226 
A First Study of the Burmese Skull 
In addition I append a table of mean measurements taken on the sagittal 
contours, but not used in construction of the sagittal type contour, as the position 
of the terminals had already been fixed by the means of other measurements. I 
have discussed already the degree of exactness with which these means from indi- 
vidual contours correspond to the values read off from the type contour. 
TABLE VI. Burmese Mean Sagittal Contour Values not used 
in Construction of Type. 
1 
Types: 
Males 
Females 
A 
B 
C 
A 
B 
C 

112-6(44) 
111-3(7) 
113-1 (8) 
106-7(39) 
106-0(17) 
108-2(18) 
\0p 
93-7 (43) 
97-3(7) 
100-2(8) 
92-8(39) 
91-8(17) 
91-2(18) 
N\ 
167-9(44) 
169-2(7) 
171-4(8) 
162-0(39) 
160-3(17) 
163-4(18) 
Lengths - 
(3 subtense ( x from .V 
86-2 (44) 
86-2(7) 
89-0 (8) 
81-5(39) 
80-4(17) 
82-5(18) 
to N\ \ 11 
72-0(44) 
70-2(7) 
70-0(8) 
68-5(39) 
68-0(17) 
68-1 (18) 
QI ... 
168-7 (44) 
171-1 (7) 
169-3(8) 
160-7(39) 
157-7 (17) 
164-1 (18) 
/3 su))tense (x- from 6' 
53-0(44) 
53-4(7) 
53-5(8) 
49-8(39) 
48-3(17) 
54-2(18) 
\ to GI \ y 
91-2(44) 
90-8(7) 
93-6 (8) 
85-3(39) 
85-7(17) 
86-5(18) 
60°-l (44) 
59°-7 (7) 
60°-5(8) 
60°-6(39) 
60°-6(17) 
59°-l (18) 
0r-7(44) 
61°-1 (7) 
61°-6(8) 
62°-2 (39) 
61°-9(17) 
60°-8(18) 
A I 
69°-6(38) 
72°-6(7) 
70°-4(7) 
71°-8(34) 
71°-6(15) 
75°-0(16) 
Angles -i 
Ni 
68°-0(38) 
66°-l (7) 
64°-8(7) 
66°-8(34) 
66°-9(15) 
64°-7 (16) 
Bi 
42°-5(38) 
41°-2(7) 
44°-8(7) 
41°-4(34) 
41°-6(15) 
40°-3(16) 
Pi 
84°-4(38) 
86"-l (7) 
83°-6(7) 
85°-0 (34) 
83°-8(15) 
87°-2(16) 
^fmll 
7°-3(42) 
8°-8(7) 
7°-6(8) 
8°-8(39) 
7°-8(17) 
8°-2(18) 
Indices 
[Frontal index if 
23-1 (44) 
21-6(7) 
22-1 (8) 
23°-6(39) 
22-5(17) 
22-9(18) 
I Occipital index i^^ ... 
26-2 (43) 
29-1 (7) 
25-8(8) 
25-2 (39) 
23-0(17) 
27-1(18) 
Before comparing our type contours we need to get some measure of the range 
of variation which would be covered by means based on other samples of the same 
size as our own, drawn from the same population. On pp. 143, 145, 147 of Crewdson 
Benington's paper on Type Contours* we have graphical representations of the limits 
between which the bulk of these variations would lie for the three contours repre- 
senting a series of 100 English skulls. Assuming the standard deviations of English 
and Burmese characters to be approximately the same (and they will certainly 
be of the same order), and taking into consideration the fact that the probable 
error of a mean varies directly in proportion to the inverse square root of the number 
of skulls on which it is based, we see that we must allow for our largest series of 
forty-four (Group A cj) a range half as big again as that shown for the English contour; 
while for our smallest series of seven (Group B c?) the range must be made about 3-8 
times as big. This demonstrates once more the unreliability of definite conclusions 
based upon short series. 
Below are given the proportions in which the range of 100 English type must 
be increased for our six series; and for the convenience of those who have not the 
Benington paper for reference we give a few measurements to indicate the extent 
of the range. 
* Binmetrika, Vol. vm. 
