ANIMAL AND PLANT BREEDING 3I5 



relationship means that some one individual ancestor has contributed 

 to both their genotypes. Thus inbreeding is to some extent the breeding 

 of Hke with Hke, and will clearly have a tendency to bring together like 

 genes and produce homozygosis. In the absence of selection, the rate of 

 approach to homozygosis is dependent mainly on the type of inbreeding 

 practised, though it is modified by mutation, linkage, etc. The most 

 rapid approach is with the closest type of inbreeding, namely self- 

 fertihzation, in which the number of heterozygous genes is halved in 

 each generation. Thus a population of self-fertiUzed plants may be 

 expected to consist of an assemblage each of which is nearly or quite 

 homozygous, and a very short period of selection of individual plants 

 will yield a set of completely homozygous types. 



The first reaUzation of this fact was by Johannsen,^ who showed that 

 a population of beans (self-fertiUzing), which showed a certain varia- 

 bihty in the weights of the individual seeds, was really composed of a 

 mixture of pure hues, or homozygous forms, in each of which the 

 variability was very much less. Moreover, within each pure hne the 

 variabihty was entirely due to random environmental fluctuations and 

 was not inherited, so that selection within a pure line remained ineffec- 

 tive. Indeed it must do so until a mutation in the desired direction 

 occurs, since in a completely pure line all factors are homozygous and 

 no hereditable variation can exist. (Fig. 131.) 



Since mutation and occasional chromosome rearrangements are 

 bound to occur, at least rarely, some selection is necessary to keep even 

 a pure hne absolutely constant;^ but granted a minimum of selection, 

 further rigour has no effect. The de Vilmorin wheats, originating from 

 single plants and self-fertiHzed, have remained substantially constant 

 for over fifty years. 



The formation of a pure line is usually an essential in the production 

 of a new variety of a sexually reproducing plant, since the variety must 

 breed true and yield uniform progeny. In plants reproducing vegeta- 

 tively or by female apospory, it is not necessary to take any steps to see 

 that the variety is homozygous, since it must breed true in any case; 

 and in fact we find that many, if not most, vegetatively reproduced 

 crop or ornamental plants are extremely heterozygous and often hybrid. 



The production of a uniform variety is simplest, as we saw, in self- 



fertihzing plants, where individual plants are usually already nearly 



homozygous. Thus the offspring of a single plant may constitute a true 



breeding variety, and several very valuable types have been produced 



^ Johannsen 1903. ^ Haldane 1936. 



