Annals op the Transvaal Museum, 
141 
edge, in front of the glenoid facet and slightly nearer to the hinder margin 
than to the front ; the other is situated near the middle line of the flat 
portion and opposite the termination of the hinder-outer border on the 
lower margin of the glenoid facet. The first mentioned is the supra- 
coracoid foramen, the last is the glenoid foramen. The glenoid foramen 
pierces the bone upwards and opens on the inner surface behind the 
lower end of the subscapular fossa. The supracoracoid foramen opens 
on the outer surface to continue for a short distance as a slight impression 
on the flat lower portion of the bone. It pierces through to the other side 
in a direction parallel with this flat portion outwards and slightly forwards 
and opening on the inside in the lower end of the subscapular fossa. 
This fossa extends in the same direction upwards. It is also nearly 
parallel with the coresponding front border of the bone. At its upper 
end it broadens to receive the inner opening of the supraglenoid foramen, 
which pierces the bone from the upper end of the margin of the outer 
supraglenoid foramen in a backwards-forwards direction. The supra- 
coracoid foramen is larger than the glenoid foramen and the supraglenoid 
foramen is the largest of all. 
The Fore-limbs. 
There are three fairly complete fore-limbs, two right ones and a left 
one and fragments of another left one (PI. XVI and XXI). 
The humerus. There is one humerus of which only a distal portion 
is broken off. Another is only represented by its distal portion and the 
impression of the proximal end. The other two are badly broken. From 
these bad specimens one would conclude that the two ends of the bone 
make an angle with each other as in Eryops for example ; but the better 
specimens show that the ends expanded in the same plane. This of course 
would be a peculiar difference from the other large Temnospondyls as 
Actinodon , Eryops , Cacops, etc. I am, however, not quite sure of this 
position although it seems very probable. 
The indifferent position of the humerus in life must have been from 
the glenoid facet slightly downwards and backwards. The articular face 
on the proximal end forms a large cavity, the edges protruding far from 
the general suiface. If the bone lies flat in its natural position, it is the 
upper edge which projects farthest. This cavity must have been filled 
with a large amount of cartilage. The upper surface of the proximal end 
is convex, the lower slightly concave. Towards the shaft the hinder 
border bends further downwards than the body of the bone. There were 
two slight projections in the centre of the upper surface. 
The distal end is broader than the proximal one. The articulation 
part is strongly convex. On the lower side the bones show a small ridge, 
visible as an impression in the matrix and running transversely over the 
bone. This ridge is distally stronger concave and defines the lower and 
inner border of the articulation facet for the radius. There is a thick, 
broad and high ridge on the upper surface extending from the distal end 
above the already mentioned articulation surface for the radius up to the 
lowest tuberosity on the proximal end, gradually diminishing in height 
in that direction and vanishing there. The distal end of this ridge 
formed part of the articulation surface for the radius. 
The shaft is slightly narrower than the proximal end and shows no 
typical features. 
There are two processes on the anterior border. One is just above 
the slighly constricted middle and has the shape of a short, truncated 
