484 On the Problem of Sexing Osteometric Material 
Hence by (xvii) to (xix) : 
8^^ = gry^ = _ ,y X -OoG.SOS = - -2810, 
g(r, = -ha.= a-x -029,809 = -1066, 
gttj = - gw, = + n X -056,308 = 15-231. 
It will be seen from these results that : 
^'=.-•0563, ^-^ = -0298, ^=-0563 
y an 
may be considered fairly small quantities, and that they justify our assumption. 
We have accordingly : 
Females. Males. \ 
Mean : 70-543 mm. 80 525 mm. 
Total Frequency : 255-27 285-73 i (C), 
Standard Deviation: 34684 mm. 3-6816 mm. 
Modal Ordinate : 29 36 30-96 j 
It is clear that the solutions (C) and (A) are for all practical purposes identical. 
Thus the short method is justified in the problem of sexing osteometric material. 
An improper extension of the method to material in which the sexes occur in very 
unequal groups may be guarded against by simply observing whether and 
are very small quantities. 
In conclusion it may be desirable to compare the values of these sex-constants 
as found mathematically witli sexing by anatomical appreciation. I owe an 
anatomical sexing of the same bones to my colleague, Dr Derry. 
The following values of the constants resulted : 
Females. Males. 
Mean: 70098 mm. 79-764 mm. 
Total Frequency : 221 320 [ (D). 
Standard Deviation : 3"514S mm. 4-1254 mm. 
Modal Ordinate : 24-55 30-95 
It will be seen that the mathematically deduced constants are not widely 
divergent from those obtained anatomically, but the accordance if fair is not ideal. 
The accompanying diagram exhibits the differences in the frequency distributions 
found by the two methods of sexing. The chief difference lies in the transfer by 
the anatomist of the larger female bones of the mathematical sexing to the male 
group. I do not propose to discuss here the relative advantages of the two 
methods, but would draw attention to a few points of interest : 
(i) The siilution (D) makes no appeal to measurement in the sexing, it is 
based purely on an anatomical appreciation. It would therefore be subject to 
