CYCLOPTYCHIUS CARBONARIUS. . 123 



;iiul " suiall " ; in ordinary s[)ccimens the ]arge ones measure about -/f inch in lengtli, 

 and the others, very niinule, are clearly seen in one specimen to occupy a more external 

 position on the edge of the jaw. 



The bones of the shoulder-girdle do not call for special remark. 



The pectoral and ventral fins are very small ; the latter are placed nearly midway 

 between the pectorals and the anal. The dorsal is placed very far back, being nearly 

 opposite the anal, Avhich it resembles both in size and in shape ; both are triangidar- 

 aciiminate, with the posterior margin rather concave. To count the fin-rays in such 

 delicate little fossil fishes is rather a difficult and uncertain matter, but I woidd 

 estimate the number in the dorsal and anal fins at al)out thirty for each, of wliicli the 

 posterior ten are vei'y fine. They do not bifurcate tdl near their terminations ; their 

 transverse articulations are rather distant, especially near the attachment of the fin ; 

 minute fulcra, rather thinly set, may be seen along the anterior margins. The caudal 

 is deeply bifurcated and inequilobate ; the fulcral scales along the anterior margin of the 

 lower lobe are more distinct and closely set than those of the other fins. About eighteen 

 or nineteen rays go to form the lower lobe ; the rest become exceedingly fine and delicate 

 as we proceed towards the tip of the upper lobe. 



As regards the scales, Prof. J. Young referred to them as being " cycloidal," but 

 remarked further that " the obliquity of their axis, from the production of their postero- 

 inferior angle, gives them in situ a rhomboidal appearance." Mr. Barkas also stated that 

 they were "cycloidal" and that they "more closely resemble those of Ccelacanthus than 

 do the scales of any other Coal Measure fishes." They are, however, in reality 

 rhomboidal, and conform in every essential respect to the type of scales seen in 

 Palffoniscus and its allies. Their principal peculiarities are their thinness, the rounding 

 of their postero-inferior angles, and their upper margins being, perhaps, rather more 

 than usually overlapped. The form of a flank-scale (PI. XXVII, fig. 5) is rhomboidal, 

 with the antero-superior angle acute and pointed, the postero-inferior rounded ; 

 from the middle of the superior margin a pointed peg projects n[)war(ls for articulation 

 with the under surface of the scale above. The ornament of the free surface consists of 

 fine, raised, thread-like ridges running in the main parallel with the anterior, inferior, 

 and posterior margins, of w^hich arrangement the drawing fig. 4 will give a better idea 

 than any further description. One ridge running along the upper part of the posterior 

 margin is generally more or less zigzagged so as to give almost a crenulated appearance 

 to that margin. The scales of the flank do not alter much in size or shape till we come 

 to the small and acutely lozenge-shaped ones which clothe the sides of the caudal body- 

 prolongation, the upper margin of which is bordered by the usual median series of V- 

 shaped scales. The scales of the back and belly are smaller than those of the sides ; 

 azygous, somewhat heart-shaped scales are often seen in the mesial dorsal line where the 

 oblique rows of each side meet, but I do not see any specially large ones in front of the 

 dorsal fin as in Palceoniscus, etc. The lateral line is very distinctly traceable from the 

 18 



