38 CAUSES WHICH FAVOUR OR [pt. i 



the river with a long frontage to it, and are not found at all on 

 the other side. 



Soil may be considered as a geographical factor in migration^ 

 inasmuch as it depends upon the geology of the country, or may 

 be considered under the next heading, of ecological factors. 

 Nothing has been a subject of greater controversy than the 

 effects of its composition upon the vegetation which it carries. 

 There is no doubt that one may observe quite different floras 

 upon, say, a chalk soil and a sihceous soil in England, and quite 

 another again upon a soil impregnated with salt. Exactly to 

 determine, however, what part of this effect is due to the 

 chemical composition of the soil, and what part to its physical 

 constitution, is a very difficult problem. My own experience 

 with tropical agriculture, extending over nearly tAventy years, 

 inclines me to lay more stress upon the physical constitution, 

 for crops will succeed almost equally well upon soils of very 

 different chemical composition, if only they be, for example, of 

 such physical consistency as to retain water Avell. Chalk soils in 

 the natural condition arc dry, and little retentive of water, sandy 

 soils even more so, while clays may retain water very well indeed. 



It is comparatively rare for any plant to be confined in its 

 growth to one kind of soil only. Feshica ovina is so abundant 

 and successful upon the chalk downs that one is tempted to 

 think it a chalk plant till one finds it almost as common upon 

 a bilberry moor in Derbyshire, or a grass moor in Scotland, 

 with peaty soil. Both chalk and peat demand in the plants that 

 grow upon them some capacity of resistance to insufiflcicncy of 

 Avater, and it may be the physical rather than the chemical 

 constitution that matters most. 



There is no doubt that if in the same climatic and other general 

 conditions there exist two belts of different soils, these will be 

 covered with floras that will be differently constituted in detail, 

 but it is comparatively rarely that a species will not occur on 

 both, though it may be common on the one and very rare on 

 the other. The only chemical constituents present in the soil 

 that really seem to have a determining effect in allowing some 

 species and excluding others are calcium carbonate (chalk or 

 limestone) and sodium chloride (salt). A good case is mentioned 

 by Drude (33) of a line of chalk-loving shrubs found running 

 through a forest on siliceous soil in France; on investigation it 

 was found that they occupied the track of an old Roman road, 

 for which chalk had been used. 



