BANTU IDIOMATIST. 



439' 



ingubo with gukh and huditm, 



urahlabati ,, ard and tin, 



ambata ,, ghata and bayad, 



hamba „ hadjdja and ghab. 



But what shall we say of inyama and lahm, isikati and waqt,. 



remembering the vagueness of the usually unwritten vowel in 

 Semitic, 



Was it worth the printing to record these pairs, which often have 

 but one consonant, and sometimes but one letter in common ? 



Note. — It is true that Suto lentsoe (word) is connected 

 with Zulu ilizwi, and both with Suto utloa, to hear, Peli kwd; 

 but here Meinhof's (iju)ngwa and its noun explains the varia- 

 tion. No such explanations are suggested by our Zuluist. 

 That our authority thought it worth while, labour and expense,, 

 is shown by this, that he often spends many lines of his page in 

 the dictionary in inserting this philologically useless material. 



Remember, please, that I am sneering at no one, least of all 

 at real idiomatic authorities, who have the courage to undertake 

 and the vigour to complete that heroic drudgery of lexicography 

 in any language. I am merely pointing out that true Bantu study 

 is so serious a matter that no one could expect the idiomatist to be 

 authoritative on comparative philology, any more than we could 

 expect the philologist to profess all, or (necessarily) any of the 

 Bantu tongues, for neither do we expect the Aryan philologist to 

 be a special authority on Welsh or Armenian. 



But to come to the comparison of Zulu with Aryan suggested 

 by our Zuluist, T will abstract a few of his pairs ''Zulu- 

 Sanskrit), placing the original Bantu, where known, to the left of 

 the Zulu and the original Aryan to the right of the Sanskrit. 

 Where the means would suggest to the normal philologist any 

 possible connection I place a query, otherwise a comparison of the 

 extremes will at once explode any hope that connection may be 

 established, any slight resemblance in the means usually disappear- 

 ing in the original forms. 



(N.B. — I use ij for the Hollander g, after the Anthropos 

 system. This symbol should appear as y with two dots.) 



