296 Practical Application [ch. 



umbels, the size of the calyx, of the involucre of bracts, 

 and many other structural features the inheritance of which 

 has not been studied in detail. 



Such a series will illustrate the ordinary practice of the 

 horticulturist who is engaged in the production of novelties. 

 He crosses together two types and picks out those novelties 

 which are produced in F^ by the re-combination of pre-ex- 

 isting factors. This has been the method which has led to 

 the creation of nearly all our modern varieties of vegetables 

 and flowers. From Mendelian discovery the practical 

 breeder learns two lessons, both of importance. The first is 

 that he must not discard the /^, generation merely because 

 it does not give him anything he wants. This generation 

 may be uniform, and indeed must always be uniform if both 

 parental types were homozygous in all their several factors. 

 It may also be quite uninteresting from the horticultural 

 standpoint, exhibiting some old-fashioned or reversionary 

 type which is reproduced because all the factors which 

 constitute it happen to have been brought together into 

 one individual. But from the appearance and properties of 

 F^ no guess can be made as to the possibilities of /%. A 

 vast amount of valuable material has again and again been 

 discarded by practical horticulturists through ignorance of 

 Mendelian principles. The uninteresting types produced 

 by crossing, though no improvement on the old and familiar 

 varieties would, if their seed had been saved, have given 

 abundant novelties in the next generation. 



Fixing the Type. 



The second practical lesson is more important. If the 

 plant is of a kind which is habitually propagated by budding 

 grafting, cuttings, or other asexual mode of division, no 

 doubt the object of the breeder is attained at once with the 

 appearance of his novelty. All that he has to do is to 

 multiply it. But if, as is generally the case with vegetables, 

 and very often nowadays with flowers also, he requires to 

 work up a strain true from seed, Mendelian analysis shows 

 him how to accomplish this in the shortest time. He must 

 h^eed from each individual separately. Take this case of 

 Primula. Suppose he wishes to get a pure strain of the 

 dark magenta {e.g. Plate VI, figs. 9, 15, 21). If he saves all 



