430 W. Thalbitzer 



in the description of Amdrup's northern discoveries. They confirm my 

 supposition, that we are able from the forms of the harpoon points to draw 

 conclusions as to the nature of the hunting methods among the earlier in- 

 habitants of north-east Greenland'). Fig. 134a is a variety of the nipparteq 

 sealing harpoons (cf. 131 /с), of quite the same form and size as the toggle 

 head which Amdrup found at Skærgaards Peninsula, where this hunting 

 method must therefore have been practised at one time. — Fig. 134 ô was 

 found in a refuse heap outside an old house; it is remarkable for the iron 

 blade being placed at right angles to the direction of the line-hole, and the 

 openings of the line-hole lying laterally. The basal barbs have been made 

 bj' an oblique section of the base, so that they lie in the same plane as the 

 under side. It is doubtless an ancient variety of toggle head; the unilateral 

 barb towards the point furnishes us no criterion for any certain period. 



General remarks on the harpoon heads. The Ammassalik 

 type of harpoon head with basal barbs facing each other and without 

 marginal barbs (figs. 131c, 135c) is not limited to this district, but it is 

 noteworthy, that there is no other district, so far as we know, where 

 this type has become predominant to such an extent over the other 

 types as here. The same type has been found in West Greenland 

 as high up as Upernawik (72° N. lat.)^), where it appears as a char- 

 acteristic variety of the more or less flat toggle heads of West Green- 

 land. Outside Greenland I am most inclined to compare the Am- 

 massalik harpoon head without marginal barbs with that known 

 from Cumberland Sound in Baffins Land ^). But the second, rarer 

 type from Ammassalik, with flat back and convex belly, and bipartite 

 basal barbs situated in the plane of the back (figs. 131 j, 135 a, b), 

 has more in common with old-time heads from Southampton Island^). 



The majority of the West Greenland heads have some essential 

 features in common with the type from Ammassalik mentioned first. 

 They are characterized by their curved line-hole with the openings 

 close together, and their concave base, simple or bipartite ; they 

 ahvays have basal barbs, often, further, marginal barbs towards the 

 point, the last barbs separated from the body by rhomboidal or 

 semisquare incisions, quite like the heads in figs. 131 e,/, ^ and 135 c/ 

 from Ammassalik^). 



1) Thalbitzer (1909) pp. 352—353, figs. 6 and 7, and pp. 359—360. 



-') Mason a 900) figs. 28, 34 and 43 to 46. 



3) Mason (1900) p. 264, figs. 58 and 60 (cf. PI. 7); Boas (1901; p. 14, lig. 4a and 

 (1888; p. 473, fig. 392. 



^) Boas (1901) p. 67, figs. 78 b, c, d. Cf. (1888) p. 491, fig. 523. Swenander (1906) p. 41. 



•'') Mason (1900; figs. 35 — 37. — There is a surprising agreement between this Am- 

 massalili tj'pe and three iron-pointed harpoon heads from Cape York('?), whicli 

 are in the Stockholm Riksmuseum, presented bj' A. E. Nordenskiöld in the year 

 1883. As Nordenskiöld in the same year also sent the Museum a collection of 

 21 heads from East Greenland, I can hardly' iielp thinking that the tiiree heads 

 in reality originate from the east coast of Greenland, perhaps Ammassaliii, and 

 lliat tliey have been by mistake mixed with his collection from Cape York 

 during the journe\' or after his return home. 



