关键词: 北京猿人遗址;热电离质谱;钟乳石类碳酸岩;铀不平衡系定年法

Abstract: The Peking Man site at Zhoukoudian in northern China presents one of the world's most important paleolithic site. A reliable chronology for the site is vital to our understanding of the physical evolution and cultural development of the human species.
Earlier efforts have provided the following time frame for the strata at the site: the bottom (13th) layer-ca. 700 ka, the lowest human fossil-containing (10th) layer一ca. 500 ka and the uppermost (1-3) layers一ca. 230 ka. However, on close examinations of the dating methods used in arriving at these age estimates, one finds that certain assumptions introduced, hence the reliability of the results, have remained problematical. Furthermore, sample BZC-3, taken from an intercalated travertine in Layers 1-2, has been reported recently to have a 230Th/ 234U age of 421+110-54ka, an age much older than the previous age assignment (Shen and Jin, 1993).
That relatively pure, compact and well-crystallized speleothem samples such as BZC-3 can be precisely dated and has been well documented. The large error limits quoted above result from the precision achievable by the α spectrometric measurements on samples of > 350ka of age. The newly developed thermal ionization mass spectrometric (TIMS) analyses of U and Th isotopes have promised to greatly reduce these error limits. With this promise in mind, we have initiated a program of TIMS dating of Zhoukoudian speleothems in hopes of refining the chronology of the Peking Man.
We report here the preliminary results on the TIMS dating of sample BZC-3. The dating was done using the facilities at Centre de Recherche en Geochimie Isotopique et en Geochronologie (GEOTOP), University of Quebec at Montreal, Canada. Ultrasonically cleaned sample grains(~0.5 cm in size) of translucent calcites, free of visible porosity and detrital contaminants, were carefully hand-picked for TIMS analyses, following procedures adapted from Edwards et al (1986/ 87). The 233U-236U-229Th spike used was calibrated against a Precambrian uraninite standard(HU-1).
The age derived from the weighted mean of four measurements is 414±13 ka. The integrity of BZC-3 as a closed-system is indicated by (1) the consistency within analytical error of 234U/ 238U and 230Th/ 234U in four subsamples (not homogenized aliquots); (2) the concordant 230Th/ 234U and 227Th/ 230Th dates measured by α spectrometry; and (3) the undetectable amount of 14C in the sample as measured by accelerator mass spectrometry.
While supporting the 421 ka age reported earlier, the present more precise age of 414±13 ka for BZC-3 points to a much greater age for the Zhoukoudian upper strata than previously thought. The No.5 skull of Peking Man, a representative of the youngest Homo enectus in China, was excavated just below these strata and has been assigned an age of 230 ka. It is mainly based on this age assigmnent that hypotheses have been formulated concerning a possible overlapping existence in China of two species of Homo (erectus and sapiens) and a possible slower evolution rate of Asian Homo erectus compared with their African counterparts. None of these hypotheses is supported by the present dating results that the youngest member of the Peking Man family should be at least 400 ka old.
The techniques of TIMS as applied to 230Th/ 234U dating have begun making significant impact on refining the Pleistocene chronology. As well, their capabilities have recently extended to the measurement of 231Pa. This paper represents the first attempt at TIMS dating of speleothem from an early paleolithic site. By careful search for suitable calcite samples from Zhoukoudian stratigraphic sequence and by stringent cross-checks between 230Th/ 234U and 231Pa/ 235U signals to be acquired by TIMS on such samples, we anticipate that temporal records of Zhoukoudian and other important early paleolithic cave sites can be deciphered with greater clarity than ever before.

Key words: Peking Man site; Thermal ionization mass spectrometry (TIMS); Speleothem; U-series disequilibrium dating methods