482 JOURNAL OF THE ROl^AL 'HORTICULTURAL SOCmTY. 



amongst the last survivors the proportion O'f doubles may be higher 

 than among samples oi fresh seed. But this method of attaining his 

 object will scarcely commend itself to the gardener, for he must still 

 sow all his seed (since he cannot distinguish by inspection between 

 the seeds which have died and those which still retain the power of 

 germination), though his return at the best will be perhaps a plant or 

 tw^o here and there, and may be nil if the lapse of time has been too 

 great. 



So far then as Stocks are concerned, the position now reached is 

 one which, if it precludes any hope of further advance in one direction, 

 at least removes uncertainty and 'fear of possible deterioration in the 

 other. If, on the one hand, we must relinquish the expectation of 

 being able to raise the average yield of doubles by pursuing special 

 methods of cultivation, we can, on the other, feel assured that an 

 individual or strain giving the maximum output of doubles for that 

 strain will continue to yield this maximum indefinitely, however 

 unfavourable the conditions, provided that intercrossing, except with 

 similar strains, is prevented. 



Whether in other genera the degree of doubling, as well as the 

 proportion of individuals showing doubleness, will be found in all cases 

 to be inherited independently of the environment, or whether cultiva- 

 tion may have some effect upon the particular grade oi doubleness 

 which is reached, is a question which cannot yet be answered with 

 certainty. 



