April 1892.] THE BOTANICAL SOCIETY OF EDINBURGH. 381 



tion. An actual fusing of the two nucleoli is seen in fig. 

 36^ ; they have met in the micropylar end of the already 

 fused nuclei, and this formation of the new nucleolus has 

 been completed in figs. 37, 38. 



Synchronously with the blending of the two nucleoli into 

 the nucleolus of the primary endosperm-nucleus a new 

 structure makes its appearance. This body (n. h.) is seen in 

 figs. 37-43, and corresponds, I believe, to the nucleolar 

 membrane of the antipodal nucleolus. It usually lies in 

 close contact with the nuclear membrane of the primary 

 endosperm-nucleus, and shows, in its earlier stages, a finely 

 dotted or granular appearance, but gradually becomes more 

 homogeneous looking. I should have traced its origin more 

 minutely but for the rapidity with which the two nucleoli 

 fuse. This fusion occurs so rapidly that the preparation 

 from which fig. 36^ was drawn only came into my possession 

 after two years of patient searching. The newly formed 

 nucleolus gradually descends towards the antipodal half of 

 the endosperm-nucleus and ultimately occupies its centre 

 (figs. 37-44). 



Before attempting to give an explanation of the phenomena 

 of fertilisation as seen during the act of conjugation, an ex- 

 planation bound up as it is in a knowledge of the minute 

 structure of the nucleus and nucleolus, let me briefly 

 summarise the stages in the formation of the endosperm- 

 cell, as follows : — 



1. Amoeboid movements of the cell-plasmata of the two 



primordial cells leading to conjugation of the cell- 

 plasmata. 



2. Approximation of nuclei and formation of para- 



nucleoli. 



3. Flattening of nuclei on contact, and absorption of 



intervening nuclear membrane. 



4. Conjugation of nuclei. 



5. Approximation of nucleoli. 



6. Conjugation of nucleoli and casting off of (male ?) 



nucleolar bag. 



Yet another body or bodies are seen a short time after the 

 completion of the conjugation, for we find in fig. 38, two 

 bodies x and y, of which the former is clear and hyaline, 



