[Herbert C. Kraft et al. (1994)] Bone is not safe in shell-matrix sites and evidences greater decay with increasing depositional time. Nichol and Wild (1984:47)calculated a half-life for the most robust bones of the snapper at 500 years in shellmatrix sites and argued that for a 10,000-year-old site, there would be one premaxilla extant for every 20,000 deposited. The paucity of fish bone in Dogan Point can be taken as evidence of great antiquity of the site. The fact that so many species at Dogan Point are represented by a single bone fragment and, consequently, a low frequency of matching elements, also suggests great age while not excluding sampling bias. Oyster Shell Seventeen raw oyster shell dates clearly place the shell accumulation at Dogan Point from 6000 to 4400 years ago and possibly as early as 7000 years ago (Table 5). Six Archaic dates have 13C corrections averaging -4.3 o/oo which can be used to age the other dates by approximately ± 330 years to derive conventional 14C dates. One may then want to calibrate these dates to reflect a reservoir age correction which must consider the specific growth locality. Elizabeth Little (1994) has explored a reservoir age correction for the Dogan Point dates and the Hudson River. Comparing the conventional dates to the associated artifacts, a delta R of between -200 and ± 500 radiocarbon years is suggested.