EXTENSION OF EVOLUTIONARY THEORIES 327 



of mountains from on the other, each species will have 

 been changed by natural selection, possibly so much that 

 they can no longer be crossed, or not with any effect. 

 Thus there will be two different species of deer in this 

 Rhine valley. 



Inundations will often cover a large area, and this 

 must have happened more frequently in former years 

 when there was no regulation of rivers. These floods 

 would certainly often drive animals into a new environ- 

 ment. If, for instance, a river floods half a wooded 

 region, and only spares the extreme corner of it which 

 borders on a wide steppe, the animals in the part of the 

 wood that has not been flooded will remain forest 

 animals, but those that live in the narrow strip of wood 

 will gradually become steppe animals, because their 

 continued expansion will drive them on to the open 

 grass. It is not necessary for the animals to be driven 

 into other regions by the flood in order to be modified ; 

 it is enough if they find a new enemy in their new home. 

 In the struggle with this enemy, the branch species will 

 be made to differ from the parental species, which will 

 not be affected by the same agency. 



If we bear in mind the continuous geological changes 

 that have taken place in the earth's surface since the 

 enormously remote period of the origin of life, we can 

 understand that the isolations must have been frequent 

 enough to account for the infinite variety of organic 

 species. Lands have sunk beneath the sea, others have 

 emerged from the waves ; at one place mountain ranges 

 have been elevated, at another great valleys scooped 

 out. Land joined on to land and water to water, then 



