— 100 



formation. In certain ovules I found stages which I have brought 

 together in fig. IX and which probably may be explained in the 

 following manner. There has been a typical reduction division, 

 but after the second division no cell wall has been formed between 

 the daughter nuclei (^4). These come nearer together (B) and 

 finally unite (C). Especially the behaviour of the nucleolus (D) 

 supports this supposition. Such figures, however, are so rare, that 

 this form of ES-formation cannot be considerad to be of great 

 importance for embryo formation in this species. I hope to be 





A 



B 



Fig. X. Polyembryony in H. excellens. 



able to explain this more definitely with more abundant and sui- 

 table fixed material. Perhaps we have here a case similar to that 

 in parthenogenetic eggs in certain insects, where the third polar 

 body remains in the egg and becomes united with the egg nucleus. 

 I have also investigated a form of H. Pilosella which is always 

 sterile in Copenhagen (Ostenfeld 10, p. 240). A cytological 

 investigation upon the ES-formation showed that a typical reduc- 

 tion division proceeds, and that the lower cell of the tetrad be- 

 comes the embryo sac in the ordinary way. In this species I have 

 never seen aposporic ES-formation. The embryo sac becomes 



