On Human Blood, Rock Art and Calcium Oxalate: Further Studies on Organic Carbon Content and Radiocarbon Age of Materials Relating to Australian Rock Art
Gillespie, Richard, Antiquity
Loy et al. (1990: 111) reported an AMS radiocarbon determination of c. 20,000 b.p. (RIDDL-1270) from a rock-art site at Laurie Creek in northern Australia. The dated sample was said to contain specific human blood proteins extracted from a painted rock surface, and this result to be a valid age for the artwork. One author of that paper, on the evidence of his chemical and isotopic analyses of new samples collected in 1990, has withdrawn his support for the date (Nelson 1993: 893-4). Loy (1994: 148) has replied by restating confidence in his own immunological testing of rock samples from the site, and in the original interpretation of the date and the dated material.
This paper presents further analyses on rock-surface fragments from the same location, including samples taken from both painted rock art and natural unpainted rock surfaces. Three new AMS radiocarbon determinations on organic matter extracted from these rock surfaces are also reported, with my interpretation of the scientific aspects. I reject the original date from Laurie Creek and the interpretation of what was measured given by Loy et al. (1990) for analytical reasons and suggest some possible directions in the dating of rock-art.
Samples for this project were collected at the Laurie Creek site in 1990 by a group including the author, T.H. Loy, D.E. Nelson, R. Jones, C. Chippindale and B. Meehan. Jones and Meehan had collected the original dated sample on a first visit to the site in 1987. My responsibilities, as a specialist radiocarbon chemist, were to perform chemical analyses for carbon and amino acids on rock-surface and other samples collected, as a guide to selecting samples that might contain sufficient organic carbon for useful AMS dating of the artwork.
Preliminary chemical testing in the field suggested traces of calcium carbonate in many of the natural rock skins; the white pigment in painted artwork of the site was shown most likely to be magnesium carbonate, rather than clay or calcium carbonate, by its delayed-effervescence response to dilute acid. Many natural and painted skins in the area contain surface algal or lichen growths; insect and bird droppings are common; mud-wasp nests or their remains occur in some of the rock shelters; charcoal particles, from bushfires that regularly pass through the area during the dry season, are present everywhere. All these materials contain significant amounts of organic carbon, which would be of more recent origin than the artwork they overlie.
A simple soil chemistry wet oxidation method(1) was used to analyse for total organic carbon. Sample sizes were 0.5-1.0 sq. cm of rock skin about 2-3 mm thick; we took fragments of flaking or exfoliating skins from the artwork which were close to falling off the walls, similar pieces were sampled from natural unpainted surfaces as controls. Results from the field analyses for carbon are shown in TABLE 1, with the immediate general observation that there is sufficient carbon for an AMS determination in most of the painted and natural unpainted samples tested, in the range 200900 micrograms per sq. cm. The relatively abundant organic carbon found - a surprise - encouraged the second series of field analyses for amino acids(2). Results from field testing for soluble and insoluble amino acids are also presented in TABLE 1, and show very low abundance of amino acids for most samples. Two painted artwork skins had levels approximating 50 micrograms amino acids in the insoluble fraction, and two soluble fractions, one on yellow painted artwork and one from adjacent to the location of the original dated sample, also indicated about 50 micrograms amino acids. All other samples tested contained at best trace amounts.
Total organic carbon measurements using similar methods on painted and unpainted skins confirmed the field analysis results. X-ray diffraction scans of four natural skins from engraved rock surfaces showed the presence of the calcium oxalate mineral whewellite at levels of 0. …