XII. CRITICAL REMARKS 



While a full critical discussion of the material contained in the 

 present volume must be reserved for another occasion, it seems desu'- 

 able to add a few remarks which indicate its character and value from 

 the points of view of contents and form. The method of collection 

 and the evidence for the reliability of the material as such were 

 briefly set forth in the introduction. On the whole, discrepancies 

 are so few in number and tho period of recording is so long that the 

 information as such evidently deserves full confidence. Further- 

 more, wherever I have been able to check it with my own inquiries 

 among various individuals belonging to various tribes, I find the 

 agreement quite satisfactory. 



The phonetic rendering of the material is not quite what it should 

 be. The wTiter is so inconsistent in the use of accents and ciuantities 

 that for all the records made in later years I have dispensed entirely 

 with accents. All the material recorded before 1901 was read to me 

 and corrected according to the reading. For some time after this 

 year I added accents according to my knowledge of the language, but 

 since the clear memory of the pronunciation gradually weakened, I 

 preferred to omit accents altogether rather than trust to my judg- 

 ment. 



In the reproduction of sounds there are a few difficulties. The 

 difference between voiced sounds and fortes is not always recognized 

 by Mr. Hunt. The sounds 6 and p!, d and t!, g' and k' !, g and Ic!, 

 g and ([!, l and l!, ds and ts! are often confused, particularly the two 

 last-named pairs. Whenever I call Mr. Hunt's attention to these 

 sounds he distinguishes them clearly, and I have sent him long vocabu- 

 laries with doubtful sounds which he has corrected, and these correc- 

 tions are undoubtedly valid. It is not necessary to quote examples 

 from lus texts to illustrate this point, because these errors appear 

 throughout the whole period, although they are much more frequent 

 in the early years than in the later ones. During the first few years 

 voiceless and voiced sounds were also sometimes confused. 



The distinction between x' , .x", and x is also uncertain. The transi- 

 tion of X into ^n, of x'* into ^w, and of x into ^x before certain sufRxes 

 makes the distinction of these sounds possible whenever appropriate 

 forms are found. I have tried to correct these sounds accordingly. 

 In a few cases the discovery of such changes was made after the first 

 part of the book was in print, so that there are a number of inconsist- 

 encies of this kind in the text. 



1467 



