EXPERIMENTS WITH COTTON BLOSSOMS. 21 



experiment 10 were pollinated when receptive ana produced »'> good 

 bolls. Those of experiment 11 were pollinated when emasculated and 

 again the next day and produced L0 small bolls. ;» of which were 

 decidedly one-sided. The flowers of experiment L2 were pollinated 

 when emasculated and produced 7 rather good bolls. 



In making the second pollination <>t' experiment 8 it was noticed 

 that the pistils of tin- flowers in experiment 7 which had not been pre- 

 maturely pollinated were longer than those of experiment 8, hut they 

 were not measured because it was then too late to determine whether 

 they had been precisely of the same Length when emasculated. To 

 determine whether premature pollination checks the growth of young 

 pistils, the pistils of experiments L0 and 1 L were accurately measured 

 at the time of emasculation, 8 of each experiment being each five- 

 eighths of an inch and 2 three-fourths of an inch in length, SO that the 

 aggregate length of the pistils was 6^ inches. The flowers of experi- 

 ment 11 were pollinated twenty-one hours before they were fully 

 open and again twenty-one hours later, when the pistils were again 

 carefully measured and found to give an aggregate length of 8i inches, 

 an increase of ltr inches since emasculation, during which time the 

 pistils of experiment 10, which were not prematurely pollinated, had 

 made an aggregate increase of 21 inches. This is to say that the 

 application of pollen to 1<» immature pistils reduced their growth 

 during twenty-one hours by five-eighths of an inch. 



The conclusion to be drawn from the cotton experiments is in favor 

 of making an extra visit to the emasculated flowers to apply the pollen 

 rather than to apply it when the flowers are emasculated. This is not 

 only because larger percentages of fruits will result, but more especially 

 because better fruits are obtained. 



It is the opinion of the writer that the variation in the results 

 obtained with Sea Island and upland cotton flowers similarily treated 

 is largely due to the fact that the pistils were closely approaching a 

 receptive condition. Of course this would cause no variation in the 

 results, providing the size of the flowers and length of time before 

 opening indicate the same stage in the development of the ovules of 

 both kinds of cotton, but this may not be the case. The one-sidedness 

 of many of the bolls resulting from early pollination indicates that only 

 a portion of the ovules were mature enough for fertilization. Perhaps 

 pollination a few hours earlier would have resulted in the falling of 

 all the flowers. Tobacco flowers pollinated one day before opening 

 give fair results, while those pollinated two days before opening are 

 killed. It is thought that the same will be found true of cotton flowers, 

 as indicated by the results of experiments 5 and 6, but more work 

 with cotton blossoms is necessary before their behavior will be well 

 understood. 



