AXOMODOXTIA. 399 



the shape and frequent presence of a spine on the scapula, and many fea- 

 tures of the limb bones in the different sub-orders, e.g. fore-limb of Therio- 

 desmus. Moreover the sqtiamosal often descends far down outside the quad- 

 rate (Fig. 223), and may contribute to the articular surface for the lower jaw. 



The vertebrae are amphicoelous and the cervical ribs are double-headed. 

 Hvpocentra are either much reduced or absent. Abdominal ribs are never 

 found. The skull has a well-marked supra-temporal fossa (except in the 

 Pareiasauria), and in one or two forms the temporal arcade is perforated 

 by a small aperture recalling the lateral temporal fossa. The quadrate is 

 fixed and frequently deduced in size. Secondary palatal plates of the pre- 

 maxillae, maxillae and palatines appear to be developed in many forms. 

 The pterygoids are large, extend back to the quadrates and meet in the 

 middle line. The occipital condyle is single or trifid, or double (Cynoq- 

 nathus}. The mandible usually shows a composition of several bones, but 

 this cannot always be seen. Pre- and post-frontals can usually be made out, 

 but sometimes, in consequence of the absence of sutures, they are indis- 

 tinguishable. In the pectoral arch there are clavicles, interclavicle and 

 ossified epicoracoids ; the scapula is frequently mammalian and has a 

 spine and all the bones are immoveably united. The pelvic girdle is 

 thoroughly mammalian ; all the bones are immoveably united and there 

 are two obturator foramina. The limbs are not well known ; but they 

 appear to have been stout pentadactyle structures. 



From the above short description and from the subjoined account of their 

 orders it is clear that the Anomodontia present both reptilian and mam- 

 malian features. Which of these preponderate it is difficult to say.* By 

 most investigators they are regarded as reptiles, and there is undoubtedly 

 much to be said for this view, which has been followed in the present 

 work ; but it must not be forgotten that our knowledge of even the coarser 

 features of their skeleton is very imperfect, and that we know nothing of 

 the finer details or of the soft parts. On the whole we think that there 

 is much to be said for the view that the Anomodontia are neither reptiles 

 nor mammals, but that they represent an independent type of structure, 

 we will not say intermediate between these two, but combining features 

 belonging to each. There is no reason that we can see for regarding 

 them as ancestral to mammals, and we do not propose to consider the 

 matter from that point of view. Having regard to the extreme incom 

 pleteness of our knowledge of their anatomy such a discussion would hardly 

 be profitable, and might draw off our attention from wider problems of 

 evolution which are perhaps more ripe for settlement. There are two 

 additional facts with regard to this group which must be borne in mind. 

 In the first place they are almost the only purely terrestrial extinct reptiles 

 known, and in the second they are known to us in a very large number of 

 cases by their skull only. ^ . 



Order 1. PARF.IASAUKIA. 



The temporal fossa is completely covered over dorsally by bone. There 

 may be a small latero-temporal fossa. Teeth conical or with a compressed, 

 cuspidate crown ; those on the margin of the jaws in a uniform series. 

 Pineal foramen large. Vertebral centra pierced for the persistent noto- 

 chord. Permian and Triassic. 



* See the " Discussion on the origin of Mammals " in the Proceedings 

 of the Fourth International Congress of Zoology, Cambridge, 1899. 



