894 LANGUAGES 



original form these possibly were "Nguniu-" and "Xgama."* It is easy 

 to see how the forms "Ngumu" and "Xgama'" can have given rise to- 

 corresponding particles which in the nominative adhered to the " gu-'' or 

 " ga-" form and in the adjectival or objective to the " mu-'' and " ma-." 



These prefixes, therefore, at one time, were mostly dissyllables ("Gumu-," 

 "•Baba-," "Ngumu-," " Ngimi-," '^Ndindi-," "Ngama-," etc., etc.), and they 

 were words which had a separate meaning of their own, either as 

 directives or demonstrative pronouns, as indications of sex, weakness, 

 littleness or greatness, and so on. In seeking, therefore, for signs of 

 relationship with the Eantu languages amongst other forms of African 

 speech, we must take into consideration what the fullest forms of these 

 prefixes probably were. 



All that can be said at the present day in regard to the relationships 

 of the Bantu tongues is that in one or two numerals and a very few 

 word-roots, in the grammatical use of prefixes, and perhaps in general 

 phonology, there are signs of approximation to the tongues which are 

 spoken on the Lower Benue and Niger, in Yoruba, at the back of the 

 (fold Coast, and even [)erhaps to the languages of Sierra Leone. There 

 are also the same faint resemblances in the Madi group which is spoken 

 within the basin of the Bahr-al-Ghazal and the Upper Welle, and across 

 the equatorial Nile. ()n the other hand, in the conjugation of the verbs,, 

 and especially in that most characteristic Bantu feature, the modification 

 of the sense of the verb by an alteration or extension of its terminal 

 syllable, there are, as a matter of fact, resemblances to the Bantu family 

 in the Hamitic languages — Somali, Gala — and even in the Semitic. At 

 the same time this feature in human speech does, no doubt, crop up 

 quite independently (in Anglo-Saxon, English, and modern French, for 

 example). Broadly speaking, it must be confessed that we have not as 

 yet found any clue to the origin of the Bantu languages. At one time I 

 was disposed to think, on account of these vague affinities with the ]Madi 

 languages, and even with .Alakarka and the languages of the Lower Benue, 

 that the original home of the Bantu Negroes was in the very heart of 

 Central Africa, in that district lying at the head- waters of the Shari, the 

 Bahr-al-Ghazal, and the Congo. I assumed that the ancestors of the 

 Bantu, driven by the attacks of other tribes from the north-west, had 

 (piitted their original home to the north of the :Mubangi River, had 

 skirted the northern limits of the great Congo Pm-est, and made their 

 first concentration somewhere between the Albert and Victoria Nyanzas 

 and the skirts of the Ruwenzori range. This may still prove to have 



* Xot " Xgunf/«-," " Xganr/(i-," as predicted by Dr. Bleek ; they may even have 

 been, as they are now, merely "Gumu-" and "(iama"; jierhaps always "Ngumu-" 

 in the third jirefix. 



