104 AFRICAN AFFINITIES, 



Mr. Wake, although possibly obtained in Im(5rina, may have 

 been that of a slave, either from one of the tribes distant 

 from the Hovas, or even from an individual with a distinctly 

 African admixture. 



Those who have lived among the IMalagasy in the central 

 province know how carefully and religiously they preserve 

 and bury the bones of their dead, so that it is one of the 

 most difficult matters possible to obtain a skull of one of the 

 free people ; while violation of a grave is looked upon as the 

 most heinous of crimes. Unless, therefore, there was most 

 unmistakable evidence that the skull in question was really 

 that of a pure Hova, little stress can be laid upon the 

 evidence of a single so-called Hova skull. Could a series 

 of crania be procured from all the principal divisions of the 

 inhabitants of Madagascar, some most valuable information 

 as to the affinities of the Malagasy would doubtless be 

 obtained ; but owing to the superstitions and habits of the 

 people there seems little hope at present of getting light on 

 this question from such a source, 



Mr. Wake has, I think, laid too little stress upon some 

 other points which tell against the supposed African affinities, 

 such as the Malagasy non-use of skins for clothing, a material 

 so universally employed in South Africa ; their use of woven 

 and beaten-out vegetable fibres, which connects them so 

 closely with the Polynesian tribes ; the use of the feather- 

 bellows found among the Malays ; their ancient knowledge 

 of iron-smelting; the employment of the brotherhood-by- 

 blood covenant, &c., &c. And the affinities which he be- 

 lieves he finds between the languages of South Africa and 

 of Madagascar seem, in the greater number of the examples 

 lie adduces, to be of so obscure and doubtful a character that 

 they have very little value as establishing any relationship. 

 I\Ir. Wake takes his illustrations from Dumont D'Urville's 

 Vocabulary, but in dictionaries of the Hova dialect, either 

 French or English, or even the Vocahulaire Sakalava et 

 Betsimisara of the Abb(5 Dalmond — much fuller and more 

 correct works than D'Urville's — some of the supposed resem- 

 blances disappear entirely, while some of the words cited as 

 connected with South African tongues — Kafir and Namaqua 



