J. Blakeman 
137 
this shrinkage is peculiar to the General Hospital Population, or extends to the 
normal population of a district. The only comparative results that we know 
Fig. 1. 
of are the regression lines calculated for A. 0. Powys' experience of New South 
Wales criminals by K. Pearson *. We have : 
English Males, Hospital Population : S = - -0378^ + G9-02 
New South Wales Males, Criminal Population : 8 = - -0337^ + 68-34 
English Females, Hospital Population : S= — '0655.4 + 66"()5 
New South Wales Females, Criminal Population: *S' = — '0361^ +63'61 
In all cases the unit of stature is an inch. 
Hence it would appear that the hospital population of males shrinks slightly 
more rapidly than the criminal population, and the hospital population of females 
very sensibly more. This is probably due to the larger proportion of male acute 
cases and of female chronic cases. We do not think the present results do more 
than confirm the conclusion of Powys' paper that there is for normally healthy 
individuals a shrinkage of a little more than ^" per ten years. 
(c) If we take in the next place head diameters we are not able to compare 
our General Hospital Population with Matiegka's data, for his paperf gives no 
clue to the ages of the individuals whose head measurements he has taken, and 
further those measurements are scarcely comparable with extei'nal measurements 
* Biometrika, Vol. i. p. 49. 
t Sitzberichte des k. bohmischen Gesellschaft d. Wissenschaften, Math. Naturw. Classe, Jahrg. 1902, 
pp. 1—75. 
Biometrika iv 18 
