ON CERTAIN CROSSOPTERYGIAXS. 1259 



the Ccelacanthids (Stensio, 1921). The pit lines seem, so far 

 as is known, to agree rather well with those in other Crosso- 

 ptery.sfians. 



POROLEPIS ? sp. 



Among the material collected by the Norwegian expeditions 

 in the Lower Devonian strata of Northern Spitzbergen, and for 

 the investigation of which I am indebted to Professor J. Kiaer, of 

 Kristiania, there is an anterior part of the cranial roof, which in 

 this connection deserves some remarks. The part has belonged 

 to a very large fish, probably referable to the same genus as 

 the scales known under the name Forolepis posnaniensis Kade 

 (c/. Woodward, 1891, p. 9; Stensio, 1918 a, pp. 66-73). 



On the upper surface of the part present (PI. I. fig. 5) no 

 sutures at all can be observed. If, however, the outer layers of 

 the bones are removed, the radiation from their ossificatory 

 centres appears, though not so distinctly as in Dictyonosteus or 

 Eusthenojiteron owing to the less favourable state of preservation 

 of the bone tissue. By means of this method, I have been able 

 to observe that the division of tiie dermal roof covering the 

 ethmoidal I'egion consists of a large number of plates, probably 

 even larger than in Dictyonosteus. The exact aiTangement and 

 extension of these plates as well as their number ca,n, however, 

 not be made out with the help of the fragmentary material at mj 

 disposal. One can only see clearly that the nasal is represented 

 by two or three plates. The supraorbital sensory canal {soc, PI. I. 

 fig. 5) ruiis exactly as in Dictyonosteus. It communicates 

 anteriorly with the anterior parts of the infraorbital canal 

 (ifc) of its side, which in its turn anastomoses Avith that of the 

 opposite side through a well-developed ethmoidal commissure 

 (ethm.com.), the conditions thus being also in these respects as in 

 Dictyonosteus. 



The Ccelacanthids. 



The Ccelacanthids are now rather completely known (Stensio, 

 1921, 1922 ffl, 1922 6; Watson, 1921), and I have here only 

 certain remarks to make on them. 



It is easily seen that the sphenoid of Diplocercides agrees well 

 with that of the Rhipidistids in most features, and there can 

 be no doubt that the soft anatomy of the region occupied bv 

 this bone in Diplocercides in the main must also have shown 

 great agreem.ent with that of the corresponding region in the 

 Rhipidistids. 



In Diplocercides (text-fig. 5) we can clearly see that the 

 trigeminus branches proper or their roots, with the exception of 

 the r. ophthalmicus profundus, which has its exit more forward, 

 have pierced the cranial w^all jusc postero-dorsally of the basi- 

 pterygoid process and at the hinder border of the alisphenoid 

 Avulst, together with the r. ophthalmicus lateralis and n. buccalis 

 lateralis (or their roots). From this place all these nei-ves have 



