Heredity in the light of recent research / by the late L. Doncaster.

  • Doncaster, L. (Leonard), 1877-1920.
Date:
1921
Catalogue details

Licence: Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)

Credit: Heredity in the light of recent research / by the late L. Doncaster. Source: Wellcome Collection.

    29/184 (page 13)
    II] VARIATION 13 as many ordinates and the curve would be nearly smooth. It is clear that a curve of this kind can be used for comparing the variability of different characters, for the greater the variability of the population the Avider will be the base ; consequently the curve for a very variable character will be relatively low and wide, that for a slightly variable one measured in the same scale will be tall and steep. A curve of this kind, which is quite similar on either side of the longest perpendicular (^median,' representing the modal value), may be obtained by plotting any measurements which vary fortuitously around a most frequent value, and such a curve is called a ' normal curve/ For example, if a large number of beans in¬ cluding equal numbers of white ones and black ones were placed in a sack, and drawn out ten at a time without selection of colour, most frequently five white and five black would be drawn, less often six of one colour and four of the other, more rarely seven and three and so on to the rarest case of ten of one colour. If the numbers of white beans in a draw are plotted along the base-line, and the ordinates represent the number of draws for each combination, a polygon approaching the normal curve will be obtained. Variation which gives a normal curve when plotted in this way is spoken of as normal variation. As mentioned above, the steepness of the curve is