Shi'if.i.di' : OsTKoi.oGY oi'- iiii'. I Ii:k()I)|()ni;>. 195 



anti-trochanterian articular surface is carried by them both as it passes 

 inwards. I^xternally this facet looks downwards and only slightly 

 outwards. 



'I'he iscJiiadic foramen is large and subelliptical, its major axis 

 being parallel to the line of the outer border of the post-acetabular 

 surface, which here arches over it. Posterior to this foramen, the 

 broad part of the ischium is roughly cjuadrilateral in outline, and for 

 the most jiart smooth and slightly concave. It is nearly at right 

 angles with the iliac surface above it. In this Heron the obturator 

 foramen is far from complete or deserving the name of a foramen. 

 Nearly its entire posterior arc is deficient, and the opening thus created 

 leads into the obturator space, which latter is found beneath the entire 

 lower margin of the ischium, being broadest in front and gradually 

 tapering off behind (Fig. 14). 



Ardcahcrodias has a blade-like pubis of nearly an equal width through- 

 out, though it is rather wider behind after it pas.ses the ischium and 

 curves mesiad towards its fellow. Just before it does this it is slightly 

 overlapped by the lower and posterior angle of that bone, or else 

 meets it in a single i)oint of tangency, or, as in the figure, does not 

 quite come in contact with it. Quite a large pneumatic foramen is 

 found beneath the projection of each ilium immediately behind the 

 anti-trochanter. 



The vertebral column may be seen in ])art through the ajjertures 

 afforded by the acetabulum and ischiadic toramen upon this lateral view. 

 Except at its sacral dilatation, the neural canal as it passes through the 

 vertebrae of the pelvis is small ; it will be remembered that we found 

 it quite so in the dorsal region also. 



My specimen of the pelvis, taken from the skeleton of Ardca caii- 

 didisiina (a bird of the year) although thoroughly herodine in all of 

 its salient points, it still differs in some of its minor details from the 

 same bone in Ardea herodias. A careful count shows that an equal 

 numl)er of vertebrae are anchylo.sed together to form the central ma.ss 

 for the support of the pelvic arch — fourteen in each case, /. c, the 

 twenty-fourth to the thirty-seventh inclusive. This obtains also in 

 the Yellow-crowned Night Heron, and in both these birds the brim of 

 the pelvic basin departs from and arrives at identically the same seg- 

 ments as described for Ardea. 



In A. candidissima, the ilia do not overreach the twenty-fourth ver- 

 tebra although otherwise these bones are comparatively longer and nar- 



