42 Journal of the Asiatic Society of Bengal. [N.S., XVI, 
blood, but otherwise can usually be recognized without precise 
diagnosis. No one could confuse a Chinaman with a Negro, or 
Peninsular India. The fact that a Burman might appear to 
belong either to the Mongolian or to the Indo-European stock 
need not concern us at present, any more than the other fact 
that the whole of mankind is not comprised in the three main 
groups thereof, even if we allow for possible admixture. 
Accepting, then, the definition of race that I have already 
given and further interpreting the phrase “‘ relatively unimpor, 
tant physical characters’’ in a restricted sense, we reach the 
gist of our discussion: Can human races be distinguished by 
measuring individuals belonging to them ?° 
Let us define a little further. What do we mean by mea- 
surements? I would here again adopt a restricted meaning 
and imply by measurements actual mensuration with callipers, 
a goniometer, a tape or other instruments of the kind. Physio- 
logical tests and the observation of what are sometimes called 
descriptive characters [ do not regard as measurements. 
ur enquiry, therefore, narrows itself still further. Can 
we distinguish between human races by the aid of instruments 
of this kind? My own answer to the question would be that 
we can not, but that measurements may be profitably em- 
ployed as a check on other observations, provided that they 
are taken with sufficient accuracy and on a sufficiently large 
number of individuals, and that we realize what we are measur- 
ing. Let us first see what these reservations imply. 
In measuring living persons we are attempting to measure 
what we cannot see, to measure bones through an opaque cover- 
ing of muscle, skin and adipose tissue. We have to find the 
points from which we measure mainly by the sense of touch. In 
a few measurements, such as those of the length and breadth of 
the head, there is no difficulty in attaining substantial accuracy, 
for the opaque covering of the bones is very thin and but 
slightly compressible. Even in such measurements we may, 
however, find considerable differences in the same individual 
in the indices calculated. The difficulty in attaining accuracy 
is greatly increased in measuring the nose, for here the points 
