32 THE VARIATION OF ANIMALS IN NATURE 



knowledge of the origin of germ cells suggests that they are not, 

 at least in their early stages, likely to be immune from in- 

 fluences affecting the somatic tissues, inasmuch as they are, 

 in many cases, morphologically indistinguishable from the 

 latter (cf. also Gatenby, 191 6). However, the fact remains 

 that no mechanism by which a true Lamarckian effect could 

 be brought about has as yet been demonstrated. It is very 

 easy to imagine that a new habit or a far-reaching somatic 

 modification involving both structural and physiological re- 

 organisation and readjustment might have a profound effect 

 on the constitution of an animal. But the proof is still 

 lacking that such readjustment would have a specific effect 

 on the hereditary material of such a kind that the original 

 somatic modification was reproduced. 



It is customary to attach very great importance to the 

 experimental evidence on this subject. Now the value to be 

 set on experiment in such a matter is open to some doubt. It 

 has the unfortunate limitation of being incapable of dealing 

 (as Caiman (1930) has pointed out) with the historical back- 

 ground of animal morphogenesis. This question becomes 

 crucial when we consider the negative evidence brought 

 forward to disprove the inheritance of induced variation. If 

 such-and-such a stimulus repeated for a few months or a few 

 years on a few generations fails to modify the germ cells, is 

 there any reason for assuming that it will have no effect if 

 the stimulus is applied, as it may well be in nature, for many 

 years and decades and over innumerable generations? We 

 cannot point to any case in which the duration of an induced 

 effect is proportionate to the time-intensity of the stimulus ; but 

 that such a contingency is possible ought not to be ignored and 

 negative results have to be accepted subject to this reservation. 



Before considering the experimental evidence we shall 

 briefly set out what appear to be the essential conditions 

 for a really convincing experiment. It is one of the mis- 

 fortunes of the controversy that so much of the evidence is 

 equivocal. The following are, we believe, the necessary 

 precautions to ensure definite results. 



1 . The Use of Inbred Stock. — In our section on natural varia- 

 tion (Chapter IV) we show how often species consist of a 

 mixture of strains. It is the universal experience of those who 

 breed animals under artificial conditions that inbreeding for 



