82 



than when the same sonnds are produced with the usual position 

 of articulation, and consequently we must suppose that those 

 articulations which I thus observed in the wide open mouth 

 are in natural speech formed somewhat farther forward. Still 

 I believe that I am not far wrong in the following hack tongue 

 positions, and I may especially emphasize that the relation 

 between them may be considered in all essentials correct. 



The Roman numerals ip. 83) indicate the individual Green- 

 landers whose positions of articulation 1 liave examined (v. p. 8). 

 When no Roman numeral is given, it means that the articula- 

 tion indicated is representative for the average of a number of 

 single observations noted on my phonetical lists. 



к before \c\ in [kcsa] ^ and similar words is often, if not 

 always, slightly palatalized or front-modified. A labialized к (with 

 rounding of the lips) I found in Upernavik, where the word 

 [ki^^'aql' was pronounced [ki'^k'^diq]. 



71 in the combination ryj may doubtless often be attracted 

 all the way to the edge of the soft palate so that the sound 

 becomes identical with the previously discussed (î^| (p. 79), 

 as in : 



[m^vTpvp'dqY^ [evri&vXune]* [гту^си-щ]-' [егэ^м^асР (егт^г^р 



It is strange that all of these words with the exception of 

 meriiorpoq are just as often to be heard in other forms with 

 rn irm) instead of rr^ , e. g. : \ernerAune] [erm-n-nq] [eriinfaq] 

 etc. In [негуИ] the 72 ^i-^s even wandered all the way from the 

 lips, since the original form is [>iermit]. These double forms 

 of the same words are no doubt used side by side in North 

 Greenland, but rii (rri) is said to be especially North Green- 

 landic, m (also) South Greenlandic. 



' finally ^ servant, housemaid ' becomes weak * immediately ^ 

 grandchild ' water, seas ^ bonds, bands, strings. 



