26 



THE POLLTXATION OF PEAR FLOWERS. 



unopened liower wns found to eontiiiu an insect which had crawled in, 

 but such cases were rare and for exi)erimental purposes were always 

 rejected. 



There is a possible chance for error in the experiments, owing to 

 the method of obtaining the pollen. The supply of pollen was taken 

 from freshly opened anthers from open flowers on the trees. It is 

 possible that a few grains of i^ollen from some other source had been 

 left on the anthers by a bee or other insect, either before or after they 

 had opened. The amount, however, which could have gotten on an 

 anther, even in the most extreme case, must have been very small in 

 comparison with the amount of its own pollen. Before bursting, the 

 anther is always smooth, and besides the stamens are incurved so that 

 they are not readily touched by the larger bees, which are the chief 

 pollen bearers, while after the stamen is erect and the anther has opened, 

 if visited by a bee it would lose most of its pollen by having it knocked 

 or x>ulled off, and would not be selected for use. It should be noted 

 that any mixture of pollen brought about in this way would have had a 

 tendency to lessen the clearness of the results secured and thus be 

 against rather than in lavor of the main conclusion. 



The crossed and hand-pollinated flowers were all covered with paper 

 bags, which were not removed till the flowers had fallen. The bags 

 were fastened and labeled in the same manner as at Brockport. For 

 convenience in carrying and safe keeping, the flowers to be used for 

 pollen Avere generally picked off' and i)laced in paper bags till used. 

 I had some fear lest the operation of emasculation, by which so much 

 of the flo\Yer Avas cut and removed, might injure the flower so that it 

 would not set fruit, but the high ])ercentage of fruits which resulted 

 when such flowers were i)roperly pollinated shows that there was no 

 foundation for the fear. 



Resvlts of the experiments. — The results of these hand-pollinations 

 were very striking. Within a week after the petals fell the young- 

 pears all over the orchard dropped off', in the great majority of cases 

 never having started to SAvell. JMost of the trees were absolutely bar- 

 ren. In a few cases two or three i)ears per tree could be found. It 

 was soon seen that wherever the flowers had been fertilized with pollen 

 other than Bartlett the fruits were developing. When on May 3 the 

 count was made it was found tliat a large proportion of the crosses 

 with other varieties had set, but not one single Bartlett flower had set 

 fruit when pollinated with Bartlett pollen, no matter what the source. 



An Anjou tree whose flowers came out several days before and whose 

 pollen was nearly all shed before the Bartlett trees surrounding it were 

 out, fruited remarkably well. A Sheldon Avhich bloomed in the same 

 way also fruited heavily. The results of the work on the Bartletts are 

 given in Chapter III on p. 39, and on the other varieties on pp. 37. 44, 

 and 52. It should be noted that although the weather was very unfavor- 

 able there was a considerable period of time for insect visits, and the 



