54 VARIATION. 



tween the two kinds of variation, because variation as induced 

 by geno- variation is something you can work with, it furnishes 

 the material for selection, and variation as induced by the 

 environment is irrelevant from the plant-breeder's view-point. 



The variability in the shape of the leaves of the ordinary dan- 

 delion, Taraxacum officinalis is enormous. This is surprising if 

 we take into account the fact that the seed in this plant is pro- 

 duced parthenogenetically or apogamously. No good pollen 

 seems to have been found in the dandelion. It seems very 

 likely, however, that occasionally fertilized seeds are produced, 

 and that new types result from such occurrences. On the other 

 hand, a verygeat part of the diversity in shape of leaves and of 

 the rozettes, must be due to variability of the non-genetic fac- 

 tors. This became apparent in some selection experiments. One 

 of us harvested seed from six plants in a garden in Berkeley af- 

 ter cutting-off the upper part of the bud with all the stigmas. 

 More than half the buds so treated produced an abundance of 

 normal seed. Two typical large leaves of each plant were har- 

 vested at the same time, and their shape was recorded by 

 making actual blue-prints directly from them. (Fig. 7). 



Four sets of plants were raised to maturity. In each of them 

 there was considerable variability in shape of the leaf. From 

 each family two extreme plants were chosen. Seed from these 

 two were harvested, and two leaves of each were blueprinted. 

 Next season a series of from twenty to twenty- five plants from 

 each lot were grown in pots under glass in the experimental 

 gardens of the firm of Vilmorin in Verrieres, France. In every 

 case, the progeny of two very different sister-plants proved to 

 be identical. Reproductions of the blue-prints will be clearer 

 than any description. As can be seen from the two series pub- 

 lished, the effect of the selection was zero. The plants with the 

 most deeply serrated leaves in the third year happened to occur 

 in the progeny of the least serrated mother-plant. 



Within one clone we are dealing exclusively with variability 

 induced by the environment, if we exclude for the present 

 cases of vegetative segregation in heterozygous individuals. As 



