— 321 — 
versed theelements and has described and figure the scapulae as overlap- 
ping coracoids, and has called the anterior and posterior portions of 
the coracoids scapula and coracoid respectively . It may also be remarked 
that Seeley’s representation of the two clavicles as fused into a single 
transverse bar is an error. Broom (1904, Pl. IX) in his restoration 
of the shoulder-girdle of the South African type has followed Seeley’s 
error and reversed the scapulo-coracoids. The scapulae he miscalls pre- 
coracoids and represents them as meeting in the middle line. The an- 
terior and posterior portions of the coracoid he names respectively sca- 
pula and coracoid, as does Seeley. These misinterpretations by Seeley 
and Broom are, of course, due to the post-mortem displacement of the 
shoulder-girdle in Seeley’s specimen. 
Pelvic Girdle. 
The elements of the pelvic girdle are almost identical in form 
with those of Stereosternum, except that the pubic foramen in that 
genus is here represented by 9 deep notch. In this character it agrees 
with the South African Mesosaurus, as figured by Broom (1904, 
Pl. IX fig., 4) and this is one of my reasons for ascribing it to the 
same genus as the African form. The presence of the pubic foramen 
in Stereosternum seems to bedefinitely established, and in the pre- 
sent material there are five pubes wich show the deep notch in the 
postero-lateral border. This pubic incisure cuts the bone obliquely, 
appearing on the outer surface as a shallow notch, while on the 
inner surface it is much deeper, with its edges almost uniting 
(Pl. Ill., fig. 3, and Pl., IV., 9, pb). 
The ilium is a short rod somewhat expanded at the upper 
or sacral end, but shorter ;than the combined extremities of the 
two sacral ribs, with ,which its union must have been rather 
loose. 
The three elements of the pelvis havea firm synarthric union, 
but are not coossified. The two specimens in which all three ele- 
ments are shown in their natural relation have, unfortunately, only 
the inner faces exposed (Pl. IV., fig. 9), so that the acetabulum 
cannot be made oul. From examination of two specimens in which 
the outer face of the ventral elements is exposed, I believe the ilium 
had the greatest share in the formation of the acetabulum. The 
pubis and ischium form a nearly solid pelvic floor, with a rather 
5569 At 
