ARCHEOLOGICAL WORK IN ARCTIC CANADA — COLLINS 517 



which was much larger than we had anticipated, had the appear- 

 ance of a flat, level meadow or pasture (pi. 6, B). Its surface 

 was covered by a sparse, dry growth of vegetation, mostly low-growing 

 saxifrages, Dryas, grasses, and lichens, in striking contrast to the lush 

 growth of grass and other vegetation at the Sadlermiut site. This in 

 itself was an indication of age, for recently abandoned habitation 

 sites, which still retain much of their organic content, always support 

 a dense plant growth. At this Dorset site, however, the scanty plant 

 cover showed that the nutritive elements had long since been absorbed 

 from the soil. At the edges of the steep bluff and at other places where 

 wind erosion had removed the vegetation the ground was littered with 

 flint chips, occasional stone and ivory artifacts, and bleached animal 

 bones, mostly of seals and birds. We soon found, from testing, that 

 shallow midden deposits extended discontinuously for an area of 

 over 30 acres. These middens, covered by scarcely more than an inch 

 of sod and vegetation, were the only indication that the site had ever 

 been occupied by man; there were no large stones, no house pits, no 

 surface irregularities of any kind. But before the summer was 

 over, our excavations had yielded over 25,000 mammal bones, addi- 

 tional thousands of bird bones, and about 3,000 artifacts that were 

 to throw new light on the whole problem of the origin and relation- 

 ships of the Dorset culture. 



We called this old Dorset site T 1, from Tunermiut, the Aivilik 

 Eskimo name for Native Point. A second, and later, Dorset site, 

 which we found buried beneath about a foot of windblown sand near 

 the Sadlermiut site, was called T 2 (pi. 7, A). In 1955 a third Dorset 

 site, T 3, only slightly younger than T 1, was found on the old beach 

 line immediately below it, at 40 feet above sea level (pi. 7, B) . 



Midden areas at different parts of T 1, designated as Middens 1, 2, 

 3, and 4, were laid off in 5-foot squares, and a number of test pits were 

 dug at other places. As the midden deposits were shallow and rested 

 on well-drained sand or gravel, they were not frozen. Permafrost was 

 encountered only at a depth of 2 feet or more in the underlying gravel. 

 The first day's digging was sufficient to show us that this was a very 

 unusual Dorset site. Flint implements were far more abundant than 

 any other artifacts, and they were small and delicately chipped, like 

 Dorset implements generally (pis. 8, 9). Most of them, however, 

 differed in form from previously known Dorset types, and some of 

 them were unlike anything known from America. The majority of 

 the blades would be described as microlithic, and some of them in 

 shape and technique resembled microlithic types found at early pre- 

 Eskimo sites in Alaska and at Mesolithic sites in the Old World. The 

 ivory harpoon heads, though basically Dorset in character, were 

 specifically different from those found at most other Dorset sites. 



