EXPERIMENTS WITH TOBACCO BLOSSOMS. 13 



experiments, for example Nos. L8 and L9, which are described in 

 Table I (p. 17), prove that the falling of the young flowers is not in 

 any way caused by the early emasculation, while in experiments 35, 

 38, and 4<> they fell soon after being prematurely pollinated, although 

 they had not been emasculated. 



A study of tigs. 1 and 2, PL I, will also show that the falling of the 

 flowers is not associated with the emasculation. For tig. I the photo- 

 graph was taken immediately after emasculating and pollinating the 

 flowers, and for tig. 2 the photograph was taken forty-eight hours later. 

 The lettered flowers on the right-hand stem of the plant were simply 

 emasculated at the same time the numbered flowers on the left-hand 

 stem were emasculated and pollinated. The prematurely pollinated 

 flowers, 1, 2, and 6, shown in these illustrations, had all fallen within 

 forty or less hours of the time they were pollinated, while those of 

 the same age. <i, C, and <l, that were emasculated and not pollinated, 

 continued to grow. These flowers were found fallen at 9 o'clock in 

 the morning, the exact time of their falling not being known; but 

 three other tests made at different times showed that in good growing 

 weather such prematurely pollinated blossoms fall in about thirty or 

 thirty-six hours after being pollinated. Flowers 3, 4, and 5 were 

 treated exactly as were flowers 1, 2, and B, but set seed instead of 

 falling, because they were almost mature when pollinated. Fig. 1, 

 PI. II, is from a photograph of the same plant as shown in PI. I, taken 

 eight days later than fig. 1, and shows the seed pods which grew from 

 flowers 3, 4, and 5, while the lettered ones that were emasculated but 

 not pollinated are just fallen, making a difference of six days between 

 the falling of those prematurely pollinated and those not pollinated. 



There was always a great difference noticeable in the manner of 

 falling of the prematurely pollinated blossoms and those that fell from 

 other causes, such as non-pollination. When the fall was the effect of 

 premature pollination, the separation of the flower from the plant was 

 rapid and complete and not accompanied by any previous wilting of 

 the flower, but invariably occurred at a joint situated at the base of 

 the peduncle shown at a, PI. Ill, fig. 3. Flowers 1 and 2, PI. Ill, fell 

 within some time less than thirty-nine hours after being pollinated, 

 while flowers 3 and 4, which were emasculated at 12 m. , February 27, 

 and not pollinated, had not yet fallen at 9 a. m., March 5, when the}' 

 were cut and photographed. 



While comparing blossoms that had been emasculated and prema- 

 turely pollinated with similar ones that had been emasculated and 

 bagged but not pollinated, the persistence with which the pistils 

 retained their fresh receptive appearance was very noticeable and led 

 to a few slight tests of the length of time the pistils could remain 

 receptive and still result in the production of seed when pollinated. 

 By referring to experiment 18 in Table I, it will be noticed that the 

 10 flowers emasculated June 7 and found receptive June 9 were not 



