FINS OP GANOIDS AND TELEOSTS. 563 



the anterior margin o£ the distal end of the first proximal seg- 

 ment has two lateral pits for the first spine, which, therefore has a 

 " peg-and-socket " articulation with that segment. Behind these 

 lateral pits there is a bony loop formed anteriorly by a process 

 of the same segment, and behind by the anterior limb of the 

 distal segment and furnishing a chain-link articulation for 

 the second spine. Posteriorly to this a second ring is formed, 

 partly by the hook-like posterior limb of the same distal segment 

 and completed by the ossified interossicular ligament which 

 extends from the extremity of the hook backwards to the distal 

 extremity of the second proximal segment this bony loop has a 

 chain-link connexion with the third and largest of the spinose rays. 

 Finally, the hooked distal segment of the second element in 

 conjunction with the mesial segment of the third forms a third 

 bony ring for a similar articulation with the fourth and last 

 spinose ray. A comparison of these spines and their relations 

 to their fin-supports renders it clear that the third spine is that 

 rightly belonging to the first radial element, and that the 

 first and second have lost their normal fin-supports and acquired 

 a secondary connexion with the first persistent, ray-bearing 

 element. 



All the soft rays are basally cleft and simply clip the distal 

 segments of their supporting radial elements. The last element, 

 however, supports two feebly developed rays. 



Beryx decadactylus, Cuv. & Val. 



In a figure of the skeleton of this species by Giinther (9 a, 

 pi. vi.) more or fewer of the radial elements in the soft-rayed 

 portions of the dorsal and anal fins are represented as triseg- 

 mental. The sutures between the proximal and mesial segments 

 are somewhat indistinct in the dorsal fin, but are quite obvious 

 in the anal fin. This is the only instance of which I am aware 

 in which the trisegmental character of the radial elements of a 

 Teleost has been previously recognized. It must be pointed 

 out, however, that Griinther gives no description of the radial 

 elements, nor does he in any way refer in the text to their 

 segmentation. 



Three vestigial elements, consisting of proximal segments only, 

 are figured in front of the first ray-bearing element of the 

 dorsal fin. 



45* 



