Page:JehuTJ 1902redux(1).pdf/8

7 is to account for this determinate variation that the theory relating to some force at work prior to selection - variously called "growth-force", "self-adaptation", etc., - has been propounded.

Galton was the first to apply the statistical method to the study of variations and his work has been further carried on by Karl Pearson, Weldon, and others. Some measurable character is taken - as stature in man - and the common or average height is noted, for instance in the recruit for the Army. The departures from the normal in the two opposite directions are counted and the results represented graphically. In a great many cases the departures from the middle form or are even or symmetrical. This is the case with stature. As the variations become great they are seen to be rarer. In fact the variations about the correspond with the laws of chance. The variability can thus be displayed graphically by a curve and a mathematical expression can be found to represent this curve. So it is possible to arrive at the measure of variability displayed by the organism. The variability is greater the flatter the curve is. Cases dealt with in this way are especially numerical cases, such as the number of stamens and carpels, the ratio of length to breadth, etc. Cases are found, however, where the curve is not symmetrical - for instance, the strength of sugar in beet-root. Whenever the curve is asymmetrical there must be something other than mere chance at work. Cases occur also in which variations