SCIENTIFIC CORRESPONDENCE the attracted E. coli cells and the proteo­ Skhul B and the hominid bearing layers in that direct dating might be a solution, but lysis of their cellular contents would Qafzeh are similar to each other and to it is unclear how they propose to do this. then produce additional attractants for Tabun C, where the isolated jaw (TIl) We are now tackling the problem indirect­ more E. coli cells to move towards the resembles the Skhul and Qafzeh fossils ly by uranium-series and ESR dating of M. xanthus colony. This primitive form referred to as early Homo sapiens. Thus elements of the fauna actually associated of ensnarement must be of significant dates in the range of 47,000/45,000 to with Skhul 5 and 9, and in the future we survival value for this very slow-moving 75 ,000/SO,000 years ago are appropriate to shall attempt direct y-ray dating4 on the microorganism. the so-called 'Tabun B-type' industry and hominid remains. WenyuanShi to the Levantine . By con­ F. McDermott* David R. Zusman* trast, the early modern H. sapiens in C.J. Hawkesworth Department of Molecular Skhul, Qafzeh and Tabun C, all with Department of Earth SCiences, and Cell Biology, similar 'Tabun C-type' artefacts, are dated Open University, University of California, to SO,000-120,OOO years ago. Thermo­ Milton Keynes MK7 6AA, UK Berkeley, luminescence dates on burnt flints from R.Griin California 94720, USA Skhul support the electron spin resonance Quaternary Dating Research Centre, readings and indicate an average age of National University, * Author for correspondence 119,000 ± IS,OOO years. GPO Box 4, Canberra 2601, Australia We agree with the conclusions 1 that C. B. Stringer Dating hominid early modern H. sapiens in the Levant Department of Palaeontology, were coeval with European Neanderthals, Natural History Museum, but believe that the former preceded the Cromwell Road, remains arrival of the latter in the near East, which London SW7 5BD, UK SIR - McDermott et al. 1 are to can perhaps be attributed to the develop­ be congratulated for obtaining mass ment of ¥lacial conditions (Emiliani stages *Presentaddress: DepartmentofGeology, University College, Dublin, Belfield, Dublin 4, Ireland. spectrometer uranium-series dates that 5c or 4) . Oter Bar-Yoset corroborate previously published early­ 1. Bar-Yosef, O. in L'Homme de Neanderthalvol. 3 (ed. Otte, uptake electron spin resonance (ESR) David Pilbeam M.) 31-38 (Universite de Liege, 1988). readings from hominid-bearing Middle Department ofAnthropology, 2. McDermott, F., GrlJn, R., Stringer, C. B. & Hawkesworth, C. J. Nature363, 252-255 (1993). Palaeolithic sites in . But the Peabody Museum, Harvard University, 3. Garrod, D. A. E. & Bate, D. M. The Stone Age of Mount authors do not mention the difficulties Cambridge, Garmelvol. 1 (Clarendon, Oxford, 1937). 4. Yokoyama, Y., Falgueres, C. & Bibron, R. In L'Hommede arising from dating isolated animal teeth Massachusetts 02138, USA Neanderthalvol. 1 (ed. Otte, M.) 135-141 (Universite de derived from Garrod's excavations in the Liege, 1988). 1930s in Tabun and Skhul. 1. McDermott, F" GrlJn, R" Stringer, C, B. & Hawkesworth, C. J. Nature363, 252-255 (1993). Exact recording of provenance was not 2. Garrod,D.A.E.&Bate,D.M. The Stone Age ofMount practised then and understanding of site­ GarmelVol. I (Clarendon, Oxford, 1937). Chemocline of formation processes was minimal. Unfor­ 3. Vandermeersch, B. Les Hommes Fossiles de Qafzeh. Israel (CNRS, Paris, 1981). tunately, the Skhul was emptied, so 4. Bar-Yosef, O. Phil. Trans. R. Soc. B337, 193-200 (1992). the Black Sea further dating of stratified samples cannot 5. Bar-Yosef, O. & Meignen, L. in The Middle : 2 3 Adaptation, Behavior and Variability (eds Dibble, H. & now be undertaken . As at Qafzeh , all Mellars, P.) 163-182 (Univ. Pennsylvania Press, SIR - Sinninghe Damste et al. I have hominid burials from Skhul were un­ Philadelphia, 1992) investigated the question of anthro­ 6. MerCier, N. etal. J. Archaeol. Sci. 20, 169-174 (1993). covered from a 1-m-thick deposit (layer 7. Bar-Yasef, O. in L'Homme de NeandertafVol. 3 pogenic impacts on chemocline depth B2; plate I of ref. 2) covered by an (ed Otte, M.) 31-38 (Universitede Liege, 1988). fluctuations in the Black Sea by examining additional 1 m of Middle Palaeolithic the molecular record of photosynthetic sediment (layer B1). Although McDer­ McDERMOTI ET AL. REPLY - All of us, bacterial pigment degradation products in l mott et al. suggest that burials 1,4 and 5 including Bar-Yosef , who work in mat­ Black Sea sediments. They interpret the might have been later in age than the rest erial from old archaeological excavations, historical presence of anoxygenic photo­ of the hominids, we cannot directly relate must make the best of the samples at our synthesis to mean that "the penetration dates derived from animal teeth to those disposal. The long-standing debate about of the photic zone by anaerobic waters is of human skeletons. whether the Tabun should be not a recent phenomenon", and suggest One way to resolve this problem would ascribed to layer B, C or even D is that "natural causes for shoaling of the be direct dating of the hominid skeletal impossible to resolve at this stage. We2 chemocline are more likely than anthro­ remains. Thus the younger dates 41 ,400 ± have preferred to stick with the strati­ pogenic ones". The issue of an anthro­ 400 and 45,500 ± 700 years ago may graphical assignment of Garrod}. Even if pogenic influence on the chemocline indicate the presence of a Late Middle the skeleton came from a different hori­ affects the coastal zones of countries Palaeolithic assemblage in the upper part zon (layer B or D in modern strati­ surrounding the Black Sea. But in my of layer B or reflect intrusions from layer graphies), so would the bulk of the associ­ view Sinninghe Damste et al. misstate A (Upper Palaeolithic). Note that these ated faunal remains in Garrod's collec­ several earlier arguments and, more dates fall within the range of the Early tion, and for this reason our uranium­ important, rely on a small and irrelevant 4 Upper Palaeolithic in the Levant . thorium age results for the skeleton are dataset. At Tabun, Garrod was also not certain robust. Murray et al. 2 reported the rapid, about the attribution to layer C of the As for Skhul, the three youngest unexpected rise of the chemocline by woman's burial (marked as Neanderthal uranium-series ages (40,000-45,000 comparing hydrographic data collected in remains in Fig. 1 of ref. 1). Even she years) almost certainly underestimate the 1969 and 1988. They concluded that suspected that it could have been interred true age because of a complex uranium­ "Natural, interannual or decadal varia­ by the producers of the Tabun B industry uptake history (they are younger than tions in climate and river runoff" drive (page 64 of ref. 2). Interestingly, both in even the early-uptake ESR ages), and as vertical fluctuations in the chemocline, Amud and Kebara similarly robust we stated in our paper2, further research is and argued that any effects caused by skeletons were uncovered with the same required to investigate whether or not two anthropogenic influences on freshwater industry as Tabun B (ref. 5). The main age popUlations are present. inputs could be "superimposed on these portion of the lithic assemblages from We agree with Bar-Yosef and Pilbeam natural variations". Water mass balance NATURE . VOL366 . 2 DECEMBER 1993 415 © 1993 Nature Publishing Group