1002 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [22] 



several primitive myotomes fused together, the septa between the lat- 

 ter having supplied the materials for the arches in excess of those de- 

 manded by the number of existing myotomes. But investigation again 

 shows that such a view is not tenable, because the number of myotomes 

 in the larva of Siphostoma are not in excess of those in the adult, so as 

 to justify the hypothesis just outlined ; so that it is evident, in this case 

 at least, that the number of homonomous appendages of a segment may 

 form a series which may be a multiple of the homonomous segment. 

 Such results throw considerable doubt upon the legitimacy of the in- 

 ference that the primitiv'e cartilaginous rays of the paired fins of Elas- 

 mobranchs are each homonomous with a single segment. 



The hypural pieces are really composed in many if not in most hetero- 

 cercal forms of several median appendicular pieces fused together, and 

 are usually not a little expanded at their distal ends, so that one of 

 these processes frequently gives a basis of support to a number of cau- 

 dal rays. In Oasterosteus and Scomber, for example, the fusion and ex- 

 pansion of the hypural ])ieces have proceeded so far that they are pres- 

 ent merely as two fan shaped pieces in the former, while in the latter 

 the last vertebra is expanded posteriorly so as to present but one proc- 

 ess which can be considered as the homologue of the series of hypural 

 bones of other forms. In still other forms, such as LopMus, the process 

 of specialization of the caudal end of the axial column proceeds so far 

 as to involve vertebral segments in advance of those supporting the 

 caudal rays. In this way it results that a number (about three) of the 

 caudal vertebrae in this genus become fused together by synostosis. 



There is a tendency in heterocercal forms for the neural arches to be- 

 come more or less suppressed or aborted, apparently for the reason that 

 the upturning of the posterior end of the chorda, when heterocercy is 

 in process of development, encroaches upon the i)osition which the pos- 

 terior neural or epural arches originally occupied in the unmodified 

 iliphycercal type of tail. This pushing upward of the end of the 

 chorda within the caudal fold seems therefore to really occur, and to 

 play an important part in the evolution of the specialized caudal of 

 Teleosts. In the Eel there are three hypural processes, which have no 

 epural homonyms. Even when the epural pieces are developed as far 

 as the cartilaginous condition, as the serial homologues of those in ad- 

 vance of them, they often have their direct connection with the chorda 

 or urochord more or less sundered, owing to the degenerative influence 

 of the upbendiug of the end of the latter, as may be seen in the course 

 of the development of mau}^ forms. Sometimes the epural arches, or 

 antitypes of the hypurals, are completely suppressed, as in Amiiinis. 



The tendency toward suppression of the epural pieces and that of 

 the hypural pieces to become distally widened are correlative, and are 

 evidently to be ascribed to the same causes. While the epural elements 

 are being suppressed by the upturning urochord. the dev^elopment of 

 their cartilaginous beginnings even being interfered with, the hypural 



