92 



PROCEEDINGS OE THE GEOLOGICAL SOCIETY. 



peninsula and Ceylon, not extending even into the Indo-Gangctic 

 plain. Thus we appear, in this family of fishes, to have two lines of 

 migration indicated from Africa into Asia ; one by the Nile Vallev 

 to Palestine, the other by the Mascarene Islands to the peninsula 

 of India, each branch terminating in types quite distinct from the 

 terminal representatives of the other, and no form of the family 

 being known to occur in Asia, except in the localities mentioned. 



In the Seychelles, to which I have already referred as the only 

 thoroughly authenticated case of oceanic islands composed of grani- 

 toid or gneissoid rocks, two Proffs and two Ca)cilians are found. The 

 latter belong to an order entirely unknown in oceanic islands 

 elsewhere, and not yet recorded from Madagascar. One species 

 pertains to a genus found also in Africa, the other to a. peculiar 

 generic type ; but the order Apoda, consisting of the Coecilians, 

 is particularly w^cll represented in Southern India. The presence 

 of the Batrachia serves to prove the former union of the Seychelles 

 to a continent ; but this might have been Africa, or Madagascar 

 when forming part of the African land. 



The land and freshwater Mollusca of the Mascarene Islands are 

 just as peculiar as the vertebrates, and exhibit the same remarkable 

 affinities ; nothing can better show^ that we are dealing with a very 

 ancient fauna. A large proportion of the molluscan genei-a are 

 peculiar, such as Ilelicoplianta^ Ampelita, and Gihbus amongst the 

 Pulmonata, Acroptijcliia, Hainesia, and Tropidopliora amongst the 

 Prosobranchiata, but perhaps the chief claim to recognition is that 

 in these islands, as in the West Indies, there is a remarkable de- 

 velopment of the Ci/clostomatidce, possibly due in both cases to the 

 preservation, under insular conditions, of the members of a family 

 exposed to too many enemies for vigorous development amongst th.o 

 modern denizens of Africa and S. America. Attention has been 

 directed by the late Mr. G. Nevill * to the connexion with the 

 Oriental fauna exhibited by the land-mollusca of the Seychelles in 

 l)articular. It would take up too much time to go into detail, and 

 therefore I will merely say that some Madagascar shells of Belinda', 

 60 closely resemble Indian forms that I suspect them to be con- 

 generic, but that without detailed knowledge of the animals' it is im- 

 l)ossible to speak with any certainty. A Comoro Glesstda and a 

 Seychelles Streptax is have decided Indian affinities, however, whilst a 

 species of CocJdosfyla , a characteristic Philippine genus, and the small 

 Indian Beli.v bandqiorensis have been obtained from Madagascar. 



* P. Z. S. 1869, p. G2. 



