ANTHROPOMETRIC LABORATORIES 253
not yet "taken on," there was spare time for inquiry into finger-prints.
I described the results in the above-mentioned lecture so far as they had then been obtained, and subsequently in a more advanced shape in a memoir read before the Royal Society in 18gi [1171. It was argued in it that these patterns had a theoretical significance, which has not, I think, even yet been adequately appreciated, which bears on discontinuity in evolution. I showed that the different classes of patterns in finger-prints might be justly compared to different genera. As, however, they had been formed without any aid from natural selection, I concluded that natural selection had no monopoly in moulding genera, but that internal conditions must be quite as important.
I have always believed that the number of positions of stability in every genus must be limited, from which moderate deviations, but not great ones, are possible without causing destruction. There are limits which, if they can be overpassed without disaster, would require a new position of stability in the organisation. Comparatively few intermediate finger-patterns are found between a " loop " and a
whorl," these representing two different and wellmarked genera or positions of stability.
The modern division of views concerning the immediate causes of evolution, whether it be due to the slow accumulation of small factors or else by the sudden mutations of de Vries, are paralleled by those held by the physicists of the fifties on the method by which a glacier adapts itself to its bed, just as if it were a viscous body, which it certainly is not