54 APPLICATION OF PEINCIPLES OF HEEEDITY TO BREEDING. 



while in cross-fertilized species we have the same combinations, but 

 not in homozygote form. In self-fertilized species, when a desired 

 combination is not at hand, we can easily produce it and get it in 

 fixed form. In cross-fertilized species, if the desired combination is 

 not at hand, we can get it by crossing, but can not be sure of keeping 

 it unless it is a form that can be propagated vegetatively, such as 

 berries, tree fruits, and potatoes. The best we can do in cross-fer- 

 tilized species which are propagated from seed is to make a cross 

 with a view to getting a certain desired combination of characters 

 and then select the desired type until we get a fairly constant strain 

 of it. The results that follow such selection have already been 

 explained under the head of Cross-fertilized species," page 36, and 

 illustrated in figure 3. 



It has already been hinted that in cross-fertilized species we 

 should not try for too great uniformity, as such uniformity in cross- 

 fertilized species usually goes with weak development. If by cross- 

 ing and subsequent selection we can get two forms, each of which is 

 homozygote for the same desirable characters but heterozygote for 

 other characters which are of no importance, and then plant them 

 so that the two forms will cross freely, we shall probably have the 

 nearest approach to the desired end attainable in such forms. 



MENDELIAN ANALYSIS OF HETEROZYGOTE RACES. 



When an individual which is heterozygote for a given pair of 

 allelomorphs is self-pollinated it breaks up in the next generation 

 into three forms, or types, two homozygote and one heterozygote. 

 Let our pair of allelomorphs be represented by Aa. With self- 

 pollination we have in the next generation one-fourth AA, one-half 

 Aa, and one-fourth aa. That is, one-fourth of the seed produced 

 is homozygote for A, one-fourth for a, and one-half of it is heterozy- 

 gote, Aa. Hence, in such species as apples, pears, strawberries, and 

 cassava, which do not ordinarily reproduce true to seed because 

 they are heterozygote, usually for many of their characters, if we 

 self-pollinate them, in the next generation we get many forms that 

 are homozygote for some characters. If the species will endure con- 

 tinued self-pollination it is clear that we could in eight or ten gen- 

 erations break up any variety of this kind into homozygote types 

 that would reproduce true to seed. In fact, there are a good many 

 varieties of some of the species just mentioned that do reproduce 

 practically true from seed. The Royal Anne cherry of Oregon 

 and Washington is a case in point. Many of its seedlings can hardly 

 be distinguished from the original variety. Where it is possible to 

 secure these homozygote forms by this process of breaking up into 



165 



