428 ANNUAL REPORT SMITHSOMAN INSTITUTION, 195 



which probably indicate a Dorset stage of culture preceding the 

 Thule on the west coast (Collins, 1937; 1940) ; and similar Dorset 

 types from Ammassalik and the Clavering Island region, illustrated 

 by Solberg (1932), Mathiassen (1933), and Larsen (1934), suggest 

 that future excavations may also reveal a Dorset stage on the 

 Greenland east coast. 



In contrast to the Thule, the Dorset culture appears to be deep- 

 rooted in the eastern Arctic. Its origin, however, is uncertain. On 

 the one hand it shows affinities with Indian culture, particularly the 

 Beothuk of Newfoundland and prehistoric cultures of the Northeast. 

 More difficult to explain but undoubtedly significant are the close 

 resemblances of some of the Dorset art motifs and stone-implement 

 types to those of the Ipiutak, Old Bering Sea, and prehistoric Aleutian 

 and Cook Inlet cultures of Alaska (pi. 1). The Dorset can hardly 

 have been derived from any of the prehistoric Alaskan Eskimo cul- 

 tures as we now know them, although a remote connection of some kind 

 is indicated. The most likely explanation, as suggested by Jenness 

 (1941), is that the Dorset has stemmed from the same parent trunk 

 as the ancient Alaskan cultures. The many and fundamental dif- 

 ferences between them, however, would indicate that the Dorset moved 

 eastward to Hudson Bay before the Ipiutak and Old Bering Sea 

 cultures had reached their full development. 



It is probably significant that recent work in Alaska to be described 

 below has revealed indications both in the interior and at Cape Den- 

 bigh on the Bering Sea coast of an ancient, apparently pre-Eskimo 

 culture or cultures with definite Asiatic affinities, characterized espe- 

 cially by burins, by small lamellar flakes, probably used as knives or 

 scrapers, and the polyhedral cores from which they were struck off 

 (Rainey, 1939; Skarland and Giddings, 1948; Giddings, 1949; 

 Solecki and Hackman, 1951). Lamellar flakes of the same kind are 

 found at many Dorset sites, and Solberg's Disko Bay collection, which 

 probably is Dorset, also includes a polyhedral core comparable to those 

 from Alaska (Solberg, 1907, p. 39). There is also a strong probabil- 

 ity that the stone burins from Giddings' Cape Denbigh site and two of 

 the early inland sites in Alaska are related to a characteristic Dorset 

 implement of somewhat similar form which De Laguna (1947, pp. 

 193-194) suggests were used as burins. 



Birnirk. — The first excavations in the western Arctic were made by 

 Stefansson in 1912 (1914). Digging in a large mound at an aban- 

 doned site called Birnirk near Point Barrow, Alaska, Stefansson 

 noted the presence of clay pottery and unusual types of harpoon heads 

 and the absence of such characteristic modern features as iron, soap- 

 stone pots, pipes, net sinkers, and net gages. Wissler (1916), who 

 described parts of Stefansson's collection, recognized the site as pre- 

 historic but did not consider it to be especially old or to represent a 



