348 
Anthropometry of Scottish Insane 
to insanity. With regard to the remaining colour characters there is no material 
difference between the sane and the insane. 
II. In the one local district where the adult sane and insane populations can 
be compared — that of Aberdeen — it is found to agree, with respect to pigmenta- 
tion, with the general conclusion just stated. With regard to measurable charac- 
ters, the local sane population is broader headed and taller than the local insane. 
III. In the entire insane population there is a group whose characters are 
affected b}^ special causes not characteristic of insanity in general. This group 
has on that account been excluded from the general analysis. The group is quite 
different iu type from the "general insane" population, is both macrocephalic and 
microcephalic in character, and thus shows excessive variability. 
IV. {«) The distributions of the various characters in the general insane 
population are distinctly skew, with the possible exception of head length, which 
may be fairly described by the normal curve. Further, the distributions are 
leptokurtic and negatively asymmetric. For long series of the same characters, 
just as great divergences from normality occur among sane populations as are here 
found in the long general insane series. Asymmetry in distribution therefore is 
not a special feature among the insane. A difference in form, however, may exist 
between sane and insane populations. If it existed, it would be detected by a 
general survey of the sane population. In any case, heterogeneity would account 
for much of the asymmetry, and heterogeneity (see V. and VI.) has been found to 
exist among the insane. 
(b) There is a probably significant departure from linearity of regression 
among the males in the pairs of characters L &c B, B H, and L ki S. Otherwise 
the regression is linear. The values of the correlation coetficients are somewhat 
higher iu the entire insane population. In the general insane population the 
values approximate to those already found for the same pairs of characters among 
the sane population. 
V. Assuming the insane population in the various districts of Scotland to be, 
with respect to measurable characters, an anthropometric sample of each district, 
we find that local populations differ from each other sensibly in many respects. 
(a) The south-west of Scotland (exclusive of Glasgow) is long-headed or 
macrocranial, the north-east is short-headed or brachycranial. The north of 
Scotland is broad-headed or platycranial. Glasgow, Edinburgh, and the populous 
centres round them are stenocranial or narrow-headed. Again, the north of 
Scotland is distinctly brachycephalic, while the south-west is distinctly dolicho- 
cephalic. In the large towns (excepting Edinburgh, which approximates to the 
mean) the population is microcephalic or small-headed. The Scottish Midlands, 
excluding towns, are macrocephalic or large-headed. The rest of the country 
approximates to the average size — the poptdation is mesocephalic. In head height, 
there are two sharply divided groups — a hypsicranial or a high-headed group in 
