— 93 — 



dance of self-fertilization. In the absence of information 

 regarding the kind of insects which visit the flower, one is 

 tempted to trj bj analogy to discover the method of fertili- 

 zation. As raay well be supposed self-pollination cannot take 

 place. The obliquity of the position of the stigma precludes 

 the possibility of lts occurrence. Protandry seems not to 

 occur. If it does occur, it is not well-marked. Insects are no 

 doubt the agents employed in fertilization. They must either 

 remove the pollen from the anthers, or have it shaken on 

 their head or body. It is possible that the insect visitant, 

 when endeavouring to reach the nectary, may pass its head 

 between the filaments where the space is widest, then press 

 it downward into the narrower part, and be so far 

 caught that an effbrt raay have to be made to get release, 

 with the result that some pollen may be shaken out of the 

 anthers. 



The next step is more easily imagined. The insect, 

 guided, let us suppose, by the bright yellow anthers, will 

 doubtless fly directly towards the centre of the flower. If 

 the style be bent in the direction of the lineofflight, the 

 stigma has much chance of being struck, it being relatively 

 inconspicuous. If, ho\vever,the style be bent in the opposite 

 direction, the stigma will be missed, and fertilization will 

 not be accomplished. Pollen may be carried off in either 

 case. 



Theelement of chance, evidently existing in the mode of 

 pollination is a matter which should be carefully kept in 

 mind. 



It was thought that an enumeration of the flowers with 

 the styles bent either to right or left might afford interesting 

 statistical information. The following table shows the pro- 



