LANGUAGES 893 



I think by a lucky accident a clue has been found to this enigma. I 

 have hit upon a discovery which would have delighted the soul of Dr. 

 Bleek — a solution for which he was groping in the early 'sixties. Until 

 quite recently no attention whatever was paid to the remarkable Bantu 

 dialects spoken on Mount Elgon and in Kavirondo — indeed, a little more 

 than a year ago it would have been denied that any Bantu dialects were 

 spoken so far to the north as the western slopes of Mount Elgon. It was 

 not even known that Bantu forms of speech practically are spoken 

 completely round the coast of the Victoria Nyanza. When in the 

 'seventies of the last century the researches of the late Mr. Wakefield 

 and of Mr. E. (x. Eavenstein revealed the existence of a Nilotic form of 

 speech in Southern Kavirondo on the north-east angle of the Victoria 

 Nyanza, it was too hastily assumed that the whole of the east coast of 

 this lake must be withdrawn from the Bantu domain. The credit of 

 upsetting this theory and of greatly enlarging our knowledge of Bantu 

 languages is due in the first instance to Mr. C. W. Hobley, the Sub-Com- 

 missioner of the Eastern Province of the Uganda Protectorate. When the 

 present writer came to Kavirondo at the beginning of 1901. Mr. Hobley 

 clrew his attention to the fact that Bantu languages of an interesting 

 type were spoken on the west side of Mount Elgon (Masaba), and like- 

 wise that the eastern coast-lands of the Victoria Nyanza were inhabited 

 by people who spoke Bantu dialects, and not Nilotic or Nandi languages. 

 Mr. Hobley showed that even in the already known Kavirondo dialects 

 the tenth prefix (one which has long disappeared from Luganda and 

 Eunyoro) still existed. This is a statement which will leave ninety-nine 

 out of my hundred readers perfectly cold. But possibly the hundredth 

 man will have a beating at the temjaes on learning this important fact 

 of the existence of the tenth prefix in the north-eastern corner of the 

 Bantu language field. My interest having been awakened by Mr. Hobley's 

 remarks, I took an early opportunity when visiting the western side of 

 Elgon to collect vocabularies of the dialect spoken there. I found amongst 

 many other interesting facts that these people employ " Ghimu-" and 

 "Gama-" as the full and commonly used forms of the " Mu-" and "Ma-" 

 prefixes. In regard to the other prefixes also they, too, had a tendency to 

 duplication which would explain the preceding vowel that so puzzled Dr. 

 Bleek. Thus the ordinary " Ba-" and " Bu-" prefixes were generally given 

 as " Baba-" and " Bubu-." From these forms, by the degeneration due 

 to the slipshod pronunciation of the Negro, it is easy to show how the 

 abbreviated "Aba-," " Ubu-," " Ba-," and " Bu-" arose, to degenerate 

 further in many Bantu dialects to " A-," and "U-." But the " Gumu-" 

 and " Garaa-" in the living speech (Lukonde) of West Elgon to-day 

 throw much light on the origin of the first and sixth prefixes. In their 



