152 LOCK : STUDIES IX PLANT BREEDING 



per cent, of blue grains — Mendelian expectation for hetero- 

 zygote plants in which blue is dominant. Those plants 

 which lie first on the list. (3) showing percentages of blue 

 well above 50 per cent, must be the produce of homozygote 

 grains in which blue is not invariably dominant. Ami 

 towards t he end of the table percentages of from 22 to 37 

 indicate the produce of heterozygote grains of the corres- 

 ponding kind (4) with dominance of blue in the case of 

 certain grains only. The only difficulty which presents itself 

 is that of definitely separating the second group of plants 

 from the third ; we may perhaps make an arbitrary division 

 at the value 55 per cent. 



The result may then be summed up in a table — 42. 



Table 42. 



Percentage Number of F- Observed behaviour of 



Grains. offspring in F3. 



Blue: 7*5 J _ .. Blue exclusively (I). 



u . 5 s 2-2 [expectation 2.-,) _ _ 0ver 5 - 0/q blue (3)< 



1 3-3 ) . . About 50 % blue (2 ). 



L9-7 V 52 {expectation 50) .. 22-37 % blue (4). 



White: 19 j .. 1-36 % blue. 



26 . . [expectation 25) . . White exclusively. 



From the arrangement of the figures in the above table it 

 is clear that in spite of the great variability in the propor- 

 tions in which the blue grains appeal in F 8j the Mendelian 

 ratio can -till be clearly traced out for V 2 ; 22 : 52 : 26 being 

 a near to I ; 2: 1 as could reasonably be expected considering 



the con 1 pa rat i\ ely small number of plants which it was possible 



to examine. 



In tit.'. 2 the black curve represents for the offspring of 

 blue grains the number of individuals which showed different 

 ..I blue, those with Kid per cent, being omitted 

 The doited curve gives the same information with regard to 



the 1 'II ipi ing of white grains ( ESxpt. 65), 1 hose which showed uo 

 blue grains beinj '.nutted. 



