516 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1956 



The Dorset culture was first described by Dr. Diamond Jenness 

 (1925) on the basis of material in the National Museum of Canada 

 excavated by Eskimos at Cape Dorset on the south coast of Baffin 

 Island and on Coats and Mansel Islands. The material described by 

 Jenness differed strikingly from that found at Thule culture sites. 

 The Dorset people had used chipped-stone instead of rubbed-slate im- 

 plements. Their harpoon heads and other bone and ivory artifacts 

 were small and delicate, and entirely different in form from those made 

 by the Thule Eskimos. The Dorset collection contained no trace of 

 such typical Thule elements as whalebone mattocks, snow shovels, bone 

 arrowheads, bow drills, ulus, harness toggles, or other evidences of 

 dog traction. Another striking difference was that the Dorset artifacts 

 were usually deeply patinated. In the past 25 years Dorset sites have 

 been excavated at a number of localities in the eastern Arctic from 

 Newfoundland to Greenland. These excavations, however, added 

 little to what Jenness had originally deduced as to the age, relation- 

 ships, and significance of the Dorset culture. It was clearly older 

 than the Thule culture, for Dorset implements, and in some cases 

 Dorset occupation levels, were found underlying Thule (Holtved, 

 1944; Collins, 1950). Though the Dorsets had ocupied the central 

 and eastern Arctic many years before the arrival of the Thule people 

 in the twelfth and thirteenth centuries, a few groups of them con- 

 tinued to live on long after that. Evidence of this is a small Dorset 

 site of post-Thule age excavated by Dr. Deric O'Bryan (1953) on 

 Mill Island to the east of Southampton. And if the hypothesis 

 mentioned above is correct, the Sadlermiut themselves may have been 

 such a remnant group, though one greatly modified by contact with 

 the Thule culture. To test this hypothesis we would need information 

 on Sadlermiut sites somewhat older than the one at Native Point, 

 sites that might reveal an earlier stage of Sadlermiut or Dorset- 

 Sadlermiut culture lacking distinctive Thule traits, particularly the 

 bow drill. 



The main Dorset culture site at which we excavated was 1 mile to the 

 east of the Sadlermiut site. It was situated on the gently sloping sur- 

 face of a TO-f oot-high headland, a hill or plateau of glacial till, which 

 had once fronted on the sea but which now lies half a mile back from 

 the present beach (pi. 6, A). Extending east and west over this now 

 elevated surface and clearly visible only from the air, are a number 

 of low, closely spaced curving ridges of sand and gravel — remnants 

 of marine bars that were formed when sea waves washed over the 

 surface during the post-glacial marine submergence that inundated 

 the Hudson Bay lowlands following retreat of the glacial ice. 



It would be difficult to imagine two Eskimo habitation sites more 

 different than this and the big Sadlermiut site. The Dorset site, 



