220 Pigmentation Survey of School Children in Scotland 
the first sample; = number in first sample, V' — (1 ~'P') = 9.''i 2/' = observed 
frequency of the class in the second sample ; n = number in second sample ; 
N = total children in first sample ; and N' = total children in second sample. 
The negative sign indicates that the proportion of the class considered was 
less in 1903 than in 1896 and the positive sign that it was greater. The above 
results seem to indicate that the school population of East Aberdeenshire became 
darker haired to an extent which must be reckoned significant, and lighter eyed 
to an extent also significaut in the eight years' interval. Making allowance 
for any difference in method of observation, and comparing parish with parish, 
the results are however very similar. The difference lies chiefly in the results 
from the two towns in the division, Peterhead and Fraserburgh. 
(/S) Scottish Adults — The Insane. The colour results of the survey of asylums 
in Scotland are not directly comparable, since the observations were made on 
adults and since the group is a selected one and is not truly representative of the 
general population. All one can do is to note in a general way the agreement or 
otherwise of the two sets of data. The author has not found it possible to spare 
the time to estimate from the juvenile data the probable distribution of the 
ordinary adult population in each division or to deal in further detail with the 
colour characters of the insane. Moreover it seems more desirable — more satis- 
factory — to wait until the adults of the normal population are directly surveyed. 
Instead of the promised detailed comparison between the two sets of data, it seems 
sufficient to point to the leading features. Both sets of data agree in showing less 
divergency in densely populated parts. The excess of dark hair in the west found 
for the asylum population has been amply confirmed by the results of this survey. 
The region of excess of dark eyes in the asylum population has proved to be the 
same region for the general school survey. Perthshire, Stirling and Forfar are the 
counties constituting this region. The excess of medium hair is in both associated 
with density. The proportions of fair hair and red hair are small compared with 
the juvenile population. Excess of light eyes is somewhat similarly distributed, 
but is not so much south-west as the juvenile distribution. The region of excess 
of red hair is quite the same. Briefly, while it would be useless to compare the 
relative frequencies of the two sets of data for the reasons ah'eady stated, still when 
the local class frequencies of each set are compared with each set's own general 
population, they show on the whole the same significance. It has been shown 
that the colour distribution of the insane population as a whole cannot represent 
the general distribution of the sane in one important particular, namely, in the 
frequency of the light-eyed class. Regions of excess of insanity from the general 
average are associated with regions of excess of light eyes, thereby increasing the 
proportion of light eyes in the general insane population beyond the expected 
amount for the general adult sane population*. 
* Biometrika, Vol. v. pp. 298—350. 
