192 
with a very gradual and slightly curved incline* (PI. ii., fig. 2 C)— 
a feature which is in marked contrast to the more abrupt and 
steeper rise of this process in Dinornis. Nevertheless, owing to 
the great lateral width of the upper extremity of the bone and 
the consequent length of the incline, the summit of the trochan- 
ter, even in its slightly abraded condition, reaches to quite as 
great a relative height above the head, as in the New Zealand 
birds. 
When a proximal view of the superior extremity is presented 
(Pl. iii., fig. 3) it will be seen that there is no projection poster- 
iorly of the hinder surface of the trochanter, such as there is 
both in the New Zealand Dinornis and in D. queenslandie. 
Thus, whereas in the last named birds and, to some extent also, in 
Dromornis, the posterior margin of the upper articular surface 
forms a well-marked indented curve. In Genyornis, however, it 
forms nearly a straight line up to the point where the contour of 
the trochanter sweeps forward (Pl. ui. fig. 3, wpper border of 
jigure). The same figure will indicate the manner in which the 
mass of the trochanter is projected forwards and outwards (8). 
In Dinornis the pre- meets the postero-external trochanteric 
surface at an acute angle, and the crest, corresponding to their 
line of union and terminating below in the ecto-trochanteric 
tuberosity, is sharp, prominent and laterally compressed. In 
Genyornis, on the other hand, the conditions may, perhaps, be 
best described as being such as are produced by the inclination 
towards one another of two plane, or, at most, very slightly 
concave, surfaces (pre- and ecto-trochanteric) at little less than a | 
right angle, the angle along the line at which these two surfaces 
meet being at the same time broadly rounded off, instead of form- 
ing a prominent crest. Thus, though the anterior production of 
the trochanter is considerable, the process lacks the lateral com- 
pression, which is a conspicuous feature in all the Dinornis femora 
to which we have access. An obscurely indicated rough surface, 
rather than a distinct ecto-trochanteric tuberosity, marks the 
subsidence upon the shaft of the anterior trochanteric prominence 
(Pl. iii, fig. 1 #7’). The pre-trochanteric surface (PI. iii., fig. 1 D) 
is nearly flat, and does not present an oblique line or ridge, nor 
the conspicuous subcircular area for muscular attachment, which 
is shown in various Dinornis femora. The latter feature may, 
however, be represented by an irregular elevation, obscurely 
represented in Pl. iti, fig. 1, at the junction of the anterior and 
inferior surfaces of the neck close to the head. 
* In PI. iii., fig. 1, the steepness of the ascent of the trochanteric part 
of the articular surface is much exaggerated owing to the foreshortening, in 
the photograph, of the anterior projection of the mass of the trochanter 
itself. 
