The recently described [30] A. africanus specimen from Sterkfontein, South Africa, Stw 505, with a currently calculated cranial capacity of 515 ml ([4] provides the required evidence (see Fig.1 and Fig. 3 and Nota Bene below).

Here, we wish to concentrate on the question of reorganization. A silastic rubber endocast was made from the original cranial portion by Dr. R. Clarke, and sent to RLH. Fig.1, Fig. 3 and Fig. 2 show that the occipital lobe possesses a strongly curving crescentic sulcus just superior to where the transverse sinus is located. The crescent is strongly concave posteriorly and medially, and is situated some 20–25 mm posterior to the sigmoid sinus. In chimpanzee brains (see Fig.1, Fig. 3 and Fig. 2), the LS is normally flush or only slightly posterior to the sigmoid sinus, and the lateral arc distance from the occipital pole (OP) to the LS is on average 35–40 mm [26]. It is important to note that the OP is always lateral to the midsagittal plane, so that the distance from the LS to the midsagittal plane is clearly greater than the 35–40 mm average in chimpanzees. On the Stw 505 specimen, the distance to where the occipital pole would be if the occipital bone were complete is less than 20 mm. No other sulcus normally found in the occipital lobe, including the lateral calcarine, inferior and lateral occipital sulci, matches the position or strong posterior crescentic concavity of this sulcus on the Stw 505 brain endocast, either in chimpanzee or Homo . The dorsal extension of the LS is only visibly present for a short distance, and yields an approximate dorsal arc significantly smaller than on chimpanzee brains of lesser volume. This value is between 30–35 mm on most chimpanzee brains, with volumes usually around 350 ml, which is very significantly less than the 515-ml volume of the Stw 505 endocast. The best estimate we have of this possible distance on the Stw 505 endocast is a maximum of 25 to 30 mm (see scales in Fig.1, Fig. 3 and Fig. 2), and that distance is from the concavity to the midsagittal plane, not the OP, which would be significantly less, that is, approximately 20 mm. It should be noted that the midsagittal plane reconstructed in Fig.1, Fig. 3 and Fig. 2 is clearly medial to where the expected occipital pole on the endocast would be, a distance which in chimpanzees is between 5–10 mm. This means that the distances suggested for the lunate sulcus are maximal since the occipital pole is not available on Stw 505. However, on the Taung endocast the OP to midsagittal plane distance is about 10 mm, and the same would be anticipated for Stw 505. It is important to note that the midline (midsagittal) morphology is present and undistorted on both dorsal and ventral portions of the frontal lobe, extending from the rostral bec for a distance of roughly 65 mm on the dorsal surface. This morphology permits a reasonably accurate midline to be extended posteriorly.

Based on a sample of 78 chimpanzee brain hemispheres (this is a recently increased sample from that reported in [26]), the correlation between the logs (base 10) of brain volume and distance from the OP to the LS is 0.678, with a slope of 0.375, indicating a normal within-species allometric relationship. This r is statistically significant at the 0.001 level. With a cranial capacity of Stw 505 thought to be about 575 ml from our first reconstruction (in prep.), the predicted OP–LS distance is calculated to be 41 mm, if there is an allometric relationship between these two variables in Pan. This is nearly double the actual distance seen on the endocranial cast of Stw 505, and visible in Fig.1, Fig. 3 and Fig. 2. The point we wish to make here is that the large size of the Stw 505 endocast would demand a larger distance of the LS to the OP if the endocast followed a Pan pattern. Clearly, the opposite is the case.

