THE DISTRIBUTION OF BUTTERFLIES IN INDIA. 595 



the west lies the Konkan stretching from the Tapti to the Goa frontier 

 and comprising the well- watered mainly forest area between the sea 

 and the crest of the Ghats. South of the Konkan from Goa to Cape 

 Comorin is the rich forest area of Malabar, in which I am including 

 Canara and the lower slopes of the Nilgiri and other hills in the south. 

 The Mysore plateau and Carnatic form a fifth region which we may call 

 Sooth Indian, differing only from the Malabar in its comparative 

 poverty of species, the result of extended cultivation and limited rainfall. 



In addition to the above there are two attached island areas, Kach and 

 Kattiawar to the north-west and Ceylon to the south. The Kach-Kattiawar 

 island or islands seem to be remarkably poor in species, but they have not 

 been thoroughly worked ; of Ceylon, of course, we have fairly accurate 

 information. 



(The fauna of the Andamans is rather Indian than Burman, but it will 

 be best not to complicate the present paper by a discussion of these 

 interesting islands.) 



The higher slopes of the mountains of Southern India and Ceylon 

 present certain peculiar features which almost warrant their being 

 erected into two additional regions. Almost all the species which are 

 peculiar to Ceylon belong strictly to the mountains, the insect fauna of tho 

 plains practically not differing from similar country north of the Straits. 



Now these regions into which the insect distribution is easily divided 

 agree, as has been shown, with the superficial and meteorological 

 characteristics of the country. A glance at a geological map will show 

 that they correspond also with its past history. 



It is a general rule in the distribution of animals that an island fauna 

 is much poorer than that of the neighbouring continent. But in the 

 Oriental region we find the Indian, or what is now the continental part, 

 much poorer than the Malayan islands. This is because throughout 

 geological history down to the tertiary period the position appears to 

 have been exactly reversed. The peninsular area was an island, the 

 Indo-Gangetic plain an open sea, the Himalayas much lower and much 

 further north than at present (they seem to have been raised to their 

 present form by lateral pressure from the north just as one might push up 

 a crease in the table cloth). 



The Arabian Sea would thus have been connected at least in cretaceous 

 times by a tolerably broad channel with the Caspian and Aral Seas 



