BOAS] HANDBOOK OF INDIAN LANGUAGES — TAKELMA 129 



45.10). Sufiixes with no distinct derivative signification drop off in 

 the frequentative (cf. ts'!iu-d- and hiil-p!- above, also §42 passim), 

 but, if they are functional elements, are put after the reduplicated 

 complex (cf . loho-n- and 7ien-d- above) ; frequentatives thus become, 

 as was indicated in the treatment of petrified suffixes, criteria for 

 the determination of the simple base. Some verbs, however, retain 

 a petrified suffix in the frequentative without apparent reason: 

 ts'lumUmt'a he boils it; ts'lumil'ts'lamCa he always boils it. 



The only use made of the aorist stem in the formation of fre- 

 quentatives is in the case of such forms as have an initial fortis 

 in the aorist as against a media in the verb-stem, mainly verbs of 

 Type 8. The aorist of the corresponding frequentative also shows 

 the initial fortis, but is not otherwise influenced by the form of the 

 aorist stem of its simplex; e. g., aorist of simplex, tloxox-, but of 

 frequentative, t!ox-o-t!ax- vnth. retained t!-. Such verbs as aorist 

 tloxotlax, non-aorist do^xdax-, are to be considered as of mixed type 

 (in this case partly 8, partly 13a). 



Verbs like odo^ad- and ogo^ag- with a secondarily developed glottal 

 catch in the aorist (see §6) seem to retain this catch in non-aorist 

 forms, a stop + the catch resultiag ia a fortis : 



aorist ogo'^ag- always give to ; non-aorist o'lcliwlag- 



A small sub-class is formed by those frequentatives that omit the 

 -a- of the repeated base (Type 13c). Such are: 



Verb-stem Repetitive 



wa-yon.Qig'wa'n I shall run after wa-jojiii-magwa'^n I used to 



him run after him 



waii'e^ I shall sleep (71.15; wayauhi(ie^ I used to sleep 

 142.14) {-h- conditioned by accent) 



he^l-jo'^na'n I shall sing a song jonoina^^n I always sing it 



(106.7) 

 waga'n I shall bring it wagao'k'wa^n I used to bring 



it ( ? = Hoagawg-, but see 4, 

 footnote) (45.6) 



A very peculiar type of frequentative formation is illustrated by : 

 lohaThifc' {a' is inorganic) they used to die (inferential) (168.9); 

 aorist stem doubtless loholhi- 

 derived from aorist lohoi- die, non-aorist loTio- (contrast aorist loho- 

 lah-an-, non-aorist lohlah-an in the causative) . The otherwise purely 

 aoristic -i- of Type 4 is here dragged into the non-aorist forms. 



3045°— Bull. 40, pt 2—12 9 § 43 



