Prof. E. I). Cope on the Systematic Belations of Fishes. 165 



is seen in many members of tliis tribe, wliile in others it re- 

 mains in its primitive condition. Among Pliysocljsti it is 

 nearly always complete, though in a few (Trichinridas &c.) it 

 remains larval. In the first development of the vertebral 

 column in fishes it forms a straight axis. The fin is repre- 

 sented by a fold of the integument which extends equally 

 round its extremity. In this membrane the rays are deve- 

 loped, and in many fishes they remain thus equally distributed. 

 In this case the caudal vertebrae remain in a straight line to the 

 extremity, and we have a termination such as is seen in Lepido- 

 siren and the eels. This form of tail may be called the isocercal. 



If, now, the radii, basal or distal, acquire a greater develop- 

 ment on the lower side of the column, those on the upjjer side 

 remaining rudimental, it will be necessary that such enlarged 

 portion should strike the water in the plane transverse to the 

 longitudinal axis of the body, in order that the Aveight of the 

 body be propelled with the least expenditure of force. This 

 will necessarily cause the distal vertebra, or end of the chorda 

 dorsalis, to be turned upward, so that the inferior rays of the 

 fin shall be brought as* near to the vertical line of the superior 

 as possible. This is the type of tail known as the heterocercal^ 

 as called by Agassiz. 



We find among the Physoclysti that the lower rays of the 

 fin are more and more strengthened, and the hamal spines 

 which support them are more and more enlarged ; consequently 

 the end of the column is more curved upwards, as seen in 

 Amia. The superior rays and neural spines are also strength- 

 ened, and the inferior so extended upwards as to pass round 

 the extremity of the colmnn and come into contact with them. 

 And now the vertebral centra are successively atrophied from 

 the extremity. Counting from the extremity to the bases of 

 the first supports of the outer rays of the caudal fin above and 

 beloAV, we find that ten vertebrae remain in the tail of Noto- 

 2')terus. In the Hyodontidaj, Albulidaj, Elopidas, Alepocepha- 

 lidee, and Salmonid^e there are but two left, while one only 

 appears in the Osteoglossida?, Aulopidje, Lutodiridse, Butyri- 

 nidse, Coregonidfe, Clupeidas, and Chirocentrida;. In most other 

 families, especially of Physoclysti, the last one has disappeared, 

 and the numerous lieemal arches are arranged like radii diverging 

 upwards and dowuAvards from the last caudal vertebra. In the 

 highest groups, as Pharyngognathi &c., they become coossi- 

 fied, and the tail has completed specialization. This is the 

 type called homocercal or diphyocercal by later writers. 



These types are thus plainly stages in the development of 

 this member, the first and second being simply arrests of de- 

 velopment of the last. Thus the young salmon commences 



