956 MR. D. M. S. WATSON OX A 



always occur in the large Stegocephalia, modified of course by 

 changes in the shape of the skull. I have seen them in : — 

 " Loxomma^^ Pteroplax, Batrachosuchtis, Cyclotosaurus, Plagio- 

 stemutn, TremMosav,rus, Mia'opholis, Capitosaurits, Aphcmierainma, 

 Bothriceps, Archegosaurus. 



In Bairachiderpeton the squamosal is united to the pterygoid, 

 and covers the back of the quadrate, not by the turning down 

 of its hinder edge but by a special descending flange, a difference 

 depending entirely on the development of the "horn" which has 

 carried the otic notch, which, in all amphibia and reptiles where 

 it is present, lies between the tabulare and the squamosal, far 

 out and backwards. The great sheet of squamosal which lies 

 behind the quadrate is in fact a new development, the original 

 posterior border being now represented by the flange. 



There is hence no doubt that the single temporal bone re- 

 maining in Batrachiderpeton is the outer of the original three. 

 That this is really the homologue of the mammalian squamosal, 

 I hope to be able to show in a subsequent paper. The other 

 bones of the top of the skull do not call for special notice. 



The palate of Bati-achiderpeton is of very great interest. I 

 pointed out recently that the tyf)e of palate found in '•^Loxomma,''' 

 with a large single basioccipital condyle, basipterygoid processes 

 on the basisphenoid, and large pterygoids, which leave only a very 

 small interpterygoid vacuity divided by a narrow parasphenoid, is 

 the primitive type of amphibian palate, because it agrees very 

 closely with that of the Crossopterygian fish, and because no 

 palate with large parasphenoid and interpterygoid vacuities is 

 known to occur in the Coal-Measures or Lower Carboniferous. 

 From known types it is possible to pick a morphological although 

 non-phylogenetic series showing exactly how the change from the 

 primitive palate to the typical palate of Capitosat(,rus has taken 

 place. Eryops, which, as shown by v. Huene's recent extremely 

 interesting paper, has a quite large basioccipital and basipterygoid 

 processes of the basisphenoid, and in which the parasphenoid and 

 interpterygoid vacuities are only of moderate size, is exactly 

 intermediate between the primitive type and that found in the 

 Triassic forms, where the basioccipital and basisphenoid are very 

 much reduced and the pterygoids are supported by sutural 

 union with the edges of the parasphenoid. The palate of 5«^ra- 

 chiderpeton agrees with the primitive type, in having large 

 pter3^goids meeting in the middle line and articulating with 

 basipterygoid processes of the basisphenoid. It diffeis, however, 

 in that it has already completely replaced the basioccipital con- 

 dyle by a pair of exoccipital condyles, just as have the later 

 large Stegocephalia. 



Discussion of the Lower Jaio. 



Except that it is not known whether an ej)icoronoid was present 

 or not, the lower jav/ of Batrachiderpeton is extremely like that 



