160 



its vocalic stem. Its inner changes {itt-u- for inn-} will no 

 doubt always be most difficult to explain. 



The following is a series of cases which I think are certain. 

 They stand as types of many similar cases — it would be im- 

 possible to give an exhaustive list of them all. Many suffixal 

 assimilations have become so stereotyped that they are no longer 

 felt as living sound-changes. Indeed it must be taken into 

 account that there are many words in the language which are 

 in reality originally suffix-formations , but can no longer be 

 recognized as such, either because the suffixes have otherwise 

 been lost to the language or because they have become such 

 integral parts of the words that they can no longer be analyzed. 

 In contrast to these words, there are others which are formed 

 on the spur of the moment, as it were, by the Eskimo as he 

 speaks. Also this kind of assimilations, which may be called 

 spontaneous, take place in accordance with the basis of articula- 

 tion of the language. It is not strange, however, if some of 

 these fall out of the general plan and must he considered as 

 rare or unique. 



Since all Greenlandic Eskimo word-stems or words end 

 either in a vowel or in one of the consonants q к t p, the 

 first element of the assimilated mass of sound must always be 

 either one of these sounds or a sound derived from one of 

 them. 



The reconstructed primitive forms are marked by an aster- 

 isk *; forms taken from the living language to show the com- 

 pletion of the assimilation are enclosed in brackets. Some few 

 of the primitive forms are still used at random instead of the 

 assimilated forms in certain parts of the country. 



