384 Measurement of Internal Capacity of Skull 
to pp. 35, 40, 55, etc. in order, repeating the process until 20 skulls had been 
extracted and the following table resulted. To this we applied P and L*, G. F. 
and B^, with the results in Table IX. 
TABLE VIII. 
Individual Crania Sampled at Random from Barnard Davis. 
Kaoe 
Barnard Davis' 
Number 
ct 
L 
B 
H 
I 
U 
S 
Q 
S' 
Q' 
(1) Ancient Roman ... 
171 (p. 18) 
70-0 
7-2 
5-6 
4-25 
76 
20-7 
14-5 
11-7 
12-5 
11-9 
(2) Anglo-Saxon 
674 (p. 34) 
75-0 
7-4 
5-5 
4-75 
74 
20-8 
14-5 
12-4 
12-5 
12-6 
(3) English 
1029 (p. 48) 
94-0 
8-0 
6-3 
5-25 
78 
22-8 
16-6 
14-4 
14-6 
14-6 
(4) Scottish Highlander 
6 '8 
19"7 
13"7 
11"7 
(5) French 
1312 (p. 80) 
74-5 
6-9 
5-4 
4-25 
78 
20-0 
14-1 
11-6 
12-1 
11-8 
(6) Italian 
1178 (p. 95) 
73-0 
7-1 
5-2 
4-35 
73 
20-2 
14-6 
11-7 
12-6 
11-9 
(7) Dutch Jew 
1328 (p. 110) 
79-5 
7-5 
5-5 
4-35 
73 
21-2 
15-0 
12-0 
13-0 
12-2 
(8) RiLssian Tartar ... 
941 (p. 125) 
76-5 
7-0 
5-7 
4-45 
81 
20-6 
14-6 
12-0 
12-6 
12-2 
(9) Hindoo 
487 (p. 140) 
6-8 
5-5 
4-55 
80 
19-5 
14-3 
12-3 
12-3 
12-5 
(10) Gond 
808 (p. 155) 
65-5 
6-8 
4-9 
4-45 
72 
18-8 
14-1 
11-7 
12-1 
11-9 
(11) Bodo 
735 (p. 169) 
68-0 
6-8 
5-5 
4-55 
80 
19-6 
14-2 
12-5 
12-2 
12-7 
(12) Egyptian Arab 
1015 (p. 185) 
67-0 
7-1 
5-4 
4-55 
70 
20-1 
14-4 
12 '0 
12-4 
12-2 
(13) Fantu 
262 (p. 200) 
67-0 
6-9 
5-2 
4-55 
75 
19-3 
14-4 
12-1 
12-4 
12-3 
(14) Kafir 
586 (p. 215) 
64-0 
7-4 
4-9 
4 '35 
66 
20-2 
14-6 
11-4 
12-6 
11-6 
(15) Vancouver Islander 
1211 (p. 229) 
85-0 
7-5 
5-7 
4-65 
76 
21-0 
15-4 
12-7 
13-4 
12-9 
(16) Aymara 
1198 (p. 244) 
82-0 
7-1 
5-6 
4-75 
79 
20-5 
14-8 
12-6 
12-8 
12-8 
(17) Australian 
99 (p. 260) 
69-0 
7-0 
5-6 
4-35 
80 
20-0 
14-3 
12-0 
12-3 
12-2 
(18) Sumatran ... 
1367 (p. 275) 
67-0 
7-0 
5-1 
4-35 
72 
19-8 
14-0 
11-4 
12-0 
11-6 
(19) Dyak 
280 (p. 290) 
60-0 
6-6 
4-9 
4-65 
74 
18-7 
13-7 
10-7 
11-7 
10-9 
(20) Papuan 
1401 (p. 305) 
73-0 
6-5 
5-4 
4-45 
83 
19-4 
13-5 
12-0 
11-5 
12-2 
Here again we reach the same general conclusions, namely the correlation 
formulae give not only better means for the whole series, but less mean errors, 
and this, although a whole series of assumptions has to be made to get any results 
from them at all, if Barnard Davis' material be used. The further point has to 
be borne in mind that B^ was 'guessed' to fit Barnard Davis' material, whereas 
the correlation formulae were calculated from data measured in an entirely 
different, and we consider, in a far more correct manner j. Any mathematician 
using a formula of Dr Beddoe's type could get immeasurably better results by 
least squares or by correlation from Barnard Davis' data, but the labour would be 
wasted, for the formula would only apply to the very doubtful measurements of 
the Thesaurus Graniorum and not to the more reliable measurements now made 
in both Germany and England. 
* If L, B, H be measured in inches the formula to give C in cubic centimetres is nearly for males : 
(7 = 406-01 + 6-52214 1/xBxiJ. This is the mean formula of Dr Lee for finding the capacity of 
individual crania. 
t Ozs. of Calais sand. To be multiphed by Topinard's factor 19-89 to reduce to cubic centimetres 
and then by 19/20 to reduce to cranial capacities as now measured. The other measurements are in 
English inches. 
J We believe the rise in the mean error from 47 to 64 when we pass from suitably measured crania 
(p. 371) to Barnard Davis' material is almost entirely due to the vagueness of the allowances to be made 
and the crudeness of Barnard Davis' measurements. 
