238 GKAMMAR OF THE KLAMATH LAN(JUAGE. 



Ill words of four syllubles or more, our ear is not satisfied by the subordi- 

 nating' of so many unaccented syllables under one syllable, but seeks relief 

 by accentuating- iinother of their number by what is called the secondary 

 accent. In the same vva}' as the spoken sentence hurries towards its end, 

 the main accent of long wo/ds will also follow this forward rush; thus the 

 secondary accent increases in strength and tends to bring down the main 

 accent to the level of the other unaccented parts, unless its quantity offers 

 resistance to this leveling ])rocess. Thus shliutuapkuga in order to shoot 

 with, with the secondary accent on -ug-, ni;iy under the influence of the 

 following words become sliliutuapkuga, the vowel i of the radix being short. 

 The same holds g-ood oi' terms like shniulatchganka to (/lance off from, 

 lulukshaluapkuga for the purpose of crciiiatinij. Shifting' of the accent can 

 also take place when proclitic and enclitic words crowd around accentu- 

 ated words, especially verbs. 



Long vowels are not always accented; that is, quantity exercises no 

 decisive influence on accentuation. Cf. viiksh^eni toirard the anyliny place, 

 where -u- is a contraction of -uya-; but in saiga;{eni to the prairie, the suffix 

 has the accent. Nishtak during the same niyht (from nishta ak) is just as 

 possible as nishtfi'k. 



The syntactic accent, as determined by the sense of the clause or sen- 

 tence, constantly interferes with the other principles of accentuation and 

 imposes its own laws There are two sorts of syntactic accentuation; one 

 lavs the stress of the voice on that term of the sentence which seems of 

 paramount weight to the speaker, while the other lets it fall on certai)i syl- 

 lables of one, two, or several words of one sentence. This latter accent is 

 the one to be studied more carefully, the other needing no commentary. 



In using the terms friendly, plurality, selfishness we think of them as 

 tvholc words only, and do not concern ourselves about the real meaning of 

 their roots or suffixes. Indeed, very few of us know, that in friendly the 

 radix fii- means to love, cherish, that -end- represents the old Saxon form of 

 the present participle, nnd tlmt -ly is (»ur like, alike to, originally lie l)ody, 

 flesh, form. But in Klamath this is different, for these and many other 

 Indians possess an intuitive if not a real knowledge of the functions of their 

 aihxes A verb like gutil;ipk;i))r'li to make turns ivhlle deseendiny suggests at 



