Toumaï shows the Orbital
and Foramen magnum Planes

FIGURE 1. Method of determining the orientation of the foramen magnum plane (FM). A straight edge is laid in contact with basion and opisthion, this line defining the FM plane which is at right angles to the sagittal plane. Specimen KC3 female Gorilla.

FIGURE 2. Examples of the FM plane in extant chimpanzees showing part of the range of variation. Note that in most specimens the FM plane intersects the toothrow. A ­ 1898.178 female; B ­ 1974.72, female; C ­ 1910.126 male; D ­ A 13924 male.

FIGURE 3. Examples of the FM plane in extant gorillas showing part of the range of variation. Males tend to have steeper FM planes than females, but note that in most individuals the FM plane intersects the tooth row. A ­ 1978.6 female; B ­ 1910.373 female; C ­ 1924.385 male; D ­ 1971.203 male.

Gorilla orbital plane
The orbital plane intercepts the M1/ in the majority of gorilla individuals (23) or is anterior to it (at the front of M1/ in 7 females). It intersects M2/ in 6 individuals. Thus the orbital plane of the gorilla has much the same relation to the tooth row as it does in chimpanzees.

Gorilla FM-OP angle
The most acute gorilla FM-OP angle measured was 42° (2 males) and the least acute was 75° (a female), the mean being 63.5°. Males (42°­72°, mean 57°) tend to have more acute FM-OP angles than females (59°­ - 75°, mean 67°) but there is overlap between the sexes. Aged males have more acute FM-OP angles than young males, and it is evident that this angle changes ontogenetically.

Human foramen magnum
In both of the humans measured (Table 1) the FM line was well above the dentition, intersecting the nasal cavity. This is considerably more elevated anteriorly than in any of the chimpanzees and gorillas. The relationship with the occlusal plane is thus diametrically opposite to that of chimpanzees and gorillas (Figure 4), the intersection of the FM plane occurring posterior to the foramen magnum rather than anterior to it as in chimpanzees and gorillas.

Human orbital plane
The orbital plane in both humans examined intersected the premolars.
Human FM-OP angle the two humans in the sample have FM-OP angles of 100° and 105°.

COMPARISONS

Foramen magnum orientation
In chimpanzees and gorillas, the plane of the foramen magnum, as defined by basion and opisthion, generally intersects the occlusal plane somewhere along the tooth row, the few specimens that intersect this plane behind the M3/ all being males (1 chimpanzee out of 20 examined, and 4 out of 31 gorillas included in the study). The specimens in which the FM plane falls below the toothrow are all fully adult to aged males, young males being more like females in the values obtained. From this it is inferred that the relationship between the FM plane and the maxilla changes ontogenetically, the inclination of the FM increasing with age relative to the occlusal plane, especially in gorillas (Delattre, Fenart 1960). It is also evident that there is a great deal of individual variation in the orientation of this plane, reflecting individual differences in the angle at which the splanchnocranium is hafted onto the neurocranium. In humans the FM plane in the two specimens examined lies above the toothrow (including the roots), intersecting the nasal cavity anteriorly and the occlusal plane well behind the foramen magnum (Figure 4). In the individual illustrated by Zollikofer et al. (2005) the FM intersects the apices of the molar roots, and thus appears to be an extreme variant among humans, a conclusion confirmed by the values published by the authors.
The FM plane in Sahelanthropus as reconstructed by Zollikofer et al. (2005), lies above the roots of the toothrow including the apex of the canine, which is extremely elongated within a hominid context, and it intersects the nasal cavity anteriorly and the occlusal plane well behind the foramen magnum as in humans.
The FM plane of the australopithecine illustrated by Zollikofer et al. (2005) is inclined at such an angle that it intersects the tooth row low down, much lower than in the Sahelanthropus reconstruction and close to the condition in Gorilla and Pan.
The published FM plane of Sahelanthropus relative to the occlusal plane is more elevated than it is in australopithecines, chimpanzees and gorillas, and falls well within the range of variation of humans.

The orbital plane
The orbital plane of chimpanzees and gorillas generally intersects the M1/, with a few individuals intersecting the M2/ behind or the contact between M1/ and P4/ anteriorly. In none of the chimpanzees examined did the OP intersect the premolars, although in some female gorillas the plane intersected the contact between the P4/ and M1/.
In humans, the OP intersects the premolars, usually the P3/ (Zollikofer et al. 2005) (Figure 4).

FIGURE 4. Examples of the FM plane in humans. Note that the plane does not intersect the toothrow. A ­ 1932.152 female (reversed); B ­ PVA 1, unrecorded sex.

