DEVELOPMENT BY NATURAL SELECTION. 107 



every part of the structure, every molecule, and every 

 atom, is involved sooner or later in the process of 

 discarding. 



In recent times most biologists have been content to 

 leave the main problem of heredity on one side as too 

 difficult for solution at present. But the attempts which 

 have been made to present some sort of solution on 

 mechanistic lines are at least instructive. 



There was first the " box - within - box ' theory, 

 according to which the complete structure of the adult 

 organism is contained hidden away in the embryo, and 

 within this hidden structure equally complete structures 

 of all future embryos. This grotesque theory need not 

 detain us further, though it is certainly the logical 

 outcome of the mechanistic theory of life. 



To come to recent times, an attempt was made by 

 Weismann to simplify the problem by denying the 

 transmission of characters acquired by the organism 

 itself, as distinguished from its germ-cells. Thus simpli- 

 fied, the problem was to see why one germ should be so 

 exactly similar to the germ from which it originated ; 

 and this gave rise to Weismann's theory of the continuity 

 of the germ-plasma. He maintained that the germ- 

 material of one generation is directly descended, without 

 intermediate forms, from the germ-material of the 

 preceding generation, and that this helps us to account 

 for germ-cells of the same strain being so similar. He 

 has to admit that the germ-material must have an 

 immensely complex structure, since it determines not 

 only the whole complexity of structure in the adult 

 organism, but also the complexity of the process of 

 development. Nevertheless he proceeds to call the 