Nota Bene. A new volume estimate for the endocranial capacity of the Stw505 A. africanus hominid is being prepared. The volume estimate of 515 ml, based on CT scanning, is too small, at least in the opinion of the first author. Indeed, Hawks and Wolpoff [13] calculated a larger volume of about 575 ml. The Stw505 endocast shows some plastic deformation, as well as several cranial breaks that must be corrected. However, the base and anterior dorsal region of the frontal lobe are undistorted and provide midline landmarks that can be extended posteriorly to the occipital lobe to provide an accurate midsagittal plane to within ±5 mm. The actual distance that should be compared with chimpanzee measurements are from the occipital pole and not the midsagittal plane. Thus the distance from most probable OP on Stw 505 to the crescentic furrow would be less than the 20–25 mm suggested in Fig.1, Fig. 3 and Fig. 2, based on the midsagittal plane. While this part of the occipital lobe bearing the OP is missing, it would certainly have been lateral to the present midsagittal plane, thereby increasing the contrast with the measurement in chimpanzees. On chimpanzee brains, the OP is usually 5–10 mm lateral to the midsagittal plane. The first author’s preliminary reconstruction with these corrections yields a volume of roughly 575 ml. The first author believes this to be accurate to within 20 ml. If the original volume of 515 ml is used, the predicted OP to lunate sulcus distance is around 39–40 mm, still considerably larger than found on the Stw505 endocast.

Clearly, this Stw 505 specimen of A. africanus has an indisputable crescentic, concave-posterior, lunate sulcus as seen in Fig. 3B and it is in a posterior, non-Pan position, indicating that at least one member of the species had undergone some cortical reorganization toward a more human-like condition. If there is some other sulcal element that has caused this strong crescentic sulcus we have not been able to identify it. No other known sulcal morphology appears in this position. We think it is now perhaps more parcimonious to regard the older specimens, such as Taung, and the A. (Paranthropus) robustus SK 1585 as having LS's in posterior positions rather than in the typical ape anterior position, because a chimpanzee-like anterior position violates the sulcal morphology available, a conclusion that applies to the Hadar AL 162–28 A. afarensis specimen as well as Taung and SK 1585. Indeed, the 15.5 mm distance between OP and LS on the Hadar specimen is some 4 SDs posterior to where the LS would lie on an average chimpanzee brain [26]. If one looks down the axis of the lambdoid suture on the Taung endocast, there does appear to be a groove that might be ascribed to the lunate sulcus, although we prefer to believe that the posterior inferior lip of the parietal bone is responsible for the groove, the puckering of the occipital lobe notwithstanding. The SK 1585 endocast shows a more pronounced puckering in this region and a complete smoothness of the endocast anterior to where the lambdoid suture is present. The Stw 505 specimen indicates that at least this early hominid brain was indeed reorganized toward a more human-like pattern despite its small brain volume, and prior to any significant cortical enlargement as seen in the genus Homo. In sum, this endocast strongly suggests that cortical reorganization preceded brain enlargement in hominid evolution as argued elsewhere [14], [15], [16] and [17]. We repeat, that in brains with strongly developed lunate sulci, the sulci are the anterior limit to PVC. We have not found evidence that this finding is violated in either humans or apes (please refer the literature discussed in [26]).

It is tempting to speculate about what selection pressures may have played a role in the evolution of this derived neural condition. Relative to the earliest hominids [37], the associated faunal and geological evidence from the later Australopithecus-bearing sites suggest more open habitats and an expansion of niches with food resources separated by increased distances. Stone tools, while showing minimal standardization but primitive patterns exist as early as 2.6 Myr [5] and [12], a date not at odds with that of Member 4 of Sterkfontein [36]. Elsewhere [14], [15], [16], [17] and [23], RLH has speculated that reorganization of the Australopithecus brain indicates an expanded posterior parietal cerebral cortex, and was most likely associated with enhanced social behavior including communication: [1] this region involves multimodal processing involving visual, auditory, and sensorimotor integration; [2] visuospatial integration related to tool use and making, throwing objects with force and accuracy, as well as more sophisticated longer-term memory of spatial locations and qualities of self, others (i.e., facial recognition), prey and predators, including objects and resources (stone, digging sticks, trees, waterholes, etc.). All of these were possibly adding to a more advanced cognitive adaptation within a changing and expanding ecological zone, and formed the basis for additional cognitive changes, such as patterned tool-making, and possibly more advanced communicative skills that accompanied the evolution of the genus Homo. We are cautioned, of course, to realize that one endocast with this unambiguous sulcal location might not be representative of the entire species, but thus far, we have seen no alternative strong evidence that there were A. africanus (or A. afarensis) individuals with lunate sulci in ape positions.