R. Crbwdson Benington 311 
not our measurement. As it cannot be corrected, it is idle to correct G 1 ; and by 
a lucky chance the index fits well for the male with what we might anticipate. 
The table is most hypothetical and its chief service must be to emphasise the 
need of precise standardisation of the palate measurements. As it stands, it may 
suffice to show that possibly in palate measurements also the Congo-Gaboon negro 
is differentiated from the Kaffir-Zulu group. The general want of systematic 
differentiation between negro and European for palate measurements may either 
prove that the palate is not a markedly racial character, or indicate that our 
hypothetical corrections have screened the true values. 
Orbits. The orbit is another character which is distinctly of racial importance, 
but which it is very hard to discuss. There is not the least doubt that in facial 
appreciation orbital form is a most important factor, and the orbital index ought 
to be most suggestive in its bearing on racial differences. Unfortunately one 
component of it, the breadth of the orbit, is almost as vague as the length of the 
palate. This point has been discussed by both Fawcett and Macdoneli*, and 
a few additional remarks may be made here. It is probable that most writers 
have taken the dacryon as the start on the nasal side of the greatest breadth. 
There are great advantages in simplicity of measurement if this be done, and 
possibly it will be best to standardise the measurement in this manner, 
but when it is done it must be admitted at once that we are only getting an 
artificial orbital index. The physiognomical effect of the orbit is far better 
expressed by the original words of the Frankfurt Concordat — the greatest breadth 
from side to side of orbit wherever found, measuring from margin to margin — , 
only as there happens to be no margin on the nasal side, we are thrust back 
on the alternative of either giving up the definition, or defining what is meant by 
the 'margin' on the nasal side. Those who give up the margin use the dacryon 
or, occasionally, a purely arbitrary point leading to endless personal equation in 
the measurement, and to perfectly idle results, as they can be compared with no 
other series of measurements. If we retain the idea of margin we must put 
in an imaginary margin on the nasal side, and measure the greatest diameter from 
this. Two possibilities occurred to the biometric workers : 
(i) To produce with a pencil the margin from the last sensible points of the 
upper and lower orbital rims as a curve on the nasal wall, starting without breach 
of continuity with the curvatures indicated by the upper and lower orbital rims, 
so that these rims meet again without breach of continuity on the nasal wall. 
The method is not difficult in practice and does not admit of as much personal 
freedom as the description may suggest to the reader. It was adopted by 
C. D. Fawcett in her measurement of the Naqada crania. I will term it the 
curvature method. 
(ii) To join with a pencil line the last sensible points on the upper and lower 
orbital rims, the line to be the curve of shortest length on the nasal wall between 
* Biometrika, Vol. i. pp. 430, 431 and Vol. m. p. 201. 
