and around tlie chalaza there is a snail franrnont of the 

 nucellus left in the ripe seed. This is never stored 

 with food materials, and so can not be called peri- 

 sperm (Fi^. 47). The macrospore is buried about as 

 deeply as in Hamanelis, It f^errninates only in the low- 

 er ovules, the upper ones never [getting typical enbryo- 

 sacs and being less developed pror;ressively toward the 

 top of the ovary. In the sterile ovules, the cells of 

 the outer integument become very much enlarged and at 

 last empty. T?ie substance of the necellus is absorbed, 

 and the ovules become polygonal bodies, resembling sa'.r- 

 dust, which fill the upper part of the ovary. 



The outer integument of the fertile ovule (Fig, 

 47) grows into a wing. The embryo is straight, and 

 bears about the same relation to the amount of endo- 

 sperm as in Hamamelis, 