In the reconstruction of Sahelanthropus, the OP intersects the M2/, which is posterior to the mean position in chimpanzees and gorillas, although it falls within the range of variation of these species. Thus the intersection of the OP and the toothrow of Sahelanthropus is markedly different from the situation in humans.
In Australopithecus, the OP published by Zollikofer et al. (2005) intersects the toothrow far back, probably about the level of M2/ (the teeth are missing in Sts 5).
Thus, if the OP of Sahelanthropus is correctly positioned in the reconstruction, then its orientation is more like that of chimpanzees and gorillas than that of humans.

FM-OP angle
The FM-OP angle reflects the position and orientation of two planes. In the reconstructed cranium of Sahelanthropus, the FM plane falls well within the range of variation of humans, whereas the OP falls into the lower part of the range of variation of gorillas and chimpanzees, well outside the range of variation of humans. As a consequence, the angle between these planes, as published by Zollikofer et al. (2005), falls close to the human range of variation, and well above that of gorillas, chimps and australopithecines.
In Australopithecus, both the FM and OP published by Zollikofer et al. (2005) fall within the range of variation of gorillas and chimpanzees, although the angle between them is slightly outside the range of variation of the African apes.
The chimpanzee FM-OP angle (57°) illustrated by Zollikofer et al. (2005) and Brunet and Allemand (2005) is more acute than any of the specimens examined in this study, and is smaller than the most acute angle (57.5°) listed by Zollikofer et al. (2005) in their text. It is 8° more acute than the value published by Brunet and Allemand (2005).
The chimpanzee FM-OP measurements in this study range from 60° 78° (mean 69°) which is uniformly less acute than the range of variation published by Zollikofer et al. (2005 ­ 63.7 +/­- 6.2°). These differences are possibly related to ways of measuring the OP, taking into account the estimated error margin of 4­6°. The chimpanzee individual illustrated by these authors falls beyond  their own estimate of the range of variation of the FM-OP angle in chimpanzees, a result confirmed by this study. In particular, the orientation of the FM plane in the illustration appears to be too steeply inclined anteriorly, the line representing the plane passing well below the tooth row. Out of the 20 chimpanzees examined during the present study, only one individual showed an FM line below the tooth row, all the others intersecting the occlusal plane, some as high as the incisors. It is thus concluded that the chimpanzee FM-OP angle illustrated by Zollikofer et al. (2005) exaggerates the acuteness of the FM-OP angle and thus magnifies the apparent difference between chimpanzees on the one hand and the other hominoids illustrated by them on the other. Examination of the image suggests the possibility that basion has been misidentified.
The FM-OP angle of Australopithecus africanus (Sts 5) measured directly from the illustration published by Zollikofer et al. (2005) is 81°. This is only 3° more obtuse than the highest value for the chimpanzee examined in this study and 11.1° greater than the most obtuse FM-OP measurement that Zollikofer et al. (2005) provided in the text (69.9°). In Australopithecus the FM plane intersects the toothrow in much the same way as in chimpanzees and gorillas, and it is likely that the range of variation of the FM-OP in australopithecines is about the same order of magnitude as it is in extant African apes. It is not known precisely what the range of variation in australopithecines is, but the chances are that Sts 5 has neither the most acute nor the most obtuse FM-OP angle among the subfamily.
It is thus possible that there is overlap between the FM- OP angle of chimpanzees and australopithecines. It is also possible that there is overlap between the FM-OP angle of female gorillas and australopithecines, the most obtuse gorilla FM-OP angle determined in this study (75°) being only 6° less than the estimated angle in A. africanus.

