118 ERIK A : SON STEXSIO 
more or less distinctly in the following- species: Coelacanthus elegans, Newberry, ( B. M. 
P. 7912), C. granulatus Agassiz (cast at Greifswald from an original at Marburg), Undina 
acutidens Reis (original specimen ofHEiNEKE’s text fig. I och PI. VI, fig. I, 1907), Undina 
penicillata Munster (orig. spec, of Reis, PI. I, fig. 9, in 1888), Libys snperbus Zittel, Coccoderma 
siievicnm (probably orig. spec, of Quenstedt, PI. 100, fig. 14 in «Der Jura» 1852; and 
orig. spec. ofHEiNEKE’s PI. VI, fig. n in 1907), Macropoma mantelli (a number of specimens 
in the British Museum, as P. 36 , P. 4219, P. 4247 and 4548), M. speciosnm Reuss (orig. 
spec, of Reis, PI. IV, fig. 2, 1888) and Mawsonia gigas ( B. M. P.,10355, orig. spec, of 
Woodward, PI. VII, fig. j; 1907 a). 
It is interesting to find, as shown by text figs. 52, 53, 54, that at least in a number 
of the geologically younger Coelacanthids the metapterygoid was comparatively low and 
more also closely connected to the pterygoid than is the case in the Palaeozoic and older 
Mesozoic forms. Even the ventral margin of this 
bone, which, contrary to what was the case in the 
Spitzbergen forms, passes into the anterior margin 
without any real boundary, seems in these younger 
forms to have been closely connected with the 
pterygoid, and in certain forms, as, for instance, 
Mawsonia (text fig. 54) this connection may be 
very intimate. 
The autopalatine has probably sometimes 
been misinterpreted as the prefrontal (in the 
sense of cartilage bone), as, for instance, by 
Heineke (1907, p. 8) in his description of Undina 
acutidens. The same bone may possibly also be 
represented by Reis’s (1888, p. 43; PI. 2, fig. 2) 
maxillary in Libys superbus. It is noteworthy that, 
at least in some forms (Macropoma), the auto¬ 
palatine has articulated with the preethmoid 
(Huxley, 1866, p. 37; PI. VIII, fig. 3 ). 
There seems to be no doubt, as Reis’s figures also show, that in certain forms a 
dermopalatine has been developed. I have observed it especially distinctly in Coccoderma 
nudum (orig. spec, of Reis, 1888, PL III, fig. 16) and in Libys superbus it is also found, 
though closely connected with the autopalatine (Reis, 1888, Pi. II, fig. 1). 
With regard to the maxillary I refer to what has been said about it above in the 
description of W. sinuosa pp. 72—72. 
In the mandibula of Mawsonia the articular has been shown by Woodward as an 
independent bone. In Macropoma, on the other hand, the same bone is said by him 
(1909, p. 176) to be fused with the supraangulo-angular; in spite of a careful examination 
of his material I can neither confirm nor deny this. Both in Coelacanthus (one specimen 
in the Berlin Museum labelled Coelacanthus sp. from Linton, Ohio) and in Coccoderma I 
have found it as a clearly independent ossification (text figs. 55, 56). 
What Reis called the hyomandibular in Libys polypterus (Reis, 1888, p. 41; PI. Ill, 
fig. 1) really consists of remains belonging to the neurocranium, either the prootico- 
Text fig. 55. Coelacanthus sp. 
Sketch of a posterior part of the mandibula from the 
medial side. Original in Museum f. Naturkunde, Berlin. 
About »/ 4 . 
Art, articular; Icor.pra, intercoronoideo-prearticular; 
Sang, ang, supraangulo-angular. 
Text fig. 56. 
Coccoderma suevicum Quenstedt. 
Sketch of the lower jaw from the medial side. Original 
in Naturalienkabinett, Stuttgart. About 3 / 4 . 
Art, articular; Icor.pra, intercoronoideo-prearticular; 
Sang, ang, supraangulo-angular. 
