198 CAUSES OF VARIATION 



unless the tendency shall be held in check by the greater attract- 

 iveness of lighter eyes, preferential mating. This would be a 

 long and slow process, but it would avail much to reduce, though 

 it could never overcome, the effects of the higher fertility of the 

 darker-eyed individuals. 



Pearson collected 4443 capsules of wild poppy. 1 They showed 

 the following distribution arranged according to the number of 

 stigmatic bands : 



The largest number of capsules (954) had 13 bands and the 

 next largest number had 12. Very few had so many as 18 or 19, 

 or so few as 5, 6, or 7. The type number of bands is then 13. 



He provided receptacles and kept the seeds of each group 

 separate. He says : 



To my great surprise, however, my receptacles for 12 and 13 were 

 filled up with the contents of very few capsules, those for 1 1 and 14 more 

 tardily, those for 10 and 15 only with emptying a great number of capsules, 

 while I could hardly get any seed at all from those capsules with very many 

 or very few bands ; they were practically sterile. The type capsules were 

 enormously fertile, [while] those with even a moderate deviation from it 

 [were] relatively or even absolutely infertile. 2 



This being true, the poppy has become about as stable as is 

 possible, for its highest fertility is with its most numerous popu- 

 lation. This plant was growing wild in nature. Obviously the 

 great bulk of seeds distributed would be of the type number, 1 3 

 or near it, and the mass of descendants would arise from seeds 

 close to the type. What chance now would there be in nature 

 for a large colony of six-or seven-banded strains to arise ? Very 

 little, unless they happened to possess some decided advantage 

 in the struggle for existence, in which case the type would 

 speedily shift in that direction; but as long as "the highest 

 fertility remained with the higher number of bands, the race 

 would be unstable. 



1 Pearson, Grammar of Science, pp. 443-444. 



2 Ibid. p. 444. 



