250 



couple of these words ending in Ä- which have a second plural 

 form, one that seems to have resulted from an assimilation of 

 -it and the final syllable of the singular form ; both iipik aud 

 tu'A'ik have such double plural forms, the first iqrft land ирщи)^ 

 the second t-yXct land to'Åiriit). Some few other words form 

 their plural only in this way: 



man'ik q^^^ plur. matvct] orpik a tree, plur. orpct\ sawik 

 a knife, plur. sawct. 



Of course -vt could have originated from -iit. the words 

 thus being inflected on the stem alone ; but щ1 too would 

 in an assimilation become c, and there is reason to assume 

 that this intermediate form has existed, when we compare the 

 doublets: iip-ct, гср-щИ with tlie sing, нр -ik. The transition 

 g > ?2 is well known. So I think we may assume that 

 the plural forms of these words have originally consisted of 

 three syllables. In the pronunciation of some people and in 

 some words, -iqi has passed into r {*orpiqit > orpi-t\ ; in the 

 pronunciation of others and in other words, it has passed 

 into irii (up-ik, plur. *ир-щи > iqrcTiit). 



With respect to the examples with final q in tiie singular, 

 the first question is, why lias the retrogressive uvularization 

 not taken place in these words. It might be suggested that 

 these words, at all events as far as their final consonant is 

 concerned, belonged to a later period of the language, so that 

 the time for uvularization was past when they began to be 

 inflected in the plural etc. But against this hypothesis there 

 is the fact that they are so wide-spread in the western dialects. 

 The truth of the matter is rather this, that there may he cer- 

 tain conditions in the base of the word which prevent uvulari- 

 zation from penetrating into it. It need not be supposed that 

 these conditions are of the same kind in all of these excep- 

 tions ; for each of these exceptions there may have been a 

 special reason why it was preserved as a trisyllabic word in 

 the plural. In the first place I have in mind the difficulty of 



