234 ANNALS OF SCOTTISH NATURAL HISTORY 



to him to differ essentially in the squamation, and was there- 

 fore probably new to science. 



The form of the fish is long and narrow, and shows 

 posteriorly two dorsal fins, behind the second of which the 

 specimen is unfortunately broken off. Its measurements are, 



From tip of snout to just behind origin of second dorsal, io| inches. 

 anterior margin of orbit . f 



,, ,, anterior margin of clavicle . 3 ,, 



,, origin of first dorsal fin . . 7^ ,, 



,, origin of second dorsal fin . . gf- ,, 



Depth at shoulder and also at middle of body . . ii ,, 



So far as exhibited, the external cranial bones are finely 

 rugose-granulated on their surfaces ; the circular orbit is very 

 distinctly marked, and is y\ inch in diameter. The gape is 

 wide but only a few small conical teeth are exhibited. The 

 opercular bones are wanting, a palatopterygoid bone of the 

 usual Rhizodont shape being brought into view. There is 

 also exhibited a clavicle of the same general form as in 

 RJiizodopsis, etc. There are some remains of pectoral rays, 

 but the ventral, anal, and caudal fins are not preserved. 



The scales are unfortunately badly preserved ; only 

 enough is seen to show that they were thin and rounded, 

 and that they exhibited the usual fine, concentric, and 

 radiating markings of the scales in the Rhizodont family. 



The head, the scales, the fins, all show that the fish 

 belongs to the family Rhizodontidas, but its generic position 

 is not so clear owing to its deficient preservation. Un- 

 fortunately, with the single exception of the Canadian genus 

 Eustlienopteron, very little is known of the Upper Devonian 

 Rhizodontidae in general, as their remains have hitherto 

 occurred only in the most fragmentary condition, like the 

 Russian remains named by Pander Polyplocodus^ and those 

 from Clashbennie in Scotland, named by Agassiz Bothriolepis 

 favosus, and now provisionally referred by Mr. A. Smith 

 Woodward to Newberry's genus Sauripterus? 



There is, however, in the collection at St. Andrews, 



1 It has been usually supposed that Pander's Polyplocodus is synonymous with 

 Agassiz's Cricodus, but I have already, in the first number of this Journal (p. 35), 

 shown that Cricodus was founded on a tooth of Dendrodont structure. 



~ 2 Cat. Foss. Fishes, British Museum, Pt. 2, p. 365. 



