Pipes (fig. 14) were not plentiful, no more than 100 fragments being found in any one deposit. The datable bowls and fragments of pipes closely followed the site's two periods as indicated by the various refuse pits; that is, examples from Pits A and B date from around 1700-1720, and those from the rest of the pits are of types loosely attributed to the period of about 1710-1780. On the evidence of association and by the use of the Harrington system of stem-hole dating, there is no reason to date any of the pipes later than the first half of the 18th century. A few deposits yielded a sufficient number of stem fragments to provide tentative dating, as follows: -------------------+--------+---------------------------+--------- |_No. of_| _Stem diameters_ | _Deposit_ | _frag-_+---------------------------+ _Date_ | _ments_| 4/64" 5/64" 6/64" 7/64"| -------------------+--------+---------------------------+--------- Pit B (T.N. 30) | 91 | 29% 60% 11% |1700-1720 Pit C (T.N. 17, 18)| 82 | 17% 78% 5% |1730-1750 Pit D (T.N. 23) | 49 | 16% 63% 21% |1730-1740 Kitchen (T.N. 1) | 55 | 57% 43% |1720-1740 -------------------+--------+---------------------------+--------- It should be noted that in all cases the samplings are too small for accuracy and that they are based on Mr. Harrington's elementary chart which he, himself, claims to be no more than a point of departure for a new approach to the dating of tobacco-pipe fragments. Nevertheless, the above results do follow fairly closely the dating of the groups arrived at on the evidence of stratigraphy and on the study of associated artifacts of all types. Since this report was first written, Lewis Binford of the University of Chicago has developed a mathematical formula based on Harrington's chart which enables one to arrive at a mean date for the deposition of a group of pipes. Audrey Noël Hume has subsequently demonstrated that a sampling of approximately 900 fragments is needed to maintain consistent results, and that the degree of accuracy rapidly falls off when dealing with groups of pipes dating earlier than 1670 and later than 1760.[117] Fortunately, the Tutter's Neck pipes, though few in number, do fall within the period of greatest accuracy. The following table illustrates the relationships between dates arrived at on the basis of all artifactual and documentary evidence (I), by the use of the Harrington chart (II), and by the Binford formula (III).