218 
A First Study of the Burmese Skull 
In Group C less projecting cheek-bones, narrower noses with higher bridges, a 
sharp edge to the lower lip of the pyrif orm aperture ; and again considerable progna- 
thism ; the whole head of a less bold and heavy character. 
Sexing divided up Group A fairly evenly into 45 males and 39 females*; in 
Groups B and C the females preponderated, there being only eight males in each to 
18 females in C and 17 in B. A number of the skulls had to be classed (J or it might 
be ? with a query, as is bound to happen in any series, but these dubious ones 
would not be numerous enough, if put into the wrong class, to modify the means 
to any considerable extent. 
It was unfortunate that two out of six adult groups should contain as few as 
eight skulls, since these numbers are far too slight to base any reUable conclusions 
upon, or even to give any very helpful suggestions. 
The whole series was accordingly examined once more some months afterwards, 
to see whether it would not be possible to divide the adults into two main types 
instead of three, putting the bulk of the intermediate group B along with C, so 
as to strengthen the numbers of the latter, and make the main categories less 
unequal. It was found that only two of the eight intermediate males could possibly 
have been put with the eight in group G, making ten in all; 'and though six of the 
seventeen B females might have been classed with the eighteen in C, this grouping 
was not very convincing since the C's were of a distinctly smaller and finer 
build. 
In short, the small gain in numbers by the broader grouping would have been 
more than counterbalanced by the heterogeneity of the groups, and it was decided 
to leave the original divisions as they were. 
It will not, however, be wise to attach significance to the results based on the 
two small male categories B and C, except in so far as they confirm the figures for 
the corresponding females. 
After these conclusions had been arrived at upon the data themselves, we put 
our doubt as to the collection being purely Burman before Colonel Caster, and he 
reported that after further consideration there might be a certain number of Hindus 
or of Karens in the material, but that it was imlikely, owing to the source from 
which it was drawn, to include any Chinese, or any Mohammedans from India. 
In any case, the bulk of the skulls were Burman. 
We therefore had little hesitation in applying the term Burman proper to our 
large groups A and A ?. Since comparison with Hindu material showed little 
affinity between our C groups and the Hindus, as will be seen later, and taking into 
account the large proportion of Karen population around Moulmein, we suspect 
our C groups to be of Karen origin, though we have not enough definite evidence 
to establish the supposition. The few isolated Karen crania at the Royal College 
of Surgeons do not suffice to fix a type. Our B groups we suppose hybrids. 
The following are the mean values obtained by dire.^t measurement in the 
six groups: 
* Not all the cr.inia could be used for each contoiir, and this accounts for the different totals given 
in Table I and Figs. I-XVIIT. 
