94 



FERTILISATION. 



Pollination takes place, probably mostly through the action 

 of the wind, though it may be that insects play a role in it. 

 The pollen is caught by a drop of fluid at the tip of the in- 

 tegumental stigma and by evaporation of this fluid it is gra- 

 dually drawn inwards until it reaches a slight cavity, a kind 

 of pollenchamber at the top of the Qucellus. This pollenchamber 

 owes its origin to the degeneration of the cells at the top of 

 the nucellus (comp. fig. 35 PI. V and fig. 18 PL III). In this way 

 a place particularly well fitted for germination is formed. A 

 number of pollengrains is here collected; they soon send their 

 poUentubes into the nucellar tissue directing them towards the 

 embryosac. The nucellar tissue is full of starch, which doubtless 

 serves as food to the growing pollentube as it gradually disap- 

 pears around and at some distance in front of it. From this 

 observation, T think, one may conclude that the pollentube 

 secretes some starch-dissolving ferment. 



The question of the origin of the pollentube-nucleus and of 

 the generative nuclei will be kept for some future paper, it 

 will sufiice here to state that before reaching the embryosac 

 three nuclei can be distinguished in it viz. one vegetative and 

 two generative ones. As little as Karsten I could distinguish 

 a generative cell, though Karsten did in the case of other spe- 

 cies and I did in pollentubes of Gnetum Mollucanum Krst. 



One or more pollentubes may finally reach the embryosac, 

 they may perforate its wall at the top (fig. 33, 34 PI. V, fig. 36, 

 PLY, fig. 37, 39, PL YI, or at the side fig. 43, PL YII), but always at 

 some place of the upper fertil part, never at the lower steril one. 



The penetration occasionally occurs, as has been allready 

 stated, at a point between two symmetrically arranged reduced 

 embryosacs (fig. 36 PL Y), these latter then produce the ap- 

 pearance of synergidae. 



There may now be formed a more or less circular hole in 

 the wall of the pollentube (fig. 42, PL YII), allowing the con- 

 tent to escape but this is rather the exception. 



