28 
ably a very large harpoon and shows part of the notch under the lowest 
barb. It is cut from a very large bone so that the obverse is convex, 
the reverse nearly flat with the marrow canal in the middle; the front 
edge is sharp, the back nearly flat. The tang is wedge-shaped, but has 
no perforation. Near the upper or broken end there are a number of 
transverse scratches arranged in two or possibly three rows along the 
highest part of the convex surface of the object. A fragment of a thin 
harpoon point made of bone, found in heap N, has a wedge-shaped base 
and a barb in one edge. Judging by the fragment, the point probably 
had one more barb, or at most not over three, but all in the same edge. 
Seventeen of the harpoon points, besides the two in the Patterson 
collection from the prehistoric cemetery, have a perforation through the 
tang. All of these are heavy and have notches cutting deeply into the 
shaft. Several of the heavy and deeply notched harpoons (Plate VI, 
figures 14, 18), of which only the point end was found, may have had no 
perforation. Of the perforated harpoons, eight are from heap A, one 
probably from heap B, five from heap D, two from the prehistoric ceme- 
tery, and one from heap L. The specimen illustrated on Plate V, figure 17, 
among simple points made of bone, is probably one of these harpoon 
points with the perforation broken out. The perforations are gouged 
out rather than drilled, and taper inwards from both openings. Except 
in one case, they are slightly off the middle of the tang, towards the barbed 
edge. The hole was apparently used for the attachment of a line to 
retrieve the point when detached by the struggles of the harpooned animal 
rather than to fasten the point to a shaft. In one specimen (Plate VI, 
figure 11) the hole is worn down towards the base, but only on the obverse 
side, and in another (figure 13) it is worn down on both sides. This wearing 
was probably done by the retrieving cord, which in the former may have 
been knotted on the reverse side, whereas in the latter it probably was 
looped or ran free. 
The first type of barb, made by shallow notches across one edge, is 
represented by only seven specimens. Six of these are from heap A and 
one from heap L (Plate VI, figure 1). The barbs are at intervals of about 
an eighth of an inch, and are notched in such a way that they are slightly 
undercut. In cross-section the shafts are crudely circular or oval except 
for the sharp edge, which gives them an almond-shaped cross-section. 
Three of these harpoons were made of solid pieces of the wall of a large 
bone, and one (Plate VI, figure 1) is made of a bone that has been only 
slightly shaped from its natural form. Two fragmentary specimens from 
heap A appear to have this type of barb. One of these is a tip, almond- 
shaped in cross-section; the other, which lacks both ends, is concave on 
one side where it was grooved in cutting it out, and has the barbed edge 
sharp and the other edge rounded. The specimen found in heap L (Plate 
VI, figure 1) is a tip, thin and oval in section, but slightly sharper on the 
barbed edge than on the back. The notches are cut into the edge at an 
angle. Four fragments found in heap A are of harpoon heads barbed by 
merely making one V-shaped notch square across a sharp edge. 
The second type of barb, made by a shallow notch on each side 
of the edge, is found on seven specimens. Six of them are from heap A 
and one from heap D. One specimen (Plate VI, figure 2) has only one 
barb, and has a much larger head than tang. It was made from a piece 
