REPRODUCTION AND HEREDITY 307 



tary substance, and then with the same precision to an 

 amazingly complex, unequal distribution of the germ-plasm. 



In his earlier writings Weismann stated that the first division 

 of the egg-cell is heterokinetic (erbungleich) , and leads to a 

 separation of germ-plasm from soma. If this were true it would 

 naturally follow that only one of the first two segmentation- 

 grooves, the stem-mother of the later sex-cells, contains the 

 entire hereditary substance, while the other, the somatic 

 cells, has already suffered a decrease in the original store of 

 determinants. If it were possible to separate these first two 

 blastomeres artificially without injuring the vitality of the cells, 

 and force them into a separate independent development, the 

 ' primordial cell ' might perhaps have prospects of developing 

 into a complete organism, but the somatic cell would at the 

 best be only able to produce partial formations. But numerous 

 experiments made with ova of medusae, echinoderms, tunicates, 

 the primitive fish Branchiostoma, and higher vertebrates show 

 that isolated segmentation-cells produce only small embryos, 

 proportionate to the smaller quantity of nutritive material. 

 Moreover, the first two blastomeres produce under favourable 

 circumstances not only two complete larvae, but in many 

 cases, for instance, in medusae and others of the lower organ- 

 isms, each of the four, eight, sixteen, and even thirty-two 

 first segmentation-cells is able to develop into a complete 

 embryo. This proves that, at any rate up to this stage, the 

 division of the hereditary substance has been equal, and that all 

 embryo-cells possess the entire hereditary substance. 



Apparently contradictory results are shown by certain experi- 

 ments made with the ova of annelids, snails, mussels, and jelly- 

 fishes. In particular in the Ctenophores, complete larvae are 

 easily distinguished from part-formations by the number of 

 ciliated ribs. In the normal larvae which have been bred from a 

 complete ovum the number of ciliated ribs is eight. If the first 

 two blastomeres are isolated they continue to develop into 

 embryos, but these possess only four ciliated ribs : and if we 

 finally succeed in inducing blastomeres of the eight-cell stage to 

 develop separately the larvae possess always only eight ciliated 

 ribs between them. 



Weismann regards this result as a proof of heterokinesis, 



