VI.] RELATION TO AFRICA. 223 



narrows itself as to the probable date of the connection between 

 the island and the continent. Now, so far as can be determined, 

 none of the European Oligocene lemuroids are referable to the 

 family Lemuridce and since both the Ethiopian and Malagasy 

 representatives of the subfamily Galagince resemble one another in 

 the peculiar structure of the ankle, or tarsus, it is pretty evident 

 that not only was the family, but likewise the subfamily differenti- 

 ated before the separation of Madagascar. Allowing time for the 

 southward migration of the Oligocene lemuroids and civets, arid 

 the modification of the former into the Galagince, it seems 

 impossible to put the separation at an earlier date than the Upper 

 Oligocene, while it might well be Miocene 1 . Confirmation of 

 this comparatively late separation of the island is afforded by some 

 observations of Dr Blanford with regard to the passage of the 

 bush-pig across the intervening strait, for it is evident that both 

 that animal and the hippopotamus must have reached Madagascar 

 by swimming, as otherwise more ungulates would assuredly have 



migrated into Africa with the lemuroids and Viverridce ; but if so, all have 

 died out. The Tertiary palseontological history of Africa or Madagascar can 

 alone decide this point ; but if ancestors of the South American extinct ungu- 

 lates reached their home by way of Africa, it is certain that primitive members 

 of that order must have first passed into that continent. 



1 It must be remembered that we are here dealing with the mammalian 

 evidence alone. In regard to the molluscan Mr A. H. Cooke (Conchologist, 

 1893, p. 131) states that this region possesses sufficient individuality from that 

 of the mainland to entitle it to separation. The Helicida are peculiar, not 

 being found in the Mascarenes, Seychelles, or Comoros. They seem to be 

 related to certain Cingalese and Australian types. Upwards of fifty-four species 

 of Cyclostoma are known, distributed over Madagascar, the Comoros, Seychel- 

 les, Mauritius, and Bourbon. The African Bulimi are represented by two 

 species, but Achatina (so common there) is scarce; and groups of Bulimi are 

 peculiar. A single species of the genus Caliella is identified with an Indian 

 form ; and unmistakable indications of Oriental affinities are afforded by the 

 freshwater molluscs. There are two species of Paludomus, Bithynia occurs, 

 and while several of the Melania are of a type common in the Indo- Malayan 

 countries, the Melanatria, which are peculiar to Madagascar, have their nearest 

 allies in Ceylon or India. Although not a single African freshwater bivalve 

 has yet been recorded from Madagascar, yet several Ethiopian genera of 

 gastropods occur there, and, in common with the land-molluscs, indicate a 

 former connection between Madagascar and Africa, and this, in Mr Cooke's 

 opinion, occurred at an immeasurably remote epoch. 



