. Anthropol. 111, 45-68.
4. Howell F.C. et al. (2004). S. Afr. J. Sci. 100, 446.
5. White T.D. (2000). A view on the science: physical anthropology at the millennium. Am. J. Phys. Anthropol. 113, 287-292.

**Note from M. Brunet et al.**

We thank Dr Beauvilain for indicating the errors in the legend to our figure, but would like to point out that the typographical error in the scale and the accidental inversions of laterality information in our legend do not affect the anatomical data and interpretation we provided on the material presented. The correct legend should read as follows :
Fig. 1. Right hemi-mandible TM266-02-154-1 of Sahelanthropus tchadensis. A, B, and C: CT scans (courtesy: University Museum, University of Tokyo) at the level of the m/3. The mandibular corpus and the retained roots of the m/3 are in light red.The crown of the third lower molar found separately and claimed to be a left one is in blue. A precise matching between the m/3 crown and the corresponding roots in the hemi-mandible can be observed.The interstitial space between the m/3 and its roots corresponds to thickness of the glue used to affix the tooth to its roots.
A, Sagittal sections with mesial side at left from left to right, CT scans are respectively shot at 3.33 mm, 4.41 mm, 7.83 mm, and 8.70 mm from the buccal edge of the tooth ;
B, transversal sections with lingual side at left from left to right, CT scans are respectively shot at 2.67 mm, 3.69 mm, 4.11 mm, and 9.36 mm from the mesial edge of the tooth;
C, sections parallel to the occlusal surface, at the cervix level and below, with mesial side at top from left to right, CT scans are respectively shot at 6.93 mm, 7.14 mm, 7.44 mm, and 7.80 mm from the occlusal edge of the tooth. Mesially, the mesio-buccal and mesio-lingual roots remaining in the corpus (see F) appear in light red and show exact matching with the m/3 crown (in blue) ;
D, occlusal viewof the complete specimen with its m/3 ;
E, occlusal view of the m/3 ;
F, occlusal view of the m/3 roots ;
G, disto-lingual view of the join (white arrow) between the m/3 and its distal root.
All scale bars are 0.5 cm.

Chad, Cradle of Humanity ?

With Australopithecus bahrelghazali (Abel) and Sahelanthropus tchadensis (Toumaï), two hominids fossils, Chad is a potential Cradle of Humanity

Toumaï Shows the Teeth

A short article may have unexpected consequences :

South African Journal of Science

South African Journal of Science, Vol 100, March/April 2004

Further details concerning fossils Sahelanthropus tchadensis (Toumaï)

Alain Beauvilain and Yves Leguellec

South African Journal of Science, Vol 100, September/October 2004

Sahelanthropus tchadensis: the facts

Michel Brunet and MPFT†

South African Journal of Science, Vol 100, September/October 2004

South African Journal of Science, Vol 101, March/April 2005

Note from M. Brunet et al.

***South African Journal of Science*, Vol 100, March/April 2004**

**Note from M. Brunet et al.**