W. R. Macdonell 
221 
percentages are maintained in a second English series seems to show that we are 
dealing with something at anyrate characteristic of the Londoner. It would be of 
interest to enquire whether the proportions of special variations are greater in 
mixed races like the English, than in races of relatively purer character. 
(9) On the Variability of the English Skull. 
The numerical values of the variations are given in Table VII together with 
their probable errors. 
Taking first the capaciti/, I extract various coefficients of variation from 
Pearson's The Chances of Death, Vol. i. pp. 328 — 349, for comparison with our 
series and the Naqadas. 
English * 
8-28 
8-68 
Parisian French 
7-36 
7-10 
Modern Italians ... 
8-34 
8-99 
Modern Germans ... 
7-74 
8-19 
Naqadas 
7-72 
6-92 
Etruscans ... 
9-58 
8-54 
Egyptian Mummies 
8-13 
8-29 
Ainos 
7-07 
6-90 
The variability of our series is high, but not quite so high as that of the 
Italians, and the female skull is slightly more variable than the male, the English 
in this respect resembling Italians and Germans. 
Comparing the coefficients of variation of lenrjth, breadth, and height of the skull 
we find : 
Race 
Len 
gth 
Breadth 
Auricular Height 
<? 
? 
? 
s 
¥ 
English 
3-31 
3-45 
3-75 
3-54 
3-73 
4-12 
Bavarian t . . . 
3-37 
3-57 
3-89 
3-39 
4-47 
3-91 . 
French % 
3-97 
3-65 
4-21 
3-67 
Naqada § ... 
3-17 
314 
3-29 
3-4.5 
3-86 
3-54 
Aino t 
3-20 
3-08 
2-76 
2-68 
3-67 
3-18 
* Mj' own results are given ; Pearson's figures were based on only 58 skulls — all that had been 
measured when he wrote. 
t Alice Lee, Phil. Trans., Vol. 19G, A, p. 230. 
X Unpublished reductions of measurements in Broca's MS. , by C. D. Fawcett. 
§ Biometrika, Vol. i. p. 438. 
