142 



Mr. W. T. Thiselton Dyer. 

 Fig. 7. 



Colour. 



There is evidence that seminal variation as regards colours occurred 

 afc least as early as 1820, but the modern forms with large coloured 

 flowers, according to Mr. Martin, originated in a different way and 

 can be traced back to the old crimson and white. That preserves 

 the crimson ring round the throat, but is otherwise an albino. There 

 is nothing remarkable in this. Any species in nature may produce- 

 white flowers ; albinism is in effect the commonest of all variations. 

 " Griant white " (figs. 4 and 5) is a pure albino, in which the crimson 

 ring has been suppressed. 



The modern coloured forms were obtained in the first instance by 

 selecting forms in which the coloured ring showed a disposition to 

 spread into the white corolla segments. The first indication would 

 be a scarcely perceptible streak. By selection from self-fertilised 

 plants the streak was widened into a stripe. Continuing the process,, 

 the stripes united, and a uniformly coloured flower was obtained. 



The more striking colours, such as that of " Vulcan," which is a 

 dark crimson, were, however, obtained not by progressive selection r 

 but amongst the progeny of cross-fertilised plants. 



I learn from Messrs. Hugh Low and Co. that coloured varieties,, 

 of course when self -fertilised, come true from seed. This is in accord- 

 ance with a well-known principle.* 



* Darwin, ' Cross and Self -fertilisation,' p. 460, 



