524 michelson; fox Indians 



'p 't and 'tc never occur in terminal syllables nor initially. The 

 surd k t p tc are unaspirated; k p tc never occur in terminal 

 syllables. 



To forestall hasty criticism it should be said that the differ- 

 ences in the quantity and quality of the vowels in such words 

 as 'I'w'' "he said," 'a'i'HcV "then he said;" 'ugima""^' "chief," 

 'utdgimdmwdwAn^' "their chief;" pemipa'd''^''" "he runs by," 

 pemipa'^owAg^^" "they run by," nepemipa""' "I run by," 

 'd'pemi'pa'uHc^' "when he ran by;" w^m'"""' "man," ne'niwAJi''' 

 (obviative of the same), neniwA'g^*' "men;" ne'sAnAgi'^f"' "I 

 had difficulty with it," sAUAge'sl'^''" "he is incorrigible" are true 

 phonetic processes and are not merely due to mishearings. ^ I 

 have had abundant opportunity to test these again and again 

 to be sure that there was no error in apperception; and Mr. 

 J. P. Harrington (of the Bureau of American Ethnology) also 

 agrees that these differences are truly phonetic and not merely 

 auditory. The phonetic laws governing these shifts have not 

 been determined, but it is clear that the position of the accent 

 and the nature of the adjacent sounds are in a large measure 

 responsible for the phenomena. The influence of the last are 

 responsible for the different results seen in the contractions''' in 

 nekVciHd'g^'"^' "he made it for me," kekVcVto'n'' "I made it 

 for thee" as compared with kekVci'ta'wipen"'" "you made it 

 for us," kVci'tawd'""^' " he made it for him." It is in the 

 belief that some time it may be possible to enunciate phonetic 

 shifts with accuracy that certain apparent fluctuations have 

 been allowed to stand. Thus for example the fact that d never 

 occurs except after w, and never in penultimate syllables sug- 

 gests that it is a modification of a due to the influence of the 

 neighboring sounds, including following ones. The fact that a never 

 occurs in penultimate syllables though both a and a do,^ tends 

 to show a phonetic shift or shifts have taken place. Hence it 

 is well to be conservative in holding that cases where a and a 



^ This will be shown in a future number of the International Journal of American 

 Linguistics. 



^ See this Journal, 9: 333. 



^ Neither does it occur before 'k, 't, 'p, 'tc, '5, 'c. Where Jones records a in such 

 cases I hear a. Similarly I hear a for a, and a' for '^. 



