I'H. XX.] CONDITIONS OF FEOGEESS. 267 



Chapter IV., tliat -wherever the integration of matter and 

 concomitant dissipation of motion are going on, there must 

 always ensue a change from indefinite uniformity to definite 

 multiformity of structure. As has already been shown, such 

 a change can be expected to take place only when a number 

 of specified circumstances concur in forwarding it. So it is 

 one of the peculiar merits of Mr. Darwin's theory of natural 

 selection, that it does not allege an unceasing or ubiquitous 

 alteration of animal and vegetal forms, but includes, in a 

 general way, all cases of persistence of type, as well as all 

 cases of progress or retrogression. One and the same general 

 theory accounts for the fact that, while some species thrive 

 in the struggle for life and acquire new capacities, others 

 dwindle in numbers or deteriorate in structure, while others 

 again maintain themselves unchanged throughout immense 

 periods. Throughout all these cases, the general truth is 

 easily discerned that the total result will depend upon a very 

 complex combination of circumstances : the difficulty is in 

 applying the general truth to the special cases that arise. 

 Probably no naturalist could point out all the specific circum- 

 stances which have caused any one race of animals to prevail 

 over another in the struggle for life. Such a task would 

 probably demand a more vast and minute knowledge of the 

 details of the organic world than it is as yet possible for the 

 most unremitting industry, inspired by the highest genius, 

 to acquire. Yet no one doubts the general principle that it 

 is natural selection which determines, not only which races 

 shall prevail, but also which races shall vary and which shall 

 remain unmodified. So in dealing with human societies, in 

 the primitive era with which the present discussion is chiefly 

 concerned, the historic data are insufficient to enable us to 

 ascertain the precise circumstances to which the prevalence 

 and the improvability of certain races are to be attributed. 

 N^evertheless we can here, too, point out sundry general 



VOL. II. S 



