32 



THE I'OI.LINATTOX OF PEAR FLOWERS. 



of experiments on different trees and at different times and places adds 

 greatly to the certainty and reliability of resnlts. The results of these 

 experiments, when reduced to percentages, must be regarded as only 

 approximately correct, and oidy the large differences need be taken 

 into acconnt. 



The numerical details of the Rochester experiments will be found 

 distributed through the synopsis. A greater number were made here 

 than at all the other places together, the general results being in 

 accordance with those secured before. The Bartlett pollen was found to 

 be almost but not absolutely sterile in producing fruits on stiguias of 

 its own variety, and very much inferior to other varieties, but was shown 

 to be perfectly good pollen when applied to other varieties. Of the 

 forty varieties worked on, about two-tliirds were nearly or quite ster- 

 ile to their own pollen. It was found that other varieties, such as Glair- 

 geau and Lawrence, w^ere even more completely sterile to their own 

 pollen than was the Bartlett. 



There were two especially new features developed by the Eochester 

 series aside from other points, (1) that the Bartlett and Anjou are not 

 absolutely sterile to their own x>ollen, but under favorable conditions 

 can set a few fruits under its influence; (2) that Angouleme, Seckel, 

 and other varieties which seem self- fertile were not always as productive 

 when limited to their own jjollen as when cross-pollinated. 



SERIES IV. — EXPERIMENTS AT GENEVA, N. Y. 



I am indebted to Mr. D. G. Fairchild for this valuable addition to the 

 experiments. The work was done in two separate orchards nearly 3 

 miles apart, and these being quite different it is desirable to give a 

 descrix)tion of each. 



The Smith orchard. — This orchard, consisting of several thousand 

 standard pear trees, is the property of E. Smith & Sons, of Geneva^ 

 and we are greatly indebted to the owners for placing it at the dis- 

 posal of the Department for the purposes of the investigation. The 

 trees are all from 20 to 25 years old. In recent years the orchard has 

 been kei)t in gVass and the trees have made very little new growth, 

 although the majority of them seem sound and healthy. There are a 

 large number of varieties and these are x^lanted so that no block of any 

 one variety of any considerable size occurs. 



Most of the trees bloomed very heavily, but as a rule they did not 

 fruit in corresponding abundance. This was doubtless due to the 

 rather low state of nutrition of the trees. Some sorts seemed to thrive 

 well in the grass, however, and set a high per cent of fruit, but the 

 resulting fruit was in most cases not more than half the size that well 

 cultivated and well-pruned trees would have borne. 



The iK'Hv scab fungus, Fmicladidni xyyyiuuDi (Lib.) Fckl.j did con- 



