44 GEOGKAPHICAL DISTRIBUTION. 



upon the island. 17 The reindeer is stated to cross the Behring 

 Straits by way of the Aleutian Islands and the frozen sea, and in a 

 somewhat similar manner the musk-ox finds its way to Melville 

 Island ; it is, however, singular that the last named, despite its long 

 ice-journeys, never manages to reach either the continent of Asia or 

 Greenland. In regions like the tropics, which support a luxuriant 

 vegetable growth, and which are subject to periodical fluminal 

 overflows, and, consequently, to the uprooting or outwashing action 

 of the inundating waters, it not infrequently happens that islands 

 or " rafts" of considerable magnitude, consisting mainly of inter- 

 laced or matted vegetation tree-trunks held together by various 

 creepers and climbers, and containing a sufficient quantity of vege- 

 table mould and soil bound together in the roots are floated down 

 stream into the open sea, where they are at once placed at the mercy 

 of the prevailing oceanic and atmospheric currents. These rafts 

 have been frequently noticed at the mouths of some of the larger 

 streams, as the Mississippi, Amazon, and Ganges, and, in the case 

 of the last named, at a distance of a hundred miles from its mouth. 

 Floating masses of wood, with upright trees growing over them, 

 were mistaken by Admiral Smyth in the Philippine seas for true 

 islands, until their motion made their real nature apparent. Such 

 floating masses not rarely harbour various forms of animal life in 

 their midst, and among these the Mammalia with arboreal hab- 

 its are not inadequately represented. The South American trav- 

 ellers Spix and Martius assert that on different occasions they ob- 

 served monkeys, tiger-cats, squirrels, crocodiles, and a variety of 

 birds, carried down stream (the Amazon) in this manner, and simi- 

 lar observations have been made by other travellers in the case of 

 the Rio Parana. It is asserted that no less than four pumas were 

 landed in one night from such rafts in the town of Montevideo. 18 

 Some of the animals thus conveyed may travel unconcernedly, and 

 without any special disadvantage arising from a change of abode ; 

 others, as the larger quadrupeds, will have been caught up and 

 transported through accident. To what distance such a floating 

 raft with its living cargo may ultimately be carried in safety, and 

 without detriment to its inhabitants, over the oceanic surface there 

 are as yet no data for determining. But there would appear to be 

 no reason for assuming that they could not be transported to a 

 distance of several hundreds of miles, seeing that the upright vege- 



