DATA FOK THE PROBLEM OF EVOLUTION IN MAN. 
263 
But this agreement does not arise from any real accordance in the formulae, but 
fi-om the fact of the close equality of the Naqada and Theban mean values for U, Y 
and C. 
To test the ajjplicability of these circumferential formulas when extended from one 
race to a second, I take the following data :— 
Organ. 
Race. 
J Aino. 
? Aino. 
French.* 
U . . . . 
522-5 
501-7 
527-6 
. . . 
328-5 
317-1 
317-9 
C . . . . 
1462 
1308 
1475 
These lead to the following results for capacity :—• 
Race. 
Actual. 
From Nacpida formula (19). 
From Theban formula(22). 
.A.ino U ■ • • 
1462 
1493 
1583 
Aino $ . . . 
1308 
1379 
1350 
French . . 
1475 
1482 
1537 
We see : 
(i.) That the Naqada and Theban formulae, although deduced from kindred races 
auid from very consideralde numbers, lead to widely divergent results. 
(ii.) That tlie Naqada, which is for Aino S and French d 1:)etter than the Theban 
formula, gives results worse than the formulae based upon L X B X H previously 
discussed. 
We conclude, therefore, that it appears unlikely that a reconstruction formula, 
based on the circumferential measurements of the skull, can be found which will 
give good results, if extended from one local race to another. 
If we apply these formulae to reconstruct the capacity within the race, (20) and 
(21) give differences much of the order of the earlier reconstruction formulae (I) 
to (8), while (22) gives results as good as (9). 
The following table gives the errors made in estimating the capacity of forty Theban 
skulls, twenty of either sex, chosen at random. It will be seen that the errors can be 
fairly large when we use circumferential measurements. 
* From the skulls of 50 French jnisoners who died at Munich 
given in the German Anthropological Catalogue. 
during the Franco-German war. 
Data 
