468 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



singles, and every such single gives in the next generation a mixture 

 of singles and doubles, the proportion being on the average three 

 singles to one double. 



Where the cross is made in the reverse way, the ever-sporting single 

 being used as the female parent and the pure-breeding type as the 

 male, the offspring are again all single, but of two kinds in respect of 

 behaviour. Rather less than half yield posterity which in each 

 generation are all single ; rather more than half give a mixture of 

 singles and doubles in the proportion on the average of about three 

 singles to one double, and these singles are again of two kinds, as 

 in the first generation derived from this cross. 



We may express these results graphically as in the Table II. on 

 p. 469:— 



The foregoing considerations have shown that it is essential for the 

 maintenance of a high percentage of doubles that the initial material 

 should consist entirely of ever-sporting individuals, since the effect of 

 intercrossing with pure-breeding or cross-bred individuals will be always 

 in the direction of lowering the proportion of doubles. And further, 

 that once such a culture of ever-sporting individuals is obtained the 

 percentage of doubles will be automatically maintained irrespective 

 of the method of cultivation so long as interbreeding with pure-breeding 

 or cross-bred singles does not occur. If these two conditions are ful- 

 filled the grower may count on the maximum return in the way of 

 doubles without further intervention on his part. The question 

 which he will naturally ask at once is — ^What is this maximum ? The 

 answer to this question presented a problem of vital and perennial 

 interest to the gardener so long as the behef remained unshaken that 

 this maximum depended upon environmental conditions, and might 

 be increased artificially by appropriate means, if they could only be 

 discovered. Now that we know that the output of doubles is definite, 

 that it is an inherent character of the plant, the solution of this problem 

 lies with the student of genetics rather than of horticulture. 



Considering the very haphazard results that must have been ob- 

 tained by the early horticulturists (see above, p. 457), it is not surprising 

 that we meet with few generalizations in their accounts of Stocks 

 as to the relative proportion of singles and doubles. In an Histoire 

 des Plantes de VEurope which appeared in 1689 it is stated under the 

 head of Leucoium album that the plants are of different colours, white, 

 red, and yellow (yellow here refers, as usual, to the Wallflower), and that 

 all three are to be found in gardens, as many doubles as singles. Here we 

 have a very close approximation to the truth. Green,* speaking of 

 the biennial forms of the Stock-GiUifiower, says that " if the seeds be 

 well chosen three-fourths of the plants will probably be double." 

 This may be merely a rough estimate, erring in the direction of a too 

 great excess ; even if, as is unlikely, it is based on counts from the 

 flower-beds, we may still, as we shall see presently, hesitate to accept 

 it, without further evidence, as a rehable indication of the actual out- 

 put of doubles. Chate f quotes 50 per cent, as usual for a number 

 * loc. cit. . t loc. cit. 



