20 INJURIOUS EFFECTS OF PREMATURE POLLINATION. 



in experiments of 10 flowers each in order to compare the results of 

 the two methods. The pollinating- was done on Sea Island cotton with 

 pollen of the Mit Afifi variety. With the exception of experiments 

 5 and 6 the cotton experiments were with flower buds that would have 

 opened naturally the day after they were emasculated. Such a bud is 

 shown by PI. IV, fig. 13. Figs. 14 and 15 show how emasculation 

 was performed. Experiments 5 and *"> were with very } 7 oung buds 

 that would have opened in four or five days. The flowers of experi- 

 ment 5 were pollinated when emasculated and those of experiment 6 

 two days later. No fruits set in either experiment. The flowers of 

 experiment 1 were pollinated when emasculated, one day before the 

 flowers would naturally have opened, and produced but one small boll; 

 while the flowers of experiment 2, as near like the others as could be 

 chosen on the same plants and emasculated at the same time, were not 

 pollinated until the next day, and produced seven good bolls. In 

 order to obtain Mit Aflti pollen that was pure and that had not been 

 mixed with other kinds of pollen by the visits of insects, the flowers 

 were collected before opening, and in each case just one day before the 

 pollen was applied to the flowers. Experiment 3 was an exact repeti- 

 tion of experiment 1 and produced no bolls; while experiment 4, an 

 exact repetition of experiment 2, produced ten good bolls. These 

 results are decidedly in favor of waiting until the pistils are fully 

 mature before applying the pollen, but work similar to the above, 

 performed in 1901 on two different varieties of upland cotton, has 

 given results so different from those obtained with Sea Island cotton 

 that further study of the cotton flower is necessary before its reaction 

 to premature pollination is understood. Although differing from the 

 results obtained with Sea Island, the work with upland varieties still 

 speaks in favor of waiting until the pistils are mature before applying 

 the pollen, not so much, however, in order to obtain a greater per- 

 centage of bolls as to obtain bolls of larger size and more symmetrical 

 shape, many bolls from the prematurely pollinated blossoms being 

 small and one-sided. 



The work with the upland varieties was conducted in the same 

 manner as that with the Sea Island, pollen of Sea Island being used in 

 experiments 7, 8, and 9. The flowers of experiment 7 on the Braddy 

 variety were pollinated when receptive and produced 10 good bolls. 

 Those of experiment 8 were pollinated when emasculated and again 

 the next day, and produced 5 bolls, 3 of which were larger on one 

 side than on the other and hardly as large as the bolls of experiment 7. 

 The flowers of experiment 9 were pollinated when emasculated and 

 produced 9 small bolls, 1 of which were decidedly larger on one side 

 than on the other. 



Experiments 10, 11, and 12 were upon the King variety of upland 

 cotton, and pollen of the same variety was used. The flowers of 



