CH. v] HINDER THE DISPERSAL OF SPECIES 39 



As a general rule, a change of soil does not cover a breadth of 

 country sufficiently wide to form an absolute barrier to the 

 passage of some species, or a special assistance to that of others. 

 If it is broad one way, it may be narrow in the direction per- 

 pendicular to that. There can be no doubt, however, and this 

 is all that matters to our present discussion, that it may readily 

 hinder or delay the passage of some species, and assist that of 

 others; and that it may distort as well as delay some species in 

 their distribution, by compelling them to go round. 



We come now to those hindrances interposed b}^ change of 

 conditions (to which plants react in different manners) either 

 from one place to another, or from one time to another, which 

 in a general Avay may be classed as ecological. The change may 

 be very sudden, as from forest to dry grassland (seen very 

 strikingly at the edge of the patanas of Ceylon; cf. 81), or from 

 a wet to a dry climate, as on the two sides of man)- mountain 

 chains; and in this case one comparativel}'^ seldom finds the 

 same individual species growing on both sides of the barrier 

 thus formed. But if the change be more gi-adual, as from warm 

 to cold in ascending a mountain, one often finds this to occur. 

 To what extent the barrier is effective, therefore, will depend 

 largely upon its sharpness of definition, as well as its width and 

 depth, and upon whether a genus on reaching it is able to form 

 new species capable of living upon the other side. This is a 

 phenomenon Avhich is very often seen, and it is in fact b\'- no 

 means certain that an ecological barrier will interrupt com- 

 pletely the progress of a genus, though it may stop a species. 

 When a genus is found confined to wet or dry, high or low, it is 

 most probably, as we shall see, because it is still comparatively 

 young in that country, and has not j'ct had time to spread 

 widely; quite possibly it has not yet even reached the actual 

 boundary. Widely distribvited genera, if they have many species 

 in the country, more usually have species on both sides of the 

 boundary. In the first hundred genera of the Ceylon flora, for 

 example, the genera which have species in both wet and dry 

 zones (which have a very different climate, cf. p. 14) are 32 

 with 141 species, or an average of 4-4 per genus, while those 

 confined to one zone are 68 with 135 species, or an average of 

 2 only. [The average for the whole flora is 2-7.] Of genera in 

 the entire flora that have over 10 species, seven only, with 21, 

 20, 13, 12, 12, 12, and 11 (average 14), are confined to one zone; 

 eight, the largest with 27 species, have one or more species 



