118 BUEEAU OF AMERICAN ETHNOLOGY [bull. 40 



one is accustomed to in certain other American languages. Of these, 

 barely one or two (a frequentative and a comitative) can be said to 

 convey anything like a material notion, the rest being of the more or 

 less formal or relational character met with in suffixes of inflective lan- 

 guages — intransitivizing elements, causative, reflexive, passive, recip- 

 rocal, and others of less easUy described signification. Those suffixes 

 that have no clearly defined value may be put in a class by them- 

 selves as ''petrified" suffixes, the justification for such a classifica- 

 tion being purely descriptive; genetically they probably form a 

 heterogeneous group. 



§ 42. PETRIFIED SUFFIXES 



In speaking of verbs of Types 2 and 3, it was pointed out that in 

 a large number of cases certain consonants that one would naturally 

 be inclined to consider part of the verb-stem could be shown by more 

 careful analysis to be really of a suffixal character. The criteria for 

 such a suffix are partly, as was there indicated, the existence of 

 evidently related forms in which the consonant is lacking, partly 

 certain phonetic features. In a considerable number of cases dif- 

 ferent suffixes are found joined to the same verbal base, yet hardly 

 ever determining so specific a meaning that their primary signification 

 can be detected. The following examples, 



fgeits'H something round lies (138.24) 



fgeyeha'^n I roll it 



V geHja'lxde^ I run around 



al-fgeye'fgiya^n I tie it around (my head) 188.5 



wi^-l-fgeye'^^Jdin he is surrounded on all sides 48.13 



evidently all contain the same radical element or base (fgey-), 

 which has reference to circular action or position. The suffixes 

 -ts'!-, -&-, and -Ic!-, however, can not be shown to be directly respon- 

 sible for the specific meanings of the different forms, these being 

 determined chiefly, it would seem, by the succeeding suffixes, the 

 prefixes, and the general form (transitive or intransitive) of the 

 verb. Similarly, the forms he^^-sgaya'pxde^ i lie down, da-sgaya- 

 na'^n i lie down, and possibly also da-sgall it lies scattered 

 ABOUT (like grain). Contain the same radical element (sga[y]-) ; but, 

 as in the examples first cited, the abstracted suffixes -p-, -n-, and 

 -Z-, refuse to yield anything tangible. The stems galh- twist and 

 gelg- twirl fire-drill are very probably related, though neither 

 § 42 



