Recognized HTML document


at the Royal Institution (which was delivered May 25, 1888) on what is briefly called " Bertillonage " ; that is, on the system devised by M. Alphonse Bertillon for identifying persons by the measurements of their bodily dimensions. The subject was attracting much interest at the time, and had received a great deal of off-hand newspaper praise. There was, however, a want of fulness in the published accounts of it, while the principle upon which extraordinarily large statistical claims to its quasicertainty had been founded was manifestly incorrect, so further information was desirable. The incorrectness lay in treating the measures of different dimensions of the same person as if they were inde,pendent variables, which they are not. For example, a tall man is much more likely to have a long arm, foot, or finger than a short one. The chances against mistake had been overrated enormously owing to this error ; still, the system was most ingenious and very interesting.

I made the acquaintance of. M. Bertillon during a short visit to Paris, and had the opportunity of seeing his system at work. Nothing could exceed the deftness of his assistants in measuring the criminals ; their methods were prompt and accurate, and all the accompanying arrangements excellently organised. But I had not means of testing its efficiency with closeness, which would have required more time and interference with current work than was permissible. I was nevertheless prepared to give an account at the Royal Institution of what I had seen, but, being desirous of introducing original work of my own, I gave to my lecture the more general