TRIBAL DIFFERENCES. 115 



tra, and na), while many words obsolete, or only used in 

 special senses, in the central provinces, are still in common 

 use among the outer tribes. 



From this fact, and also from other circumstances, as well 

 as tradition, there seems much reason to believe that the 

 Hovas (and perhaps some of the lighter-coloured eastern 

 tribes also) are the latest immigrants into the country, and 

 that the other tribes have been inhabitants of Madagascar for 

 a longer period. It seems certain that the Hovas have come 

 into Imerina within a time to which tradition points back, 

 and that they displaced an aboriginal race, the Vazimba (of 

 whom more presently). 



Here, however, come in two or three perplexing inquiries : 

 First, — Although there are considerable differences between 

 the various tribes in colour, physique, &c, yet does not 

 the substantial unity of language point, if not to a common 

 origin, at least to a common region, from which they all came, 

 although perhaps at somewhat widely-separated intervals of 

 time ? Is it possible that if they are of a radically distinct 

 stock, there would not be very clear indications of at least 

 two different languages in Madagascar ? Yet we do not find 

 such different languages ; for although the dialects of some 

 tribes are puzzling enough at first both to natives and 

 Europeans from a distant part of the country, this arises 

 chiefly from various pronunciations, the employment of words 

 either obsolete or used in another sense elsewhere, and other 

 minor differences, and certainly not because a distinctly dif- 

 ferent language is employed by such tribes. A glance at 

 the personal names employed among the Sakalava, in whose 

 dialect probably the greatest differences from Hova exist, 

 shows that most of them are essentially from roots common 

 both to them and to the Hovas. So it is also with the 

 names of places ; the structure of the language is the same ; 

 and although there is a small proportion of words in the 

 Sakalava vocabulary which are not employed by the Hovas, 

 it has not yet been shown that this is of African origin. At 

 the same time there may very likely be some African infu- 

 sion in the west- coast dialects, and possibly extending from 

 thence into the interior. Eemembering the proximity to 



