72 WKALDEX AND PURBECK FOSSIL FISHES. 



with opercular apparatus somewhat exceeding maximum depth of trunk, which is 

 twice as great as depth of caudal pedicle and contained about five times in total 

 length of fish. External head-bones coarsely tuberculated or rugose; marginal 

 teeth long and slender. Dorsal fin arising at the end of the anterior third of the 

 back, half as long as the trunk, comprising not less than 25 rays, mostly bifurcated, 

 of which the longest do not equal the depth of the trunk at their point of insertion ; 

 pelvic fins arising slightly in advance of the middle point between the pectorals 

 and the caudal. Scales smooth, with finely serrated hinder border except towards 

 the hinder end of the caudal region where they are entirely smooth. 



Description of Specimens. — The type specimen (PI. XVI, fig. 1) is contained in 

 Purbeck stone of uncertain origin, and exhibits most of the specific characters of 

 the fish. Though not much distorted, it is considerably fractured by crushing; 

 and it is represented in a rather diagrammatic manner in the original drawing- 

 published by Agassiz. The head is probably not lengthened by distortion, but the 

 cranial roof is pushed over to the left side, so that the right squamosal and some 

 other bones are seen from below. One fragment of bone displays the coarse 

 ornament of tubercles of ganoine, which are partially fused together. Slender 

 styliform teeth, with a sharply pointed and somewhat incurved apex (fig. 1 a), 

 occur in both jaws. The left operculum, exposed from within, is about two-thirds 

 as wide as deep. Below it are the comparatively small suboperculum and inter- 

 operculum, and some remains of branchiostegal rays. Above it may be recog- 

 nised the left side of the pair of supratemporal plates, and a much larger post- 

 temporal which is in contact with an elongated supraclavicle, passing below to the 

 much-arched smooth clavicle. The axial skeleton of the trunk must have been 

 imperfectly ossified, but there are traces of vertebral centra and stout neural arches 

 just behind the occiput, and more distinct remains (though not so clear as figured 

 by Agassiz) also occur in the hinder half of the tail. The caudal vertebral rings 

 seem to have been longer than deep, with stout neural and hasmal arches ; and the 

 upturned end of the vertebral axis is distinct. The two pectoral fins are crushed 

 together, each probably consisting of about twelve rays, which are closely articu- 

 lated and bifurcating in more than their distal half. They do not exhibit fulcra. 

 The pelvic fins, also crushed together, are much smaller, and seem to bear a few 

 slender fulcra. The extended dorsal fin shows about five basal fulcra in front, and 

 its well-spaced stout rays are all articulated and bifurcated distally. The complete 

 length of its anterior rays seems to be exhibited. The anal fin is represented by a 

 mere fragment, with some of its supports. The large and strong caudal fin, with 

 slender fulcra along its upper border, must have been forked. The greater part 

 of the squamation of the left side, is undisturbed and exposed from its inner face, 

 while traces of the scales of the right side, in outer view, are seen near the ventral 

 margin. There are enlarged post-clavicular scales, at least just above the insertion 

 of the pectoral fin. None of the flank-scales seem to have been deeper than broad, 



