156 



pronounce and they sound foreign to the language. Wlien two 

 different consonants happen to meet in Greenlandic in the 

 formation of suffix-compounds etc., they are completely assim- 

 ilated, as will be shown in the next paragrapli by examples. 

 If it were possible to examine this language historically, it 

 would surely be found thai many, perhaps most, of the long 

 consonants are assimilations of two different consonants. 



There is only one consonant, the uvular fricative r, which 

 can occur before mo.^t of the other consonants, but this sound 

 is akin to the vowels, and in such groups it is rather connected 

 with the vowel preceding, than with the consonant following il. 

 It never follows directly after another consonant, but, like other 

 consonants, it is often found isolated between two vowels. 



With respect to vowel combinations, all pseudodiph- 

 thongs are very common, as: 



\ia] [io\ [ill] \ru\ \oa\ fyaj \ua] \nr] [ui]. 



Every page of an Eskimo text contains plenty of examples 

 of these groups. 



Proper diphthongs, as in German: mein, haus, häuser, 

 in English: high, how, boy, in Danish: fej, hav, høj, do 

 not occur in Greenlandic. 



When a -\- another vowel occurs in Greenlandic in the 

 formation of suffix-compounds or in declension, there occurs 

 an assimilation whose first part consists of a long \a\ or [(/| 

 while the last part is a glide toward the other vowel (generally 

 / or H). but the other vowel is not articulated. A mere sug- 

 gestion of it appears at the close of the long open «-vowel. 

 Beside the examples of such vowel assimilations which I ha\e 

 occasionally written down in my notes, I made direct experi- 

 ments (With XI and Xlli, and I give here some of my plionetical 

 transcriptions, although they but roughly reflect these fine 

 shades of .sound. They show clearly enough that all real 



