112 Relationship of Intelligence to Size and Shape of Head 
Naturally I do not insist on any particular part of this scale. The numbers 
are entirely round numbers, and are based on the "normal" distribution of the 
frequency of intelligence. Still, it is deduced from three series covering the classi- 
fication of between 4000 and 5000 cases, and the three separate results are in general 
accord. It will, I think, be possibly useful for other enquirers, and it endeavours 
to give quantitative expression to our verbal definitions of the intellectual cate- 
gories. One or two points are suggestive. While the specially able men, the 
first class in academic judgment, are fewer than either the capable men or the 
fairly intelligent men, they cover a range double the extent of that of both these 
two classes added together. In other words, the differences among the specially 
able are far more marked than in tlie case of " intellisjent " men. This is of course 
a universal experience, but it is of interest to see its approximate quantitative 
value. 
In the same way the Very Dull minds, although only one-third as numerous 
as the Slow Dull, and hardly more than one-eighth of the Slow minds, yet occupy 
a range 50 per cent, greater than that of either of these groups. Thus we see 
again the source of the great differences in mental stupidity. The apparent want 
of continuity in the ranges of genius or of imbecility, which must arise when few 
individuals are spread over a large range, thus enables us to comprehend how it is 
possible to look upon these things as anomalies and mutations*. 
(3) On the Relation of Intelligence to Age. 
On the hypothesis that intelligence is very sensibly correlated with either 
brain-weight or head-size, we might not unnaturally anticipate that growth, which 
modifies largely the physical characters, would influence the intellectual f. It seemed 
of importance accordingly to enter very fully into the relationship of age to intelli- 
gence. Unfortunately the material provided by the Cambridge graduates does 
not lend itself to an age investigation. It is true that we have the ages provided, 
but the material consists in great bulk of young men measured in their 19th or 
20th years. A certain number of resident "dons" are available in the Cambridge 
Anthropometric Committee's material, but this gro>ip of older men is a stringently 
selected group, only the men of considerable intellectual achievement from the 
academic standpoint remaining in residence. Neither in the Cambridge data nor 
elsewhere did it seem possible to find the material requisite to settle the problenti 
of the relationship of intelligence to age in adults. We are forced, therefore, 
to confine our attention to the influence of age on general intelligence in 
childhood. 
* For tbe actual quantitative treatment of this form of discontinuity, see Biomctrika, Vol. i. 
pp. 385—399. 
t When is the intellectual ' prime ' in man ? His prime in stature is about 27 ; his prime in head 
measurements probably two or three years earlier ; his prime in brainweight in the teens and probably 
early in the teens. We know how many inches of stature, how many mm. in head diameters and how 
many grs. iu brainweight he loses yearly after his physical primes. How many mentaces does he lose 
each year after his intellectual prime ? We have no knowledge at present. Yet these very sensible 
changes which follow physically after the different primes are of the same magnitude as those differences 
between individuals upon which relations between intelligence and physique are based; they are, how- 
ever, tacitly put on one side by writers on this subject. 
