EXPERIMENTAL EVOLUTION 



107 



Fig. S3- — Photograph of beans rolling down an inclined plane and 

 accumulating at the base in compartments, which are closed in front by 

 glass. The exposure was long enough to cause the, moving beans to appear 

 as caterpillar-like objects hopping along the board. If we assume that 

 the irregularity of shape of the beans is such that each may make jumps 

 either toward the right or toward the left in rolling down the board, the 

 laws of chance lead us to expect that in very few cases will these jumps 

 be all in the same direction, as indicated by the few beans collected in the 

 compartments at the extreme right and left. Rather the beans will tend 

 to jump in both right and left directions, the most probable condition 

 being that in which the beans make an equal number of jumps to the right 

 and to the left, as shown by the large number accumulated in the central 

 compartment. If the board be tilted to one side, the curve of beans would 

 be altered by this one-sided influence. In like fashion, a series of factors — 

 either of environment or of heredity — if acting equally in both favorable 

 and unfavorable directions, will cause a collection of ears of corn to assume 

 a similar variability curve, when classified according to their relative size. 

 Such curves, called Qu6telet's curves, are used by biometricians in classi- 

 fjang and studying variations in plants and animals. (Photo by A. F. 

 Blakeslee. Legend slightly modified from Journal of Heredity, June, 

 1916.) 



