ON CERTAIN CKOSSOPTEEYGIANS. 1257 



Dermal Bones of the Cheek. 



The dermal bones of the cheek posterior of the orbital entrar.ce 

 seem in the main to be con-ectly desciibecl by Bryant. In my 

 specimens I can clearly observe the postorbital, the jugal, the 

 squamosal, and the preopercnlnr element. The jugal ought, 

 perhaps, to have a larger extension forward beneath the orbital 

 entrance than Brj'ant has given it in his restoration (his text- 

 fig. 2). 



Visceral Skeleton. 



The visceral skeleton is badly pi'eserved in my specimens, but 

 T find, however, that Bryant is correct in distinguishing a dermo- 

 palatine (termed palatine by him), an ectopterygoid, and an 

 entopterygoid . 



The Sensory Canals of the Head. 



The sensory canals proper form closed canals in the bones, and 

 open outwards Avith tnbuli in the normal way. As the bones are 

 usually thickened on the inner side along the sensory canals, the 

 course of these canals can be rather well ti'aced without prepara- 

 tion in those cases where the bones display their inner surfaces. 

 This is partly finely shown by certain of Bryant's figures, e. g. 

 pi. iii. fig. 2 ; pi. v. figs. 2, 3. At least most of the structures 

 described by Bryant as sensory canals proper are grooves indi- 

 cating the position of lines of pit organs, as we shall see from the 

 subsequent account. 



The supraorbital canal [soc, text-fig. 4) is very well preserved 

 in my specimen nearly throughout its length. Its posterior end 

 is situated in the dermosphenotic, where it anastomoses with the 

 infraorbital canal. From there it runs forward thi'ough the 

 lateral part of the fi'ontal. and then pierces the nasal in a longi- 

 tudinal direction. From the nasal it must have entered one of 

 the lateral rostral plates, in which it probably anastomosed with 

 the infraorbital canal of its side. The sensoiy organs in it have, 

 a,s far as one can judge, been compni-atively few, as the foramina, 

 for nerve-branches thi'ough its lower wall are comparatively few 

 {cf. Allis, 1889, ph xlii."; Cole, 1898, pi. xxii. ; Herrick, 1899, 

 pi. xvii. ; 1901. pi. xiv., etc.). 



The infraorbital canal {{ifc, text-fig. 4) : defined as in my 

 memoir on the Triassic fishes from Spitzbergen ; Stensio, 1921, 

 p. 103] has its posterior end at the boundary between the 

 supratemporal and intertemporal. In its course forward it 

 pierces first the intertemporal, then enters the deimosphenotic, 

 in which it anastomoses with the posterior end of the supraorbital 

 canal, as mentioned in the description of this region. Just at this 

 anastomosis it turns laterally, and continues laterally and down- 

 wards on the cheek, where it pierces the postorbital, the jugal, 

 and certainly also the lacrymal. It probably was in communi- 

 cation with its fellow of the opposite side through an ethmoidal 



