EXPERIMENTAL EVOLUTION 



103 



83. Two Kinds of Variation. One of the first results 

 of de Vries's painstaking work was the demonstration of 

 what he believed to be a fundamental difference between 

 two distinct kinds of variation continuous (or fluctuating) 

 and discontinuous (or saltative, i.e., leaping). 



84. Continuous Variation. Continuous variation is 

 quantitative a case merely of more or less. It deals with 

 averages. Some flowers on a red-flowered plant may be 

 lighter or darker red, but, in a series of generations, the 



FIG. 50. Fluctuating variation in the leaves of an oak (Quercus chry- 

 solepis), a, all the leaves of a twig; b, younger leaves of a twig; c, con- 

 secutive leaves; d, some leaves on one season's growth of a twig. (After 

 Copeland.) 



average of a large number in each generation does not 

 vary, and the departure from the average never exceeds 

 certain limits. The flowers of a given species may have a 

 certain characteristic odor, but the odor may be stronger 

 in some flowers than in others, or in some individual 

 plants than in others. The plants grown from a handful of 

 beans of the same variety may vary in height within 

 limits, but the average^ height of a large number will not 

 vary in successive generations, and will be characteristic 



