ON CERTAIN CROSSOPTERYGIANS. 1267 



is knowa about it, lout presumably it accompanied the trigeminus 

 branches." * 



" Tlie myodome in the Coelacanthids as described here has 

 pro-otic, pituitary, and prepituitary parts (c/. Allis, 1919). The 

 pro-otic pai't of it is represented by the fossa situated between 

 the basisphenoid (pars basisphenoidea of the sphenoid in Diflo- 

 cercides) and the parasphenoid. The pituitary pa.rt is situated 

 beneath the fossa hypophyseos, and is separated from this fossa 

 by a roof of bone pierced only by a small opening for the 

 hypophysis. The prepituitary part is very short, and is only 

 homologous with the median portion of the prepituitary part of 

 the myodome in the Teleosts. Homologues to the lateral portions 

 of the prepituitary part of the myodome in the Teleosts are 

 totally lacking, even as potential." 



" As at least in the fossil state of preservation the myodome 

 in the Coelacanthids is bounded directly by the parasphenoid on 

 the ventral side, it might be supposed to consist of both a dorsal 

 and a ventral compartment, like the myodome in the Teleosts, 

 If this were actually the case, the ventral compartment must, 

 however, at least in certain species {Axelia, Stensio, 1921, p. 95), 

 have been very short and limited only to the anterior end of the 

 myodome, as the parasphenoid in these species (Axelia) beneath 

 the middle and posterior parts of the myodome undoubtedly 

 seems to comprise parts ossified in the ventral wall of the 

 primordial neuroaranium, and the parts so ossified there form 

 the floor of the myodome." t 



'•The myodome in the Coelacanthids can easily be thought to 

 have ai'isen from the conditions in the Rhipidistids in the 

 following way. If in the Rhipidistids the postero- ventral parts 

 of the orbitotemporal region and the adjacent part of the 

 labyrinth region became much compressed from the sides and 

 thinned by an increase in the size of the eye-balls, the canalis 

 transversus would of course be ver}'" much shortened. If, then, 

 the musculi recti extern! invaded the opening of this shortened 

 canal, they would rather soon meet with their origins in the 

 median line, causing a considerable widening of the whole canal. 



* In 1113'' paper ou the Devonian Ciieliicantliids from Wildungeu (Scensio, 1922 a, 

 p. 183j, I was of the opinion that the n. abducens in Diplocercides ous'it to have 

 I'uu postero-ventvally of the so-called basipteiygoid process of the tish. The opinion 

 maintained in the present work seems, however, more in agreement with the general 

 conditions than my earlier one. 



t ''From the conditions described here in Dlctyonosteus it is evident that the 

 objections made by Allis, 1922 (pp. 149-152) against my interpretation of certain 

 parts of the neurocraniura in the CaBiacanthids must be untenable. There can be 

 no doubt that the dorsal part of the basipterygoid process of Dicti/onosteus is homo- 

 logous with the corresponding part of the process of JSitsthenopteron and the whole 

 process of the primitive Ccelacanthids {Diplocercides), and tliat accordingly also the 

 jugular vein in this form has run dorsall}' of the process as it has done in Dictt/)- 

 nosteus. Among the post-Devonian Ccelacanthids we have in Axelia a form whicii 

 with regard to the position of the basipterygoid process shows rather intermediate 

 conditions between the primitive Coelacanthids {Diplocercidss) and TFimania, in 

 which latter the process issues very high above the cranial base." 



