LANGUAGES 897 



Ruwenzori range. The affinities of Kuamba are about equally divided 

 between the West and East African Bantu. Although the language is 

 spoken in actual proximity to the Kunyoro and Lukonjo, it is remarkable 

 to notice that it has absolutely no more affinities with those forms of 

 speech than it displays to the Bantu languages in general. The same 

 may be said about Libvanuma, which, however, is only spoken on the 

 western side of the Semliki River and on the borderland of the Congo 

 Forest. Libvanuma has practically the same word for the numeral "ten" 

 as the Bambute Dwarfs {mini, mine). But it has apparently entirely 

 lost, or has never possessed, the widespread Bantu kumi. The Libvanuma 

 has as a negative particle "Si." On the other hand, Kuamba uses 

 Ka-, K-\ Both these negative particles, together with the variant 

 Ta-, T-, must have co-existed in the original group from which the 

 Bantu tongues started. 



I took advantage of the presence in the Uganda Protectorate of porters 

 and ex-soldiers who had wandered thither from the regions of the Upper 

 Congo to write down vocabularies of the Mangala, Ilingi, Upoto, and 

 other languages of the extreme Upper Congo — that is to say, of the 

 countries where the Congo reaches its most northern bend. It is curious 

 to remark that in this case as we go westwards we improve in the typical 

 Bantu character of the language — that is to say, Mangala is less corrupt 

 than the other dialects which lie between it and the archaic Bantu 

 languages of Ruwenzori and the Albertine Rift. But then the language 

 of the Lower Congo from Stanley Pool to the coast was formerly more 

 archaic, purer Bantu than is the Mangala language to-day. It is possible, 

 however, that the languages of the Lower Congo and Loango came thither 

 from the south-west, curled northwards from that archaic stock at the 

 head-waters of the Zambezi, from which Ochi-herero (Damara) also emerged. 

 But the iNIangala language may have reached its present site from the 

 east or north-east. 



Next to the Lukonde and Lusokwia dialects of West Elgon the most 

 interesting Bantu language which 1 have here illustrated for the first time 

 is Lukonjo of Southern Ruwenzori. This language in many respects answers 

 to the parent stock from which the Lukonjo and Runyoro dialects diverged. 

 I would draw the reader's attention to its possession of the tenth prefix 

 (Esi-, Esia-), which has long since died out in Luganda and Runyoro. 

 On the other hand, Lukonjo apparently retains no trace of the archaic 

 Ka- negative particle. It only uses the unvarying Si~ as a negative 

 prefix. It has acquired a curious dislike to the Ka- prefi.x, replacing it 

 often in the infinitives by Eri (fifth prefix). Here and there, however, 

 it retains old Bantu roots which have been lost in Luganda and Runyoro. 

 Of the two, Runyoro is more archaic than Layanda as regards its 



