ON PHYSIOLOGICAL MORPHOLOGY. 51 



grown together. Of course, it often happened that, by 

 accident or rapid movement, the twins were separated, and 

 they then developed into perfectly normal single embryos. 

 Since we cannot assume that in every case the same part 

 of the protoplasm escapes, we must conclude that every 

 part of the protoplasm may give 

 rise to fully developed embryos 

 without regard to preformed 

 germ-regions. In many eggs a 

 repeated outflow of the proto- 

 plasm takes place (Fig. 15). In 

 such cases each of the drops 

 of the protoplasm may give rise 



to an embryo, and I obtained not only double embryos, 

 but triplets and quadruplets all grown together. 



In order to understand these experiments more fully, let us 

 follow out the history of development in these double and 

 multiple monstrosities. The ova were put into the diluted 

 sea-water before any segmentation had taken place and while 

 they had still but one nucleus. When parts of the protoplasm 

 flowed out, the nucleus either remained in the protoplasm, 

 inside of the membrane, or passed out with the part that was 

 ejected. Therefore, at first, only one part of the protoplasm 

 contained a nucleus. The other part obtained its nucleus by 

 the cleavage which took place, as follows : In case the nucleus 

 had remained inside the membrane (Fig. 16), the first cleavage 



took place inside the membrane, the 

 cleavage plane being always at right 

 angles to the common diameter of 

 the two protoplasmic drops (Fig. 16). 

 (Only when the nuclear-spindle had 

 already formed, at the time of the 

 FlG> l6 * bursting of the membrane, were 



there exceptions to this rule.) A peculiarity of the first 

 cleavage was that the protoplasm was always divided into a 

 larger sphere (I, Fig. 16), and a smaller sphere (II) which was 

 connected with the extra-ovate. 



