ON THE CAUSES OF VARIATION. 491 



fare of the species. In short, Darwin's views kept pace with the 

 investigations of his day, and tended in the direction of restrict- 

 ing rather than widening the influence of natural selection. But, 

 as Romanes, and especially Spencer, in his " Factors of Organic 

 Evolution," have fully shown Darwin's position on this subject, I 

 may pass over the detail. 



Internal Conditions. — Physiological. — Genesis itself is the 

 first and most fundamental of all causes of variation. The phi- 

 losophy of sex may, indeed, be sought in this differentiation, as 

 the accumulated qualities in separate entities when suddenly con- 

 joined or commingled inevitably lead to aggregation and hetero- 

 geneity — in other words, to plasticity or capacity to vary. Genesis, 

 as a fundamental factor in evolution, may be more intelligently 

 considered under some of its subordinate phases, as heredity, phys- 

 iological selection, sexual selection, primogenital selection, sexual 

 differentiation, including philoprogeneity, hybridity, etc. 



Heredity, as expounded by the ablest biologists, and as ex- 

 emplified in life, is a puissant factor in evolution, and, though 

 essentially conservative, must, through the marvelous power of 

 atavism, tend to increase individual variability. The subject has 

 been too well considered by Darwin and his followers to justify 

 further discussion of it here. As a cause of variation, heredity 

 must, however, have less and less influence as we go back in the 

 scale of organized beings ; for it can not well come into play in 

 agamic or fissiparous reproduction, a fact which has given the 

 abiogenisists one of their strongest arguments, since it is difiicult 

 to understand how, for instance, the monera of to-day could have 

 descended without change from the primordial form. 



Physiological Selection. — Physiological selection, as suggested 

 by Mr. Catchpool and as expounded by Romanes, is undoubtedly 

 a veritable factor in evolution, and, while giving us another link 

 in the chain of evidence as to the causes of differentiation, lessens 

 in but very slight degree the overwhelming force of the argument 

 for natural selection. It adds, rather, an important element in the 

 evidence therefor, and may be classed as a subordinate cause of 

 differentiation. Romanes's theory is based upon the argument 

 that differences, such as constitute varieties and species in their 

 commencement, would not be preserved by natural selection un- 

 less useful, but would be lost again by cross-breeding with forms 

 like the parent, and which had not varied, except upon some 

 hypothesis like that of physiological selection. This could not 

 be prevented except by migration. This difficulty is a general 

 one, was argued by Darwin himself, and has been felt by all Dar- 

 winians. The reproductive organs are extremely variable, and 

 sterility may occur not only between species, but between races 

 and varieties, and often between individuals. Physiological selec- 



