90 



PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



of the birds is afforded by the large fruit-eating bats {Pteropidce) *. 

 The only African genus belonging to the family is Epoynopliorus^ 

 which is confined to the continent, whilst throughout the Mascarene 

 archipelago, and even in the Comoro Islands in the Mozambique 

 channel, the typically Oriental genus Pteropus occurs and is repre- 

 sented in various islands by 5 species, one or two of them only distin- 

 guished by critical characters from the common " flying-fox " of the 

 Indian Peninsula. 



So far as the Pteropus and birds are concerned, the explanation 

 afforded by Mr. Wallace seems fully to meet the case. He points 

 out that Madagascar was probably connected with Africa in Middle 

 Tertiary times, before the present mammal and bird fauna of Africa, 

 which in Miocene (and in Greece in Pliocene) times inhabited the 

 Palsearctic region, had been driven south by the approach of the 

 colder Pliocene and Glacial epochs t, and that the connexion with 

 Madagascar was severed before the southward migration of the 

 palsearctic fauna took place, leaving in Madagascar the old African 

 forms which have since undergone no great modification. He, 

 however, points out that the areas now occupied by the Laccadive, 

 Maldive and Chagos atolls, the Saya de Malha and Cargados reefs, are 

 clearly the remains of great islands now depressed beneath the sea, 

 but w^hich must have existed in late Tertiary times, and have 

 afforded means of migration to bats and birds. In the case of 

 Pteropus, wVvah. is a powerful flier J, though I should think cer- 

 tainly incapable of winging its way from India to the nearest Mas- 

 carene Islands, this explanation is highly probable, and it applies 

 to such cases as Copsi/chus, but as regards Hypsipetes or, rather, 

 Ixochida, and Tylas, the derivation from India may be rather more 

 ancient. It should be remembered, however, that distinct genera 



* Dobson, Cat. Cliiropt. B. M. Introduction, p. xxxii. 



t It is, however, important to notice that Mr. Wallace's account of the wide 

 sea occupying the Sahara and Northern India in Miocene times is founded on 

 geological views once current, but now, I think, shown to have been erroneous. 

 There is, as Zittel has shown, no reason to believe that any part of the Sahara 

 has been sea since the Cretaceous period, and there is no evidence that marine 

 conditions prevailed at any geological epoch whatever in the plain of Northern 

 India from Agra to the Brahmaputra (' Manual Geol. India,' i. p. 393), Another 

 error into which Mr. Wallace has been led by geological writers is that of sup- 

 posing the Pikermi and Siwalik faunas to be Miocene instead of Pliocene. The 

 fauna which was Pliocene in Greece may not have reached South Africa till 

 Pleistocene times, as stated above. 



J The Indian flying-fox, P. medius, has been captured 200 miles from land. 

 Sterndale, Nat. Hist. Mamm. India, p. 39. 



