422 " DESCHIl'TIVK COMlM.KMKNT ' IN SIHoNGA 



noisa; treble reduplication in lataia, to beat mucli, where the 

 treble reduplication may serve to intensify, if that were possible, 

 the meaning of the presumably more primitive ho re ta, to finish. 

 Several instances, in Zulu, of what may be abbreviation of a 

 \(!tb, 01-, as 1 think, tnore possibly more primitive forms than 

 the verbs in question, will be noted below. 



SeSotho ehous multiplication of the last consonant of the 

 word in tsimrt. to run, and tuonri, to be many, numerous. 



In all three languages, instances occur of descriptive comple- 

 ments >with prolonged final vowels: 



In SiHonga, tikod, to be quiet; gaa; fall flat on the back. 

 SeSotho, fuu^ to be quite drunk; hoaa, to be white. 

 Zulu: bu, ask something; bi, spin round. 



SeSotho has Instances of doublets, with differing final vowels; 

 chala, pass quickly, and chali, pass very quickly; chole, go in 

 quickly; and choU, go in quickly. Compare, in Zulu, cite and 

 (■Hi, both meaning to scatter. In Zulu, also, are examples of 

 descriptive complements differing from a regular verb in an 

 internal vowef only, e.g., buluJ,-asha, same meaning as verb bulu' 

 kasha, to lay sonrethmg down in a long mass. Note that another 

 des'criptive complement of the same meaning is buluhufihu. 



InSiRonga, as we shall see later, it may be that descriptive 

 complements have given rise to nouns, verbs, and other descrip- 

 tive complements. The same is true of SeSotho. The question 

 of the relation of many Zulu descriptive complements to verbs 

 will come before us later. 



For the moment we may close this comparison with the 

 remark that v^hile, in 8eSotho, the descriptive complements have 

 often two or more different meanings, in the other two languages 

 this seems rarel_y to be the case. 



IV. — Are these ii<i>r(]s inntilfive or derivative ? 



If one could accept the position of Endemann (SeSotho), who 

 calls them interjections; of M. Junod, who says some may be 

 <interject'ons ; and of (iiout, who, in his Zulu Grammar, classes 

 at least five of them as interjections, without, 1 think, recog- 

 nizing them at all as a separate class of word, then we might 

 well regard them as being primitive Miss Werner, too, suggests 

 thcname " interjectional roots. ' The instance of the coinage of 

 one of these words, and the formation of a verb from it by its 

 coiner, mentioned by M. Junod. and alluded to above, would also 

 point in this direction. Further, the fact that some, as will, 

 1 think, appear from the lists given above for SiRonga, seem to be 

 genuine onomatopoeias^as Bishop Colenso (in one place), the lat« 

 Snr. Torre do Valle, Dr. Hetherwick. the late Mr. Stapleton, and 

 Father Torrend have all thought was the case with some of those 

 they have recorded, would seem to argue a primitive origin. On 

 the other hand must be placed .the widely-held opinon that they 

 are " verbal j)articles " (Bryant), " fragments of verbs 

 (rolenso) 



