HETEROSIS IN A NEW POPULATION 449 



It remains to be shown whether more uniform hybrids actually would be 

 superior to hybrids made with Si lines in a country such as Mexico. Variations 

 in climate from year to year in any one valley in Mexico are usually extreme. 

 Under such conditions, a high degree of uniformity in a hybrid may actually 

 be a detriment over a period of years. 



The problem of what tester or testers to use in isolating lines of high com- 

 bining ability continues to be a difficult one. Usually the tester chosen de- 

 pends upon the use to be made of the lines. Judging from the segregants ob- 

 tained upon inbreeding in some of the races of maize, a point may have been 

 overlooked in the selection of lines and testers. This appeared to be especially 

 true in those races where it was necessary to recombine lines from the same 

 race to obtain immediate improvement. In some of the races upon inbreed- 

 ing, especially Bolita, Chalqueno, and Vandeno, inbred line segregants often 

 appeared which were very similar to the putative parents of the particular 

 race (Wellhausen et al. in collaboration with Mangelsdorf, Fig. 98). If these 

 races had reached equilibrium on an individual gene loci basis, one would not 

 expect to get the parental types in subsequent inbred generations from 500 

 ears selfed at random in the original population. It appears, therefore, that 

 many of the modern races in Mexico are not in equilibrium on an individual 

 gene loci basis, but consist of blocks of genes in equilibrium with each other. 

 Although it is difficult to estimate the age of some of the modern varieties, 

 these gene blocks obtained from the various ancestors seem to have persisted 

 more or less intact through many generations. 



If blocks of germplasm as received from various ancestors are still intact 

 in some of the modern high yielding races, then it may not be as difficult as it 

 once seemed to reconstruct a hybrid that would approach the yield of the 

 ideal plant in a particular variety by the recombination of inbred lines from 

 that variety. Isolation of good lines for such recombination may involve dif- 

 ferent procedures. A method based on selection for origin and type, with sub- 

 sequent crossing to an unrelated variety or varieties for the determination of 

 combining ability, may not be the best procedure. 



Selection for vigor and type in an environment best suited to a race such 

 as Chalqueno, which probably originated from the hybridization of two dif- 

 ferent races neither of which is adapted to the environment best suited for 

 Chalqueiio, would eliminate those segregants in the direction of either one of 

 the putative parents. It is probable that with the elimination of such segre- 

 gants, many genes are discarded that are needed to reconstruct the ideal 

 chance hybrids which often appear in a particular variety or race through 

 open pollination. Selection for vigor and type also would tend to select those 

 genotypes which are similar, and more nearly like those of the variety from 

 which they came, than the extreme segregants. 



Tests for combining ability of a group of lines from the same variety, based 

 on crosses with an unrelated variety or varieties, tend to select those geno- 



