NATURAL SELECTION 305 



adaptive. It is a familiar fact that most closely allied verte- 

 brate species (or races) occur either in different habitats or in 

 different geographical areas. In insects and some of the other 

 small arthropods it appears that numerous quite closely allied 

 species may occur in one habitat, often filling, as far as we can 

 see, the same ecological niche ; in other cases allied species 

 occur in different habitats, as in vertebrates, but it is not yet 

 possible to estimate which condition is most frequent. 



The factors determining the habitat of an animal appear 

 to be exceedingly complex. In the higher vertebrates a con- 

 siderable and, at present, incalculable psychological element 

 is certainly important. In some of the smaller arthropods, 

 where psychological considerations are less likely to have 

 weight, it is highly probable that the observed habitat range 

 is due to an interaction between not only the responses of the 

 animal to edaphic conditions, but also to the nature of its food, 

 of its enemies and of its parasites. We have, therefore, in- 

 sufficient knowledge to discuss any species in much detail. 

 Certain general principles, however, can perhaps be elucidated. 



A very close parallel may be drawn between species- 

 differences in food and in habitat. And the greater part of the 

 argument on p. 302 could be repeated here with a few merely 

 verbal alterations. We are, in fact, faced on the one hand with 

 the query as to whether the enlargement of the habitat range 

 by certain individuals of a species will not benefit the species 

 as a whole rather than those individuals. On the other hand, 

 if we imagine a species with a wide habitat range separating 

 into two or more races (or incipient species), each with a 

 restricted range, then adaptation requires that each race should 

 be better fitted to live in its particular habitat than in those 

 of its allies. It is seldom, if ever, possible to demonstrate such 

 ' goodness of fit ' between race (or species) and habitat. We 

 can often indicate one factor which is predominant in deter- 

 mining why one species occurs in one habitat and an allied 

 species in another — e.g. the distribution of certain species of 

 tiger-beetles is partly governed by the nature of the soil 

 available for oviposition (Shelford, 1907, 1909). But to 

 exhibit the mechanism by which an animal appears to recognise 

 or restrict itself to its normal habitat is not the same as showing 

 that the animal is really better adapted to that habitat than 

 to any other. Close adaptation to the whole complex of 



