M. L. TiLDBSLEY 249 
The following are the results: 
Races ^ 
No. of 
Characters 
C.R.L. 
Probable 
Error 
17th Century Londoners and 
Long Barrow British 
23 
3-8 
±10 
Bavarians and Wiirtcmbcrgers 
2-t 
121 
±10 - 
17 th Century Londoners and 
- Modern French 
28 - 
24-5 
±•09 
Prehistoric Egyptians and 
Congo Negroes 
27 
27-3 
±■09 
Here we find the justification of MacdonneU's statement that the Long Barrow 
British have a close resemblance to seventeenth century Londoners*. 
It is also clear that the critics who asserted a large negroid element in our 
Prehistoric Egyptians had httle basis for their criticism. French and English are 
shown to be almost as far apart racially as Egyptians and Negroes. And if 
Bavarians and Wiirtembergers are not as racially distinct as French and seven- 
teenth century Londoners, they are much further apart than are the latter from 
the Long Barrow British. 
I turn now to the question with which this paper is more immediately concerned, 
and see how far this method will throw light upon my Burmese mixed population, 
which mere appreciative analysis of characters led us to differentiate into three 
types: Type A, the Burman proper; C, another type — possibly Karen; and Type B, 
a supposed hybrid. 
Races 
Sex 
No. of 
characters 
C.R.L. 
Probable 
error 
A and C 
<? 
30 
312 
±■09 
>> 
? 
30 
G-44 
±■09 
A and B 
<? 
30 
0-47 
±■09 
? 
30 
1-16 
±■09 
C and B 
(? 
30 
1-30 
±09 
? 
30 
2-87 
±■09 
These results indicate that A and C are as racially distinct as Long Barrow 
British and seventeenth century Londoners, but not as distinct as Bavarians and 
Wiirtembergers; they also show that both A and C are closer to the supposed 
hybrid than to each other. 
The method therefore provides a numerical evaluation of a purely appreciative 
judgment, a judgment, however, which anyone who had examined a large number 
of cranial groups from the racial standpoint would be likely to make. 
Before making use of this coefficient to elucidate the problem of the relative 
affinities between the Burmese groups and the adjacent racial series with which 
they have been compared, it will be well to measure the relation of these groups 
to one another. 
* Biometrika, Vol. iii, pp. 243-4; Vol. v. p. 104. 
