INTRODUCTION. xiii 



to the fresh-water species is a tract of salt water, or a narrow arm of the sea. It is absolutely essential for 

 the migrations of these fishes that fresh-water continuity should be unbroken, while to admit this it seems 

 reasonable to insist that a land connection must have been present for the existence of the fresh-water to be 

 possible. If, therefore, forms of fre^b-water fishes are found in the islands of the Indian Ocean identical 

 with those on the mainland, it does not seem unreasonable* to suppose that the two localities have been 

 connected one with the other at some antecedent period. 



Of the 87 Genera of fresh-water fishes to which I have to allude, we find the distribution to be as 

 follow : — 



5 restricted to the Himalayas or Tartarian forms. 



69 in Hindustan ; of these 38 extend to the Malay Archipelago. 



8 restricted to Ceylon, of which 4 are found in Burma and 5 in the Malay Archipelago. 



5 in Burma, 4 of which extend to the Malay Archipelago. 



Out of those genera which are extended from India to the Malay Archipelago, 12 are common to 

 Africa, India, and the Malay Archipelago : while 2 only are restricted to Africa and India, and both these 

 forms are found in the intervening Mediterraneo- Persic sub-region. In the Indian fresh-water fishes of the 

 present period, the Malayan element is far more developed than the African. 



Of the 14 Genera at present existing in both Africa and India, we find that 9 are strictly confined 

 to fresh waters, as in them there are no species which reside in the seas or estuaries : of these 7 are lilcewise 

 found in the Mediterraneo-Persic sub-region, and it is therefore reasonable to suppose that it may have 

 been by that route that they obtained access into India. 



Of the sub-regions of India and Burma we have first, the Geylonese, extending from Goa down the 

 Western coast of India including Canara and Malabar, with the Western ghauts, to Ceylon : passing along 

 the Neilgherries, its piscifauna joins with that of the Hindustan sub-region in Mysore, while in the 

 Carnatic°it in like manner passes as far as the river Kistna. 27 Genera of fresh-water fishes have been 

 found in the island of Ceylon, all but one of which (Ghanna) are common to the mainland. 



This Geylonese sub-region possesses several elements in it which appear to point out the advisability 

 of separating it from that of the plains of India. Thus we find the Genera Pristulepis and Sicydium in 

 Malabar or Canara, absent from the plains of Hindustan, but reappearing in Burma: the Ophiocephalus 

 micropeltes and the siluroid Glarias Dussumieri, have a nearly similar distribution. Ghanna is restricted 

 to Ceylon and China. Polyacanthus is found in this sub-region, but elsewhere nowhere nearer than the 

 Malay Archipelago ; in fact one species is restricted to the island of Ceylon and to Java. Silurus in the 

 Western ghauts and Himalayas, but not in the Hindustan sub-region. The Cyprinoid genus Hoinaloptera 

 has two species on the Western ghauts identical with those on the Himalayas, but the Genus is absent from 

 the intervening country, being otherwise restricted to Burma and the Malay Archipelago. Genus Scaphiodon 

 extends from "Syria to Siud and along the Western ghauts of India. The siUiroid Genus Englyptosternum 

 is found both in Syria and along the base of the Himalayas. The Genus Etroplus is restricted to this 

 sub-reo-ion, but a nearly allied form has been found in Madagascar. These few facts may be variously 

 interpreted, but they appear to show that from Syria and the Mediterraneo-Persic region we have 

 Northern forms extended along the Western ghauts of India and likewise along the base of the Himalayas. 

 They also seem to demonstrate that some close connection must have existed in times gone by between 

 the Geylonese and Malayan sub-regions. 



The Hindustan sub-region is that large alluvial tract which extends from the Himalayas and 

 the mountain ranges of Sind southwards to meet the Geylonese sub-region, it also embraces the valley 

 of the Brahmaputra. Out of the 69 Genera of fresh-water fishes found in this area 64 extend to Burma, 

 and 38 of these to the Malay Archipelago, while 15 are common to the Patearctic region ; and there is, 

 as might be expected, a larger proportion of Burmese forms in Assam than elsewhere in India ; while the 

 Genera restricted to the area of this sub-region are no more than 5, and, with the exception of Sisor, 

 it possesses no well marked local forms. . . 



The Himalayan sub-region cannot be included with that of the plains of India; it is true that 

 some varieties of fish ascend there to breed or even to reside there. Finding tropical valleys in the.se 

 elevated districts suitable to their necessities, it is not surprising that following their prey up the streams, 

 they may have been cut off from a return to the plains. Some of the tropical forms which are found 

 there have become modified for a life in hill torrents as Pseudecheneis (p. 500), which possesses a sucker 

 formed of transverse folds situated on its chest between its pectoral fins and by aid of which it keeps 

 itself from being washed away : Ghiptosternum (p. 496) is another .siluroid genus, also with an adhesive 

 sucker on the chest, but with longitudinal folds : the fish of this last genus (although some are found in 

 the Himalayas) seem to be adapted more for rapid rivers of the plains : Exostoma (p. 501) is another 

 remarkable form extending along the Himalayas to Thibet and China, as well as the spur or continuation 

 southwards through Burma and Siam. But once near the summits of the Himalayas, we come 

 across the true Tartarian or Turkestan forms. Here there are carps which have but little relationship 



* 1 omit from this di.^cnssinn the oft-repeated tale of geese swallowing fish egg?, which were subsequently voided 

 uninjured and with ihcir vitality imimpMi.ed ! It mav be that aquatic birds have had tlieir maws filled with fasli ova and Hying 

 some short dislauce have disgorged such unimpaired, their vitality still remaining. The action of wh rlwinds I have already 

 adverted to. 



