100 BUREAU OF AMEEICAN ETHNOLOGY [bull. 40 



united with other elements of the word with which they came in 

 contact through the disappearance of one of the vowels or by their 

 contraction. It is, however, possible that from the beginning of the 

 language they have had this meager form. Second, some of these 

 elements, while existing as independent syllables, express relations 

 or subordinate ideas which do not seem to arise in the mind of the 

 Hupa when these syllables by themselves are uttered, but which 

 readily arise when the syllables are uttered in their accustomed con- 

 nection. Both of these statements are true of some of the monosyl- 

 labic elements of spoken English. The difference is not one of kind, 

 but of degree. 



Besides these older and largely conjectural phonetic changes which 

 join together the parts of the word, there are other more simple and 

 apparent modifications of the root by the suffix, or of the suffix by the 

 root, bringing the whole into greater phonetic harmony. These 

 changes are quite infrequent, and never great enough to obscure the 

 root or suffix. 



§ 7. Changes in the Phonetic Character of the Root 



There are definite and regular changes in the phonetic character of 

 the roots which cannot be explained as being due to the influence of 

 morphological additions. These are of two kinds: 



(1) Changes in the terminal consonant. 



(2) Changes in the character and length of the vowel. 



Changes in the Terminal Consonant. — One of the most common 

 changes of the terminal consonant of the root is that of n to n. This 

 is a change of series, the nature of the sound remaining the same. 

 The roots in which this change occurs have n in the forms expressing 

 past definite, customary, and negative future action, and n elsewhere. 

 A modification of the character of the sound, not in the place of its 

 formation, is found in the case of / and l. The first sound is found in 

 the forms expressing past definite, customary, and negative future 

 action. The change in this case is from surd to sonant. Of a similar 

 nature is the series of three sounds, ?, £, and L. The first (/) is 

 found in forms expressing customary and negative future action ; the 

 second (l) is employed with the forms of the present and imperative; 

 and the third (Z) with forms expressing definite action, whether past, 

 present, or future. A few verbs have roots ending in s or the corre- 



