58o NATURAL SCIENCE. oct.. 



by what means it has been limited, by what means the hmitation has 

 been progressively narrowed, and by what means the narrow limita- 

 tion is now maintained. Heredity, in fact, I propose to treat merely 

 as the narrow limitation of variation, and I hope to be able to show 

 that if no such phenomenon as heredity occurred in early stages of 

 evolution, it would of necessity arise by degrees through the action 

 of well-known influences, and without the intervention of any 

 unknown force or gemmule. 



At the outset, I am hampered by a difficulty in terminology. 

 " Natural Selection " is a term so familiar in the Darwinian sense, 

 that were it not that I believe a change in the meaning of the term 

 would add to its usefulness, I should not venture to use it at all. One 

 essential factor of " Natural Selection " (in the Darwinian sense) is 

 heredity ; and hence natural selection in that sense cannot rightly be 

 invoked to explain the origin of one of its own factors. 



Heredity being established, variation being an observed 

 phenomenon, and the struggle for existence being in progress, then 

 " Natural Selection," in the Darwinian sense, will account for the 

 origin of new species. Such is, I believe, essentially the view held 

 by all Darwinians, but I find it necessary, for present purposes, to 

 eliminate the element of heredity entirely from the meaning of the 

 term, and to use the term, so altered, to express the survival of the 

 fittest, even before heredity existed, and independently of heredity. 

 The change in the meaning consists in what is, I believe, called by 

 logicians a limitation of its intension, and consequent enlargement 

 of its extension : it renders the term applicable to a group of 

 phenomena not previously included within its signification, but in no 

 way affects its application to the phenomenon usuallj' suggested by 

 it, except, perhaps, to give it increased accuracy. 



It is to Natural Selection in this wider sense, to the survival of 

 the fit before heredity existed, that I now seek to refer the origin of 

 heredity, to the survival of the fittest that I seek to refer the main- 

 tenance of those limitations of variation which collectively constitute 

 heredity. 



The terms " offspring," " generation," and " parent," for con- 

 venience, may be applied even in the case where multiplication is 

 effected solely by fission. I shall speak of the two Aviceb(S resulting 

 from the fission of a pre-existing one as the " offspring " of that from 

 which they arose, and the latter I shall speak of as the " parent " 

 and as belonging to the previous " generation." 



In the earliest stages of the evolution of living things, whatever 

 those stages may have been, there is no necessity, and, it appears to 

 me, no justification for the aesumption that heredity existed or 

 occurred. A living mass might well differ in its various parts, and, 

 if so, the separation of two successive portions from one larger parent 

 mass need not necessarily involve the likeness of those separated 

 portions to each other in every detail of structure and constitution, 



