XIII] Intermediates 239 



nature of these cases. To the ordinary observer the 

 *' Painted Lady " Sweet Pea with its whitish wings and red 

 standard might appear intermediate between a white and a 

 totally red type. But as experiment shows, the bicolour 

 type with its whitish wings is dominant to the self-coloured 

 forms, and its wings are white, not as the Dutch rabbit is, 

 owing to the omission of something, but on account of the 

 addition of another distinct element which suppresses the 

 pink colour in that particular part more or less completely. 

 The real white is of course due to the absence of one or 

 more of the colour-factors, an altogether different cause, and 

 it is therefore evident that to speak of the bicolour form as 

 in any way intermediate between self-colour and white 

 would be a complete misrepresentation of the facts. 



Similarly in the Rabbit, Hurst has shown that the 

 curious type known as '* English " pattern — something like 

 a Dalmatian dog but for the more or less regular disposition 

 of the spots — is a dominant to self-colour, though a careless 

 observer might guess it to be a transition-stage towards 

 albinism. 



In such cases again there is every likelihood that careful 

 selection might succeed in isolating subordinate types in 

 which the suppression attains particular degrees of com- 

 pleteness ranging within well defined limits, but the evidence 

 distinctly negatives any attempt to treat these several forms 

 as a continuous series in which any member is capable of 

 reproducing any other among its offspring. 



4. Intermediates as Jluctuational forms. Lastly there 

 are intermediates due to the disturbing effects of many 

 small causes not of genetic but presumably of environmental 

 origin. Such are the fluctuations in the weight of individual 

 seeds which Johannsen has studied with success in beans 

 i^Phaseolus). 



When the weights of the seeds of a single variety of 

 bean are marshalled statistically they arrange themselves in 

 a normal curve round the weight of greatest frequency. 

 Similarly when the seeds are sown and the seeds from the 

 self-fertilised individual plants which grow from them are 

 harvested separately, the crops from each individual again 

 can be grouped according to their weights in normal curves 



