horizontal and near enough to the surface to permit the opening 

 of the flowers in the air. 



Turning now to the cleistogamous pollination one finds a rela- 

 tion of parts and a series of events of considerable interest. As 

 previously reported, seeds are produced abundantly by flowers 

 at any depth below the surface. Dissection of the mature but 

 unopen flowers show^ed that the pollen tubes from spores that 

 had germinated in the anthers had already entered the stigma. 

 Sections show that all the pollen usually germinates though one 

 or two cases were found in which the content of but one anther 

 had germinated at the time the material was collected. In every 

 flower examined at the time it Avas opening in the air it was found 

 that self pollination had already taken place. Such a condition 

 makes cross pollination impossible in any event and means that 

 the whole floral display is now meaningless, — a memory of other 

 days. 



Longitudinal sections through the flower show the relations of 

 all parts concerned. During the development of the flower the 

 tip of the stigma remains considerably below the upper end of 

 the anthers (Fig. 1) though it lies in contact all the time with 

 the inner side of these parts. As the flower matures the style 

 elongates and the stigma is shoved up forcibly into contact with 

 the inner side of the anthers at their upper ends (Fig. 2). In 

 this region occurs a slight break in the wall on each side of the 

 anther through which the stigmatic hairs come into contact with 

 a few of the pollen grains. These openings are limited to the 

 upper one-tenth of the anther at this stage, and may be due to 

 downward pressure on the anther-beaks as these press against the 

 closed sepals. The preserved material Avas slightly shrunken so 

 that one could not be sure from the sections what parts might 

 come into contact or what tensions might be set up in the unopen 

 flower. 



The uppermost pollen grains flrst germinate and the stimulus 

 is gradually transmitted downward until all the spores in the 

 given pollen sac have formed tubes. Some of these pollen tubes 

 pass out through the openings at the upper end of the anther 

 and enter the stylar canals. The tubes are numerous and in 

 favorable sections look like a rope of strongly staining threads 

 as they lie in these cavities. The anthers are flrmly cemented to 

 the stigma by these pollen tubes ; the entire content of the pollen 



52 



