164 



preceding vowel , which becomes uvularized. -aqt- becomes 

 properly f-a7-]. 



The phenomenon quite corresponds to the assimilation of 

 the labial + consonant, which occurs in a number of words: 

 tcq^ w)i^ wq, wk^ ivn, wÅ, wt, ws. wip. Here too the labialization 

 affects the preceding vowel at the same time as the following 

 consonant is geminated, -aivt- becomes [-аЧ-\ etc. 



In the case of \wq]^ either w gets the upper hand as in 

 [a^q-iitY^ or q as in \d.^q-utY^ the uvularization thus making its 

 way through the iv. 



\w^\ in awi^M'Xorio becomes 2'' : [а'^^'х'^рАоцо]^ ov\ay:)pXo7io\. 

 In all cases, the labialization is easily lost in careless enuncia- 

 tion, so that the consonant-groups under consideration simply 

 become \q- ц- у к- ir Å' t s- <p-] instead of f""^' "'>[• 'у-\ etc. 



As more uncertain examples of assimilation the following 

 may be mentioned : 



The long initial consonants in the suffixes [ffunq]'^ and 

 [-s'Udiq] * may be best explained as original assimilations with 

 the final consonant of the independent word. 



The suffix /[нщ is not affixed to consonant stems in the 

 present language, but penetrates into the word up to the last 

 vowel: qinreq^ becomes qinri//-iinq^. It is a question if we 

 have to do with an assimilation here ferî^J > fr//"]. 



More plausible is the assimilation *qisuktiuuq > [qisoîi'UAq]"' . 

 -8'ищ would together with qisnk give \qisus•г^щ\ , with qinreq, 

 \qim-ers-uaq\. So the combination rs can be better endured 

 than r//. 



aput^^ сщиЬ^ with the addition of the suffix s-uaq become 

 [ариз'ищ] \аг1и8-ищ]. But perhaps s- has come from ks rather 

 than from ts . which at present either does not undergo any 

 change or else (especially in North Greenland) passes into t\ 



* a way "* parting it in several pieces, cutting it out. ^ little * l)ig 

 ^ a dog ^ a little dog ' a little piece of wood. ^ snow on the ground 

 ' man. 



