DATA FOR THE PROBLEM OF EVOLUTION IN MAN. 
229 
exists vanes enormously, even in sign, as we pass from one local race to another. 
This want of correlation, or want of steady correlation, in the parts of the skull, as 
compared with the correlation exhibited by the long bones, or by parts of the hand, is 
extremely interesting from the standpoint of evolution. It would appear to be much 
easier to modify a single character of the skidl by selection without altering other 
characters than can he the case with other parts of the skeleton. 
The measurements considered in this paper are ; L tlie greatest length, B tlie 
greatest breadth of tlie skull, H the height measured from the auricular line, I tlie 
cephalic index = B iL, and C tlie capacity. In choosing the material several points 
had to be borne in mind : 
(i.) A sufficiently large series must he used. 
As a matter of fact, 50 to 100 skulls are considered by craniologists to be a fair 
series, hut such numbers are small from the mathematician’s standpoint. 
(ii.) Material must be dravm from as widely different races as possible, if we are to 
measure the legitimacy of applying results obtained from one local race to another. 
(hi.) The capacities must have been determined by competent observers using 
approximately like methods of measurement. 
The data which seemed to me to ajiproximately fulfil these conditions are the 
following :— 
(n.) A series of Bavarian [AU-Baierische) skulls measured by Professor J. PvANKe, 
and given in his ‘ Beitriige zur physischen Anthropologie der Bayern.’ In 
this case there were 100 d and 99 ? Avith L, B, H, I, and C available. 
(6.) A series of Aino skulls measured by Koganei, a craniologist trained in 
German schools, and given in the ‘ Mittheilungen aus der Medicinischen 
Facultat der Kaiserlich-Japanischen Universitat,’ Tokio, Bd. ii., 1894. In 
this case L, B, H, I, and C are giA^en for 76 d and 52 $ skulls, and there 
are lid and 11 ? skulls for AAhiich L, B, H, and I only are giAmn. 
(c.) A series of Naqada skulls discovered in Egypt by Professor Flinders Petrie, 
and measured by Miss C. D. Fawcett, B.Sc., on the basis of the ‘ Frank¬ 
furter Verstaudigung.’ I have to thank her for alloAAung me to use her 
results liefore publication. In this case L, B, H, and C AA^ere available for 
69 d and 98 ? skidls, and L, B, H only for 76 d and 100 ? skulls. 
As supplementary and test series, I have used primarily— 
(d.) 201 d and 96 ? skidls of ancient Egyptians. This series consists of mummies 
from Thebes in the Mook collection at Leipzig. 
(c.) 76 d and 23 ? skulls of modern Egyptians in a PrivaUSammltmg at Leipzig. 
The details of both these series are taken from the great craniological catalogue ot 
the German Anthropological Society."^ 
(5.) Starting with the series (a) and (6), I have obtained for their means and 
* The parts are published separately as off-prints from the ‘ Archiv fiir Anthropologie.’ 
