My chapters have become successively more removed from daily affairs. The early numbers printed by enthusiasts and bureaux generated the idea of statistical laws. Ideas about causation were revised. New content was given to the notion of normalcy. I have increasingly moved from practical matters to abstract ones. I shall conclude with the statistical epistemology and metaphysics of C.S. Peirce, a high-powered speculative philosopher if ever there was one. But the numbers that set these steps in motion were intended to be administrative tools. Lest we forget that, let us return to an example. I began this book with two anodyne moments in Prussian statistics: here is a third and more problematic one.

The ‘agitation’ of the epigraph was the wave of antisemitism that peaked in the new German Empire during 1879–81. We are here concerned with only one tiny aspect of it: the use or abuse of statistical data. As Salomon Neumann went on to complain ‘“the mass immigration of Jews across the Eastern frontier of the German Empire” has been quite simply erected into a statistical axiom. For the masses it summons up a nightmare, but it is no less effective in higher society, even in the learned world, where it is dressed up in economic or ethnological clothing, or some similar garb.’ Neumann subtitled his pamphlet ‘a chapter from Prussian statistics’.