56 H. K. HAYES 



There seems to be some difference of opinion regarding the selection proc- 

 ess in its application to maize improvement. One school of thought practices 

 a somewhat similar method of breeding selfed lines as is used in self-pollinated 

 plants, with the viewpoint that controlled selection makes it possible to iso- 

 late in the inbred lines the genes for characters needed in the hybrids. Ap- 

 parently the relationship between the characters of inbreds and their Fi 

 crosses will become greater as inbred lines themselves improve. The other 

 extreme of viewpoint (Hull, 1945a) is that the greater part of hybrid vigor is 

 due to interallelic interaction of genes to such an extent that selection based 

 on appearance may be harmful. In a recurrent selection program Hull, 

 therefore, does not recommend selection for vigor of growth, although he 

 states that plants showing pest or weather damage should be avoided. 



It is probable that differences between these two so-called schools may 

 have been overstated. Both believe that the actual test for combining ability 

 in hybrid combination is necessary. The stage in the breeding program when 

 such test should be made will depend on the material worked with and the 

 nature of the breeding program. In both cross- and self-pollinated plants an 

 actual trial will be needed to determine the combination that excels in 

 heterosis. 



Where clonal lines can be propagated vegetatively, a method of selecting 

 for heterosis in alfalfa was suggested by Tysdal, Kiesselbach, and Westover 

 (1942), by means of polycross trials. The method is being used extensively 

 today with perennial forage crops that normally are cross-pollinated. The 

 writer is studying the method with early generation selfed lines of rye. With 

 perennial crop plants, selection for combining ability is made for heterozy- 

 gous parent clones. Where disease and insect resistance or winter hardiness 

 are important, it may be essential to insure that the clones used in the poly- 

 cross trials excel for these characters. Polycross seed is produced on selected 

 clones under open-pollinated conditions where the clones are planted together 

 at random under isolation. 



In one study of progenies of eight clones by Tysdal and Crandall (1948) 

 yields were determined from polycross seed in comparison with top cross seed 

 when each of the clones was planted in isolation with Arizona common alfalfa 

 (see Table 3.7). The agreement for combining ability was relatively good in 

 the two trials. 



An early suggestion of utilization of heterosis in alfalfa was by double 

 crosses, from single crosses between vegetatively propagated clones, without 

 entire control of cross-pollination. Synthetic varieties also have been sug- 

 gested as a means of the partial utilization of heterosis. In one comparison 

 the progeny of a synthetic combination of four clones of high combining 

 ability yielded 11 per cent more forage than a similar combination of four 

 clones of low yielding ability. A recent comparison of eight synthetics led 

 Tysdal and Crandall to conclude that the first synthetic and second syn- 



