138 Mr. W. T. Thiselton Dyer. 



Fig. 2. 



conditions of a single species. Further, in the following statement, 

 I have confined myself to the result of continued self-fertilisation, 

 and have not thought it necessary to investigate the results of 

 crossing races which have assumed characters more or less dis- 

 tinct. 



Size. 



Mr. Martin strongly insists on the principle laid down by Mr. 

 Darwin from De Vilmorin, that "the first step is to get the plant 

 to vary in any manner whatever." * As Mr. Martin puts it, " the 

 breeder must work with nature." It is his practice to seize the 

 smallest deviation, even so small an indication as the slightest differ- 

 ence in a cotyledon of a germinating seed. The first direction of 

 work would, however, for commercial purposes, be to develop the 

 size of the corolla. Figs. 3 and 4 show two stages which have been 

 reached by progressive selection from " crimson and white." 

 Messrs. Sutton have sent me photographs of the largest flowers 

 hitherto produced by them. Fig. 5 is copied from one of these. 

 The vertical depth is 3 in. This is more than double that of the 

 form with which they started ; the increase in breadth of the 

 segments is at least six times. This represents the continuous 

 work of forty years. As the work was not done for a scientific 

 purpose, the whole of the progressive steps have not been preserved 

 or recorded. Only saleable stages have survived. But Mr. Martin 

 emphatically denies that they have been attained by other than pro- 

 gressive ^selection or that they have been reached by leaps and 

 bounds. } In developing any particular character it is, to use his 

 own words, always done by a "ladder," i.e., continuous self-fertilisa- 



* ' Animals and Plants under Domestication,' vol. 2, p. 262. 



