J. F. Tocher 
307 
axis round which the skull swings when suspended from the uppermost points of 
the upper rims of the auricular orifices. This height is taken by some to the 
bregma. 
(3) Basi-bregmatic height, i.e. the height of the skull measured from the 
basion to the bregma. 
In (1) and (2) the highest point in the vault of the skull is determined by the 
German plane, and therefore these measurements may not be quite so satisfactory 
as (3), the basion and the bregma being two fairly definite anatomical positions. 
But in the living head, the conditions of (1), (2), and (3) are never reached. The 
centre or mid points of the ear holes are not positions so definitely ascertained as 
the uppermost position on the temporal bone of the external auditory canal, as 
indicated by the suspension of a skull on two pointers. The soft tissue of the ear 
yields readily to the slightest pressure, and, therefore, with an instrument having 
blunt or spherical ends for the ear passages, the greatest care must be exercised in 
order to avoid drawing the ear up. Any error arising through this however in 
the asylum survey would be small, as the greatest care was exercised in deter- 
mining the position. Probably the error is small also at the upper limit, although 
with the greatest care, one cannot expect the same precision as is obtainable 
with length and breadth, when it is remembered that the upper limit is " the top 
of the head, measured in a vertical plane when the eyes are directed to the 
horizon." What seem more important and real are the thickness of the scalp, 
the hair, and the slight variations in the pressure on the instrument. Thus it is 
difficult to say whether the positive skevvness in the distribution of head heights 
is due severally or jointly to (1) nature of the measurement, (2) nature of 
the instrument, and (3) to the observer, or (4) whether the positive skewness 
belongs to the character itself and is really in the nature of its distribution. 
To summarise, it is clear that the distributions of the various characters, 
whether the " entire insane " or the " general insane " population is considered, 
may be described with fair accuracy by skew curves, with the exception of L 
(</ and % "general insane") which may be fairly described by the normal 
curve. (See Diagrams I. to VII.) 
Further, it has just been shown that for long series, just as great divergencies 
from normality as exist among the " general insane," occur among the sane 
population, although greater divergencies are shown when the " entire insane " 
population is considered. It is not, however, established that there is not a differ- 
ence in the form of distribution between the sane and the " general insane." So 
far as Scotland is concerned this cannot be definitely determined until a corre- 
sponding general survey is carried out. This analysis and discussion thereon 
merely show that when long series are pitted against long series, quite as great a 
divergence from normality, as measured by the kurtosis and asymmetry, occurs 
among the sane as among the insane. The question whether there is really a 
difference in the form of distribution between the sane and insane must be left an 
open one when it is remembered that, after the striking " exceptionals " are 
