DIVISION OF THE SUBJECT. 39 



forming influences we have next to none. I therefore prefer to 

 adopt an apparently arbitrary and illogical division, classing the 

 external influences as (a) those that belong to inorganic or inani- 

 mate nature, and (5) those which are due to living organisms, 

 and above all to living animals of other species. To the first class 

 naturally belong all the relations which originate in the need 

 of animals for inorganic nourishment ; and this, though it is 

 not unfrequently consumed in the form of living animals, is 

 not able to exert its specific influence until they are dead. 



This division is, as I have observed, somewhat illogical. 

 But, irrespective of the impossibility, at present, of adopting 

 any other, it has this advantage — that it indicates at once the 

 fundamental differences between the two groups. The influ- 

 ences of the first group may be both selective and transforming, 

 while those of the second are exclusively selective. However, 

 this division is not altogether sharp and accurate, as will be 

 seen. 



