R. Grün, P. Beaumont1, P.V. Tobias1 and S. Eggins
1 McGregor Museum, South Africa
2 Sterkfontein Research Unit, School of Anatomical Sciences, University of the Witwatersrand Medical School, South Africa

The archaeological site of Border Cave, Kwazulu Natal, South Africa, has yielded a number of modern hominid remains, labelled BC1 to BC8. BC5 is a fairly complete lower jaw (Figure 1) which was recovered by C. Powell in 1974 from the northwest edge of square T20, while she and one of us (PBB) were collecting sediment samples, at the request of K.W. Butzer, from the south face of Excavation Area 3A. It came from the 3 WA (the site shows an succession of white and brown sediments which have been termed "white ashes" (WA) and "brown soils" (BS), numbered from top to bottom), about 0.25 m below its intact surface, and immediately adjacent to a previously mapped and photographed depression, the base of which cut by up to 0.15 m into the upper part of the underlying 4BS.

Border Cave has been the subject of several dating studies, by radiocarbon on charcoal, ESR on faunal teeth, amino acid racemisation of ostrich eggshells and thermoluminescence on burnt flint. Comparison of the dating results found slightly younger mean ESR dates (in the range of 5 to 10%) than those of the other dating techniques. This has been attributed to a small fading component (see Grün and Ward 2002). There are unfortunately no definite amino acid and TL results available on layers 3WA and 4BS. If BC5 was contemporaneous with the faunal teeth found in 3WA its age would be about 66±2 ka (average ESR age of layer 3WA) and younger than 82±2 ka (average ESR age of Layer 4BS).

Sillen and Morris (1996) have published and provided one of us (PBB) with splitting factor (SF) and nitrogen (N) assays on BC1 to BC7 and on faunal cortical fragments from the Excavations 3A and 4A sequence that were submitted to Sillen between 1990 and 1993. The interpretation by Sillen and Morris (1996) of the Border Cave data (Figure 3) was based on Elands Bay Cave (EBC) results where SF increases with depth to a limit at about 20 ka BP. However, no comparable trend is evident at Border Cave, except for increases from about 2.9 to 3.7 in the 1BS.UP (Iron Age) spits and from about 3.4 to 5.3 in the basal 6BS (arrows in Figure 3). Sillen and Morris (1996) concluded that "until the differences [of SF factors] between BC3 and BC5 on the one hand, and the MSA fauna on the other can be explained, these hominids cannot be connected to the MSA period with confidence".

The following questions arise from the study of Sillen and Morris (1996):
* Are BC3 and BC5 significantly younger than the layers in which they were found?
* Is the measurement of lower SF factors in hominid material proof of their claimed Holocene provenance?

However, all arguments above are circumstantial and conclusive proof for the age of BC5 can come only from the direct dating of this specimen (see also Grün and Beaumont 2001). In 2002, we obtained permission to sample a small tooth fragment of BC5 for ESR dating. Using a small screwdriver, a 4.6 mg fragment was detached from the loose, partial crown of the mandibular right third molar and can be fitted back on to the stump of that tooth which is still in the alveolar part of the mandible. The tooth from which the chip was detached by Dr. R.J. Clarke is moderately worn with only a small island of dentine exposure, in contrast with the more heavily attrited second molar of the same side, which is in position in the jaw. The degree of wear of the second and third molars is compatible with this jaw having belonged to an adult individual. The average thickness of the fragment was 600±100 µm. First, we used laser ablation ICP-MS on the mirror surface of the remaining tooth to obtain U-profiles of the enamel and dentine (for details of this technique, see Eggins et al. in press). The uranium concentration in the enamel, 1 to 10 ppb, is close to modern values, whereas the uranium concentration in the dentine, 220±20 ppb, is slightly elevated and very uniformly distributed. For measurement of the dose value we used the non-destructive procedures developed at ANU and have been reported in earlier annual reports. A 4.6 mg tooth fragment yielded an average dose value of 150±5 Gy. Using the gamma dose rate values of an earlier detailed survey as well as neutron activation results of a representative sediment sample, we obtain an age of 76±5 ka. Because of the low U-concentrations in enamel and dentine, the age estimate is not sensitive to the mode of U-uptake. When compared to the results of the previous ESR dating study on faunal material, the age of BC5 fits exactly into the sequence of previously obtained age estimates (Figure 2).

Our dating result demonstrates conclusively that BC5 was buried at the beginning of the deposition of layer 3 WA. Its best age estimate is 76±5 ka. We conclude from our dating result on BC5 and the above discussion that measurements of SF and N assays are not particularly well suited to derive age assignments for the fossil hominids at Border Cave. The age of the other specimens are also best obtained by chronometric studies. At present, it seems, ESR cannot provide any further age information for the Border Cave hominids, as the only other specimen with teeth, BC3, has vanished. It may, however, be possible to date enigmatic fossils BC1 and BC2 by U-series dating. On the other hand, non destructive ESR dating may help to establish the ages of other important fossils whose age is debatable, e.g. the mandible of Banyoles.

References
Eggins, S., Grün, R., Pike, A., Shelley, A. and Taylor, L. (in press) 238U, 232Th profiling and U-series isotope analysis of fossil teeth by laser ablation ICPMS. Quaternary Science Reviews.
Grün, R. and Beaumont, P. (2001) Border Cave revisited: A revised ESR chronology. Journal of Human Evolution 40: 467-482.
Sillen, A. and Morris, A.G. (1996). Diagenesis of bone from Border Cave: implications for the age of the Border Cave hominids. Journal of Human Evolution 31, 499-506.