1900.] on Facts of Inheritance. 349 



Discovering no natural way of accounting for this, the majority fell 

 back upou a hypothesis of hyperphysical agencies, that is to say, they 

 abandoned the scientific method and drew cheques upon that bank 

 where credit is unlimited as long as credulity endures. 



An attempt to solve the difficulty which confronted the preforma- 

 tionists — the difficulty of accounting for the complex organisation 

 presumed to exist in the germ-cell — is expressed in a theory which 

 seems to have occurred at intervals in the long period between 

 Democritus and Darwin, the theory of pangenesis. On this theory, 

 the cells of the body are supposed to give off characteristic and 

 representative gemmules : these are supposed to find their way to the 

 reproductive elements, which thus come to contain, as it were, 

 concentrated samples of the different components of the body, and 

 are therefore able to develop into an offspring like the parent. The 

 theory involves many hypotheses, and is avowedly un verifiable in 

 direct sense-experience, but the same might be said about many 

 other theories. It is perhaps more to the point to notice that there 

 is another theory of heredity which is, on the whole, simpler, which 

 seems, on the whole, to fit the facts better, especially the fact that our 

 experience does not warrant the conclusion that the modifications or 

 acquired characters of the body of the parent affect in any specific 

 and representative way the inheritance of the offspring. 



As is well known, the view which many, if not most biologists 

 now take of the uniqueness of the germ-cells is rather different from 

 that of pangenesis. It is expressed in the phrase " germinal con- 

 tinuity," and has been independently suggested by several thinkers, 

 though Weismann has the credit of working it out into a theory. 

 Let me recall its purport. There is a sense, as Mr. Galton says, in 

 which the child is as old as the parent, for when the parent's body is 

 developing from the fertilised ovum, a residue of unaltered germinal 

 material is kept apart to form the future reproductive cells, one of 

 which may become the starting-point of a child. In many cases, 

 scattered through the animal kingdom, from worms to fishes, the 

 beginning of the lineage of germ-cells is demonstrable in very early 

 stages before the differentiation of the body-cells has more than 

 begun. In the development of the threadworm of the horse, 

 according to Boveri, the very first cleavage divides the fertilised 

 ovum into two cells, one of which is the ancestor of all the body- 

 cells, and the other the ancestor of all the germ-cells. In other 

 cases, particularly among plants, the segregation of germ-cells is 

 not demonstrable until a relatively late stage. Weismann, generalis- 

 ing from cases where it seems to be visibly demonstrable, maintains 

 that in all cases the germinal material which starts an offspring, 

 owes its virtue to being materially continuous with the germinal 

 material from which the parent or parents arose. But it is not on a 

 continuous lineage of recoguisable germ-cells that Weismann insists, 

 for this is often unrecognisable, but on the continuity of the germ- 

 plasm — that is of a specific substance of definite chemical and 



