THE NUMBERS OF ANIMALS 115 



if the cats on Tristan da Cunha had possessed poison fangs 

 like a cobra they might have been able to maintain themselves 

 with a small population. On the other hand, if the cats on 

 Berlenga Island had possessed chloroplasts like Euglena^ they 

 might have been able to exist permanently, without eating out 

 their food-supply. Again, the density possible for a species 

 depends partly upon the size of the animals. Given the same 

 food-supply and other things being equal, a small species can 

 be more abundant than a large one. This has a certain im- 

 portance in ecology, since there are a great many examples of 

 species in the same genus, and with the same sort of food 

 habits, differing in size to a very marked extent. The common 

 shrew {Sorex araneus) and the pygmy shrew {Sorex minutus), 

 the small and large cabbage white butterflies (Pieris rupee and 

 brasstcce), the smooth newt {Molge vulgaris) and the crested 

 newt {Molge cristatd), are instances. Of course if the size- 

 differences are too great they often automatically involve 

 different food habits, so that the two animals cannot be 

 compared closely. 



22. The principle of optimum density applies equally to 

 any herbivorous animal. Under normal circumstances the 

 numbers of deer are kept down by two big factors — enemies 

 and disease. Recently the deer in a sanctuary in Arizona were 

 left to themselves for some years. Owing to the absence of 

 their usual carnivorous enemies {e.g. cougars or wolves) they 

 increased so much that they began to over-eat their food-supply, 

 and there was a serious danger of the whole population of deer 

 starving or becoming so weakened in condition as to be unable 

 to withstand the winter successfully. The numbers were 

 accordingly reduced by shooting, with the result that the re- 

 maining herds were able to regain their normal condition.^i 

 Here it was clear that the absence ^of their usual enemies was 

 disastrous to the deer, that the former are in fact only hostile 

 in a certain sense, in so far as they are enemies to individual 

 deer ; for the deer as a whole depend on them to preserve their 

 optimum numbers and to prevent them from over-eating their 

 food-supply. 



23. One more example may be given. Carpenter, when 



