302 POMOLOGY 



testing 169 cultivated varieties (and many seedlings), he 

 classified them as follows: 



Class I. — Self-fertile varieties having perfect clusters or 

 clusters varying from perfect to somewhat loose, 38 vari- 

 eties (21.8%). 



Class II. — Self -fertile varieties having clusters loose but 

 marketable, 66 varieties (39.0%). 



Class III. — Varieties which are so imperfectly self-fertile 

 that the self-fertilized clusters are generally too loose to be 

 marketable, 28 varieties (16.5%). 



Class IV.— Self-sterile varieties, 37 varieties (21.8%). 

 He also noted that varieties with short or recurved stamens 

 are always self-sterile or nearly so. The explanation offered 

 at that time for sterility was "a lack of affinity between the 

 pollen and pistils of the same variety." ^ 



Dorsey - illustrates the method of testing sterility from 

 the work of Beach as follows: 



''When 143 clusters of Brighton were covered with bags 

 and self -pollinated, the average rating of the clusters formed, 

 counting 100 as a perfect cluster, was approximately one, 

 and when thirty-two clusters distributed among eight other 

 varieties were pollinated with Brighton pollen, the average 

 rating was three, showing Brighton, for those varieties used, 

 as well as for itself, to be a poor poUenizer. On the other 

 hand, when 116 clusters of the Catawba were selfed, the av- 

 erage rating on the same basis as above was eighty-six, as 

 compared with one in Brighton. When the thirty-three 

 clusters of eight other varieties were pollinated with pollen 

 from Catawba, the average rating was sixty-seven, showing 

 a marked difference between the Brighton pollen and the 

 Catawba pollen when used either in selfing or crossing." 



1 N. Y. Agr. Exp. Sta. Bull. 157. 1898. 



2 Dorsey, M. J. Jour. Heredity, 6: 1915. p. 243. Minn. Agr. Exp. 

 Sta. Bull. 144. 1914. 



