e;xpe;riments in bre;e;ding swe;e;t corn. 305 



to a large degree upon the maintenance of "broad breeding," 

 i. e., of a heterozygous condition in the strain. From the prac- 

 tical standpoint it seems to us that the aim of corn breeding 

 should be to get rid of poor genotypes and leave the good. 

 There will under all ordinary conditions be enough of these 

 latter to insure the heterozygote condition in the strain, par- 

 ticularly if, on the one hand, the all selected corn is planted on 

 the ear-to-row system in a single plot and no detasseling is prac- 

 ticed, and, on the other hand, a deliberate attempt is not made 

 in the selection to reduce to the lowest number possible the 

 female lines involved in the pedigrees, as by taking the seed 

 ears for a subsequent year's ear-to-row plot all from the same 

 row of a plot planted on the same system, and keeping up this 

 practice through several years. We are very much inclined to 

 believe that in actual practice substantially as good results may 

 be obtained by a general ear-to-row selection method, relaxed 

 in intensity after a few years, as by the use of the more elabo- 

 rate and costly "pure line method of corn breeding" of Shull. 

 The latter method is probably right in principle, but the former 

 method, in a much cruder and less precise way, really makes 

 use of the same principle. Continued ear-to-row breeding 

 (without too close pedigree selection) is continually testing out 

 different heterozygote types and rejecting the undesirable ones. 

 Presently a point will be reached where the great majority, if 

 not all of the more strikingly undesirable genotypes (which 

 when crossed produce the undesirable heterozygotes) will have 

 been automatically eliminated. We shall then have reached by 

 a much slower route the same goal which Shull attains quickly 

 and directly by his "pure-line" method. 



The rejection of undesirable genotypes can be most readily 

 accomplished practically through the ear-to-row system of plant- 

 ing. Naturally one will never expect to get such rapid results 

 following selection with an open fertilized plant like corn as 

 with a self fertilized plant like the bean. One can not so quickly 

 get rid of the influence of all poor genotypes. 



9. Without wishing in any way to be dogmatic in the matter 

 it appears to the writers that the results obtained in these selec- 

 tion experiments with com indicate that inheritance in this form 

 is fundamentally in accord with the "pure line" or genotype 



