K. Pearson 
119 
diagram), then it falls at the terminals slightly outside it, or very intelligent 
and very dull girls are more frequent than the corresponding boy classes. This 
greater relative variability of the girls appears confirmed by a comparison of 
Table V, where in each intellectual class there is more variety in the girls' than 
the boys' ages. To further test this, I have drawn up a table of the standard 
deviations in intelligence of each age group of boys and girls. These standard 
deviations have been calculated in terms of the Intelligent + Slow Intelligent 
range as the equivalent of 180 mentaces. Here again the girls appear on the 
whole slightly more variable. 
TABLE VII. 
Variation of each Age Group in Ability. 
Age 
Group 
Standard Deviation 
in Mentaces 
Boys 
Girls 
3— 7 
8— 9 
10—11 
12—13 
14—15 
16—17 
18—20 
103 
99 
94 
90 
89 
97 
84 
95 
88 
83 
95 
98 
100 
119 
Mean of 
7 groups 
93-7 

96-9 
Personally, I should lay no stress whatever on this difference, except to assert 
the important point, that no inequality exists between the mental variability 
of girls and boys. Boys are not more variable intellectually than girls. I have 
tried the fundamental data by all sorts of processes — not here recoi'ded — and 
always with the result, that in the mass there is no sensible difference in 
intellectual variability between boys and girls. But the age classes do seem to 
indicate a difference in the distribution of this variability. While both sexes fall 
in variability from the earliest years of life, the girls much more than regain the 
lost variability by 20 ; on the other hand, the boys may pos.sibly regain it — if we 
exclude the 18-20 group as not fairly representing boys of this age, many of 
whom have left school — yet they do not appear to increase rapidly beyond the 
infantile variability, as the girls do. The matter is deserving of a further special 
investigation. 
(4) On the Relation of Ability to the Size of the Head. 
Having shown that to a first approximation the grade of intelligence as 
estimated in the school observations is not markedly affected by age, I turn now 
to the relation between ability and the size of the head. In the case of the 
