596 STONE IMPLEMENTS FROM SHORES OF LAKE MICHIGAN. 



convex surfaces, as shown in the characteristic profile of the flake, also 

 serves to locate the point at which the initial blow fell. 



Grindiug has served more in this class of implements than in others 

 as a method of shaping. It was applied to the edge of the rough sur- 

 face for a space of half an inch on the average, and at such an angle as 

 to remove a considerable portion of the thin, natural edge. At the same 

 time it served to bring the edge more exactly into the plane of the 

 implement. 



The waterworn surface was rarely treated, as grinding, even in slight 

 amount on the convex side, would carry the edge farther from the 

 median plane of the implement. 



The functions of these celt forms, as indicated by worn edges, were 

 the same as those of the simple flakes. A length of 3| inches may be 

 taken as an average for these implements. 



Fig. 1. 



In fig. 1 two implements are sketched within the outlines of flakes 

 that might readily have served as blanks for such specializations. The 

 outlines have been traced from common forms of outer flakes selected as 

 far as practicable to correspond in proportions with the included imple- 

 ments. In each case the section through the long axis shows the 

 nearly unaltered flake section still present in the implement. Elabo- 

 rated outlines are thus clearly seen to be specializations of the flake. 

 Tlie length of the flake and the length of the implement correspond 

 in the specimen «, while in b the width of the flake has become the 

 long axis of the imj)lement. This is shown by the location of the per- 

 cussion notch [x) and in the character of the curves of the section. 

 An oblique diameter often became the long axis of the implement 

 when the blank oflered some irregularity of thickness or outline. 

 Alignment of the edge by grinding is shown in both specimens. 



Bejectage. — Tiejectage which has necessarily occurred during special- 

 izing operations is not particularly abundant. The number of rejects 



