166 GANOID FISHES OF THE CARBONIFEROUS FORMATION 



higher than broad ; ganoid externally, though I see no very distinct sculpture, while 

 internally they show traces of the well-known "peg and socket" articulation. Then, in 

 the middle line of the hack just in front of the dorsal fin, is a row of scales, oval in 

 shape, slightly imbricating in arrangement, and distinctly sculptured with ridges, which 

 are approximately concentric in disposition. The next scales we meet with are those 

 clothing the prolongation of the body-axis along the upper lobe of the caudal fin, 

 which are exactly as in any other Palceoniscoid fish, and as solid and well-preserved as in 

 any other "ganoid " of similar size from Carboniferous beds. Along the upper margin 

 of this part we find the usual median row of pointed imbricating V-scales, and simul- 

 taneously with this median row there commences, on each side, a band of lateral ones 

 covering the right and left aspect of the prolonged body-axis, these being minute, acutely 

 lozenge-shaped, and, like the median scales, delicately sculptured with ridges and 

 grooves; this band of lateral scales does not, however, extend to the origin of the 

 caudal fin-rays until the base of the lower lobe is passed. A few imbricating, similarly 



s. cl. p.i. 



Fig. 9. — Re^oration of Lhe skeleton of Phanerosleon miralile, Traquiiir. Sliglitlv t-nlinged. b.»., body-scales; 

 b>\, branchiostegal rais; i.cl., iufraclavicle ; inn., mandible ; mx,, maxilla; op., operculum; or., orbit; /'.<■/., 

 postclavicle ; p.i., {ost-temporal ; pop., preoperculuin ; s.cl., supraelaviclc ; s.op., subopcreuhun. 



pointed and sculptured median scales may also be seen just in front of the commencement 

 of the lower lobe of the caudal fin. 



The almost complete absence of body-scales reveals the internal skeleton in a manner 

 unusually distinct for fishes of this family. There is no trace of vertebral centra, the 

 position of the persistent notochord being indicated by an empty space. This is 

 succeeded on the neural aspect of the anterior part of the body by a series of neura- 

 pophyses forming the neural arches, each of which is surmounted by a slender 

 backwardly-inclined neural spine. Posteriorly the neural arches and spines are united 

 into one piece, bifurcated proximally, corresponding haemal pieces being seen on the 

 opposite aspect. Behind the anal fin, where the haemal spines are seen supporting 

 the lower lobe of the caudal, they are laterally flattened and dilated at their extremities ; 

 further on they are concealed from view by the scales of the caudal body-prolongation. 

 There are no ribs in the abdominal region, nor have I seen with certainty any haemal 

 pieces or htemapophyses in this region, though I once supposed I had done so. 



The contour of the pectoral fin is not seen in any specimen, as its rays are always 



