186 



and where tlie ice is uncertain in the winter. North of Uper- ■ 

 nawik similar conditions seem to prevail. Melville Bay has 

 not yet been entirely explored, but at all events the northern- 

 most inhabitants of Upernawik never come so far north. 

 Such conditions must necessarily give rise to linguistic group- 

 ings or dialects. 



Slight dialectic differentiations are indeed noticeable every- 

 where even vs'ithin limited areas. In the region about Disko Bay, 

 for instance, a man from Jakobshavn does not talk with the same 

 accent as a man from Godhavn or one from Egedesminde. 

 Around the Oommaymaq Fjord, there is a difference between 

 the language in its northeastern and in its southwestern corner. 



One of the sharpest lines of linguistic division on the west 

 coast seems to lie betv/een the districts of Oommannaq and 

 Upernawik^ that is on both sides of Svarlenhuk Peninsula at 

 about 72° N. lat. When I came there from the south 1 met 

 with a characteristic change in the stock of sounds, which 

 seems to correspond to similar conditions on the east coast in 

 the Ammassalik language. It is tempting to consider it more 

 than accidental that there should be this phonetical conformity 

 between two groups of languages so widely separated from 

 each other, and to assume that there is a closer genetic con- 

 nection between tbem than between either one of them and tin* 

 other dialects in Greenland. 



The following survey of the dialect forms from the various 

 parts of Greenland, so far as I have been able to get informa- 

 tion about them*), shows a series of sound-changes which all 

 are in perfect agreement with the character of the language on 



') With respect to the South Greenlandic forms I am perfectly aware that 

 many, perhaps most of them, are applicable only for a part of South 

 Greenland, in reality л^е can no doubt here too distinguish between 

 several dialects, although the distinguishing marks may only consist of 

 some comparatively unimportant phonetical diilerences: but as to the 

 nature of these differences and the boundaries between them, we have 

 not vet sufficient information. 



