FINS OF GAKOTDS AND TELEOSTS. 587 



more primitive anal fin cannot readily be compared wifcli those 

 of other G-anoids. The dorsal and ventral segments of each 

 element may correspond to the proximal and mesial segments of 

 other Ganoids, the distal segment having been suppressed, but it is 

 by no means clear that this is the correct interpretation. I am 

 inclined to think that the counterpart of this type of fin-support 

 must be looked for in older and more primitive forms. Com- 

 parison with the simple bisegmental radial elements of the dorsal 

 fins of such ancient Elasmobranchs as Cladoselache and Pleura- 

 canthus, or of such Arthrodira as Coccosteus, reveals a very close 

 agreement with Polypterus, and suggests that the latter has 

 retained in its anal fin a more primitive type of fin-support than 

 any living fish except, perhaps, the Marsipobranchs. Further 

 indications of the primitive character of the anal fin of Polypterus 

 are to be found in the absence of the characteristic articulation 

 between contiguous radial elements which is so marked a feature 

 in Amia and Lepidosteus, and in the fact that the dermal fin-rays 

 are twice as numerous as their supporting elements. 



The radial elements of the dorsal fin present a striking contrast 

 to those of the ana] fin. That their simple unsegmented condition 

 is not due to the retention of a primitive character, but, on the 

 'contrary, is the result of specialization, is suggested by the size 

 of the structures they support. The spines of the anterior part 

 of the fin, and even the multiarticulate branched rays of the 

 hinder part, are exceptionally massive, and the segmentation of 

 the supporting elements would obviously detract somewhat from 

 their value as skeletal supports for the former. Hence, whatever 

 may have been the primitive condition of the fin-supports, and 

 the probability is that they resembled those of the anal fin, it 

 seems legitimate to infer that the reduction of each element to a 

 single segment is correlated with their function as supports for 

 exceptionally large dermal fin-rays. A precisely similar modifi- 

 cation and reduction is frequently associated with the develop- 

 ment of unusually large spines in many Teleosts *. But if this 

 explanation be correct, it might reasonably be anticipated that 

 fossil Crossopterygidae with soft fin-rays would throw some light 

 on the primitive character of the fin-supports in this group ; but 

 unfortunately the evidence available from this source, although 

 not opposed to the suggestion, is by no means conclusive. In 



* See also Aulostoma chinense, where a modification very similar to that 

 referred to in Polypterus has taken place in the anterior dorsal fin. 



