the Ccelacanth Fish. 335 



identical position in all Osteolepids, and by itself goes far 

 to establish a derivation of the CoeUicanths from that group. 



In a paper now ready for publication I have pointed out 

 that the long uiiossitied region which separates tlie basi- 

 sphenoid and basioccipital of Osteolepids, as shown in Bryant's 

 description of Euslheiiopteron, is functionally connected with 

 the hinge in the dor.«al roof of the head. 



The neural cranium of Coelacanths is very difficult to 

 interpret. Dr. Stensio is, of course, perfectly correct in his 

 iiiier])retation of the "sphenoid.'" It is probable that the 

 most dorsal foramen passing through tliat bone is for the 

 o[)thalmicus profundus, and that the loAver and most anterior 

 of these transmitted the third or sixth nerve. The remaining 

 foramen may have transmitted the trochlearius. 



I ;im very doubtful of Stensio's interpretation of the space 

 between the basisphenoid and parasphenoid as a myodonie. 

 In i\Iacro2)oma the basisphenoid fades away gradually until 

 it is represented only by bony spicules. In different indi- 

 vi'iuiils there is some evidence that there is a progressive 

 groAvih of bone into this region, and the conclusion seems 

 obvi )us that it was in life occupied by curtilage which failed 

 to ossify. The basisphenoid-i of Macropoma, Wiinania, and 

 Axelia are niuch less couipletely ossified than those of the 

 Coal Measure and Upper Permian Coelacanths. In these 

 forms the hinder surface of the basisphenoid is formed by a 

 large, nearly circular, concave condyle, extended laterally 

 and upward by small additional faces ; it is entirely identical 

 with that of Megalichthys. In these forms it is most 

 probable that the lower surface of the basisphenoid is in 

 direct contact with the parasphenoid. 



In any case, the position of this hole is not that of the 

 Piilaeoniscid myodonie which lies dorsal to the basi[)terygoid 

 process and lateral of the body of the basisphenoid. 



Tlie large bone called by Stensio the prootico-opisthotic 

 ccrre>ponds most accurately with the prootic in Eustheno- 

 pferon (Bryant, 1919). The two bones agree in articulating 

 witii the lateral borders of the basioccipital, and stretching- 

 forward in front of that bone along the sides of the great 

 unossitied tract of the basis cranii. In Osteolepids they do 

 not reach the basisphenoid as they do in Coelacanths. In 

 both £oi;ms there are anterior and posterior flanges from the 

 skull-roof which meet or nearly meet the upper edge of the 

 prootic. The two bones differ, however, in the abjence in 

 Eusllteuopteron ot the pocket wliich occupies the iiinder part 

 of the side of the Ccelacanth prootic. Tlie function of this 



