678 CRIMINAL ANTHROPOLOGY. 



The priuciple of the classification oi'M. Bertillon is to divide each one 

 of these measurements into three classes: tiie large, the small, and the 

 medium. This classification, beginning with the length of the head, 

 then to its width, extends through all the measurements indicated, and 

 ends in a division containing about ten cards, but which must not ex- 

 ceed twenty. The lines of demarcation between these divisions are 

 made arbitrarily and with the sole intent to make each division ap- 

 proximately equal in point of numbers. So he has found the numbers 

 for line of division for the length of skull to be at 184 and 189 millime- 

 tres. All heads the length of which fell between these two numbers 

 inclusive, constituted the middle divisi(m; all of 183 and less formed 

 the division of short heads, while all of 190 and more constituted the 

 division of long heads. 



For the breadth of the skull the two dividing figures were 153 to 156, 

 and these formed the middle division. Those 152 and less formed the 

 shortest, and those 159 and over formed the broadest division ; and this 

 system was continued throughout all other measurements. 



It was found in practice that this slight difference of 5 millimetres, 

 being only about one-fifth part of an inch, taken, as it were, out of the 

 middle of head measurements, would contain about an equal number 

 with those in the other two divisions. 



The divisions made by the measurement of the middle finger of the 

 left hand established for the medium class from 110 to 115; all middle 

 fingers from 109 and under are classed with the short ; from 110 to 115 

 with the medium, and 116 and over with the long fingers. So also with 

 the length of the foot, the spread of the arms, and, as I have said, by 

 the color of the eyes. 



In practice the 60,000 photographs would be first divided according 

 to the length of the head, large, medium, and small ; and this would 

 separate them into three divisions of 20,000 each, in the case of 

 drawers. The width of the head would again divide each one of these 

 20,000 into large, small, and medium, which would give practically 

 6,000. The three divisions arising from the spread of the arms and the 

 length of the middle finger will reduce it to 600. The length of the 

 foot will again reduce it to 63, and the further reduction by the color of 

 the eyes of seven classes to 9 photographs in each division. The prin- 

 cipal divisions are made in the cases of drawers, while the smaller are 

 made within the drawers themselves. 



The anthropometric establishment under M. Bertillon does not abolish 

 the use of photography. The photographs are taken in double, a full 

 face and a profile, and this should always be done. The change of face 

 arising either from accident or intention on the part of the subject is 

 much less easily controlled by him in i^rofile view than of the full face. 

 He can at best only change the lower part of his face, and in making 

 comparisons by photographs, where such a change is suspected, it is 

 well to cover the lower part of the face on the photograph by a spot of 



