138 



JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



Turning again to the original cross Dorothy Eckford x Countess 

 Spencer, we find that this results in the generation in four families 

 having the following composition : 



I. CcYyXX II. CcYyXx 

 III. CcyyXX IV. CcyyXx 



Assuming each family to contain 64 individuals, this gives us a total 

 of 256 individuals of which one quarter will be homozygous for c and con- 

 sequently white. 



The others work out as follows : 



15 YYXX 1 



Y^X 



I. 6YYXx>51Y;X 

 30 YyXX J 



These are ' blue pinks,' Countess of Northbrook, and form 20 per cent, 

 of the whole. 



12YyXx (1)^ 



These are the salmon-pinks, Audrey Crier, and form 49 per cent, of the 

 whole. 



III. ISyyxx 



These are Helen Lewis (Thoday yyDDxx), and form 6 per cent, of the 

 whole. 



IV. 64 cc. 



Whites of various composition, Etta Dyke, forming 25 per cent, of 

 the whole. ■ 



The percentage then is as follows, out of every hundred plants on an 

 average we have 49 Audrey Crier, 20 Countess of Northbrook, 6 Helen 

 Lewis, and 25 Etta Dyke. 



So far this is all theory and book-work, and of no value unless borne out 

 by the actual results. Let us see, then, how my theoretical figures com- 

 pare with those sent me by Mr. Breadmoke as being the actual results in 

 the field. They are as follows : 



Estimated Actual 



Audrey Crier 49 per cent. 50 per cent. 



Countess of Northbrook . . 20 ,, ,, 10 ,, „ 



Helen Lewis 6 ,, ,, "a few " 



Etta Dyke 25 „ „ 25 per cent. 



I think Mr. Breadmore must have under-estimated the number of 

 Countess of Northbrook, otherwise his few " Helen Lewis will amount to 

 15 per cent, of the whole, or more than there are of Countess of Northbrook. 

 My 6 per cent, corresponds better with the expression a few," and from 

 my own experience I should say there were more than 10 per cent, pale 

 pink in an average sample of Audrey Crier. When the actual figures given 

 by Mr. Breadmore are compared with my estimates, one must admit that 

 there is a strong probability, at least, that my deductions are correct and 



