158 



offspring ; and that a race could, in this way, be improved and 

 confined, as it were, to a few of the best individuals, the large 

 number of moderate, or inferior, individuals usually occurring 

 in the race being thus excluded. To a great extent this is 

 found to be the case, and modern horticulture and agriculture 

 depend largely on these two principles of (1) the selection of the 

 most suitable race or sub-species and (2) the improvement of the 

 selected strain by breeding only from the best individuals. In 

 cases of fluctuating variability the greater the number of indivi- 

 duals examined the more chance there is of finding an indivi- 

 dual which exhibits a particular character in an extreme degree 

 and for this reason large cultures are often resorted to in horti- 

 culture. Professor Hugo De Vries records a case in which 

 40,000 plants were cultivated, only a single individual being 

 finally selected from this multitude for further propagation, all 

 others being destroyed. Such an individual, when found, can 

 only be propagated with certainty by asexual reproduction. 

 Many plants, however, cannot easily be propagated in this way 

 and such large cultures are very inconvenient. In addition to 

 this the asexual method of reproduction affords no means of 

 effecting further considerable improvement of the race. It is, 

 however, found 'that such desirable variations can also be 

 obtained and to a certain extent be propagated by sexual 

 methods. Experience shows that if individuals exhibiting a 

 certain character in a marked degree are chosen as parents, 

 although the average of the offspring raised from their seed do 

 not exhibit this character in so high a degree as their parents, 

 they do show a distinct improvement in this respect and are 

 superior to the average of the race from which they have arisen. 

 By continuing this method for some generations the same 

 degree of improvement can be .obtained with comparatively few 

 individuals, which could only be obtained in one generation 

 by cultivating a very much larger number of plants. Moreover 

 the sexual method possesses the great advantage of enabling us 

 to continue the work of improvement. This principle is a very 

 important one in modern agriculture. The fact that not one, 

 but several, characters have almost invariably to be taken 

 into consideration is one of the great difficulties in the way 

 of successful practical work in this line. If, when selecting 

 certain individuals in a race of wheat, we were only to pay 

 attention to the quantity or quality of grain yielded, we might 

 select those plants which are also very susceptible to disease or 

 climatic influences, and the selection would thus effect no 

 practical improvement. It .is generally recognised that the 



