1910-11.] Observations on Australian Aboriginal Crania. 5 
Pearson (10) defines his conception of “ pure,” as applied to race, thus : 
“ I doubt whether anything corresponding to a pure race exists in man, if 
by that term is meant a group absolutely without ‘ Blutmischung ’ or 
mixture. Such a view would mean an indefinite number of special 
creations or independent evolutions of man. The ‘ purest race,’ as I have 
said elsewhere (11), is for me the one which has been isolated, intrabred, 
and selected for the longest period. It may well, in the dim past, have been 
a blend of the most diverse elements.” 
The purer the race, therefore, the more homogeneous will it be, and 
vice versa. 
According to Fawcett and Lee (1), “It is very hard to obtain a homo- 
geneous group of skulls, even fifty in number ; and these, again, must be 
distributed between the two sexes; the probable errors therefore of con- 
stants determined from such series are proportionately large.” 
Macdonnell (2), in commenting on Turner’s paper on Scottish Skulls 
(18), says, “that as they range from quite unknown antiquity to modern 
specimens from the dissecting-room, they cannot be considered a homo- 
geneous series ; and this conclusion is confirmed by an examination of their 
variability, e.g. the standard deviation of the glabello-occipital length in 
these Scottish skulls is 7 ’41 8 for males and 7T51 for females, which shows 
a much higher variability than is found in fairly contemporary individuals 
of homogeneous races.” 
As to the general principles concerning homogeneity or heterogeneity 
to be deduced from figures, Pearson (11) says: “I think they are these. 
The heterogeneity of any series, the variability of which for skull lengths 
exceeds 6‘5, or for skull breadths exceeds about the same quantity, should 
be suspected, and the series subjected to a close examination. If the vari- 
ability of the skull lengths be less than 5*5, or of the skull breadths less 
than 3*3, then we must suspect that the series is a rather stringently 
selected sample. This rule will generally enable us to distinguish between 
heterogeneity due to a mixture of crania from diverse races and the homo- 
geneity of a single race, which indeed may be the product of a number of 
generations of cross breeding, such as we may assert of modern English, 
French, or Germans, but hardly with the same certainty of Ainos or 
Bengal castes.” 
According to Macdonnell (2) and Pearson (11) one must conclude, 
therefore, that if we get a high standard deviation, e.g. as in Turner’s 
Scottish skulls (18), the material cannot be looked upon as homogeneous. 
On the other hand, Fawcett and Lee (1) get a high standard deviation 
for what they regard as a homogeneous series of French skulls, viz. 
