Page:Proceedings of the Royal Society of London Vol 69.djvu/217

Rh gather the impression that he regards these differentiations as largely recognisable and capable of exclusion. He may hope too that by increasing the area of his statistics these orderly disturbances may cancel each other. This appears to me highly improbable.

Order in occurrence is generally the only indication of differentiation, and when the order is obscured, differentiation may pass wholly unobserved. But the presence of such differentiation will vitiate the result, even if the area of statistics be indefinitely increased.

The only answer which seems open is that though it may be impossible to define precisely in words which examples should be reckoned in determining average homotypic correlation, and which must be excluded as showing differentiation, yet in practice the difficulty is not a real one, and that divers features (e.g., regularity of occurrence) enable us to detect sensible differentiation.

Such an answer is far from covering the whole ground of the objections I have indicated, as may readily be seen by attempting to apply it in the practical illustrations given above. But besides this, to take that ground would be to turn back from that appeal to rigid numerical treatment, which Professor Pearson has told us should be the sole test of these hypotheses.

I may further point out that if it were suggested that the distinction between differentiation and variation may be left to the judgment of the observer ; we might by a similar exercise of judgment attempt a distinction of variations into evolutionary or specific, and normal.

This is far more than a merely logical point. I am disposed to think that such a rough classification boldly made and carried out for a number of familiar forms might greatly promote the study of evolution, even though no precise criterion can yet be provided. This suggestion will be abhorrent to many naturalists, though for want of such a distinction much of the statistical work produced by Professor Pearson and his followers has, I believe, gone wide of its mark, if that aim is the elucidation of Evolution. More fitly might this work be described as " Mathematical Contributions to a Theory of Normality."

In the treatises I have referred to Professor Pearson is seeking for a statistical conception of Species through an examination of miscellaneous variations. The impression left on my mind by such imperfect study of his works as I have been able to make, and especially by the present paper, is that the evidence points to some conception of nor- mality to be otherwise attained, a conception more finite and concrete than any we have yet reached.

By the one word Variation we are attempting to express a great diversity of phenomena in their essence distinct though merging in-