ON VARIATIONS. 365 



produced it in the offspring also. This could not have 

 been the case in WestphaPs experiments, however, as 

 in them no wound at all was made. 



How, then, can this apparent transmission of ac- 

 quired characters be accounted for? Our hypothesis 

 of internal secretions supplies a very simple explana- 

 tion. Thus the secretions from the brain of an epi- 

 leptic guinea-pig, no matter how this epilepsy had been 

 produced, would almost certainly be abnormal. Even 

 supposing that they were without effect on the " de- 

 terminants ' ' of the nervous tissues in the germ-plasm, 

 it is a very probable supposition that they might so 

 affect the growth of the nervous tissues of the offspring, 

 during intra-uterine development, as to provoke a 

 similar abnormal condition in them. 



In mammals and other viviparous animals, it is prob- 

 able that changed conditions of life produce part of 

 their cumulative action during the period in which the 

 embryo is under the influence of the maternal fluids. 

 It is of course possible that all of the cumulative effect 

 is then produced, though in such a case we should have 

 to find some other explanation than that given above of 

 the cumulative effects noticed in oviparous animals as 

 Polyommatus phlceas. In the case of the gradual de- 

 generation of the pure bred dog under an Indian 

 climate, for instance, the environment may so act upon 

 the maternal parent as to produce slight changes in the 

 body tissues, and also to alter the character of the 

 secretions and excretions. These, acting on the off- 

 spring during their embryonic development when, 

 as we have seen in a previous chapter, the tissues 

 are extraordinarily sensitive to their environment 



