CH. V] HINDER THE DISPERSAL OF SPECIES 37 



as elevations of the ground, they would probably make com- 

 paratively little difference to the existence or to the migrations 

 of plants, unless very high or very steep, but their presence 

 usually mvolves change of climate from one side to the other 

 and trom bottom to top, so that they mav produce cxreat effects 

 upon the composition of the vegetation, whether as seen in 

 simply ascendmg them, or in crossing to the other side The 

 climate usually becomes cooler and damper in ascendino- until 

 the cloud belt is passed at high elevations; and if the ra'nge be 

 transverse to a damp air-current, as so often happens owing to 

 the fact that ranges are frequently parallel to the sea, much rain 

 will be precipitated on the nearer side, and the farther side will 

 have a much drier climate. This effect can be well seen in the 

 mountains of Scandinavia, of Portugal, of New Zealand, in the 

 VVestern Ghats of India, the northern Rocky Mountains the 

 Cascades, etc. If the change is very great, the flora mav be 

 almost totally different on the two sides of a range. 



Mountains may also serve as agencies facilitating migration 

 of species, inasmuch as they may enable the passage into or 

 through a country, otherwise unsuitable in whole or in part, of 

 the plants of cooler or moister climates, or of herbs of open 

 gTound. They are also fa^^ourable to rapid migration because 

 the frequently occurring landslips may open appreciable areas 

 ot new soil not covered by vegetation, upon which plants may 

 at once take hold, without having to wait to secure a spot 

 temporarily free, or struggling to effect an entrance into a closed 

 association of plants. Such plants will probablv be mostly herbs 

 or small shrubs, inasmuch as landslips will be more common at 

 the higher elevations, which are above the tree line in many 

 cases. Owing to the fact that changes of climate have often 

 taken place in a north and south direction, mountain chains 

 running east and west have been of especial importance. 



As a general rule, a river hardly seems to be of sufficient 

 width to offer a very formidable obstacle to migration, though 

 It will doubtless delay it considerably. The only river that really 

 seems large enough to be, possibly, an actual boundary to 

 migrntinn in some eoscs is the Amazon in the lower half of its 

 course, from Manaos, where it is joined by the Rio Negro, to 

 the sea, and where it may be several miles wide. Owing to the 

 density and enormous size of the forests, howe\xr, we do not 

 yet know enough of the local distribution of the plants of that 

 region to be able to say whether or not any species really meet 



