295 



case left us no alternative but to follow the outer line of coast. 

 Thus it is by no means unthinkable that a more thorough 

 investigation would have disclosed traces of Eskimo settle- 

 ments. But, considering the nature of the coast, I am 

 convinced that only very scattered traces, if any, would be 

 discovered. The coast from Кар Dalton to Scoresby 

 Sund presents rather more favourable conditions for Eskimo 

 settlements, as it possesses several straits and islands. And 

 in fact we discovered the isolated ruins of a house in 

 Stewart Ö, and an entire little settlement in the island of 

 Dunholm. 



The following table gives an exact conspectus of the 

 house ruins, tent encampments, graves etc., discovered by us 

 (see pag. 296). 



Before 1 now proceed to give a detailed description of the 

 different settlements, [ must begin by stating that, at any rate 

 as regards the settlements examined by myself in person, the 

 time seldom permitted of a thorough-going excavation of the 

 houses. As a rule, we had to confine ourselves to excavating 

 round the place where the lamp had stood, this being the most 

 likely spot for small objects to have got lost for good, men 

 and women having been sitting round here at their work. And 

 in houses where the inhabitants had not died out, we could 

 only expect to dig up objects which had accidentally disap- 

 peared. This part of the work fell, as a rule, to Søren 

 Nielsen, as I myself was generally occupied in surveying. 

 After the excavation we examined in common the objects that 

 had been found and deliberated upon what to take and what 

 to reject, considerations of space in the boat compelling us 

 to be very particular in our selection. In this regard I received 

 invaluable assistance from Søren Nielsen, who had lived 

 three years among the Eskimo at Angmagsalik before he 

 joined our expedition, and thus had an opportunity of becoming 

 thoroughly acquainted with the life and ways of this tribe. 



