HEREDITY. 73 



An early isolation of reproductive cells, directly con- 

 tinuous and therefore presumably identical with the original 

 ovum, has been observed in the development of- some 

 " worm types " (Sagitta^ Thread-worms, Leeches, Polyzoa), 

 and of some Arthropods (e.g. Moina among Crustaceans, 

 Chironomus among Insects, Phalangidae among Spiders), 

 in Micrometrus aggregatus among Teleostean fishes, and 

 with less distinctness in some other animals. A cell which 

 will give rise to the germ-cells can be recognised in the 

 gastrula stage of Cyclops, and in the very first segmentation 

 stages of the thread-worm Ascaris. 



In many cases, however, the reproductive cells are not 

 recognisable until a relatively late stage in development, 

 after differentiation has made considerable progress. 

 Weismann gets over this difficulty by supposing that the 

 continuity is sustained by a specific nuclear substance 

 the germ-plasm which remains unaltered in spite of the 

 differentiation in the body. It is perhaps enough to say 

 that, as all the cells are descendants of the fertilised ovum, 

 the reproductive cells are those which retain intact the 

 qualities of that fertilised ovum, and that this is the reason 

 why they are able to develop into offspring like the 

 parent. 



Finally, it may be noticed in connection with heredity, 

 that there is great doubt to what extent the " body " can 

 definitely influence its own reproductive cells. Animals 

 acquire individual bodily peculiarities in the course of 

 their life, as the result of what they do or refrain from 

 doing, or as dints from external forces. The "body" is 

 thus changed, but there is much doubt whether the repro- 

 ductive cells within the " body " are affected specifically by 

 such changes. Weismann denies the transmissibility of 

 any characters except those inherent in the fertilised egg- 

 cell, and therefore denies that the influences of function 

 and environment are, or have been, of direct importance 

 in the evolution of many-celled animals. Such influences 

 affect the body, and produce what are technically called 

 "modifications? but these modifications do not affect the 

 reproductive cells at least not in a specific representative 

 way. Therefore modifications are not likely to be trans- 

 mitted, and there seems no good evidence to show that 



