Page:Darwinism by Alfred Wallace 1889.djvu/84

62 sufficient, which, however, is not often the case. The accompanying diagram exhibits the actual differences of size in five organs which occur in five species taken almost at random from this catalogue. Here, again, we perceive that the variation is decidedly large, even among a very small number of specimens; while the facts all show that there is no ground whatever for the common assumption that natural species consist of individuals which are nearly all alike, or that the variations which occur are "infinitesimal" or even "small."

The notion that variation is a comparatively exceptional phenomenon, and that in any case considerable variations occur very rarely in proportion to the number of individuals which do not vary, is so deeply rooted that it is necessary to show by every possible method of illustration how completely opposed it is to the facts of nature. I have therefore prepared some diagrams in which each of the individual birds measured is represented by a spot, placed at a proportionate distance, right and left, from the median line accordingly as it varies in excess or defect of the mean length as regards the particular part compared. As the object in this set of diagrams is to show the number of individuals which vary considerably in proportion to those which vary little or not at all, the scale has been enlarged in order to allow room for placing the spots without overlapping each other.

In the diagram opposite twenty males of Icterus Baltimore are registered, so as to exhibit to the eye the proportionate number of specimens which vary, to a greater or less amount, in the length of the tail, wing, tarsus, middle toe, hind toe, and bill. It will be noticed that there is usually no very great accumulation of dots about the median line which shows the average dimensions, but that a considerable number are spread at varying distances on each side of it.

In the next diagram (Fig. 10), showing the variation among forty males of Agelæus phœniceus, this approach to an equable spreading of the variations is still more apparent; while in Fig. 12, where fifty-eight specimens of Cardinalis