104 AMERICAN BREEDERS MAGAZINE. 



supports this view is that reciprocal hybrids 6 give essentially equal 

 results. Thus, strains A and B, between which reciprocal hybrids 

 were reported last year as yielding 74.4 and 78.6 bushels per acre, 

 respectively, produced this year, in reciprocal hybrid families, exactly 

 equal yields, namely, 79.8 bushels per acre. More conclusive still is 

 the result from my best hybrid combination of the past season which 

 has also been tested in reciprocal crosses; see figure. These crosses 

 were made between a self-fertilized strain which had been selected 

 continually to 16 rows and another which had been continually se- 

 lected to 20 rows of grains on the ear. When the 16-rowed type was 

 used as the mother, a yield of 98.4 bushels per acre was produced; 

 when the 20-rowed type was used as the mother, a yield of 96.1 

 bushels per acre was produced. If the production of 98.4 bushels 

 per acre had been purely a chance result which might by equal chance 

 have appeared in any other strain, it is scarcely conceivable that the 

 reciprocal should have so nearly approached the same extreme yield. 

 The lower of these two yields, namely, 96.1 bushels, is 8 bushels per 

 acre above the best yield produced during the same season by any 

 continually cross-bred family of corn in my cultures. From all the 

 results reported in this paragraph, it may be safely concluded that the 

 production of the highest yield requires simply the finding of the best 

 combination of parents and then repeating this combination year 

 after year. 



Several new evidences of the correctness of my view regarding 

 the hybrid character of any ordinary vigorous corn plant have resulted 

 from the past season's work. The assumption that self-fertilization 

 results in the isolation of pure homozygous strains or biotypes and 

 that the real purpose of cross-breeding is to secure the stimulus which 

 comes from the heterozygous^ association of alternative qualities from 

 the two parents, requires that the first generation of the cross between 

 two pure self-fertilized strains be relatively uniform, and that the 

 second generation, in which these various hybrid qualities are re- 

 arranged in every possible combination, shall show greater diversity. 

 I have now reared two families representing the second generation 

 of such a cross between strains A and B. The variation in the number 

 of FOWS in self-fertilized strains, in F x hybrids, and in F 2 hybrids, 

 are shown in the following table : 



e Formed by using one variety as the male parent in one cross and in an- 

 other cross between the same varieties using the other variety as the male parent ; 

 thus A x B and B x A are reciprocal crosses, and their progenies are reciprocal hybrids. 



f Having each separate characteristic derived from only one of the two parents. 



