294 



THE CONTINUITY OF THE RACE 



It is evident that there are far too many individuals like the original 

 parents and not nearly enough of the two new combinations. There 

 are, however, 783 (390 + 393) individuals that do show the new com- 

 binations, and they rule out the possibility that purple and long and 

 red and round are invariably associated with one another. 1 



A sweet pea that was homozygous for purple flowers and round pollen 

 grains was then crossed with one that was homozygous for red flowers 

 and long pollen grains. (Note that this second cross starts with the other 

 possible combination of flower color and pollen-grain shape.) Again, as 

 would now be expected, the F\ had purple flowers and long pollen grains, 

 but once again, when the actual ratios of the F 2 were compared with the 

 expectation for independent assortment, there were too many individuals 

 with the combinations shown by the parents — in this case, red and long, 

 and purple and round — and too few with the new combinations — purple 

 and long, and red and round. The actual and the expected ratios for the 

 419 F 2 individuals are given below: 



No satisfactory explanation of the F 2 ratios obtained by Bateson and 

 Punnett was found until after 1910. In that year Morgan and his col- 

 leagues at Columbia University began to find a similar lack of independ- 

 ent assortment in a number of crosses in the fruit fly Drosophila. Their 

 method of attack was different from that of Bateson and Punnett in 

 that they resorted to appropriate backcrosses to obtain a more direct 

 test of independent assortment and more readily to permit measurements 

 of the extent to which any kind of assortment might occur. They were 

 also particularly fortunate in selecting Drosophila as their experimental 

 organism. 



The backcross as a test of independent assortment. As we have 

 already seen, the direct effect of independent assortment is that the 

 dihybrid or polyhybrid Fi individual produces equal numbers of all the 

 possible sorts of combinations of genes. If we return for a moment to 

 Mendel's original dihybrid, it will be remembered that the F\ yellow- 

 round pea had the formula Gg Ww and that to give a 9:3:3:1 F 2 ratio, 

 it must have produced an equal number of GW, Gw, gW, and gw gametes. 



1 In such case, the 6,952 F 2 individuals should have shown a ratio of 5,214 purple- 

 long and 1,738 red-round and no others. 



