OF THE IMPROVEMENT OF RACES. 311 



ly applied to the introduction into the world of a na- 

 tural form which has probably never before existed 

 in it. In constitution the mixed offspring appears to 

 partake of the habits of both parents ; that is to say, 

 it will be less hardy than the one of its parents which 

 bears the greatest exposure, and not so delicate as the 

 other ; but, if one of the parents is quite hardy, and 

 the other not quite able to support our winters, the 

 probability is, that the offspring will support them, 

 though it may suffer from a very unusual depression 

 of the thermometer, or excess of moisture, which 

 would not destroy its hardier parent." 



In the many successful attempts made by Mr. 

 Knight to improve the quality of fruit trees by rais- 

 ing new varieties, his method was to obtain cross- 

 breds by fertilising the stigma of one variety of 

 known habits with the pollen of another also of 

 known habits. But, in doing this, his experiments 

 were not conducted at random, and without due con- 

 sideration ; on the contrary, we learn from himself 

 that he was very careful in selecting the parents from 

 which his crossbreds were obtained. He found that 

 the general opinion, that the offspring of crossbred 

 plants as well as crossbred animals usually presents 

 great irregularity of character, is unfounded ; and 

 that if a male of permanent habits, and of course not 

 crossbred, be selected, that will completely overrule 

 the disposition to sport, "the permanent character 

 always controlling and prevailing over the variable." 

 He tells us that he usually propagated from the seeds 

 o£ such varieties as are sufficiently hardy to bear and 



