PAIiATOQUADRATE WITH NEUROCRANIUM IN CCELACANTHIDS. 51 



cranium was pi'imaril}^, in all pi^obability, that of a processus 

 basalis metapterygoidei with the lateral edge of the basis cranii 

 at or near the hind end of the trabecula. The processus basalis 

 has, in certain of them, later there fused with the basis cranii, 

 while still retaining its connection with the palatoquadrate. In 

 others it has apparently lost its connection with the palatoquadrate 

 but retained that with the neurocranium and then separated into 

 two parts, thus giving rise to a meniscus metapterygoideus and 

 a processus basipterygoideus. The metapterygoid cartilage is 

 always here lined ventrally, and supported by a dermal pterygoid 

 bone, the processus basipterygoideus being similarly lined and 

 supported by a lateral process of the dermal parasphenoid, and in 

 those animals in which the metapterygoid and basipterygoid 

 cartilages are resorbed, or fail to be developed, the two dermal 

 bones replace them and acquire sutural connection with each 

 other. 



The conditions in the three Ooelacanthidse so fully and so well 

 described by Stensio (1921, 1922) may now be considered, and 

 the question is : Is the postorbital ai'ticulation of the palato- 

 quadrate of these fishes with the cranium the homologue of that 

 in Heptanchus and Hexanchus, or of that in Lepidosteus, 

 Osteoglossitm, and higher vertebrates ? 



The metapterygoid of these fishes is a substituent bone, as it is 

 in most recent fishes, and at either end of its dorsal edge there is 

 a pointed process. Between these two processes the dorsal edge 

 of the bone is concave, and articulates with a process of the 

 neurocranium that Stensio considers to be a processus basi- 

 pterygoideus and which he refers to as the process e. In both 

 Winiania and Axelia this latter process is a process of what 

 Sbensio considers to be a median basisphenoid bone. It is 

 directed dorso-antero-laterally, and in Wimania extends so far 

 dorsally that its dorso-anterior corner is shown in contact with 

 the dermal bones of the roof of the skull. In Axelia it is shorter, 

 but still extends above the middle line of the lateral surface of 

 the cranium. In Diplocercides it is still shorter, and is a process 

 of a median sphenoid bone and not of a basisphenoid. The 

 metapterygoid would seem, from the figures given, to have 

 articulated either with the la,teral surface of the process e 

 immediately beneath its laterally projecting dorsal edge, or with 

 that edge itself, and as the dorsal edge of the metapterygoid is 

 longer than the process e is wide, the dorso-anterior end of the 

 metapterygoid would seem to have projected anteriorly somewhat 

 beyond the process e. 



The metapterygoid has an even outer surface, without process 

 of any kind, and there is no posteriorly directed process, or other 

 feature, either on this bone or at any place along the hind edge 

 of the palatoquadrate that would indicate a rigid attachment 

 to the hyomandibula. This latter attachment was there- 

 fore probably by ligament only, and in this, as also in the 

 general configuration of this part of the palatoquadrate, there is 



4*' 



