314 LEROY POWERS 



fruit ripe. The number of major gene pairs found to be differentiating the 

 parents was eight. Due to the magnitude of the work involved it was not 

 possible to measure yield of fruit, but in all probability the hybrid popula- 

 tions of this cross would have shown heterosis for yield of ripe fruit per plant. 

 In such an event it seems highly probable that some and perhaps a con- 

 siderable amount of the increase in yield attributable to heterosis could be 

 obtained in inbred lines through recombination of genes. 



Considering the data for all the crosses listed in Table 19.8 the informa- 

 tion may be summarized as follows: In the Danmark X Red Currant cross 

 a large number of gene pairs differentiates the parents and individually the 

 genes have minor effects. The same is true of the Johannisfeuer X Red Cur- 

 rant cross with the exception that two or three pairs of genes have major 

 effects. In both the Danmark X Johannisfeuer and the Porter X Ponderosa 

 crosses weight per locule is differentiated by a comparatively few pairs of 

 genes having major effects. It is apparent that in the Porter X Ponderosa 

 cross it should be possible by selection in the segregating populations to ob- 

 tain by recombination of genes inbred lines equaling if not excelling the Fi 

 fruits in weight per locule. 



The discussions treating weight per locule and number of days from seed- 

 ing to first fruit ripe as component characters of yield of ripe fruit per plant 

 reveal that the recombination of genes to retain some or all of the advantages 

 of the Fi hybrid is analogous to recombination of genes for the purpose of 

 combining desirable characters. 



Linkage Relations 



Linkage may be an aid or a hindrance to gene recombination. The data 

 in Table 19.9 were computed to facilitate a consideration of the manner in 

 which different linkage relations may affect recombination of genes. 



Certain assumptions were essential to a calculation of the data. First, it 

 was assumed that the coefficient of coincidence is 1. Since in most cases there 

 is interference, to assume a coefficient of coincidence of 1 is to err on the 

 conservative side. For example, all the values given in the second row head- 

 ing (with the exception of the first and last) would increase as the coefficient 

 of coincidence became smaller. The reverse is true of the figures in the third 

 and fourth columns. The frequencies listed in the second, third, and fourth 

 columns of Table 19.9 are the theoretical number of individuals in the F2 pop- 

 ulation carrying the 12 plus genes in the homozygous condition. The cross- 

 over values expressed as decimal fractions are assumed to be equal for the 

 different sections of the chromosomes delimited by any two adjacent genes. 



The conclusions to be drawn from the theoretical data of Table 19.9 are not 

 invalidated by these assumptions. They merely serve the purpose of allowing 

 the calculation of theoretical values for illustrative purposes. Other assump- 

 tions such as different values of crossing over for the various sections of the 



