462 JOURNAL OF THE ROYAL HORTICULTURAL. SOCIETY. 



season were raised from seed nine years old ; and my experience has 

 convinced me that new seed seldom produces double flov/ers." 



Recent experiments designed to test this point in Stocks show, 

 however, that although in one sense there appears to be some truth 

 in this idea, this was evidently not the sense in which these observers 

 intended their statements to be interpreted. As with lapse of time 

 more and more of the seeds lost the power of germination, these 

 writers concluded from their observations that more and more of 

 those that still survived became capable of producing double-flowered 

 plants.* Whereas, what appears to be the case is that seeds destined 

 to give rise to singles lose the power of germination somewhat earlier, 

 on the whole, than seeds yielding double-flowered plants. Hence 

 any increase in the proportion of doubles obtained from old seed 

 is to be attributed, not to any change gradually effected in course 

 of time in the surviving seed, but to the shghtly greater viability 

 of the double-producing seed. The capacity of a seed to develop 

 into a double or a single, as already stated, is a fixed quality, not 

 alterable by environment or treatment. 



7. Reduction of the number of pods which are allowed to mature 

 either by removal of the weaker flowering branches (advocated by 

 Chate),| or as the result of castration (method of Messer.J quoted 

 by Bronn § and Verlot and referred to by Chate,|| who emphasizes 

 the impracticabihty of carrying out this latter method on a large 

 scale, while Bronn adds that Messer's results were not confirmed 

 by Kress 



8. Cultivation under conditions which prevent luxuriant growth, 

 as, e.g., by the method of pot culture with a minimum amount of 

 watering. (Method adopted by German cultivators.) 



Both these methods (7 and 8) aim at the same result — a limited 

 production of seed, but achieve it by diametrically opposite means. 

 In the one case the maximum amount of nourishment is distributed 

 among a diminished number of pods, and in the other the production 

 of a small number of well-ripened pods is obtained by a method of 

 starvation. Both are found, however, to be equally without effect 

 on the production of doubles which, as, previously stated, is 

 independent of methods of nutrition. 



9. Selection for sowing of seed from locahzed regions of the mother- 

 plant, as, e.g., from the lower part of the pod, the upper portion being 

 discarded before the seed is extracted (advocated by Chate) ; or 

 from siHquas which are clustered close together in twos or threes 

 on the axis, in preference to those which occur singly and far apart. 



♦ See Chate, loc. cit. p. 73. 

 t loc. cit. p. 72. 



X " Die Kunst unfehlbar gefiillte Levkoyen zu ziehen." Vevhandl. d. Preuss. 

 Gartenbaii-Vereins, v. 1829, p. iSi. 

 § loc. cit. p. 77. 

 II loc. cit. p. 64. 



•jj Vevhandl. d. Pveuss. Gartenbau-Vereins, viii. 1832, p. 232. 

 ** loc. cit. pp. 73, 74, and So.j 



