DOUBLE FLOWEKS. 



477 



Definite proof of the correctness or otherwise of this view will be 

 forthcoming when we know the results of the reciprocal cross^ 

 double 5 x single ^ , evidence which has hitherto not been available 

 owing to the difficulty of procuring seed-bearing doubles, but which 

 I now hope in a short time to obtain. The exact proportion of doubles 

 obtainable from the singles is also not yet certainly established, though 

 there is some amount of evidence pointing to a proportion approximat- 

 ing in some cases to three single to one double, in others to nine singles 

 to seven doubles, the doubles appearing . in every case to be in the 

 minority. 



It is in Stocks, however, that we meet with the greatest com- 

 plexity as regards the inter-relationship of the factors concerned in 

 doubhng. Here the doubles, which contain an enormously increased 

 number of petaloid structures, are so double as to be completely sterile, 

 neither pollen nor ovules being present. In other genera where this 

 is the case the only means of propagation is by vegetative methods. 

 But the case of the Stock appears to be unique, in that the doubles 

 are obtained from the seed of pure-bred singles. The single and double 

 forms are sharply separated; there is no series of intermediate grades 

 to bridge the gulf between them. It is true that an individual may 

 occasionally be found on which, among the many flowers borne in a 

 season, one here and there may show one or even two extra petals, 

 but it is open to question whether this condition is connected with true 

 doubling. It does not, at all events, appear to have any hereditary 

 significance. For the seeds obtained from such flowers give the same 

 results as those from the normal flowers on the same individual. If 

 the individual has been bred from a double-throwing strain, then seeds 

 from either normal flowers or flowers with an extra petal will produce 

 some doubles. If, on the other hand, the individual belongs to a strain 

 producing only singles, then the seeds of all the flowers, whether 

 normal or not, produce only singles. 



My attention was first seriously directed to the investigation of 

 the inheritance of doubling in Stocks when, after breeding these 

 plants for several years for the purpose of determining the inheritance 

 of hoariness and flower colour, and always obtaining numerous 

 |;ioubles, I found one season, to my surprise, that among the whole 

 '^rop there was not a solitary double to be found. But, if on the same 

 ^ound, and under what appeared to be similar conditions, sowings 

 la some years yielded numerous doubles and in others none, it seemed 

 anlikely that conditions of nutrition or environment, as commonly 

 jield, should be the determining factor.'^ The sudden total absence of 

 loubles convinced me that current tradition concerning the appearance 



)f doubles in Stocks would prove to be without foundation, and that 

 Houbling, like the other characters previously investigated (see above), 

 ,YOuld be found to be inherited in accordance with definite laws, 



vhich breeding experiments, either by the pure-line method of culture 



' * Later it became clear tliat this result was due to the fact that the 

 amilies raised in this year happened all to be derived from those matings in 

 ^hich non-double-throwing individuals had been used. 



i 



