20 Dr. JR. H. Traquair — Devonian Fishes of Canada. 



account, I did not feel quite certain as to whether I should consider 

 the postaxial processes of certain mesomeres as pararoeres or not. 

 The internal skeleton of the pectoral in the Carboniferous Ehizodopsis, 

 as I have observed it in specimens belonging to Mr. Ward. F.G.S., 

 of Longton, is also constructed on the very same plan. 



It follows, then, that the skeleton of the paired fins in the Rhizodon- 

 tidaj is an abbreviate uniserial " archipterygium," and if, as seems 

 probable, the Holoptychiidaa are the more archaic group, then the 

 Ceratodus-like pectoral of Holoptychius is a more primitive form 

 of limb. This paired-fin skeleton in Eusthenopteron and its allies 

 may therefore, along with the corresjoonding conditions in Fletiracan- 

 thus and Cladodus, be considered of special interest in connection 

 with the "archipterygium question." 



PaLuEONISCID^S. 



Cheirolepis Canadensis, Whiteaves. — Unfortunately the Edinburgh 

 Museum contains only a few fragments of this species, so that I have 

 no means of comparing it thoroughly with the Scottish specimens of 

 the genus, all of which, in my opinion, are referable to one species 

 only, Ch. Trailli, Ag. 



II. Fishes from the Lower Devonian of Campbelltoivn. 



Coccosteid^;. 



Fhlyctcenius, nov. gen. Ph. Acadicus, "Whiteaves, sp. — On examining 



several pretty good specimens of this curious Coccostean, named by 



Mr. Whiteaves Coccosteus Acadicus, I find that it exhibits certain 



characters which are neither in accordance with those of the genus 



Coccosteus, nor with Mr. Whiteaves's diagram of its cranial shield. 



Allied to Coccosteus it is, as the arrangement of the sensory grooves, 



correctly indicated in Mr. Whiteaves's figure, cleanly shows. But 



the sutures between the bones are only seen with the greatest 



difficulty ; indeed Mr. Whiteaves admits that the dotted lines in his 



figure only represent their " supposed outlines." Accordingly he 



has indicated by means of those dotted lines certain plates having 



the same general outline as the median-occipital, lateral-occipital, 



and central plates in Coccosteus decipiens, Ag., whereas the real 



outline of these plates seems to me to be very different. The 



median-occipital, instead of being trapezoidal, with long posterior 



margin, shorter anterior one, and convergent sides, appears elongated 



and five-sided, there being an anterior acute angle which is received 



in a notch between the two centrals in front, which are themselves 



also elongated and more or less of a six-sided contour. Laterally 



the external occipital, marginal, and postorbital may be easily made 



out, and I think the preorbitals are also apparent enough. The 



orbital portion of the shield (= the part between the letters d 



and e in Whiteaves' figure) are rather more anterior and look more 



forwards than in Coccosteus. None of our specimens show Mr. 



Whiteaves's " rostral " plate, which is evidently the equivalent of 



that which I have called anterior ethmoidal in Coccosteus decipiens. 1 



1 Homosteus compared with Coccosteus, Geol. Mag. Dec. III. Vol. VI. Jan. 

 1889, PL I. Fig. 2 a, e. 



