460 Variation and Correlation of the Human Skull 
Facial and Palate Measurements. 
The correlation of breadth and height of orbit is fairly constant, say '45, for 
both eyes in both sexes, the female orbit has its parts, however, somewhat more 
highly correlated. The same sexual advantage occurs more markedly in the length 
and breadth of palate and significantly, if less markedly, in height and breadth of 
face. On tlie other hand the male breadth and height of nose are more closely 
correlated. 
We have absolutely no data available at present for comparative purposes. 
Index Gorrela,tions. 
Here again the results of the present paper embrace nearly all we at present 
know*. But they lead to several suggestive hints for further investigation. We 
may note that : 
(a) In both sexes chamaecephaly is associated with plat3''rrhiny. 
(h) In the male chamaecephaly is also associated with chamaeconchic and 
brachystaphyline chaiacters, but in the female there is no really sensible relation- 
ship between the shape of palate or orbit and the height-length cephalic index. 
What little relationship there is tends to mark an association between chamae- 
cephaly and hypsiconchic and leptostaphyline characters. 
(c) In both sexes brachycephaly is associated with the leptorrhine and 
brachystaphyline characters. 
(o^) In the male brachycephaly is associated with hypsiconchic, in the 
female with the chamaeconchic character ; in the latter sex the association is 
much smaller. 
(e) In both sexes there is a quite sensible association of platyrrhiny with 
chamaeconchy, or when the nose is flat the eye is oval, 
(/') In the male platyrrhiny is associated with the leptostaphyline character, 
but in the female there is practically no relationship of the nose and palate 
characters. 
(g) In both sexes there is sensible correlation between the palate and 
orbital indices, hypsiconchy being associated with brachystaphyline characters. 
All these results, it must be borne in mind, are relations between the characters 
of individuals within the race, and indicate how, if an individual differ from the 
mean m one character, he or she will be likely to differ from it in a second. They 
must not be extended without further consideration from association of deviations 
within the race to association of racial characters. Still a consideration of Table 
XIV. shows that they may be suggestive also in this direction. The correlation 
* Dr Lee's correlations of cephalic index and capacity for Germans, Ainos, Thebans, Copts, Malays, 
Etruscans and French are the only other cases we know of: see Phil. Trans. Vol. 196, A, p. 232. 
