14 INJURIOUS EFFECTS OF PREMATURE POLLINATION. 



pollinated till June 14, at which time the corrollas of the flowers were 

 dead or badly withered, while the pistils were still fresh and receptive. 

 On June 20 live pods were found growing, but they did not appear as 

 thrifty as those of experiment 16 in which the pistils were pollinated at 

 the proper time. In experiment 9 the pistils became receptive May 13, 

 but were not pollinated until the next day, when pollen that had been 

 left in the laboratory for three days was applied. The stigmas at the 

 time of pollination were covered with a large drop of stigmatic fluid, 

 in which the pollen seemed to dissolve and disappear from view. The 

 experiment resulted in the setting of but three pods. It is thought 

 that the reason for so poor a percentage of seed pods is the placing of 

 the dry pollen in so large a drop of stigmatic fluid, perhaps causing the 

 pollen grains to swell and burst; but this is only conjecture, for no 

 attempt was made to ascertain the true behavior of the pollen grains. 



In experiment 37 the pollen was placed on about one-fourth the 

 entire surface of receptive stigmas and on only one side of the small 

 suture tnat divides the stigma into equal parts. While the pods were 

 still green four or five showed a marked one-sidedness, but when they 

 had ripened and were cut for examination the} T all seemed as S} T m- 

 metrical as the general crop of pods. On cutting them, however, the 

 good plump seeds were almost altogether in one-half of the ovary, 

 the other half containing light, chaffy, undeveloped seeds. 



It being fully determined that pollen applied to the entire stigmatic 

 surface of premature pistils results in the falling of the flowers, some 

 tests were made to determine the effects of smaller quantities of pollen. 

 In experiment 11 fresh pollen was applied to one-fourth of the stig- 

 matic surfaces of pistils that would have been receptive in three days, 

 and the flowers were bagged without emasculation. Four days later 

 but one flower had fallen, while it is safe to say that had the stigmas 

 been entirely covered with pollen 90 per cent would have fallen. They 

 were naturally self-pollinated and 8 set pods. No explanation is known 

 why an occasional prematurely pollinated flower failed to fall (experi- 

 ments 39 and 10) unless it be that accidentally a portion of the stig- 

 matic surface escaped pollination and the flowers continued to grow 

 because there was still a chance for them to set some seeds. Experi- 

 ments 41 and 46 (see table) were treated as was experiment 41. except 

 that the flowers were emasculated. But two pods set in experiment 

 44 and four in experiment 46; the other flowers withered and fell, not 

 because a small quantity of pollen was applied prematurely, but because 

 of a lack of pollination after becoming receptive. In experiment 37 

 the pollen was placed on only one side of the medial suture and seeds 

 were found mostly in one-half of the pods, while in experiments 44 

 and 46 it was placed across the stigma in a line at right angles to the 

 suture, and the well-developed seeds were found scattered among the 

 undeveloped ones in both halves of the seed pod. From the partition 

 which divides the ovary (PI. Ill, rig. T) a sharp line can be traced up 



