494 
The Measurements of the Pelvis 
(A) Obstetric conjugate. Equations were obtained for deriving this from each 
of the diameters with which it is most closely correlated, viz. the diagonal con- 
jugate (-91), and the antero-posterior (-30). These equations are 
G = -978Z>- 1-553 + '299 (a), 
0 = -365^ + 6-081 + -705 (b), 
the first of which has a much smaller probable error (-299) than the second ( - 705). 
A single equation was also obtained, using both diameters 
G = 9727) + -023.4 - 1-743 ± 298 (c) 
with an extremely slight improvement of the probable error as compared with (a). 
The diameters next in order of correlation with the obstetric conjugate are the 
posterior sagittal (-21) and the inter-tuberal (-19). The posterior sagittal is not 
measurable on the living subject. The inter-tuberal was measured in a special 
way on the dry pelvis by Emmons and is not necessarily the same as the inter- 
tuberal diameter as understood by the obstetrician. The method is not applicable 
to the living subject, but this diameter was employed here to test whether it was 
of sufficient importance for its measurement to be followed up further. An 
equation was formed by taking it with the diagonal conjugate and the antero- 
posterior diameter 
C = -972Z) + -023il + -0017- 1-745 + -299 (d). 
On comparing (c) with (d) it is apparent that the addition of the inter-tuberal 
diameter makes little difference, and that little to the bad, as is shown by the 
probable errors. This diameter will therefore not be considered further. 
It has been pointed out previously that the pubic height is sometimes used 
in calculating the obstetric conjugate but that its correlation with this diameter 
is low ('05). The equation connecting the obstetric conjugate with the pubic 
height is 
G = -HOP + 10-011 + -738 (e), 
having the largest probable error ("738) of the equations yet obtained. The result 
of combining the pubic height with the diagonal conjugate is, however, good : 
C= -9897)- -165P- 1-229 ±-293 (/) 
with a probable error lower than that of (c). This is due to the fact that, although 
the obstetric conjugate is more highly correlated with the antero-posterior diameter 
(-30) than with the pubic height ("05), the former of these two is more highly 
correlated with the diagonal conjugate ('31) than is the latter ('14), and it has 
been found that when the crossed correlation between the variables is high the 
resulting equation is not so good as when the crossed correlation is low. The 
pubic height, whether measured externally or internally, involves exposure of or 
inconvenience to the patient, so that it is desirable to avoid its use if just as good 
or nearly as good results can be obtained by other means. 
