THE OBSERVATIONAL METHOD 
2 43 
to substitute for the scales and the yardstick his own 
unverified observation. It is impossible to build up a 
science of observation in any field except by getting away 
from observation as such, and supplementing it with 
scientific tests or measures like those which have been 
already discussed. 
But, even if observation were accurate and reliable, 
it would still be under a great handicap. For example, 
what a great change in the shape of a man’s head and the 
height of his forehead is made by a hair cut. A man who, 
to the observer, looks like a ferocious round-headed 
simian one day, may become a mild-featured, sedentary, 
long-headed bookkeeper the next—after a hair cut and 
a shave. What a remarkable difference may be wrought 
in the texture of the skin by a hot bath! How compara 
tively easy it is to govern one’s appearance and to act the 
part for which one is aspiring. The practical significance 
of this contention is shown by the substitution of the 
Bertillon finger-print method for the photographic method 
of identifying individuals. The former is far the more 
accurate. 
Even if the three weaknesses outlined above did not 
exist, there would still remain the following great difficulty. 
The method which we have been discussing judges char 
acter by analyzing and comparing the parts of an in 
dividual with other parts of the same individual. Is the 
head long in proportion to its width? Is it high in propor 
tion to its length? These and similar questions show how 
the individual is compared with himself instead of with 
other individuals. There is no standard of shapes and 
sizes to which the observer can compare individuals and 
"with reference to which he can form his conclusions. There 
is nothing to correspond with the exact standards set by