"Blossom Hosts Jjui Insect Guests 



Fig. 5. 



where the pollen ripens before the stigma matures. 

 In many cases, too, the pollen-bearing stamens are all 

 in one flower, while the pistils are in 

 others, as in the cucumber and Indian 

 corn, and in some instances, as in the 

 palms and willows, on different plants. 

 A purely mechanical method of pre- 

 ventinof self-fertilization is shown in 

 Fig. 5. These phenomena are, however, 

 all of one class. When the method of 

 fertilization in one case is understood, it may be 

 applied to all. 



Under the conditions presented at E and F, if the 

 insect did deposit the flower's pollen on its own 

 stigma, the pollen would fail in its mission, since 

 the stigma would not yet have reached the point 

 where it could make use of the gift. The reverse 

 problem also faced him, the pollen often being 

 retained within the anthers until the stigma withers. 

 In spite of these facts, he saw that this class of 

 flowers continued to exist and to produce seed, but 

 their method of doing it steadily eluded him. 



In addition to this weakness in SprengeFs theory, 

 there was yet another defect. Sprengel could give 

 no reason why insect fertilization was more desirable 

 than self-fertilization. Why should the flowers be 



