July IS. 1915 Inheritance of Certam Characters of Grapes 317 



Certain results secured in the summer of 1914 seem to indicate that this 

 method is perhaps open to criticism. While emasculating clusters of the 

 Janesville variety it was found that, although the cap had not split, 

 the pollen in the anthers seemed to be mature, and, as the anthers were 

 ruptured during the emasculation, there was a possibility of self-fertiliza- 

 tion taking place. Several clusters were emasculated and bagged without 

 being pollinated. These set nearly the full quota of berries with seeds 

 that have every appearance of being viable. With two other varieties, 

 clusters emasculated and not pollinated matured a few plump seeds, 

 though the clusters were much below normal. A somewhat similar in- 

 stance is reported by Beach (i),^ the variety being the Mills. This point 

 deserves careful study, for if it is found that serious danger of self-polli- 

 nation exists before the calyx cap splits it will be necessary to change 

 the method — at least to the extent of emasculating the clusters several 

 days before the cap is ready to come off. 



All data regarding size and shape have been recorded in comparative 

 terms, instead of the actual measurements being taken. With a limited 

 number of observers and thousands of seedlings of the various fruits to 

 be studied each year, it was a physical impossibility to take measure- 

 ments and would not have increased the value of the records to any 

 extent from a horticultural point of view, though it would, of course, 

 have furnished interesting material for a statistical study. The value of 

 data reported in comparative terms depends upon the accuracy of the 

 recorder. The work with the grape has always been done by members 

 of the scientific staff, and the observations have usually been checked 

 during several seasons. 



GENERAL RESULTS OF THE STUDY OF VARIETIES OF GRAPES 



One of the surprises in the study of varieties of grapes was the failure 

 of many of our commercial sorts to transmit desirable qualities to their 

 progeny. Seedlings of Concord, Niagara, Worden, Delaware, and Ca- 

 tawba grapes have so far proved only disappointments. The best results 

 have been secured from such little-grown varieties as the Ross, Collier, 

 Mills, Jefferson, Diamond, and Winchell. This has made it seem desir- 

 able to test all varieties that show any promise. The first step, then, 

 was to secure as many varieties as possible which were of any value and 

 which could be grown in northeastern United States. More than 400 

 such varieties have fruited in the Station's vineyards and have been 

 described. About 200 of these have been used to a greater or less extent 

 in the breeding work. 



As an aid in studying the breeding possibilities of grape varieties the 

 Station has grown nearly 3,000 selfed, or pure, seedlings, using as parents 

 most of those varieties which have entered into the crosses that have been 



' Reference is made by number to " Literature cited" p. 329-330. 



