'seminal sports' 309 



peculiar characters as the parent are allowed to produce 

 seed, all others being pulled out and discarded. The seeds of 

 this first selected generation are then sown, and a further 

 selection and sowing of those possessing the new attributes 

 is made. This process is repeated for several generations until 

 no weeding out is found necessary, that is until the new char- 

 acters become constant in all the offspring, after which the 

 variety is said to be ^ fixed'' and 'comes true' from seed. The 

 time taken to 'fix' a variety in this manner depends upon the 

 power which the plant possesses of transmitting its characters 

 to its offspring. This power is exceedingly variable and no 

 rules can be laid down in regard to it; in some instances 50 

 per cent, or more of the first generation may resemble the parent, 

 and, on sowing the seeds of these, 90 per cent, of the seedlings 

 may ' come true ' ; in such cases fixation of a new variety is 

 tolerably easy and may be effected in three or four generations. 

 In other cases the number of plants true to type in each succeed- 

 ing generation may be very small, and even after selection has 

 been carried on for many generations a large proportion of the 

 plants obtained at each sowing may possess none of the char- 

 acters of the variety which the plant-breeder wished to establish. 



H. Vilmorin stated that some of his hybrid varieties of wheat 

 took six or seven years of cultivation and selection before they 

 were of sufficiently fixed character to be put on the market for trial. 



The process applied to five or six generations of plants is 

 generally found to be sufficient to 'fix' many new varieties of 

 cereals, beans, peas, cabbages, turnips, tomatoes and other 

 annual and biennial plants ; probably the raising and selection 

 of a similar number of succeeding generations of plants would 

 be needed to make a variety of a perennial 'plant ' come true ' 

 from seed. However, on account of the fact that several years 

 often elapse before seed is produced by many seedling perennials, 

 the process of fixing new varieties of such plants by selecting and 

 propagating in the above manner has rarely been carried out ; 



