598 PEOCEEDTXGS OF THE GEOLOGICAL SOCIETY. [June 20, 



zodopsis is separated by the presence of teeth of two sizes, by the 

 thinner scales, and by the regularity of the striate ornament over 

 their whole surface, contrasting with the coarse tuberculation of 

 the posterior and smoothness of the anterior area of the scales in 

 the Old Red Sandstone genus. Prom Rhizodus the absence of tren- 

 chant teeth is a broad distinction ; the character of the scale-orna- 

 ment of the former, as described below, is equally diagnostic. 



The specimens which have been selected as typical of the genus 

 average 5 inches in length. For them the specific name B. sau- 

 roides has been used, as in Prof. Williamson's paper, which contains 

 the only published description in England of the scales. Other 

 specimens, whose length is estimated at 12 or 15 inches, are imper- 

 fectly preserved ; it remains to be seen whether they are specifically 

 distinct, or whether the greater thickness of their bony scales is 

 merely a concomitant of the greater size of the fish. The same 

 doubt exists in the case of the small flexible scales, whose possessors 

 probably did not exceed 2 inches in length. 



In a paper " TJeber das Yorkommen von Rhizodus Hibherti, Owen, 

 Megcdichthys Hibherti, Agassiz and Hibbert, in den Schieferthonen 

 des Steinkohlengebirges von. Volpersdorf in der Grafschaft Glatz " 

 (Zeitschr. deutsch. geol. Gesellsch. p. 272, 1865), Roemer figures 

 at least one scale presenting distinctly the characters of Rhizodopsis. 

 The concentric and radial sculpturing is described nearly as above, 

 with the minor difference that the concentric seem of equal promi- 

 nence with the radial ridges. But there is some confusion in the 

 author's mind — there is at least in his remarks — about the identity 

 of his specimens with the Scottish genera. For in speaking of the 

 teeth of Rhizodus he omits notice of their distinctive feature, the 

 double trenchant edges and the elliptical form of the transverse sec- 

 tion of the tooth ; both are figured (taf.vi. f. 5), but only the plicate base 

 is described. Again, it is said that the thick Scottish scales (which 

 are not horny, as Roemer says, but osseous) show, when carefuUy 

 split off in layers, the same ornament as those which he has figured 

 as Rhizodus. Admitting the hazard of criticism without sight of the 

 specimens, I cannot help suspecting that Roemer has fallen into the not 

 uncommon error of mistaking internal structure for superficial orna- 

 ment. The suspicion is further strengthened by his statement of 

 the occasional occurrence of a prominent boss on the outer surface of 

 his Rhizodus-sGsles. That such a boss is invariably present on the 

 under surface is well known ; but I have never seen a scale in this 

 or any other ganoid on which the ornamentation has such a structure 

 for its centre. Abrasion of the under surfaces of thick scales in which 

 the boss remains intact while the deeper structure of the scale is ex- 

 posed, would give somewhat of the appearance described, and is of 

 not unfrequent occurrence. But while the presence of Rhizodus, 

 Owen, in the German coal-strata must be held as unproved, the in- 

 teresting fact still remains that the genus Rhizodopsis is unquestion- 

 ably common to the English and German coal-fields : on the identity 

 of the species it is impossible to form an opinion. 



