374 . BOTANICAL GAZETTE [November 



cell plates appear on the spindle fibers but do not persist, so that the 



remain free m the cytoplasm 

 a bv two successive divisions fr 



. 10, n). 



in which the heterotypic prophases occur, they are to be regarded 

 as megaspore nuclei, and any one of them is thus the morphological 

 equivalent of the nucleus shown in fig. 3. By one further division 



megaspore 



from a single megaspore. An 

 : is organized; the antipodal 



may or may 



meet in the vicinity 



The various stages described in the foregoing paragraph may 

 distinguished from those in the development of a sac from a 

 3jle megaspore by the absence of disorganized cells at the micro- 



pylar end. 



become indistingui 



from the disorganized contents of the epidermal cells of the ! 

 so that it is then unsafe to use them as evidence, but there 



ling against the assumption that the fate of the embi 

 same whether it has been derived from one megaspore 



from 



This 



In this Epidendrutn, as in Epipactis, two megaspores evid 

 take part in the formation of the embryo sac in a few cases, 

 condition results when the division of the megaspore mother cell is 

 very unequal and that of the inner daughter cell equal, the separat- 

 ing wall at the second mitosis being ephemeral. 



The pollen tube enters the sac, disorganizes the two synergids, 



male 



the 



erm 



fuses with the two polars (fig. 13)- The 

 nucleus formed by the latter fusion undergoes no 

 division, but degenerates along with the three antipodal nuclei 

 (fig. 14). 



The first few divisions of the fertilized egg are transv 



f 



embryo of a varying number 



come in. and for a time 



very regul 



Fig. 1 6 represents 



verrncosum, in which the number 

 very high, forming a filament of a 



