HYBRIDIZATION AND SELECTION. 



51 



The relative proportion of each of these nine types for a series of 

 generations, from the second to the tenth, is shown in the diagram 

 of figure 4. For instance, the proportion of type WWCC is indicated 

 by the space between the top horizontal line of the diagram and the 

 upper one of the curved lines. The generations are indicated by 

 the figures at the bottom of the diagram. In generation two the 

 space for type WWCC is narrow, constituting only one-sixteenth of 

 the second generation. But this type increases from generation to 

 generation until by the tenth generation it is practically one-fourth 

 of the population; that is, when there is no selection to type. 



Type WWCc is seen to decrease from generation to generation, 

 as indicated by the space between the two upper curved lines. This 

 space gradually becomes narrower, so that it has practically dis- 

 appeared by the tenth generation. The space beginning opposite 

 each type formula shows what happens to that type. It is seen that 

 the four homozygote types WWCC, WWcc, wwCC, and wwcc gradually 

 increase in proportion while all the heterozygote types decrease. By 

 the tenth generation the whole population consists practically of the 

 four homozygote types, each of them constituting practically one- 

 fourth of the population. Only small amounts of any of the heter- 

 ozygote types remain in the tenth generation. Of these four ho- 

 mozygote types, two of them will be just like the two parents, as far 

 as the characters we are considering are concerned. The other two 

 will represent new combinations of the characters under consider- 

 ation; the new types are (1) winter character with club heads and 

 (2) spring character with long heads. We may therefore in such 

 cases (i. e., with self -fertilized crops) secure our hybrid and plant 

 its s.eed for several years without any selection at all, then select out 

 the type we want and it will be almost entirely pure; that is, 

 nearly all the plants selected will reproduce true to type as far as 

 the characters wanted are concerned. Then selecting individual 

 plants of the type wanted we can quickly get plants that are ho- 

 mozygote with reference to practically all their characters by plant- 

 ing the seed of each plant separately and observing which of them 

 do reproduce true to type. 



These fixed forms which occur in the progeny of hybrids are 

 sometimes mistakenly called ''mutations." They are in no sense 

 mutations of the sort comprehended by that much misused term 

 as it is at present understood. They are simply recombinations of 

 characters which, before the hybridization occurred, existed in 

 different combinations. 



After these fixed forms are obtained the same laws apply to their 

 selection as have already been described under the effect of selection 

 on close-fertilized forms. Generally speaking, we can not modify 

 165 , 



