136 BUREAU OF AMERICAN ETHNOLOGY [Bull. 192 



barbs. The last (fig. 15, b) is from Old Town II and has two barbs. 



Two fragmentary heads from Old Town II (pi. 13, d) and III 

 (pi. 13, /) are probably also for harpoon arrows. There is also a 

 tang from Old Town II, presumably broken from a similar head. It 

 differs only in being set off by a shght shoulder. 



Harpoon arrows, especially for sea otter, are characteristic of the 

 Aleut, Pacific Eskimo, Bering Strait Eskimo, Tanaina, and Eyak. 

 The Ingalik Tena used the same type for land otter, beaver, and large 

 fish (Osgood, 1940, pp. 203 ff.). Although all his Northwest Coast 

 informants denied to Drucker (1950, p. 234) that the harpoon arrow 

 was aboriginal, presumably modern examples have been reported 

 from the Tlingit, Tsimshian, and Kwakiutl (Birket-Smith and de 

 Laguna, 1938, p. 432). Small barbed heads, which I interpreted as 

 harpoon arrowheads, were found in KachemakBay I, although Birket- 

 Smith (1953, p. 218) warns us that "size alone is not decisive, for 

 from Kodiak we have sea-otter harpoons with heads which are in 

 no way bigger than those intended for harpoon arrows," and he sug- 

 gests that the latter may be a relatively late invention. I believe 

 that it originated in prehistoric times in the Aleut-Pacific Eskimo 

 area and did not spread to the Northwest Coast southeast of Yakutat 

 until the great days of the fur trade. Here it is closely associated 

 with the surround method of hunting sea otter, a technique known 

 in modern times from the Tlingit, Haida, Clayoquot Nootka, and 

 Kwakiutl. Drucker (1950, Trait 184, p. 243) suggests that "it may 

 have been an historic innovation over a wider part of the coast than 

 the entries here show." It was probably introduced among the 

 Northwest Coast tribes by the Russians and their Aleut and Pacific 

 Eskimo hunters. On the other hand, I would expect it to have been 

 long practiced at Yakutat. 



Drucker's identification of detachable barbed heads (type IV), 

 "usually under 5 inches long," from the Tlingit and Haida as har- 

 poon arrowheads (Drucker, 1943, p. 37), is, therefore, suspect, es- 

 pecially since most of them must have been considerably longer than 

 Aleut and Pacific Eskimo specimens. No small barbed heads sug- 

 gestive of harpoon arrows were found at Tlingit sites in the Angoon 

 area, even at the early historic site at Daxatkanada where sea otter 

 bones were more numerous than those of any other animal. Our 

 Angoon informants, in describing the surround, said that sea otters 

 were shot with ordinary unbarbed arrows, of the same type em- 

 ployed for land animals. Although the heads were detachable, the 

 arrows were not harpoons (de Laguna, 1960, p. 112). Drucker's 

 (1950, Traits 181, 182) informants among the Tlingit, Haida, and 

 Nootka also reported that sea otters were shot with an ordinary 



