no HEREDITY AND EVOLUTION IN PLANTS 



may spring at one leap, not by gradual changes, from those 

 that are fragrant; in one generation the factors controlling 

 height are so altered that, in successive generations, the 

 average of height may change by either more or less, so 

 that the heights of the individuals fluctuate about a new 

 mean. In other words, we recognize a second type of 

 variation — not the fluctuation of individuals about an 

 unchanging mean, but the appearance of a new mean, 

 about which the given character in individuals may 

 fluctuate (Fig. 54.) 



When discontinuous variation proceeds along a definite 

 line through several successive generations, each step being 

 an intensification of the preceding one, it is designated 

 " orthogenetic saltation" (Fig. 55). 



88. Illustration of the Pendtilum. — The difference be- 

 tween discontinuous and fluctuating variation may be 

 aptly illustrated by a swinging pendulum (Fig. 56). The 

 vertical position, assumed when at rest, is i he mean of all 

 positions that may be assumed as the pendulum swings; 

 the oscillation about this mean illustrates continuous or 

 fluctuating variation. 



But we may conceive that the point of suspension of the 

 pendulum changes, as shown in the figure. The pendulum 

 continues to oscillate, but now about a new mean position; 

 a new character has been introduced, with its own fluctua- 

 tions of more or less. 



89. Mutations. — Darwin, as well as others before and 

 after him, recognized both kinds of variation, but de Vries 

 was the first to work out in detail the hypothesis that 

 discontinuous variations furnish the material for natural 

 selection. Discontinuous variations he called mutations; 

 plants which give rise to or "throw" them are said to 