DISCUSSION AND CONCLUSIONS
It is generally agreed that australopithecines were bipedal, whereas chimpanzees and gorillas are quadrupedal, albeit with the ability to stand and move on two feet for short periods and short distances. Australopithecines show skeletal morphology that reflects their bipedal stance and locomotion, whereas chimpanzees and gorillas do not. The fact that the FM-OP angles of chimpanzees and gorillas are close to those of australopithecines, with the possibility of overlap between the ranges of variation, indicates that this angle does not provide a clear idea of locomotor or postural potential. at least for these groups.
Humans in contrast possess a considerably more obtuse FM-OP than either australopithecines or African apes (103.2 +/­- 6.9° according to Zollikofer et al. 2005). There is no overlap between the ranges of variation of chimpanzees and gorillas on the one hand (maximum 78°) and humans (minimum 96.3°) on the other. The FM-OP value for Sahelanthropus is estimated to be 96° by Zollikofer et al. (2005), which, if correctly determined, would imply that it is just below the range of variation of humans.
Given the observation that the FM plane of the reconstructed Sahelanthropus falls within the range of variation of humans, whereas its OP falls well outside the human range, it is pertinent to ask whether the value of the FM-OP angle illustrated by Zollikofer et al. (2005) is reliable. The proximity of values between the FM-OP of humans and Sahelanthropus has been used to infer the strong possibility of bipedal posture in the genus (Brunet and Allemand 2005). If the angle is correctly determined, and if the FM-OP angle reflects posture and locomotor repertoire as claimed, then it is difficult to evade the conclusion that Sahelanthropus was likely to have been better adapted for bipedal posture than australopithecines were, as the FM- OP of Australopithecus is significantly more acute (81°) than that of Sahelanthropus (96°), and is only marginally more obtuse than that of chimpanzees (most obtuse angle is 78°). However, since the OP of Sahelanthropus is quite unlike that of humans, it is legitimate to ask whether its FM plane has been accurately reconstructed.
The original specimen of Sahelanthropus tchadensis is crushed and warped (Brunet et al. 2002), making any direct measurement of the foramen magnum and orbital planes, as well as the relationship between them, impossible. For this reason, Zollikofer et al. (2005) based their measurements on a virtual reconstruction of the skull. The basicranium was particularly badly damaged, with the foramen magnum fractured into two, one part with basion, the other with opisthion (Brunet et al. 2002). Furthermore, the relationships between the basicranium on the one hand and the rest of the neurocranium and splanchnocranium on the other were heavily distorted, with displacement of fragments in three dimensions, as well as rotation and translation of the various pieces relative to each other. There could even be some plastic deformation, although this is not immediately apparent from the published images. Given that the basicranium is far removed from the splanchnocranium (taking into account size differences, the distance between the foramen magnum and the rear of the tooth rows is as great in Sahelanthropus as it is in gorillas), then the possibility of obtaining an erroneous relationship between the foramen magnum and the orbits is rather high.
With crania oriented in the Frankfurt horizonal, the dorsal profile of the skulls of chimpanzees, gorillas, australopithecines and humans show a rise in the profile of the braincase behind the orbits. In the virtual reconstruction of Sahelanthropus this part of the cranium is almost flat, suggesting that the neurocranium may be hafted onto the splanchnocranium at too horizontal an angle. If this is so, then all the posterior elements of the  neurocranium may have been rotated ventrally with respect to their life position (i.e. in right lateral view, counterclockwise rotation of the neurocranium relative to the splanchnocranium). There are several breaks in the neurocranium about which rotation could have taken place. Removal of this rotation would make the nuchal area steeper and would make the plane of the foramen magnum more acute relative to the orbital plane. A change in angle of the FM plane of only 5° would result in a value of the FM-OP angle of ca 89°, a change of 10° would result in an angle of ca 84°, not very different from the maximal figure for chimpanzees and gorillas. Given the uncertainties of working with crushed and  distorted fossils, and considering the ape-like orientation and position of the orbital plane in Sahelanthropus, the presence of a fully human-like FM plane in it is surprising.
Because, if Sahelanthropus is a hominid which gave rise to humans via australopithecines, as envisaged by Brunet et al. (2002), Zollikofer et al. (2005) and Brunet and Allemand (2005), then the FM plane would change from human-like in Sahelanthropus tchadensis to ape-like in Australopithecus and back to human-like in the genus Homo, at the same time that its OP remained ape-like until the human stage was reached.
In view of these contradictions it is considered that the FM plane of Sahelanthropus has been incorrectly estimated by Zollikofer et al. (2005). In the author's view, the FM of Sahelanthropus would probably have intersected the toothrow as in chimpanzees, gorillas and australopithecines, and if so, then its FM-OP angle would have been appreciably more acute than the reconstructed figure published by Zollikofer et al. (2005). As a consequence it is considered premature to deduce bipedal posture and locomotion for Sahelanthropus on the basis of its reconstructed FM-OP angle. This conclusion reinforces that of Wolpoff et al. (2002) that Sahelanthropus tchadensis was not a hominid in the strict sense of the term.

ACKNOWLEDGEMENTS
I thank Y. Coppens, B. Senut, A. Beauvilain, E. Gitonga, C. de Muizon and S. Sen for support. Thanks to C. Lefevre for access to skulls of African apes in her care. 

REFERENCES
- Brunet M., Allemandf L., 2005 : Un nouveau grade d'évolution humaine. La Recherche 387: 36­43.
- Brunet M., Guy F., Pilbeam D., Taïsso Mackaye H., 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 Léon M., Rage J-C., Sapanet M., Schuster M., Sudre J., Tassy P., Valentin X., Vignaud P., Viriot L., Zazzo A., Zollikofer C., 2002 : A new hominid from the Upper Miocene of Chad, Central Africa. Nature 418 : 145­-151.
- Delattre A., Fenart R., 1960 : L'Hominisation du crâne. Editions CNRS, Paris VII. 418 pp.
- Wolpoff M., Senut B., Pickford M., Hawks J., 2002 : Palaeoanthropology (communication arising): Sahelanthropus or Sahelpithecus? Nature 419 : 581­-582.
- Zollikofer C., Ponce de Léon M., Lieberman D., Guy F., Pilbeam D., Likius A., Mackaye H., Vignaud P., Brunet M., 2005 : Virtual  reconstruction of Sahelanthropus tchadensis. Nature 434: 755­-759.