.VND CLASSIFICATION OF THE FOSSIL REPTILIA. 
255 
sacrum and pelvis referred by Sir PticiiARD Owen to Platijpodosaurus rohustus as 
the missing fossil. I have attempted to fit the specimens together ; but the caudal 
vertebrae are larger than might have been expected from the size of tlie sacrum, and 
a vertebra or two must be missing if the specimens are parts of one animal. On this 
point there appears to be no doubt, since Mr. William Davies, F.G.S., who 
superintended the development, states that he remembers the association of this tail 
with the remains which were subsequently referred to Platypodosaurus. The generic 
difference of this pelvis from the forms which have been attributed to Dicynodom is 
obvious ; but it is not improbable that a skull already referred to Dicynodon may be 
associated with the remains. 
The Scapidar Arch. 
In all Anomodonts the scapular arch probably includes the same elements, which 
are:—an inter-clavicle, clavicle, scapula, pre-coracoid, and coracoid. In some types there 
is a sternum also, but there is no reason for supposing that this bone is always ossified. 
In mode of grouping and arrangement of the bones there is a close resemblance to 
the Monotremata, the only group in which the pre-coracoid is similarly distinct. 
But in Procolophon it will be shown that tliere is a sutural union between pre-coracoid 
and coracoid ; not unlike that which has been demonstrated in other Anomodonts, and 
this leads me to believe that the ordinary Beptilian type of coracoid, which is per¬ 
forated in many types exactly as is the pre-coracoid, is probably the result of the 
obliteration of that suture, so that the coracoid in Reptiles may be held, especially when 
perforated, to comprise both bones ; and, therefore, the persistence of the pre-coracoid 
suture in Anomodonts may rather indicate a line of descent than a direct affinit_y ; 
and, judging from absence of the suture in some Amphibians, like Cryptohranchus, in 
which this part of the skeleton may be unossified, I am not disposed to regard 
division of this element into pre-coracoid and coracoid as conclusive against Reptilian 
affinity, or as showing affinity with Monotremes, wdiich might at first have been 
surmised. The difference from both Reptile and Mammal is, however, of an ordinal 
kind; and, so far, the Anomodont characters help to show that only one more ordinal 
type is required to complete the gradation between these classes. The united pre¬ 
coracoid and coracoid in Procolophon make a bone elongated in the antero-posterior 
direction, which may be compared in length and form with the coracoid of Plesiosaurs 
and certain Ichthyosaurs, and among existing Reptiles with IPitteria. 
On the Anomodont Scapula. 
The scapula appears to be more variable than any other bone in the Anomodont 
skeleton. In Keirognatlms cordylus I have described an elongated slender type, like 
the bone in Kistecephalus. The specimen No. 36,272 is imperfect at its junction with 
the coracoid (Plate 15, fig. 4), but shows a distinct constriction or neck, external to the 
articulation, because a moderate acromion process is developed on the anteiior margin ; 
