of the head changes, and we are compelled to reduce our measurements to what they would be if the children were all of one age. Taking the length of head, Dr. Alice Lee formed a correlation table between length of head and age for 1856 boys, ranging from 4 to 19 years of age. The mean head-length was taken for each year of age, and thus a curve of average size of head of boys from 4 to 19 obtained. This curve presented several marked features, notably an apparent period of rest in growth during the 12th year.* These features will be more fully discussed on another occasion, but the rest in the 12th year is referred to here as a reason why we have chosen the 12th year as a standard age to which we may refer all our measurements. The growth of the average boy from every year up to the 12th year, and from the 12th up to every later year, was now determined. These growths were then added to or subtracted from the head-length measured on the individual in order to deduce from the observed size of the boy's head his approximate size of head in the 12th year. This, of course, is far from legitimate ; it supposes that each individual boy grows like the average boy, which we know is not the fact. Still it will not lead us very far astray for the purpose of the broad classification we are proposing.!
A precise!}' similar process was adopted for the breadth of head. The 12th year was still chosen for the standard age, although in this case the growth-curve from 4 to 19 years exhibits no period of rest, being almost exactly a straight line. %
In our school observations the following classification of intelligence was adopted (it having been shown by experiment to give fairly con- cordant results when the same children were classed by independent observers) : Quick-Intelligent, Intelligent, Slow-Intelligent, Slow, Slow- Dull, and Very Dull. For the purposes of our present investigation,, we grouped the first two categories together and the remaining four together. We will call the first group Intelligent and the second Slotc, and these may be taken to roughly correspond to our first division of Honours and Poll men when dealing with Cambridge graduates, results on this point obtained by Dr. W. Pfitzner, ' Zeitschrift fur Morphologic u. Anthropologie,' vol. 1, 1899, p. 372.
- A like but less marked rest in the 12th year is indicated in W. T. Porter's
curves for growth of head in boys : see "The Growth of St. Louis Children," ' Trans. Acad. of Science of St. Louis,' vol. 6, pp. 2G4-380, Plate 32.
t An examination of Mr. Porter's curves of growth for the 25th, 50th, and 75th percentile grades (loc. cit., Plate 32) seem to indicate (by the parallelism of these curves) how small really is the correlation between amount of growth and actual length of head.
J This is again in keeping with Mr. Porter's results for ages 6 to 17 ; see loc. cit., Plate 33. His percentile curves are again approximately parallel, and thus- favour the view that amounts of growth are largely independent of size. Why there is rest in growth of length but not in growth of breadth is an interesting problem.
