COAL-MEASURE STEGOCEPHALIAN". 959 



Urocordylus reticulatus. These three specimens agree very closely 

 in measurements and in all charactei'S which can be observed ; 

 they are undoubtedly cospecific. They also agree exactly with 

 Andrews' and Smith Woodward's specimens in the proportions of 

 the skull and body, and in size, but differ from Andrews' specimen 

 very markedly in the character of their cranial ornament, which 

 is almost entirely composed of ridges instead of pits. They mu«t 

 hence be regarded as specifically distinct. 



In text-fig. 167 (p. 955) I have given a drawing of the top of 

 the head with such sutures as are clearly visible in one of the 

 Newcastle specimens preserved in counterpart ; this should be 

 compared with Dr. Andrews' figure. It is certain that the ptery- 

 goids met in the middle line, as in Batrachideiyeion, although 

 the detailed structure of the palate cannot be made out. 



One important feature, shown clearly by the type-specimen of 

 " Urocordylus " reticttlatus, is the T-shaped cleithrum, also shown 

 in Dr. Woodward's figure, where it is interpreted as scapula. 

 Such a shaped bone is known in no other animal except Batrachi- 

 derpeton, and seems to show definitely that " Urocordylus " 

 reticidatiijS and Batrachiderpeton are closely related. The very 

 remarkable clavicle is also identical in the two types. 



Whether Urocordylus reticulatus H. & A. really belongs to the 

 genus Ceraterpeton is uncertain, but it is probable that it does ; 

 in any case, by calling it Ceraterpeton reticidatum in full, no 

 trouble will arise ; it seems probable that it is not Urocordylus. 



One of the most marked features of Ceraterpeton is the ex- 

 pansion and corrugation of the distal ends of the neural and 

 hjBmal spines ; the meaning of this condition, which occurs 

 throughout the column, is obscure. It also occurs in Ceraterpeton 

 galvani, Urocordylus wandesfordii, Scincosaurtis crassus, Ftyonius 

 cvstocephalus, and other types. As there is no definite evidence 

 that these types cannot be related, it is natural to assume that 

 they are ; but it must be remembered that any such relation 

 can only be slight, for if Fritsch's restoration of the skull of 

 Scincosaurus crassus is at all correct, it difiers veiy greatly from 

 C. reticidatum and Batrachiderpeton. 



Another type which appears to be related is Dicer atosaiiriis 

 jninctolineatus Cope, recently described by Jsekel. The palate 

 of this type is extremely like that of Batrachiderjjeton, diff'ering 

 in minor features of the dentition, in the larger interpterygoid 

 vacuity, and in the presence of a transverse bone. 



If we suppose, as is quite pi-obable, that some of the sutures 

 of the cranial roof were invisible, it is probable that the skull 

 structure was very like that of Ceraterpeton retictdaticm. On the 

 other hand, this type has not the T-shaped cleithrum and peculiar 

 neural arches of the latter form. 



Assuming, as seems justifiable, that Batrachiderpeton is allied 

 to Geraterjjeton reticidatum, it is of interest to see the differences 

 betweeir the two types. Ceraterpeton reticulatum is obviously the 



64* 



