PRESERVATION OF RACES BY SEED. 303 



flower, and such plants, have been frequently selected 

 on account of their perfect characters, and been care- 

 fully planted in gardens for a stock of seed, but their 

 produce has as frequently been of the worst descrip- 

 tion, bearing no resemblance to the parent. In such 

 cases as these, it would seem as if bees and other 

 insects were attracted from all quarters by the gay 

 colours, or odour, of such isolated individuals, and, 

 arriving from a hundred flowers which they had pre- 

 viously visited, bring with them so many sources of 

 contamination. 



When, however, the action of other flowers can be 

 prevented, as in the Melon and other unisexual 

 plants, by "setting," the largest, healthiest, and most 

 cultivated varieties will yield seed of the purest and 

 finest quality. The tendency of Persian Melons to 

 degenerate in this country was remarked soon after 

 their introduction; and, for a long time, it was 

 thought impossible to preserve them for many 

 generations. Mr. Knight, in his numberless experi- 

 ments upon this fruit, found that to be the case ; for 

 his fruit, at one time, became less in bulk and weight, 

 and deteriorated in taste and flavour. But when he 

 came to consider that " every large and excellent 

 variety of Melon must necessarily have been the pro- 

 duction of high culture and abundant food, and that 

 a continuance of the same measures which raised 

 it to its highly improved state must be necessary 

 to prevent its receding, in successive generations, 

 from that excellence," the cause of his Persian Melons 

 deteriorating became apparent ; and he found that by 



