i38 
ERIK A : SON STENSIO 
In primitive Actinopterygians we still find, as we shall see, the myodome in a very 
primitive stage of development. 
The parasphenoid of the primitive Actinopterygii , unlike that of the Coelacanthids 
and the Rhipidistids, is characterized by the strong and high processus ascendentes. 
Several of the membrane bones that were independent in the Rhipidistids are 
found fused with one another in the Actinopterygians. These fusion have, how¬ 
ever, taken place partly in another way than in the Coelacanthids. Thus while the 
membrane bone-plates on the dorsal side of the ethmoidal region in the Coelacanthids 
have comparatively well preserved their primitive arrangement and to a certain extent 
their mutual independence as well, they exhibit in Actinopterygians in general a stronger 
degree of specialisation. It is however, not convenient in this connection to give a 
detailed account of the homologues of the elements on the ethmoidal region of 
the Actinopterygii, but I refer the reader to the description of Palaeonicids, Cato- 
pterids and Saurichthyids given below in this work, where the recent forms are 
also dealt with. 
Further it seems to be evident that the so-called dermopterotic («squamosal») of 
the Actinopterygii or its equivalent (the dermal component in the so-called pterotic of 
the Teleosteans) corresponds both to the supratemporal and the intertemporal of the 
Rhipidistids (cf. Regan, 1904, p. 33 g; Allis, 1919 b, pp. y 3 —77). On the other hand it 
is more difficult to be quite certain as to the homologues of the cheek-plates of the 
Actinopterygians as the course of the sensory canals and the so-called «pit-lines» in the 
primitive forms among them is inadequately known. It seems, however, to be certain that 
at least the posterior part of the large «preopercular» in the Palaeoniscids, Platysomids 
and Catopterids (cf. my description of these forms below, and Gregory, 1915, p. 332 ) 
and the whole of the bone that is termed the preopercular in higher Ganoids and 
Teleosts are homologuous to the similarly termed bone in the Rhipidistids. In addition 
we also find, as will be seen from the account given below in this work, in Saurichthyids, 
Chondrosteids and Acipenserids an independent element probably corresponding to the 
quadratojugal of the Rhipidistids. It seems probable too that this element is contained 
in the posterior part of the maxillary if Palaeoniscids, Platysomids and Catopterids, 
while in other Actinopterygians it may either be reduced (certain higher Ganoids and 
at least the majority of Teleosts) or represented by* one or more of the ventral 
postorbital- (suborbital-) plates (Semionotids, Eugnathids and other higher Ganoids). 
In higher Ganoids the post- and suborbital sections of the infraorbital canal may 
generally run through the so-called «circumorbitals», which to judge at least from the 
conditions in Semionotids and Lepidosteus (Allis, 1905, pp. 407 —417), seems to me most 
probable. 1 ) Then the circumorbital plates situated behind, beneath and in front of the 
orbit should together correspond to the postorbital, jugal and lacrymal of the Rhipidistids. 
For the same reason the bones in the infraorbital chain of the Teleostei ought together to 
correspond to some extent to the bones mentioned in the Rhipidistids. 
z ) According to Cqllinge (1894 a - PI. I, fig- 4) the sections in question of. the infraorbital canal in Dapedius 
pass off through the postorbitals (suborbitals) and the maxillary. This view does not, however, seem to be supported 
by the knowledge we now possess of the course of this canal, e. g. in Lepidosteus (Allis, 1905). The course of 
the anterior part of the supraorbital canal in Dapedius as given by Collinge seems also very doubtful. 
