!Prof. Agassiz on Fossil Fishes. 123 



Ifectly similar to those of the body. The operculum itself was 

 much higher than long, and formed a trapezoid with rounded pos- 

 terior angles. Its free margin is thin, but entirely smooth, which 

 is also the case with that of the preoperculum. The thoracic 

 girdle is extremely strong ; it forms, hindwards toward the throat, 

 a rounded angle, in front of which, situated in a hollow, is the arti- 

 culation of the pectoral fin, which was rather small, at least so it 

 appears, but of which I am able to say nothing further, never hav- 

 ing seen it preserved entire. 



The ventral fins were placed beneath the throat, perhaps even a 

 little more in front than the pectorals. 



The dorsal fin begins directly behind the nape with very strong 

 and long spines ; it appears to terminate at the commencement of 

 the last third of the total length. I presume that its last rays were 

 soft, and that there was no separation between the two kinds of rays 

 in the fin. 



The anal commences near the middle of the body ; it is narrow 

 but long, and may have possessed about fifteen rays, the first three 

 of which were spinous. 



The caudal is not yet known in detail ; its rays are covered at 

 the base by very close minute scales. The lateral line describes a 

 curve parallel to that at the back, occupying at top the first third of 

 the total height of the body. The scales which cover the entire of 

 the body are somewhat large and very thin ; so that the posterior 

 margin is rarely preserved. Examined under the microscope, these 

 scales present numerous concentric lines, very close upon each other, 

 and furnished in their anterior portion by a dozen grooves arranged 

 like a fan, which are visible to the naked eye. The concentric 

 lines disappear on the hinder portion of the scale, where we see 

 small granulations, which become excessively minute denticulations 

 on the free margin of the scale, and which fell off easily during life, 

 for I have never found them preserved except on some few scales. 



In fact, Scisenurus Bowerhankii is an Acanthopterygian thoracic 

 Ctenoid, having scaly cheeks, the hinder margin of the opercular 

 pieces smooth, and the jaws furnished with crooked and equal teeth ; 

 the bones of the cranium rather solid with thin crests. A peculiar 

 character is found in the enormous suborbital, and in the presence 

 of a single dorsal and of a single anal. 



Now, if we endeavour to determine the place of this fish in the 

 present system. of classification, we find but a single family of Acan- 

 thopterygian Ctenoids with which it can be associated, that of the 

 Sparoidce, which, while they have smooth opercular margins, pos- 

 sess, in other respects, the characters of the Percoidee. The fol- 

 lowing, in fact, are the characters assigned by Cuvier to his Spa- 

 toidte : — " The opercular pieces are not furnished with denticula- 

 tions or spines ; the bones of the head are solid, but not hollow, as 

 in the Scieenoidse. The palate is not furnished with teeth ; the 



