This study is devoted to, brings new data to bear on, and seeks to contribute to, a solution for the problem of the prehistoric origin or origins of the Dorset culture in the eastern Canadian Arctic. I carried out archaeological field work for the National Museum of Canada, Ottawa, and was concerned primarily with that problem, during the summers of 1957-58-59. The analytic work was largely completed by June 1960. This study draws principally on pertinent data available up to that date, but occasionally incorporated subsequently available evidence. As so often must happen, the task of completing this manuscript, having done the analysis, was often delayed by the more immediate demands of other museum duties. So that the reader may be forewarned, if not forearmed, a history of the work on Dorset culture and the question of its origin follows as an introduction to this study.

The 1957–58–59 archaeological reconnaissance in northwestern Ungava and adjacent islands sought to find sites of the Dorset and Pre-Dorset occupations and to test them extensively. I hoped to collect data that would contribute to a solution of two questions — first, the origin or origins of the Cape Dorset culture, and second, the chronological development of that culture. This meant the work must produce large enough site samples from very early Dorset stations for comparison with the other areas, and enough site samples from throughout the Dorset time span for seriation analysis. Unfortunately the eastern Arctic is only in rare cases archaeologically prolific, either in number of sites or abundance of artifacts in a site. Also anyone working in Arctic prehistory must resign himself to a lavish expenditure of time on travel, camp maintenance, and poor weather.

The Arnapik site, JlGu-9, is located about two miles east (Fig. 1) of Amulet Creek (61°51′15″N., 79°27′40″W.) on the central east coast of Mansel Island in the northeastern extremity of Hudson Bay. Mansel Island, 36 miles west of the Ungava mainland at its closest point, is a low, generally barren island measuring roughly 60 by 34 miles. Despite its size its maximum elevation is under 500 feet. The island, formed of limestone and limestone gravels, geologically is a member of the Arctic lowlands. The vegetation, arctic grasses, sedges, mosses, lichens, willow and arctic flowers, is extremely sparse for this latitude, a marked contrast to the hardrock terrain of the nearby mainland, but much like such high Arctic locales as Bathurst and Comwallis Islands near 75° north latitude (Polunin 1948: 248).

The Tyara site, KkFb-7 in the National Museum catalogue and site file, faces the north coast of the Ungava mainland and rests on the west shore of Sugluk Island (Fig. 1). That island stands about five hundred yards from the mainland and from Sugluk Inlet, one of the few good harbors on that coast. This handsome little island, about one and one-half miles long and as wide, consists of rounded, rugged, hardrock hills that shelter well-vegetated, generally flat-floored valleys. The valleys often contain marshy patches. The shore, of variable incline, is quite jagged, a result of abrupt rock outcrops projecting seaward from brief stretches of sandy beach. The shore facing the mainland is, therefore, quite convenient for small boat use. Dark grey gneisses seem to predominate, although they are often cut by dykes and veins of lighter material, notably quartz. The dense, green valley and hillside vegetation includes willows, mosses, grasses, lichens, and a pleasant profusion of arctic wild flowers (Polunin 1948, Pt. III). I was told at Sugluk that at the head of the inlet, willows, growing in protected situations, reach the thickness of a man's wrist.

Before noting traits shared by the Arnipik and Tyara assemblages, a few words of caution and explanation may prove useful. I hope they will remain alive in the reader's mind as he examines the three closely related conclusions presented in this chapter. The three conclusions are, in order, a preliminary conclusion of restricted application, an intermediate conclusion of wider import, and a final conclusion of general consequence to this study. These conclusions' ranges of consequence increase as the discussion expands to encompass broader reaches of evidence and thinking pertinent to the question of Dorset culture origins. First, I have assumed, for previously presented or self-evident reasons, that the Tyara and Arnapik assemblages are pure assemblages, that is, that neither contains artifacts made by peoples of other cultures. Second, since the Arnapik material comprises only stone artifacts, comparison between that site and the bone, antler, and ivory artifacts from Tyara is precluded.