| Abstract: | Frequency distributions of the lengths of the bone fragments are plotted for a number of archaeological sites. Down to a certain critical length, the distributions prove to be very similar for all the sites. Below this length, failure to recover fragments during excavation becomes the predominant factor and the frequencies decline. Distributions of diagnostic bones of sheep/goat, pig and cattle are plotted separately and the preferential recovery of the bones of larger animals is demonstrated. A method is suggested for using data from water-sieved sites to estimate and correct the effect of this preferential recovery on the figures for the relative proportions of different species at unsieved sites. |
