464 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



equal conditions, even as between seeds grown in the same pot and 

 apparently all equally well ripened, is a difficult matter ; when this 

 is taken into account, as well as the fact that seed even in the same pod 

 is not always equally " good," it will be clear that differences in the 

 rate of germination are likely to occur quite apart from the single- or 

 double-producing quality already pre-existing in the seed. Further, it 

 may often be noticed that in a pot full of seedlings there are perhaps 

 only one or two laggards, whereas nearly half the total number in the 

 pot will develop into singles. Various sources of error would have to 

 be very carefully eliminated before statistics on this point could be 

 regarded as proving an inherent difference in germinating energy 

 between seeds giving rise to singles and doubles respectively. To 

 the possibiHty, however, that doubles are sometimes, at some stage in 

 their development, more precocious than the singles is a point to which 

 I shall return later. 



12. The method of essimplage practised by French gardeners 

 (advocated by Chat^). This method is based on the view that it 

 is possible to distinguish singles and doubles when quite young by 

 their vegetative characters. By this means it is claimed that a con- 

 siderable saving to the cultivator is effected, since he need not grow 

 so many plants to maturity which are useless from the market point 

 of view. According to CHAxfe * the leaves of the singles are dark 

 green, shining in the wallflower-leaved forms, and rounded at the 

 apex ; the heart, i.e. the centre of the leaf rosette, is of shuttlecock 

 form, and the plant is of a stumpy and compact habit. In the doubles 

 the leaves are very long, of a pale green, and curled at the edges. The 

 central leaves are whitish and rolled, and form a closed heart. He 

 adds that these characters are well marked, and with practice they 

 become easily recognizable at the first glance. It is far m.ore probable, 

 however, that these differences are features which distinguish one 

 strain from another rather than the singles from the doubles in any 

 one strain. We know this to be so in the case of one of the characters 

 cited, viz. that appearance of the leaf surface which is quite un- 

 connected with singleness and doubleness, as high a proportion of 

 doubles being obtained from the wallflower-leaved forms with their 

 glossy deep green leaves as from the hoary types which have leaves 

 of a greyish green. And the same is almost certainly the case as 

 regards leaf -shape and the general habit of the plant. Moreover, as 

 ViLMORiN t has pointed out, the method is not practicable where 

 many different strains are cultivated, each of which may be 

 characterized, in the singles and doubles equally, by certain sHght 

 differences in habit and leaf form. For my own part, I may 

 add that, after many years' experience, I do not find myself 

 able by this method to distinguish with certainty between the future 

 singles ;'and doubles. In the absence of any experimental data in 



* loc. cit. p, 67, 



f Les Fleurs de pleine terra. (See under Giroflee.) 



