TRIASSIC FISHES FROM SPITZBERGEN 
77 
that it may correspond, at least partly, to the hyomandibular canal and partly to the 
mandibular line of pit organs of the Selachii (cf. Ewart, 1892, pp. 71—72; Allis, 1902, 
p. 108; Johnson, 1917, pp. 20—21). In Amia (cf. Allis, 1889, p. 506, PI. XLII; 1905, 
pp. 406—407) and Polypterus (Allis, 1900 a, p. 445) it seems to be represented by the 
horizontal cheek line of «pit organs» together with the ventral part of the preopercular 
canal and in the Stegocephalians by the «jugal canal» of Modie’s description (Modie, 
1908, p. 515; 1915 a, p. 320 ). 
As far as is known its agreement with the latter canal is very great and on this 
ground I have in the subsequent exposition used the term jugal canal for it (cf. Wiman, 
1914 b, PI. Ill, fig. 3 ; PI. V, fig. 2I 
While in W. sinuosa the jugal canal has, as we have seen, anastomosed with the 
infraorbital canal between the lacrymo-jugal and plate x, in Osteolepis (Goodrich, 1919) 
this anastomosis on the contrary takes place in the jugal bone. Consequently, as far as 
one can judge, the lacrymo-jugal in W. sinuosa extends not so far dorsally in its postorbital 
part as the jugal in Osteolepis, and the part that is lacking corresponds to the plate x. 
If, however, the restorations of Coelacanthus (Wellburn, 1902, p. 483), Diplurus (Dean, 1895, 
p. 154, fig. 156), Undina (Reis, 1888, PI. V, fig. 12; Woodward, 1898a, p. 80, fig.6o) x ) and 
Macropoma (Woodward, 1909, p. 173) are correct, the jugal part of the lacrymo-jugal ought 
in these forms to reach the postorbital and thus in this respect to agree with the jugal 
of the Rhipidistia, i. e. the plate x ought to be comprised in the posterior part of the 
lacrymo-jugal, of which it has probably been originally an integral part. 
As I have shown, the lacrymo-jugal extends in W. sinuosa a good way in front of 
the orbita along the lateral surface of the ethmoidal region and this is also the case in 
other Coelacanthids. This makes it probable that the lacrymal bone of the Rhipidistids 
(Watson and Day, 1916; Goodrich, 1919), which is normally situated under the anterior 
part of the orbital opening, is contained in the lacrymo-jugal of the Coelacanthids in 
general together with the jugal, as I wished to indicate in my terminology here used. 
The postorbital of W. sinuosa seems, on account of the course of the supraorbital 
canal and also because of its extent, to be undoubtedly the homologue at least of the 
postorbital of the Rhipidistids (Watson and Day, 1916), which also may be the case 
with the postorbital in the carboniferous Coelacanthus- species, if Wellburn’s (1902) view 
is correct, while, on the other hand, in Diplurus (Dean, 1895), Undina and other Jurassic 
forms (Reis, 1888, 1892, Woodward, 1898 b) and Macropoma (Woodward, 1909) the same 
plate would extend backwards right to the preopercular. In this connection it is interesting 
to find how the upper part of the squamoso-preopercular in the Carboniferous forms 
projects upwards between the postorbital and the opercular and thus shows exactly the 
same conditions as the dorsal part of the squamosal sometimes does in the Rhipidistids. 
On the other hand it is also noteworthy when one finds, for instance in Sassenia 
tuberculata described below, remains of a small independent plate situated behind the 
postorbital, and the squamoso-preopercular at the same time does not reach up dorsally 
between the latter and the opercular. It is also remarkable that this independent plate 
l ) In 1892 Reis gives another restoration of Undina (PI. I, fig. 1; PI. II, fig. 14) that differs rather considerably 
in this respect from that published by him in 1888. If this new restoration were correct the homologue of the plate 
x would probably be included in the squamoso-preopercular. 
