INFLECTION FOR NUMBER. 433 



tua i wak gi-uapkug tebl shanaholil ivliat do you want to do with the talk ? 



Gi to act also composes some verba denominativa, as nka'kgi, nkasligi, 

 mentioned above. 



(/) To say, to speak. Gi is used in this sense (instead of hemkanka) 

 only when the spoken words are quoted either verbatim or in part. This 

 use of gi has evolved from gi to do, to act, viz., "to do by words", and in 

 French we often hear il fit instead of il dit. 



na-asht gi, na'shtk, na'shtg so I say, said ; so he said etc. 

 tsi sha hiiu gi so they said ; hataktk there he said. 

 nu gi'tki gi / say they must become. 



INFLECTION FOR NUMBER. 



There are some grammatic categories wliicli have remained in a state 

 of rudimentary development in the mind of the Maklaks Indian, and seem 

 to have been too abstract for him. Anu)ng these is the category of number, 

 or what we call, graumiatically, the sinyular, dual, and plural; for these do 

 not exist here in the sense expressed in European tongues. As to verbal 

 inflection, tliis would necessitate the incorporation of the pronoun into the 

 body of the verb, or of particles marking plurality. This we observe in 

 many American languages, but not in Klamath, where only a faint com- 

 mencement was made toward incorporating personal pronouns into the 

 verb. But this language uses several other means to express number in a 

 more indirect manner. One of these is the use of a different radix when 

 the subject or object changes from the singular to the dual or plural ; but 

 this is not verbal inflection for number, for the term inflection implies rule 

 and regularity extending over all verbs, whereas here the choice of the radix 

 is sometimes arbitrary. Such a change in the radix always implies also a 

 change in signification, liowever small; and if this change is no longer per- 

 ceptible, it was so in the earlier history of the language. Moreover, the 

 assumed term for dual does not mean here two subjects only, as with us. 



At an early period the genius of this upland tongue seems to have left 

 unnoticed the expression of nu'nb^r in verbs, as well as in nouns, and found 

 no more necessity to define it than to define sex. Only a little more atten- 

 tion was paid to the categories of mode and tense, for what was done in all 

 28 



