TRIASSIC FISHES FROM SPITZBERGEN 
io3 
on the anterior parts of the canal above the ethmoidal region, although the state of 
preservation makes it impossible to give a more detailed account of the conditions there. 
With regard to the extent of the infraorbital canal I have here given it in the 
definition of Ewart (1892), Ewart and Mitchell (1892) and Cole (1896; 1898), i. e. I restrict 
it to the parts which are innervated by the r. oticus lateralis and n. buccalis lateralis. 
The canal in the supratemporal part of the supratemporo-extrascapular, which was 
probably innervated by the r. dorsalis glossopharyngei, comes in this way to be assigned 
to the cranial section of the main lateral line of the body. 
The mo^t anterior part of the infraorbital canal is incompletely preserved in this 
species. There are, however, remains that indicate that an ethmoidal commissure was 
present, situated in the rostrals. From this commissure the canal has certainly continued 
backwards through the membrane bones on the lateral surface of the ethmoidal region 
ventrally of the nasal capsule, but in the material present there is nothing now preserved 
of this part. The sub- and postorbital parts, on the other hand, are preserved; and it is 
clearly seen (cf. PI. 15, fig. 1) that they were enclosed in a tube of bone. When leaving - 
the cheek the infraorbital canal enters the fronto-dermosphenotic at about the posterior 
margin of the most posterior supraorbital plate. From here the canal passes off medially 
and backwards in the dermosphenotic component of the fronto-dermosphenotic, then 
turning straight backwards. Just at the point where it takes the last-mentioned direction 
it anastomoses with the supraorbital canal. Continuing in a caudal direction, it reaches 
the suture between the fronto-dermosphenotic and the parieto-intertemporal. In its further 
course backward it was first quite surrounded by membranous tissue on the ventral 
side of the lateral part of the parieto-intertemporal, then entered the lateral part of the 
bone, finally passing over into the ventral lamella furthest backwards. Guided chieflv 
by the exits of the cranial nerves I place its posterior end at the suture between the 
parieto-intertemporal and the supratemporo-extrascapular. 
The foramen trigemini, as we have seen above, is so situated that the r. oticus 
lateralis must have been sent off from n. buccalis lateralis beneath about the median 
part of the parieto-intertemporal; on the other hand, the n. glossopharyngeus ought to 
have had its exit in such a position that its ramus dorsalis may be supposed to have 
met the anterior half of the supratemporo-extrascapular, if it ascended somewhat vertically. 
As far as one can see from this, the region of innervation for r. oticus lateralis would 
thus be restricted to a part of the sensory canal in the parieto-intertemporal, while 
r. dorsalis glossopharyngei would innervate the anterior part of the canal in the supra¬ 
temporo-extrascapular (cf. Allis, 1889, p.516, PI. XLII; 1900 a, p.449; 1905, pp. 405, 
482; 1919, p. 73; Herrick, 1901, p.207; Pollard 1892 a, p. 397; 1892 b, pp. 529, 532, 
5 3 5 , 5 36 , 5 38 ). 
The number and position of the pores of the infraorbital canal are not yet known. 
The jugal canal (jc, PI. 15, fig. 1) has the same course as in Wimania simiosa. 
As we have seen, the cheek plates , in this region are reduced to small scale-like plates 
and the canal is preserved only because it was enclosed in a tube of bone. 
The mandibular canal (me, PI. 15, figs. 1, 2) passes through the whole length of 
the supraangulo-angular and then continues forward in the splenial. In its passage 
through the former bone it has opened outwards with five large pores (p), which are 
