384 THOMAS THOMSEN 



rence either of time or locality; thus the harpoon heads figured in 

 PI. XIII, 3 6, all from the same house, include both types. It must 

 be assumed that the two types were used for different purposes; 

 but archaeological material alone is not sufficient to solve the 

 question. True, these differences of type prevail at Smith Sound, 

 but then there is at the same time a fairly great difference in the 

 size of the implements, which is not the case in North-east Greenland. 



Type III, with the line hole perpendicular to the horizontal 

 plane and a single line groove along the one plane, is represented 

 but sparingly. It implies the use of a single line only, one end of 

 which passes through the hole and is fastened with a knot; the line 

 then follows the line groove on the opposite plane. By this method 

 of fastening, the pull of the line acts exclusively in one direc- 

 tion, and the head is thus apt to be detached before the proper 

 moment. In the above-mentioned specimen from SkaBrgaards Halv0 

 a very clever remedy has been found for this defect ; 1 it has two 

 holes, one below the other, and the line is carried first through the 

 upper hole and then back through the lower one so that it runs 

 along the same plane on which the knot rests; hereby the pull is 

 distributed in both directions. 



In West Greenland, harpoon heads of this type with one line 

 are fixed to the ends of the three-pronged salmon fork. Similar 

 specimens have been found on Southampton Island, a locality which, 

 on the whole, is of great interest in its bearing on the present question. 



Slate blades of Weapons, 

 a. Trilateral. 



It has been mentioned above that one of the types of harpoon 

 head has usually had a separately-made blade; it was therefore to 

 be expected that the collection would contain such blades. Two of 

 the specimens (Figs. 5 and 6) still have bone blades nailed to them. 

 These are the only bone blades in the collection; of slate blades, 

 on the other hand, there are no less than thirty with the base 

 cut off square. The majority of these should doubtless be regar- 

 ded as harpoon blades, though some are no doubt intended for 

 lances and arrows; however, since no definite line of demarcation 

 can be drawn between them, it is best to deal with them all here 

 (cf. PI. XII, 215 and PL XIV, 1-11). 



The blades, which are rather thin, are trilateral in form, the 

 one side shaped, by grinding, into three faces, so that three 



1 THALBITZER I, p. 348; BOAS, III, p. 446, Fig. 248 I, o and p, cf. k and m from 

 Iglulik. 



