888 LANGUAGES 



in common between the Nilotic and some of the Masai and Nandi languages 

 is the stopped terminal consonant, chiefly a " k " or a "t." This has a sound 

 similar to the " silent 'kaf'" in the ]\lalay language. (Such place-names, 

 for instance, as Perak and Sard,wak are really pronounced Pera', Sarawa'.) 

 The explosive in the terminal "k" or "t" is not pronounced. In Masai, 

 and in a lesser degree in the languages of the same group, plurals are 

 formed by altering the termination of the words (generally into a suffix 

 ending in "-k," "-t," or "-in"), and also by a distinguishing article. In 

 some of the Nilotic languages there is apparently no way of indicating 

 the plural except by the use of a numeral, or the word " many," or by the 

 employment of a distinct word in a plural sense. But in some of the 

 Nilotic languages plurals are obtained by altering the termination of 

 the word. 



The Madi group of languages offers that peculiar feature of West 

 African phonology, the guttural labial, the "kp," "gb." They also share 

 with the Manbettu, Momfu, Lendu, and other independent stocks (with 

 which they have not the slightest connection otherwise) a peculiar trilling 

 of the " d " which is generally expressed by " dr." Sometimes this peculiar 

 combination is best rendered by " dd," the last of the " d's " being pronounced 

 like the Arabic ,>. In other respects their phonology offers a deceptive 

 resemblance to the Bantu tongues, but any attempt to trace a resemblance 

 or connection in the vocabularies is practically without result. It should, 

 however, be pointed out that they do offer marked approximation to the 

 Bantu group in their limited use of prefixes to indicate the plural sense of 

 words. These prefixes, however, appear to be confined to two forms — " ba-" 

 and " ma-." There is no prefix used in a singular sense. " Mva " is a 

 child; " mamva," children. " Dilimbi " is a finger; " madilimbi," fingers. 

 In its affinities the Madi seems to be distantly related to languages of 

 West Central Africa lying between the region of the Bahr-al-Ghazal and 

 the watershed of Lake Chad. It is, in fact, what I should call a West 

 African language, strongly resembling in its phonology the tongues of the 

 Lower Niger. 



MuNDU is one of those absolutely isolated languages which are at^ 

 present the despair of the philologist in Africa. I can trace no re- 

 semblance in its words to any other known African tongue. Jluch the 

 same may be said about Makarka (Nyam-Nyam), Lendu, and Mbuba- 

 MoMFU. These, together with Manbettu, often suggest the Bantu 

 languages in their phonology, but offer little or no support to the theory 

 of any past resemblance by their approximation in word-roots, or numerals,, 

 or grammar. 



As has been already related, the Pygmies of the Congo Forest have no 

 special Pygmy language of their own, and it is an open question whether 



