REDUCTION OF VARIABILITY. 133 



vespertina. After two years the oaks had crowded out all the 

 Lychnis plants from this field, and it was impossible to find any 

 similar individuals in new fields in the neighbourhood. One day, 

 however, plants may originate which would not differ from one 

 of these two species by having the colour, or the hairiness oi 

 the other, but in a new physiological character, a constitution 

 which made them fit to live in marshy spots. Such plants 

 would almost certainly find the exact environment to which 

 they were adapted by chance, and in that environment they 

 might live for a time. If their constitution happened to be fav- 

 ourable, they might even continue to exist and a new species 

 with its own genotype and set of characters would have origin- 

 ated. 



As Bateson pointed out, there need not be anything in the 

 striking characters wihch distinguish diurna from vespertina, 

 in the red colour of the flowers, in the curvature of the capsule- 

 teeth in the absence of glands, which make the former better- 

 fitted to a dry and sunny environment than vespertina. All these 

 things are simply consequences of the genotype for which the 

 plants happen to be pure. 



Is it true that a new form, which happens to be far in the mi- 

 nority will disappear into the multitude? 



Let us take case of a dioecious plant like Lychnis vespertina. 

 We find a female plant which has a gene less than the common 

 plants. Her daughters will be heterozygous for the gene, for aD 

 the males have it. And the children of these daughters will be 

 pure and heterozygous in equal numbers. If every plant has 

 one daughter, the daugther will be pure for half the number oi 

 genes for which the mother was still heterozygous. In every 

 following generation the number of genes for which a plant is 

 still impure, is reduced by one half. The same holds true for 

 genes which a plant posesses more than the common run oi 

 plants of the group. Therefore we can say, that if in a popula- 

 tion of habitually crossing plants (or animals) an individual is 

 introduced which differs from the multitude in a number of 

 genes, 2q, the population will again be pure in n generations. 



