178 
On Ci'lminal Anthropometry 
PART I. 
Material and Methods. 
(1) The object of this memoir is threefold : 
(i) To test to what extent the criminal classes diverge in physical characters 
from other classes of the community. 
(ii) To consider how far the shorter methods recently proposed by Professor 
Karl Pearson for finding the variability and correlation of characters in the case of 
normal frequency may be applied to some of the chief anthropometric measure- 
ments now customarily made, and 
(iii) To determine what is the best manner in which these measurements can 
be applied to the identification of criminals. 
I shall first consider the material I have had at my disposal ; I shall then 
indicate the methods I have used for the determination of its metrical constants, 
and finally apply my results to the consideration of the above three problems. In 
the course of my work I shall have to consider the important point of the homo- 
geneity and normality of my material, and apply three separate tests: 
(1) If broken up into groups, the statistical constants of a sub-group ought 
to remain, within the limits of random sampling, the same as the bulk of the 
observations. 
(ii) Frequency distributions, if plotted and fitted with frequency curves, 
ought to give nearly normal distributions. 
(iii) The regression lines in type cases ought to be closely represented by 
straight lines. 
(2) Nature of material used. The data on which the memoir is based were 
obtained, through the kindness of Dr Garson, from the Central Metric Office, New 
Scotland Yard, where the register of habitual criminals is kept, and their identifi- 
cation effected. In an interesting paper published in the Journal of the Anthro- 
pological Institute, Vol. xxx. 1900, July — Dec, Dr Garson has explained in 
detail the metric system of identification in force in England, and the sort of 
prisoners whose metric description is registered at the Central Office, and whom 
we may call briefly " habitual " criminals. 
But in addition to the official forms on which the descriptions of habitual 
criminals are recorded, the Central Office possesses a number of "practice" forms 
which record the metric description of a less pronounced type of criminal than the 
"habituals" — prisoners whose crimes and sentences are comparatively slight, and 
who may be called " non-habitual." These practice forms are filled up by warders 
