1910-11.] Observations on Australian Aboriginal Crania. 3 
and by Boas (14), both of whom dealt with the correlation of cranial 
capacity with length, breadth, and height, and with the cephalic 
index. 
Fawcett and Lee (1) state (a) “that craniometry cannot in future 
content itself with either the raw measurements, tables of mere average, 
or graphical^ exhibition of correlation results, but must adopt the methods 
of modern statistical investigation, tabulating means, variabilities, cor- 
relations, and their probable errors, in order to draw safe inferences and 
make racial comparisons. ( b ) The relationship between cranial characters, 
as exhibited by their coefficients of correlation in the case of Naquada 
and other races, is seen to be low, and to vary much from race to race. 
It is therefore very doubtful how far it is legitimate to press results 
found for individuals of one race upon those of another. We cannot pass 
from intra-racial to inter-racial conclusions, but we must work towards 
a knowledge of inter-racial correlation ; and the first step in the direction 
should be to obtain the average values of some forty or fifty characters 
in fifty or one hundred races, measured on some uniform plan. Only on 
such inter-racial correlations will it be possible to establish a properly 
founded statistical theory of race in man.” 
From the foregoing opinions of the leading and most recent in- 
vestigators of craniological data it is obviously imperative that all 
investigations into the characteristics of any group of crania should be 
conducted upon modern biometric plans, and this work embraces the 
application of the “ Rational method ” of dealing with the craniological 
data of the Australian aboriginal with the following objects : — 
1. The location of the position of the Australian aboriginal with the 
races of man. 
2. The recording of the results for future use, as has been suggested by 
the investigators already named, and by the Wistar Institute of Anatomy 
and Biology, Philadelphia (21). 
3. The comparison of the results obtained with those for other races 
that have been already worked out. 
The correct sexual separation of skulls is a matter of great difficulty, 
and has been referred to by many investigators ; e.g. Parsons (16) speaks 
thus : “ As long as the sexing of the skulls remains a question of individual 
judgment, nothing approaching mathematical accuracy can be expected 
from the results of measurements, though I find that by shifting a block 
of 100 of more or less doubtful skulls from one sex to the other, I have 
only succeeded in altering the average length index 0T per cent.” 
Pearson (7) says: “While in observations made on the living, or in the 
