120 THE RATE OF GROWTH [ch. 



niits, the florets of a daisy, the stripes of a zebra, the- nearness of 

 shots to the bull's eye*. It thereby illustrates one of the most 

 far-reaching, some say one of the most fundamental, of nature's 

 laws. 



We find the curve of error manifesting itself in the departures 

 from a mean value, which seems itself to be merely accidental — 

 as, for instance, the mean height or weight of ten-year-old English 

 boys; but we find it no less well displayed when a certain definite 

 or normal number is indicated by the nature of the case. For 

 instance the Medusae, or jelly-fishes, have a "radiate symmetry" 

 of eight nodes and internodes. But even so, the number eight is 

 subject to variation, and the instances of more or less graup them- 

 selves in a Gaussian curve. 



Number of " tentaculocysts'' in Medusae {Ephyra and Aurelia) 

 [Data from E. T. Browne, QJ.M.S. xxxvii, p. 245, 1895) 



The curve of error is a "bell-shaped curve," a courbe en cloche. It 

 rises to a maximum, falls away on either side, has neither beginning 

 nor end. It is (normally) symmetrical, for lack of cause to make it 

 otherwise; it falls off faster and then slower the farther it departs 

 from the mean or middle line; it has a "point of inflexion," of 

 necessity on either side, where it changes its curvature and from 

 being concave to the middle line spreads out to become convex 



* "I know of scarcely anything (says Galton) so apt to impress the imagination 

 as the wonderful form of cosmic order expressed by the Law of Frequency of 

 Error. ... It reigns with serenity and in complete self-effacement amidst the 

 wildest confusion" {Natural Inheritance, p. 62). Observe that Galton calls it the 

 "law of frequency o/ error," which is indeed its older and proper name. Cf. (int. al.) 

 P. G. Tait, Trans R.S.E. xxiv, pp. 139-145, 1867. 



i 



