188 



Life and Letters of Francis Galton 



measured fall into classes in which there are 7 to 24 repetitions*. But even 

 the group of 24 individuals could be separated out by taking finer divisions 

 of the head measurements than the three classes and introducing seven eye- 

 colour classes. I think Galton was not unnaturally critical of bertillonage, 

 because it started by theoretically asserting the independence of measurements 

 which he knew to be correlated^ ; it did in fact overlook one of his greatest 

 discoveries, the quantitative measurement of the correlation of bodily measure- 

 ments. Nevertheless Galton is fair to the results of the system : 



" It would appear from these and other data, that a purely anthropometric classification, 

 irrespective of bodily marks and photographs, would enable an expert to deal with registers of 

 considerable size... it seems probable that with comparatively few exceptions, at least two 

 thousand adults of the same sex might be individualised, merely by means of twelve careful 

 measures, on the Bertillon system, making reasonable allowances for that small change of pro- 

 portions that occurs after a lapse of a few years, and for inaccuracies of measurement. This 

 estimate may be far below the truth, but more cannot be safely inferred from the above very 

 limited experiment." (p. 163.) 



It may be remarked that Bertillon does not appear to have made even 

 such a limited experiment before he started his vast collection on the basis 

 of his "independence" dogma! 



Some account is then given of an American system of identification in the 

 case of recruits and deserters. It seems to be based on height, age (how 

 judged ?), hair and eye colours for indexing purposes and then on a careful 

 record of the body-marks placed on outline figures. Body-marks form of 

 course an important factor of bertillonage (pp. 164-5). Galton remarks that 

 no system he knows of appears to take account of the teeth. If teeth are 

 absent when a man is first examined, they will be absent when he is ex- 

 amined a second time. He may have lost others in addition, but the fact of 

 his having lost certain specified teeth prevents his being mistaken for a man 

 who still possesses them (p. 166). 



M. Herbette, speaking at the International Prison Congress in Rome, 

 remarked of bertillonage : 



" In one word, to 6x the human personality, to give to each human being an identity, an 

 individuality which can be depended upon with certainty, lasting, unchangeable, always 

 recognisable and easily adduced, this appears to be in the largest sense the aim of the new 

 method." 



Galton fitly remarks that these perspicacious words are even more ap- 

 plicable to the method of finger-prints than to that of anthropometry. 

 Bertillonage can rarely supply more than grounds for very strong suspicion, 

 finger-prints alone are amply sufficient to produce absolute conviction of 

 identity. 



t Some of the Bertillon measurements are indeed highly correlated. See Macdonell, 

 Biometrika, Vol. i, pp 202, 212. 



