OF THE IMPROVEMENT OF BAOES. 809 



ing several successive seasons, by abstracting all tlie 

 flowers and fruit, and bj' such other means as may- 

 suggest themselves ; and then to obtain the most per- 

 fect seed possible by a destruction of the tubers dur- 

 ing the season when seed is finally to be saved. Mr. 

 Knight found, in raising new varieties of the Peach, 

 that, when one stone contained two seeds, the plants 

 these afforded were inferior to others. The largest 

 seeds obtained from the finest fruit, and from that 

 which ripens most perfectl}'- and most early, should 

 always be selected {Hort. Trans., i. 39) ; and, in his 

 incessant efforts to obtain new varieties of fruit of 

 other genera, he had reason to conclude that the trees, 

 from blossoms and seeds of which it is proposed 

 to propagate, should have grown at least two years in 

 mould of the best quality ; that during that period 

 they should not be allowed to exhaust themselves by 

 bearing any considerable crop of fruit ; and that the 

 wood of the preceding year should be thoroughly 

 ripened (by artifici'^l heat when necessary) at an early 

 period in the autumn ; and, if early maturity in the 

 fruit of the new seedling plant is required, that the 

 fruit, within which the seed grows, should be made to 

 acquire maturity within as short a period as is consis- 

 tent with its attaining its full size and perfect flavour. 

 Those qualities ought also to be sought in the parent 

 fruits, which are desired in the offspring; and he 

 found that the most perfect and vigorous progeny was 

 obtained, of plants as of animals, when the male and 

 female parent were not closely related to each other. 

 (See the Horticultural Traiisaciions, i. 165.) 



