[23] THE EVOLUTION OF THE FINS OF FISHES. 1003 



elements have been given more room for development in an antero- 

 posterior direction distally. In consequence of this they have been fre- 

 quently much widened at the distal extremities. This expansion of the 

 hypural cartilages has then influenced the subsequent development of 

 the membrane in which ossification occurs and also extended or ex- 

 jjanded its development. 



In Amiurus there is a small somewhat bent cartilaginous nodule found 

 just at the apex of the urostyle, Fig. 1, PI. IV, op, or at the extreme tip 

 of the chorda in a still younger stage. This nodule is evidently an 

 almost suppressed member of the once more extended series of haemal 

 arches. It is separated by a considerable interval from the last hypural 

 piece in the young fifteen days old, and belongs to a system of arches 

 which were doubtless well developed farther back in time in the ances- 

 try of this form. 



The position of this nodule in relation to the hypural pieces and to 

 the end of the chorda and the extremity of the spinal cord, which here 

 extends backward beyond the end of the notochord, would seem to prove 

 (hat there probably was a time during the phyletic history of the Cat- 

 fishes when the tail was longer and more nearly approached the most 

 primitive diphycercal condition. The extension backward of the spinal 

 cord beyond the end of the urostyle is also strong evidence in favor of 

 the probability of the existence of a diphycercal condition which pre- 

 ceded the present one. 



The sigmoid flexure of the end of the spinal cord ins, in Fig. 1, PI. 

 IV, is also evidence in favor of the theory of the upbending of the 

 chorda which was defended above. The suprachordal portion, over the 

 urostyle, has been bent up, but the part extending behind the end of 

 the urostyle has not been bent upward at tlie same angle, but is very 

 much less inclined to the plane of the axis of the body than the portion 

 immediately in front of it. So it appears that the upturning of the 

 chorda seems to affect the upbending of the spinal cord to the same 

 extent as itself only as far as the two structures coincide in their rela- 

 tive positions of parallelism with each other. 



The exserted end of the spinal cord in Aoniuf-its rests veutrally upon 

 the opisthural piece op ; it, like the end of the cord, has not been so ex- 

 tensively pushed upward distally as the more anterior, ultimate, hy- 

 pural piece, the axis of which has been shifted through at least ninety 

 degrees from the position it occupied in the old diphycercal type in 

 which it first appeared. 



V. — Development of the jiedian and paired fins and the 



EFFECTS OF CONCRESCENCE. 



The Structure of the tail and of the median fiuvsof fishes has been de- 

 scribed by a number of eminent authorities, but it is only within a few 

 years that we have been furnished with the kind of embryological 

 knowledge which has enabled investigators to reach any general and 



