128 DARWINISM AND THE PROBLEMS OF LIFE 



nothing, and we will keep clear of all attempts to 

 explain things in this way. 



Once the armadilloes disappeared it would be all 

 over with the machaerodi as well, since they lived 

 exclusively on them ; their long teeth prevented them 

 from tearing up or eating other animals like the other 

 carnivores. But would it not be possible for the 

 teeth to be gradually reduced by natural selection .■* 

 No ; because the teeth were too long for the slight 

 reductions that variation might afford to enable the 

 animals to adopt a different diet, and so avert famine. 



These two species, mutually affecting each other in 

 their modification, could only be destroyed by some 

 accident or other extirpating one of them. We can 

 only suppose that, as a rule, it is external events 

 that come to affect and destroy the relation of two 

 species. If, for instance, in our old illustration of the 

 fox and the hare, the mice increased enormously from 

 some circumstance or other, the foxes, which feed on 

 mice also, would propagate more freely owing to the 

 abundance of food. If the mice then died off from 

 some disease, we should find the numerous foxes 

 reduced entirely to eating hares, and they would soon 

 extinguish them altogether. 



This illustration shows clearly what must happen 

 for a species to be entirely rooted out — namely, some 

 sudden accident. Dangers that arise gradually can 

 be met by a species, which will be gradually 

 modified. The danger must always be just great 

 enough for individual variations of the menaced 

 animals to escape it ; and in the next generation only 



