82 
Annals of the Transvaal Museum 
of 5 mm. The articulation surface of this bone has a breadth of io mm. The 
scattered haemapophyses belong to the missing vertebrae between the second 
caudal and the series of eight. It seems as if the vertebrae in the mass near 
the end of the series belong to the proximal portion of the tail, but no re- 
markable features are visible. 
Shoulder Girdle. 
The scapulae are completely embedded in the rock. The right scapula only 
shows an indeterminable edge and the left only what are probably its articu- 
lation surfaces. One of these surfaces has a more or less triangular contour 
and this is probably the articulation surface for the humerus. The other is for 
the contact with the coracoid. Both surfaces are concave and they meet at 
an obtuse angle. 
Both coracoids are present. Half of the left coracoid, however, is hidden 
by the one of the right side. Both bones only show their lateral surface, and 
therefore nothing can be 
made out of their thickness 
or their marginal surfaces. 
The contour of the right 
coracoid is complete. The 
scapular border of the cora- 
coid is not straight, for slight- 
ly above its junction with 
the glenoidal border it is 
deeply indented. This in- 
dentation has a length of 
12 mm. and a maximum 
depth of 3 mm. Further 
upwards the contour line pjg 2 . Aristosauruserectus. Right and left coracoid 
bends slightly forwards. The and proximal surface of left scapula. The border 
upper border is not well pre- behind the supracoracoidal foramen in the left 
served but seems to be hori- coracoid is not clearly visible. About nat. size, 
zontal ; it would then make a nearly right angle with the front border. The upper 
portion of the front border is straight; below the level of the middle of the 
bone it curves backwards; the lower front border of the bone is straight. The 
glenoidal border is fairly straight and meets the lower portion of the front 
border practically at right angles. The surface of this coracoid is fairly flat 
and one would scarcely take notice of some slight undulations if the left cora- 
coid were not present. The visible portion of that bone is not flat for it shows 
a rather deep concavity on its outer surface, running from the hinder end of 
its glenoidal border upwards. The supracoracoidal foramen is situated in this 
concavity, which continues past this opening. The lower corner of both inner 
and outer surface of this coracoid bulges outwards. Both peculiarities are also 
seen in the right coracoid, but in a much less degree. As the bones are now 
reduced to the thinness of paper, it is not impossible that the described un- 
dulations are due to original differences in the thickness of the bone. However, 
it must be remembered, that the bulge in these coracoids is in practically the 
same position as the tuberositas coracobrachialis described by v. Huene on 
the coracoid of Pachysaurus magnus ( 3 , p. 149). 
The upper and front portion of the bone is much thinner than the lower 
and hinder portion. The greatest height of the right coracoid is 43 mm. and 
its greatest breadth 32 mm. Both coracoids have a foramen supracoracoideum. 
It is fairly large and situated close to the middle of the scapular border. Its 
