96 
This is a part of the right os innominatum, p 3 (PI. XX. fig. 5.), including the posterior 
and inferior angle of the acetabulum, the origins of the pubis and ischium, which form 
the obturator notch, and a fractured continuation of the latter bone. The fragment has 
belonged to a pelvis intermediate in size between p 2 and p 4, but is nearer the former. 
From tbis it differs in the concavity of the upper boundary of the ischiadic notch, and 
the descending process forming its posterior boundary which almost touches the pubis. 
The posterior margin of the wall of the acetabulum is straight, and ascends at a right 
angle with the horizontal ischium. In the larger pelvis, p 2, as in the smaller one, p 4, 
this margin curves back at less than a right angle. The ischium is thinner and less 
convex internally. 
The pelvis p 4 agrees in the proportions of its acetabula and the form of the posterior 
articular protuberance with the femora of Dinornis dromeoides ; and the smaller pelvis 
p 5 offers the same correspondence with the femora of the Dinornis didiformis. 
If a species of Dinornis intermediate in size between the D. struthoides and D. didi- 
formis had not been indicated by the femora f 6 and f 16, I must have been led to the 
same conclusion by the two pelves, p 4 and p 5, as to the existence of such a species, 
Since the foregoing description of the pelvic bones was put into type J have been 
favoured by William Cotton, Esq., F.R.S., with the view of some specimens of the bones 
of the Dinornis, very recently transmitted by his son the Rev. Wm. Cotton, M.A,, from 
New Zealand, one of which is a fractured pelvis', corresponding in length and in so 
many other characters with p 5, as to lead to the conclusion that it belongs to the same 
species, and that the differences between them are attributable to sex. 
These differences are the following. The bodies of the first two sacral vertebre in 
Mr. Cotton’s specimen, which I shall call p 6, are flatter on their under surface and 
broader ; the form of the anterior articular surface of the first sacral vertebra more 
nearly resembles that in the largest fragment, p1, The spines of the seven anterior 
vertebre and the co-ascending plates of the iliac bones are less elevated in p 6. Thus, 
from the end of the transverse process of the sixth sacral vertebra to the summit of the 
ilium in p 5, is three inches ten lines, whilst in p 6 it is only two inches nine lines. The 
length of the part of the pelvis formed by the first seven sacral vertebre is precisely the 
same in both, But whilst the height of p 5 is greater its breadth is rather less, espe- 
cially immediately behind the acetabulum. The expanded horizontal thin plate of bone 
at the back part of the pelvis, between the diverging iliac bones, is well preserved in 
Mr, Cotton’s specimen. 
The breadth of this at the back part of the acetabula is five inches five lines ; the 
breadth at the hinder end of the pelvis is three inches nine lines. 
Of the two pelves which correspond in length and in most of the characters by which 
the one first described differs from p 4, we may regard the higher and narrower speci- 
men, p 5, as belonging to the male, and the lower and broader one, p 6, to the female. 
1 Pl. XX a. fig, 1. 
