274 INHERITANCE OF ACQUIRED CHARACTERS. 



tion reared under normal conditions, then the changes must be 

 definitely hereditary. It is claimed that the altered conditions 

 acted directly on the germ-plasm, which then altered the soma ; 

 such an argument is inadmissable, since the two substances are 

 already segregated at the time when the experiments com- 

 menced. 



There is. moreover, a most important theoretical oljjection to 

 the theory of germinal continuity — one which Weismann has 

 already pointed out — viz., the difficulty of explaining why the 

 germ-plasm should alternate between periods of germ-cell for- 

 mation and periods of soma formation. Weismann established 

 a theory as to the nature of polar body extrusions to account 

 for this cyclical development, but he had to abandon his theory 

 when it Avas demonstrated that spermatozoa passed through an 

 homologous process. 



It may also be added that the theor}- of germinal continuity 

 is quite devoid of experimental proof. I am not aware that 

 anyone has already pointed out in what manner such jjroof may 

 be given, but it seems to me that the truth or otherwise of the 

 theory should be capable of demonstration. If, for example, 

 an organism be brought under new conditions and it ac(|uire 

 new characters which are not 'hereditary, will the hereditary 

 characters appearing at some later stage be in the same direc- 

 tion as the acquired character, or will they be unconnected with 

 such characters? We should expect that a certain environment 

 acting on the soma will result in something different from what 

 AA'Ould he produced by the action of that same environment on 

 a different substance, viz., the (jcnn- plasm. Has such proof 

 been advanced? On the contrary, apart from sudden mutations, 

 not referable to changes in the environment, it is always the 

 acc|uired character Avhich later on is found to be fixed in the race 

 {vide experiments mentioned above). 



It may perhaps be thought that the occurrence of mutations 

 in nature is a case such as is called for above ; that an animal 

 or plant living under certain conditions varies about a mean 

 as a result of the interaction between the environment and the 

 soma, but that such animals or plants occasionally throw sports 

 Avhich may be considered to be the result of the interaction of 

 the environment and the germ-plasm. To such a view one must 

 answer that occurrences in nature cannot be adduced in i:)lace 

 of controlled experiments, since we know too little about the 

 varying conditions in the former case ; that mutations apj^arently 

 only dift'er from fluctuations in that they result from early inter- 

 ference with normal development; that the necessary experi- 

 ments should be made on successive generations at the same stage 

 of development. 



But if we are to suppose that the substance or substances in 

 virtue of which development takes place is not continuous — that 

 is, that it is formed an.ew for each generation — how are we to 

 imagine that it occurs? 



The germ-cells are without doubt highly differentiated cells. 



