ON SOME HYBRIDISATION EXPERIMENTS 179 



Hybridisation II. 



502 a, white, X 501 a, lilac. 

 Fi: 40 plants, all with lilac flowers. 

 F2: 11—9: lilac 16, white 19. 



Hybridisation III. 



502 c, white, X 501 a, lilac. 

 Fl'. 2 plants, both of them with lilac flowers. 

 Fo-. no plants. 



Totals. 



Fi: 84 plants, all with lilac flowers. 

 F2: lilac 27, white 24. 



All of the lilac-flowered plants had red stems; the stems of the 

 white-flowered ones were green or slightly tinged with red at the base 

 of the stem and at the nodes. The figure in the parenthesis in the 

 hybridisation I gives the plants that flowered, the other figure gives 

 the number of plants with red-coloured stem, which surely would 

 have shown lilac flowers had not some of them died before the flowering. 



As all the 84 Fi-plants in the three hybridisations were lilac- 

 flowered, the lilac colour behaves as dominant to white. The ratio 

 of the lilac-flowered plants to the white-flowered ones in the three Fo- 

 generations, taken together, was 27 : 24 = 8,47 : 7,53. On the 3 : 1 

 ratio the theoretical numbers would be 38,25 : 12,75 and the probable 

 error dtz 3,092. The difference between the calculated and the actual 

 numbers being + 11,25 and thus nearly four times the probable error, 

 it is not probable, though not quite impossible, that the segregation 

 followed this ratio. However, the numbers strongly suggest another 

 theoretical ratio, namely that of 9:7. On this ratio the theoretical 

 numbers would be 28,6875 and 22,3i25 with the probable error dt 3,543. 

 As the difference between the actual and the theoretical numbers is 

 rb 1,6875 and thus not even half the probable error the facts are very 

 well in accordance with the theoretical values. 



The ratio 9 : 7 is explained through the hypothesis of two genes, 

 A and B, which when both are present produce lilac colour in the 

 flowers. When only one of them is present no visible effect in the 

 flower-colour is produced, and the flowers become white. The lilac- 

 coloured parent plants would then have been AABB, the white ones 

 probably aabb. However, it is probable that one of the genes, let it 

 be A, when alone has a visible effect on the stem colour, as some 



