I.] BARRIERS TO DISPERSAL. 13 



the narrow strait separating the islands of Bali and Lombok, and 

 its northward continuation, the Makassar Strait, dividing Borneo 

 from Celebes. Although the two former islands are extremely 

 close together, while Celebes is much less widely separated from 

 Borneo than is the latter from Sumatra, yet the faunas of Lombok 

 and Celebes are markedly distinct from those of the islands lying 

 to the north and west of Wallace's line. Soundings show that the 

 Makassar Strait, and likewise the Bali-Lombok Strait are of greater 

 depth than the channels separating the other islands of the archi- 

 pelago; and consequently that Wallace's line indicates a very 

 old barrier which has long been impassable to the majority of 

 mammals. 



That continental islands have received the great bulk of their 

 mammalian fauna by means of a more or less complete land- 

 connection with the mainland, is perfectly evident. Nevertheless, 

 there are cases where certain mammals have crossed the inter- 

 vening channel, either by swimming, or by having been carried 

 across on natural rafts of some kind ; an instance of this nature 

 being exemplified by the occurrence of an African type of pig in 

 Madagascar. It thus becomes a question of considerable interest 

 to ascertain what stretches of sea large mammals are capable of 

 crossing. It is stated that the jaguar has been known to swim 

 across the Rio de la Plata, which at its mouth is something like 

 eighty miles across ; and a polar bear has been observed swim- 

 ming at a distance of twenty miles from land in Bering Strait. 

 The tiger frequently crosses the narrower channels in the Sandar- 

 bans of Lower Bengal; and both deer, pigs, and elephants are 

 good swimmers. The latter animals have, indeed, been known to 

 swim for six hours at a stretch, and, with a rest, for upwards of 

 nine ; but their rate of progress is extremely slow, and probably 

 exceeds but little, if at all, a mile an hour. The Palk Strait, 

 which is considerably less than forty miles in width at its narrowest 

 part, has formed an effectual barrier to the passage of the tiger 

 from India into Ceylon; and it may accordingly be assumed 

 that about twenty miles is the utmost limit which mammals are 

 likely to cross by swimming, even when favoured by currents. 



Such passages as these must, however, be of very rare occur- 

 rence, for a terrestrial mammal is not likely to take it into its head 



