Sampling

Samples for laboratory analysis were collected from two different parts of the site. Those upon which most of the present conclusions are based were taken in continuous two-inch intervals down a vertical face in Square M11 ([see Figs. 7] and [10]), very near the back wall of the rockshelter. They were not collected until after much of the site had been excavated and it was only near the rear wall that a relatively complete section could be obtained. Rodents had disturbed the upper levels and only the profile from 38-102 inches below baseline was studied. Because the location of this sample might very well reflect the influence of the overhang on the particle size distribution of the deposits, another sample section was taken in front of the shelter. Stratigraphic relations and similarities in the archaeological materials suggested that the deposits extended essentially horizontally from the front to the back of the site, and that correlations for six-inch horizontal levels were generally warranted. The samples from Square J8 were collected in continuous six-inch intervals from a point 38 inches below baseline. The expected horizontal variation with respect to particle size distribution was, in fact, not borne out by the results of this second sampling. A comparison of the median diameter and sorting coefficients of the samples suggests that both samples are generally representative of the deposits and of variations within them ([see Figs. 67], [68]):

Median diameter of 7 samples from
the back of the shelter (M11)0.61-0.75mm.
Median diameter of 4 samples from
the front of the site (J8)0.60-0.70mm.
Average sorting coefficients of
7 samples from M112.20
(relatively well-sorted)
Average sorting coefficients of
4 samples from J82.18

One would not expect that the results of the chemical analyses would be appreciably affected by the location of the sample.