120 BUREAU OF AMERICAN ETHNOLOGY [Bull. 182 



or protuberances on their surface as though the stone did not chip 

 evenly but formed these uneven surfaces through faults in the rock 

 itself. In others these humps appear to have been left intentionally, 

 but we cannot explain their reason for leaving them. 



The Stallings Island, or Savannah, pomts from Virginia are char- 

 acterized as : 



Straight stemmed points : 



Right angle shoulders. 



One shoulder more prominent than the other. 



Only one shoulder indicated. 



Outline: Like a triangle with either a straight central stem, a straight ofE- 

 center stem, or a moderately converging stem with rounded base. 



Lateral edges: Vary from straight to slightly convex. 



Base in dominant type: Straight to moderately concave. 



Base in less dominant type: Slightly rounded to well-rounded. 



Stem: Varies from 50 to 75 percent as wide as the blade and less than 30 

 percent of the total length of the point. 



Length: Varies from 111 mm. to 62 mm. 



Width: Varies from 34 mm. to 41 mm. 



Thickness: A''aries from 11 mm. to 15 mm. 



Distrihution: Occurs in Late Archaic and Early Woodland horizons along 

 the eastern coastal area and well into the western sections of the east. 

 Similar types have been reported from the Plains and occur in all levels at 

 Signal Butte (Strong, 1935). Comparable points have been reported from 

 the prehistoric Piney Branch quarries in Rock Creek Park, Washington, 

 D.C. (Holmes, 1907), in the Seldon Island site (Slattery, 1946), in the 

 Marcey Creek site (Manson, 1948), as well as at numerous nonpottery- 

 bearing sites along the Potomac, the James, and the Roanoke and Meherrin 

 River valleys. They have not only been noted in numerous nonpottery- 

 bearing sites in North and South Carolina, Georgia, Tennessee, and Ken- 

 tucky, but they occur in similar sites to the northeast of the area running 

 well up into New York State. 



Several types are represented by single specimens which are unique 

 in shape. Since none of these have been reported from other areas 

 and all are surface finds, it will be difficult to indicate cultural link- 

 ages. The only positive fact is that they were all derived from 

 prepottery sites ; hence, they must be either Archaic or paleo-American 

 forms. Practically all of them are made of chert or rhyolite, which in 

 itself would tend to place them in either of these categories. It is 

 true that some of these forms were carried over, in modified form, 

 into the Transitional and Early Woodland Periods, but they did not 

 last very long for they soon disappeared from the horizon and the 

 notched and stemmed forms assumed dominance. Working techniques 

 consisted of either pressure flaking or rapping, with very little em- 

 phasis placed on percussion chipping. 



In this group are what may be called a "Hopewellian turkey tail 

 point" from site 44Mc75. This rhyolite form is characterized by being 

 lanceolate in outline, with small basal notches near the terminus, 

 which were formed by constricting the sides to form a short, straight- 



