Liberia <•- 



extends over the western portion of the French Ivory Coast, 

 especially in the littoral region. It includes, no doubt, Padebu, 

 and all dialects which use as a tribal designation the suffix -bo, 

 -bu, -pwe, -po, -wo, etc. (The original name of the Kru people 

 seems to have been Krawo.) 



All the Kru languages are very nasal in pronunciation ; 

 they employ an aspirate before several consonants, rather clip 

 their vowel sounds, and evince a liking for the conjunction of 

 a trilled r with a dental, guttural, or labial without an inter- 

 vening vowel. In these two last features their phonology differs 

 from the Mandingo and Kpwesi groups. 



But all the known Liberian languages (Mandingo, Kpwesi, 

 GoRA, and Kru) except, perhaps, Kisi, differ from the Fula, 

 BuLOM.and Wolof families, and agree amongst themselves (i) in 

 their use of suffixes exclusively in construction, (2) in their pre- 

 fxitig the possessive pronouns, and (3) in their objection to 

 ending any word with a consonant (a nasal « excepted). Kisi 

 shares some of these features, but, from the very little we know 

 about it, seems related to the Bulom group in its vocabulary. 



Kisi is spoken in Western Liberia. It would seem to be 

 an outlying member of the Temne-Bulom group, so far as the 

 roots of its words are concerned, though in grammatical 

 structure it approximates more to the suffix-using groups such 

 as Mandingo.^ 



We are feeling our way gradually, slowly, with difficulty, 

 towards a classification of the Negro languages. To the present 

 writer they seem to be falling into groups somewhat after this 

 order : 



1. The Bushman languages of South-west Africa. 



2. The Hottentot group in the same region. Possibly 



' The present writer has not been able to obtain a vocabulary of Kisi. The 

 only one existing is given by Koelle in his Polyglotta Africana. 



1 100 



