226 SECTIONAL ADDRESSES. 
age of six at the age of fifteen. Since beyond the stage of puberty 
inborn intelligence does not develop to an appreciable extent (another 
startling paradox of psychological testing), such a person will never 
rise above the six-year level, and will remain mentally defective for 
the rest of his life. 
From the numerous results obtained from the widespread employ- 
ment of intelligence-scales, one fact of deep social significance emerges— 
the vast range of innate individual differences. A famous clause in the 
American Declaration of Independence proclaims that ‘all men are 
created equal.’ In the psychological sense as distinct from the political, 
not only are men created unequal, but the extent of the inequality sur- 
passes anything before conjectured. Ina survey carried out upon all the 
children in a representative London borough—a census covering more 
than 30,000 cases—it was found that, within the elementary schools, 
the mental ratios might vary from below 50 per cent. to above 150 per 
cent. ; that is to say, the brightest child at the age of ten had the mental 
level of an average child of fifteen, while the dullest had the mental 
level of a little child of only five." 
Over this vast scale the distribution of intelligence is neither flat 
nor yet irregular; it follows a simple mathematical law. Its frequency 
conforms to the so-called * normal curve,’ and the abnormal and defec- 
tive are found to constitute no isolated types, but to be simply the tail- 
end of a chance distribution. Probably all or most of our mental capa- 
cities are distributed in the same fashion. This fact, if it be a fact, 
greatly simplifies the problem of mental measurement. It should be a 
recognised maxim of procedure to measure people, not by arbitrary 
marks between a conventional zero and an equally conventional maxi- 
mum, but by the degree of divergence above or below the average or 
middle line (much as we measure the depth of the ocean or the altitude 
of the hills from the intervening sea-level), the divergence being calcu- 
lated in terms of the standard deviation. This is a technical hint of 
special value in estimating qualities that lend themselves to no obvious 
quantitative units like mental ages or additive marks. 
Since variations in intelligence are so wide and so continuous, it 
becomes convenient to divide the entire population into about six or 
eight separate classes or layers. A classification of this kind, worked 
out empirically, for children, is already implicitly embodied in the 
organisation of our various schools. A second classification can be 
drawn up, on an analogous basis, for adults, and will be found, in the 
main, to reflect the amount of difficulty and responsibility entailed by’ 
their several occupations. It is interesting to find that the proportionate 
number cf individuals falling into the parallel sections tallies pretty 
closely both for adults and for children (see Table I).*° Tere, there- 
fore, lies a simple aim alike for educational administration and for 
19 The Distribution and Relations of Educational Abilities (London County 
Council Reports, 1917). Mental and Scholastic Tests (London County Council 
Reports, 1922). 
20 Compiled partly from data published in the L.C.C. Reports (Joc. cit. sup.), 
and partly from a table recently included in a paper on The Principles of 
Vocational Guidance (VIIth Int. Congr. Psych., 1923). The figures and categories 
given in the present table are round approximations only. 
