REPRODUCTION AND HEREDITY 309 



burdened with the heavy food, is, according to the laws of gravi- 

 tation, always pointing downwards, whilst the dark animal 

 pole, containing the light formative yolk, is pointing upwards. 

 The entire yolk substance is, therefore, distributed in the ovum 

 bilatero-symmetrically, and, according to all appearances, this 

 arrangement regulates the course of the ' division planes.' 



As long as this semi-lateral distribution of the yolk is pre- 

 served in the surviving segmentation-cell by the pressure of the 

 egg-half which has been killed, it seems that only a half-forma- 

 tion can develop from it. But if we turn the segmentation-cell 

 in such a manner that the yolk-substances become arranged 

 as they were in the normal ovum, it develops into a complete 

 embryo. Herlitzka's experiments with ova of the salamander 

 point to the fact that in the semi-lateral distribution of the yolk 

 is found the real reason for the imperfect development of the 

 embryo. By means of an ingeniously constructed apparatus 

 Herlitzka succeeded at the two-cell stage in constricting the 

 ovum in the ' division plane,' by means of a fine cocoon thread, 

 and making both blastomeres completely independent of one 

 another. They were afterwards able to round off, the yolk dis- 

 tributed itself in them according to its specific gravity exactly as 

 in the complete ovum, and the two isolated blastomeres developed 

 consequently into two normal embryos. 



These experiments proved unambiguously that at least in 

 these cases it is not a deficiency in the nuclear substance but the 

 structure of the egg-plasm which must be held responsible for 

 the appearance of part*formations. That an analogous process 

 goes on in the case of the ova of the Ctenophores is shown by 

 experiments made by Driesch and Morgan, in which larvae with 

 a diminished number of ribs were obtained from an unsegmented 

 ovum merely by removing a considerable part of the outer plasm. 

 Although, therefore, the ova possessed their full nuclear material, 

 and consequently the ' determinants ' for the full number of ribs, 

 these ' rudiments ' (Anlagen) were unable to develop owing to the 

 deficiency in plasm. 



Weismann himself has recently admitted that in many animals 

 the first divisions of the ovum are homoeokinetic, and that only 

 later the ' germ-paths ' become separated from the soma. 



A certain amount of corroboration of the doctrine of hetero- 



