GENE RECOMBINATION AND HETEROSIS 313 



ing such a polygenic system conditioning weight per locule, provided greater 

 weight per locule was at least partially dominant, and provided the genes 

 for increased weight per locule and shorter period from seeding to first fruit 

 ripe were divided between the two parents. The analyses and discussions in 

 the immediately preceding paragraphs show that in such an event it would 

 be almost impossible to obtain inbred lines which through gene recombina- 

 tion would retain any appreciable amount of the yield of the Fi hybrid. 



On the arithmetic scale the Johannisfeuer X Red Currant populations 

 show partial phenotypic dominance of smaller weight per locule with the 

 exception of the Bi to P2 which indicates no dominance. The parents of the 

 Johannisfeuer X Red Currant hybrid populations were found to be differ- 

 entiated by a large number of gene pairs (probably more than 40) each of 

 which individually had minor effects and in addition by a few gene pairs 

 (probably 2 or 3) having major effects. In these hybrid populations the total 

 effect of the minor genes was greater than the total effect of the major genes. 

 Again the shorter period from seeding to first fruit ripe showed heterosis. 



With the number and type of gene pairs conditioning weight per locule 

 found for the Johannisfeuer X Red Currant hybrid, and provided the genes 

 differentiating weight per locule exhibited at least partial dominance, as is 

 indicated for the Danmark X Johannisfeuer populations, certain parental 

 combinations of the genes would result in the hybrid populations showing 

 heterosis for increased yield of fruit per plant. Since comparatively few ma- 

 jor gene pairs differentiate weight per locule, it should be possible by re- 

 combination of genes through selection in F2 and backcross populations of 

 such a cross to combine into inbred lines some of the increased yield at- 

 tributable to heterosis. 



The Danmark X Johannisfeuer hybrid populations show partial pheno- 

 typic dominance for greater weight per locule, and complete dominance for 

 shorter period from seeding to first fruit ripe. Two or three major gene pairs 

 were found to be differentiating weight per locule. For weight per locule and 

 number of days from seeding to first fruit ripe, dominance is such that had the 

 genes tending to increase each of these two characters been divided between 

 the two parents, the hybrid populations would have shown heterosis for both 

 component characters. Likewise, if the above conditions had been fulfilled, 

 yield of ripe fruit per plant would have shown heterosis in the hybrid popu- 

 lations. 



The Porter X Ponderosa hybrid populations showed at least partial genie 

 dominance for weight per locule (Powers, Locke, and Garrett, 1950). The 

 parents were found to be differentiated by three pairs of genes and the genes 

 tending to increase weight per locule were distributed between the two par- 

 ents. As was to be expected, the hybrid populations showed heterosis for in- 

 creased weight per locule. Period from seeding to first fruit ripe showed al- 

 most if not complete dominance for the shorter period from seeding to first 



