OCR Text |
Show 1906.] DIAPTOSAURIAN REPTILE. 599 Sphenodon,but the imperforate condition of the coracoid is a feature found in a few Diaptosaurians. The pelvis is not much modified from tlie type found in most Diaptosaurians. The plate-like pubis and ischium are found in Rhynchosaurus. In the Phytosaur Erytlirosuchus the pelvis is very similar in structure; in Sphenodon the pelvic structure is very different. It is impossible to compare the tarsus with that of most allied forms, as little is known of the tarsus in the Gnatliodonts or Phytosaurs. There is, however, an undoubted resemblance between the tarsus of Howesia and that of Sphenodon, and this is more marked when that of the embryo is considered. The similarity is so close, that it is difficult to doubt that it indicates a relationship. In Rrocolophon we also see some affinity. It is interesting to note the similarity of the mode of articulation of the 5th metatarsa1. Taking the various points into consideration, I conclude that Howesia is a very slightly specialised Gnathodont. It seems to show, moreover, that the Gnatliodonts are not true Rliynchocephalians, but are probably more nearly related to the Phytosaurs. I incline to agree with Osborn in placing them in a distinct order, no doubt showing some points of resemblance to the Rhynchocephalians, but not so nearly related to them as is generally held. In the following table an attempt is made to indicate the phylogenetic relationships of the Diaptosaurian orders:- (COTYLOS AURTAN ANCESTOR) (R, P, PI, Permian). P r l y c o s a u r i a Pi; o r o r o s a u r i a (F, P, PI, Permian). (F, A, PI, Permian). R h y n c i io c e p h a l i a G n a t h o d o n t ia P h y t o s atjria (F, A, T, ? Triassic). (F, A, PI, Triassic). (F, A, PI, Triassic). P r o t e r o s u c i i ia (F, ?, ?, Triassic). In the above table the first letter of the formula, R or F, indi- |