ignificant fact here is that this normal
variability exhibited by the traits of living organisms follows
closely the laws of chance or probability. That is to say, the number
of individuals occurring in any class which has a certain deviation
above or below the average, is directly related to, or dependent upon
(in mathematical terms, "is a function of"), the extent of the
deviation of the value of that class from the average of the whole
group. The significance of this is that the precise fluctuation which
we find in any individual is the result of the operation of a large
number of causes or factors, each contributing slightly and variably
to the total result.

[Illustration: FIG. 3.--Recorded measurements of the stature
of 1,052 mothers. The height of each rectangle is
proportional to the number of individuals of each given
height. The curve connecting the tops of the rectangles is
the normal frequency curve. The most frequent height is
between 62 and 63 inches. Average height--62.5 inches.
Standard deviation, 2.39 inches. Coefficient of variability,
3.8 (2.39 = 3.8+ % of 62.5 inches). (From Pearson.)]

Many of the most important facts about variability can be illustrated
by a simple model such as that suggested by Galton. This is a
modification of the familiar bagatelle board, covered with glass and
arranged as shown in Fig. 4. A funnel-shaped container at the top of
the board is filled with peas or similar objects (Fig. 4, _A_). Below
this is a regular series of obstacles symmetrically arranged, and
below these, at the bottom of the board, is a row of vertical
compartments also arranged symmetrically with reference to the chief
axis of the whole system. If we allow the peas to escape from the
bottom of the container and to fall among the obstacles into the
compartments below we find that their distribution there follows
certain laws capable of precise mathematical description, so that it
might be predicted with fair accuracy (Fig. 4, _B_). The middle
compartment will receive the most; the compartments next the middle
somewhat fewer; those farther from the middle still fewer; and the end
compartments fewest. If we connect the top of each column of peas by a
curved line we get just such a curve as that given by the stature
measurements above (Fig. 3), i. e., the normal frequency curve. A
curve of the same essential character would result from plotting the
dimensions of a thousand co