VARIATION AND HEREDITY 199 



with other individuals. These may be described 

 and catalogued, but words alone will hardly suffice 

 to discriminate the finer shades of distinction be- 

 tween so many classes. To seek order in such a 

 chaos, some sort of mathematical basis must be 

 devised. 



If we study a group of a hundred men with regard 

 to a single character, such as stature, we find, of 

 course, that all the individuals fall within rather 

 definite limits, from the shortest man to the tallest, 

 and we might classify them by arranging them in a 

 row in the order of height. The line connecting the 

 tops of the heads of such a row of men should be 

 irregular and jagged and would defy analysis. Sup- 

 pose, however, that we group such a lot of men in 

 classes corresponding to the various statures, and 

 place a representative of each class with his heels 

 on a base-line. Then, grouping all of a class together 

 (see fig. 71), one in front of another, we would find 

 that the line connecting their heads, when viewed 

 from above, is of a very different sort compared with 

 the former one. Briefly, the shortest rows, that is, 

 the fewest individuals, would be found in the shortest 

 and tallest classes, and the longest rows in the inter- 

 mediate classes. Viewed from above, the outline 

 marked by their heads would describe a fairly regular 

 curve, reaching its highest point in the middle, and 

 curving down to the base-line in both directions. If 

 a thousand individuals instead of a hundred were 

 thus" arranged, the line would be more even, since 

 individual differences would tend to merge in the 



