NON-MENDELIAN RESULTS 383 



seeds) and with white seeds (161). The mottled seeds when sown 

 gave rise to " coloured " and white-seeded beans. Of the 

 coloured seeds 52 per cent, were of the mottled " tortoiseshell " 

 appearance and 48 per cent, were of a uniform bronze colour. 

 This, again, is not a Mendelian result. The white seeds bred 

 true, and in the fourth generation all the mottled seeds bred 

 true, — the mottling having become apparently a dominant 

 character. 



There seems at present no reason to believe that the 

 Mendelian formula has more than a limited appUcation, though 

 it is of course possible that apparent exceptions may 

 eventually turn out to be less formidable than they seem. There 

 seems no reason why there should not be several formulas of 

 inheritance — each applicable to particular sets of cases, e.g. to 

 cases where blending does occur and to cases where it never 

 occurs. As the method of experiment is obviously the surest 

 line of progress, the more it is prosecuted the sooner will the 

 mists surrounding heredity disappear, but progress cannot be 

 secured by ignoring difficult cases or by straining the formula 

 in the eager desire to universalise it. 



Johannsen's Experiments on " Pure Lines." — Prof. 

 Johannsen of Copenhagen has made experiments on the inheri- 

 tance of quantitative characters, e.g. weight of seeds, in "pure 

 lines " of barley, beans, and other plants. By a " pure line " 

 he means all the descendants of a single plant, the mode of 

 reproduction being by self-fertilisation. The members of such 

 a pure hne showed " normal variability " in the weights of their 

 seeds, that is the variations in the weights, when plotted out, 

 gave the normal curve of frequency. 



Selecting a markedly divergent member of a particular pure 

 line, Johannsen bred from it, and found that its offspring showed 

 regression to the mean of the line from which it had been selected, 

 but not to the mean of the general population of beans. It 

 seems as if the inbreeding established a kind of sub-stock to the 



