104 THEOllY OK THE EARTH. 



highly interesting in itself, under a variety of 

 relations, and yet one that has heen hitherto very 

 little attended to. 



Before entering upon this inquiry, it is proper 

 to define what is meant by a species, so that the 

 definition may serve to regulate the employment 

 of the term. A species, therefore, may be defin- 

 ed, as comprehending the individuals which de- 

 scend from each other ) or from common parents, 

 and those which resemble them as much as they 

 resemble each other. Thus, we consider as va- 

 rieties of a species, only the races more or less dif- 

 ferent which may have sprung from it by genera- 

 tion. Our observations, therefore, regarding the 

 differences between the ancestors and descen- 

 dants, afford us the only certain rule by which 

 we can judge on this subject ; all other considera- 

 tions leading to hypothetical conclusions desti- 

 tute of proof. Now, considering the varieties in 

 this view, we observe that the differences which 

 constitute it, depend upon determinate circum- 

 stances, and that their extent increases in propor- 

 tion to the intensity of these circumstances. 



Thus, the most superficial characters are the 

 most variable : the colour depends much upon the 

 light ; the thickness of the fur upon the heat ; the 

 size, upon the abundance of food. But in a wild 

 animal, even these varieties are greatly limited 

 by the natural habits of the animal itself, which 



