128 R(latioiis<hij> of liitdUf/ciwc to Size ami Shape of Head 
future investigations on special characters. Judging the series as a whole, it seems 
impossible to use any of the physical measurements to estimate intelligence from. 
Hair colour is practically as good as head length or breadth, and eye colour as good 
as auricular height, and even all these are more important than the age influence. 
Health and temper have more relation to intelligence than any of the physical 
measurements we have made, while the intelligent child is athletic, popular and 
above all markedly conscientious. Handwriting is doubly as good a test of in- 
telligence as any head measurement. If it be argued that this is merely a school- 
master's measure of intelligence, then the reply must be that this remains to be 
proved*. If good handwriting be the schoolmaster's standard of intelligence, it 
appears also to be — as will be shown on another occasion — his standard of health 
and popularity. For handwriting, we find, is fairly closely correlated with a 
number of mental and physical characters. It is interesting to observe that, as far 
as our data go, the handwriting character-readers ought to be able to predict more 
closely than the anthropometers not only the amount of intelligence in an indi- 
vidual but also his grade in a variety of other mental and moral characters ! 
Looked at broadly our table seems to justify fully current common-sense 
methods of estimating intelligence. Give weight to health, temper, physique, 
popularity, handwriting and above all conscientiousness, in seeking friend, assistant 
or servant, and in doing this you will most probably obtain intelligence also. If 
you wish to take anthropometric characters into account — and they are not worth 
much — hair and eye colour will be as valuable as head measurements, and you 
need not produce the callipers in order to observe them ! I am not denying that 
in the future other anthropometric characters may possibly be discovered which will 
be found to be more closely correlated with intelligence. By all means let them 
be sought for and investigated biometrically ; let all types of head measurements 
and indices be taken and correlated with ability and achievement ; it is worth 
doing even if it leads to purely negative results. But let us hesitate on the 
ground of slender, or worse than slender, unscientific evidence to proclaim clo.se 
association between intelligence and external physical measurements -f-. So far 
there is nothing to encourage belief in such association ; and if we are consistent 
and apply any of the dogmatic views currently held to the problem of interracial 
* As far as the non-expert can judge, the classification of the liandwritings is a fair one. It is 
proposed to place the 5000 — 6000 specimens of handwriting with the ages of the children before an 
expert and obtain his classili cation of the whole material. 
t Some years ago I was struck by the widespread medical opinion that mentally defective children 
have peculiarly shaped palates. I asked an exponent of this view for the statistics bearing upon the 
subject, but I could not find that there had ever been a thorough study of the palate in mentally 
normal children. In tlie American Journal of Innanitij, Vol. Lxi. pp. 687 — 097 will be found a preliminary 
report of Drs Walter Channiug and Clark Wissler : "Comparative Measurements of the Hard Palate in 
Normal and Feeble-Minded Individuals." They show biometrically that " the absolute size of the 
palate as measured by the three specitied dimensions [height, length and breadth from casts] seems to 
be the same for feeble-minded and normal individuals," p. 095. It is most unfortunate that quantitative 
tests so rarely precede the spread and acceptance of very dogmatic opinions in a certain section of the 
medical profession. 
