[NYSAA (1986)] artifacts, lay below the dark brown soil. This sterile deposit also existed in some spots on the southern side, but often the dark brown layer rested on bedrock in this sector. Projectile points and other lithic artifacts were concentrated at a depth of 80 to 140 cm below datum, 60 to 110 cm below the surface. Squares 5 and 6 yielded the greatest quantities of artifacts, bone, and shell. It is likely that Case did little if any digging in the area of these squares, which lie just beyond the mouth of the roofed chamber. A small heap of oyster shells was encountered in Square 6, and a Levanna point was found lying in contact with the shells, at a depth of 1 meter below datum. Several Late Archaic points were found at lesser depths; on the other hand, one Woodland potsherd lay at a depth of 138 cm below datum. Obviously, these objects were not found in their expected stratigraphic relationship. The apparent disturbance of the deposits can be attributed to the activities of roots, worms, insects, and burrowing rodents. Because of this disturbance, it is impossible to assign typologically indistinguishable artifacts or faunal remains to particular periods of occupation. Thirty three whole or nearly whole projectile points were found, as well as 8 tips, 9 bases, and 1 mid-section piece. These artifacts are listed here in approximate chronological order, according to type. The numbers in parentheses correspond to the artifacts' numbers in the accompanying photographs (Figure 3); other data record excavation unit, depth of specimens found in situ, and lithic material. Figure 5. Pre-forms from the cache. A stone with a roughly pecked encircling central groove may have been used as a netweight (Figure 7). Very few such objects have been found in other Hudson Valley rockshelters (Funk 1976). Two broken tips of bone tools, perhaps awls, were found at the site (Figure 6, #s 33 and 34). An antler tine, bearing a few whittling marks, evidently was modified for some use, perhaps as a projectile tip (#32). Several other pieces of bone (e.g., #35) and a fragment of turtle shell display polished surfaces or edges, but are too small and amorphous to permit determination of their original form and function. A total of 162 ceramic potsherds were found at the site. These could be divided into 2 lots, on the basis of their temper. One lot consists of 74 sherds tempered with crushed oyster shell and vegetal fibers; the other includes 88 sherds tempered with small chunks (maximum size, 2-3 mm) of a black mineral that has been identified as pyroxenite. The grit-tempered sherds are cord-marked on the exterior, smooth on the interior. Their color ranges from light yellowish brown (Munsell 10 YR 6/4) to reddish brown (5 YR 4/4). The sherds range in thickness from 5 to 11 mm. Sixteen sherds derived from shoulders or necks of vessels bear clear impressions, in parallel rows, of z-twist knots. Faint traces, observed in molds of these impressions (Figure 9), of regularly spaced strands connecting the rows of knots, suggest that basketry or coarse fabric was used to make the impressions. One of the pots represented by these sherds had a slightly raised collar, decorated with diagonal rows of knot impressions (Figure 8); below the collar, the neck bore at least 5 closely spaced horizontal rows of such impressions, about 1 mm apart. Other decorated sherds, which have rows separated by 2 to 4 mm, may be derived from another vessel. The vessels represented by these cord-impressed sherds appear to be related to the Owasco Corded Horizontal type. Comparable vessels of this type are illustrated by Ritchie and Funk (1973: plates 135-7), and Funk (1976:120, no. 5) shows a similar potsherd from the Claverack Rockshelter. C-14 dates associated with this ware at the Bates site indicate a date of c. A.D. 1100-1200. However, the raised collar and diagonal impressions of the Ossining 38 Figure 6. Tools and pre-forms. Figure 7. Netsinker (obverse and reverse). THE BULLETIN 39 40 THE BULLETIN Figure 9. Negative impression of cord-marking, sherd #1. Figure 10. Shell-tempered pot, Ossining Rockshelter. Drawing from sketch by Case 1929. sherds are also seen in a sherd that Ritchie and Funk assign to the type, Jack's Reef Corded Collar (1973:164, plate 80, no. 6). This is a Middle Woodland type, associated with a C-14 date of A.D. 630±100 at the Kipp Island site. So, the Ossining sherds probably date to some time between A.D. 600 and 1200. Occupation of the rockshelter during this period is confirmed by the presence of a Levanna point. As mentioned above, Case discovered shell-tempered sherds that he arranged to form a nearly complete vessel (Figure 10). This pot measured 14 inches in diameter and 16 inches high. Its rim was slightly flared and was