910 REPORT OF NATIONAL MUSEUM, 1897. 



Hues of base and edjie. The etlj^es may be convex or concave and 

 the base with an exaggerated concavity, the two corners forming barbs, 

 the arrow shaft the stem (Plate 32, figs. 3, 8, 20, 23, 2G). Some of these 

 ini[)lenients are extremely rude, especially those of quartz and of jas- 

 per, which are refractory material, but many of these have been deli 

 cately and finely chipped. 



Triangular arrowpoints, while found in great profusion in some local- 

 ities, are not nearly so numerous throughout the country as other divi- 

 sions. They ai^pear in greater numbers on the Atlantic coast than in 

 the interior. Dr. Abbott says that in a series of 3,300 arrowpoints from 

 Mercer County, New Jersey, 1,428 were triangular. Although this 

 may be the simplest form of arrowpoint, yet the author doubts if that 

 be evidence of its having had any precedence in manufacture, or that 

 there was any evolution from it to other forms. That there may 

 have been relationship is granted. The arrow maker may have made 

 indifferently the triangular and leaf-shaped, and he may have changed 

 from one to the other, dependent upon the peculiarities of the material 

 and the success with which he was able to work it, and the question of 

 fashion and custom can not be ignored. It is thought these reasons 

 are sufficient to account for the infinite variety of shape in arrowpoints. 



The author has laid down no hard and fast lines of division in this 

 classification. Some of the leaf-shaped may have had their bases and 

 edges straightened (Plate 30, fig. 8), and the triangular had their 

 corners rounded until the two divisions came together (Plate 32, figs. 1, 

 6); so also with the leaf-shaped and the stemmed. Some of the former 

 have been notched near the base and thus been changed to stemmed, 

 and so on through the entire system. This classification is made for 

 the student and for convenience of description; therefore there will be 

 overlapping of the dividing lines between the classes, as will be read- 

 ily seen by referring to Plate 32. This must be accepted unless we 

 would make infinitesimal divisions and every slight difference in form 

 make a separate class. So each division includes all forms wliich 

 approach nearest to it, even if they have peculiarities which make it 

 difficult to harmonize. Some of the peculiarities in the triangular 

 division are to be noted. One is where the convexity of the edges 

 continued to the base brings a close resemblance to Division I, leaf- 

 shaped. Class B, (Plate 30, figs. 1, 6). Another is the widening just at 

 the base, by which the implement takes on a slight bell shape (Plate 

 31, fig. 10) ; another is where the edges of the triangle do not come in a 

 straight line nor yet in a curved line from the point to the base, but 

 make an angle midway between the two and give the imj^lement a pen 

 tagonal form rather than strictly triangular (fig. 178). A few of the 

 triangular forms have serrated or beveled edges, though this is rare. 

 Occasionally the barbs on one side are longer than the other. There is 

 no rule for the concavity of the l)ase: it varies from almost a straight 

 line to a depth equal to one-third of the length of the implement. 



