140 
Annals of the Transvaal Museum. 
covered with ridges, radiatirg from the narrow end ; the inner surface is 
smooth but not fiat. The relation of this part to the scapula is not quite 
clear, although it seems to touch the front edge of the scapula. There 
may have been an opening between the anterior superior angle of the 
scapula and the bend of the clavicle. The upper portion of the bone, 
which is preserved in one specimen only, forms an angle of some 
130 degrees with the lower portion. It is long and slender, much broader 
than thick and about as long as the triangular portion. The outer surface 
of this part is smooth. There is only one specimen showing the relation 
of clavicle, cleithrum, and scapula. This shows the clavicle and scapula 
separated by the cleithrum. It seems to me that this is the natural 
relation of these bones. According to Case (3) the same relation occurs 
for example in Dissqrophus (p. 118), but it seems to be different in 
Eryops (p. 100). 
The cleithrum is a long slender bone, which covers nearly the whole 
top of the scapula but is not suturally united with it. The anterior end 
of this bone lies slightly in front of the superior anterior angle of the 
scapula. There it lies against the hinder part of the outer surface of the 
cleithral end of the clavicle. Higher up the cleithrum bends behind this 
bone. Clavicle and cleithrum are not suturally united. 
The scapula-coracoid is a large bone with no indication of a suture 
between the composing parts. The upper part is broad and thin and 
slightly bent inwards. The inner surface of this portion is transversely 
convex, while its outer may have been correspondingly concave. Its 
upper border is convex behind and slightly concave in front. As these 
properties are only concluded from impressions, nothing can be said of 
this border with regard to cartilage. If this has been present it can only 
have been a small quantity, as the cleithrum fits nicely into the slight 
concavity and over part of the convex border. The anterior border is thin, 
and still represented by small fragments. The posterior border is thick 
and its lower part is still present together with the whole lower portion of 
the bone in a left scapula-coracoid. 
The posterior border divides in its lower half to include the supra- 
glenoid fossa (PI. XXIY, fig. 2). The outer portion of this border does 
not extend so far backwards as the inner ; it is much thinner than this 
and continuing downwards, curves backwards to terminate at the upper 
and outer end of the glenoid facet. The inner and much thicker portion 
of the hinder border makes a much stronger curve downwards, inwards, 
and backwards, towards the inner end of the glenoid facet. This facet is 
very long and comparatively narrow, mostly so beneath the lower end of 
the supraglenoid fossa. The position of the facet is from outward to 
inward and backward looking slightly downward. Its edges project from 
the general surface, thereby denoting that the facet has been covered by 
cartilage. The connection between the humerus and the scapula-coracoid 
was through the medium of cartilage, as the proximal end of the humerus 
is built on the same principle. The outer end of the glenoid facet marks 
two different portions of the bone in as much as the upper part is bent 
along a line running from this end horizontally forward. These portions 
enclose an angle of about 90 degrees. The lower portion is thin, long, 
and moderately broad (PI. XXIY, fig. 1). Its inner border is elliptical, 
while its hinder-outer border is only slightly convex and curves concavely 
into the lower margin of the glenoid facet where this is narrowest. The 
extreme front of the part where the two portions meet is lost. The 
lower portion contains two foramina ; the one is situated near the outer 
