E. T. NEWTON ON FISHES' TAILS. Si 



supported. As ossification proceeds the vertebra? become segmented 

 off, but that part of the notochord which turns upwards becomes 

 gradually enclosed in a bony sheath, and in the adult is seen lying 

 along the upper edge of the uppermost triangular bone, fig. 2, nch, 

 so that nearly the whole of the tail is attached to the under side of 

 the termination of the vertebral column. The tail of the stickleback 

 therefore, although externally homo cereal, is internally extremely 

 hetero cereal. 



There is yet one other kind of fish tail to which attention must 

 be drawn — it is that which has both the external form and the in- 

 ternal structure arranged symmetrically with regard to the end of 

 the vertebral column. The Ceratodus of Australia has a tail of this 

 description. The notochord in this fish is never ossified, but pass- 

 ing directly backwards, ends in the middle of the tail, and the fin 

 rays are arranged as much above as below it. 



When it became known that some of the externally homocercal 

 tails were really extremely heterocercal, it became necessary to make 

 some distinction between those which were really homocercal and 

 those which were only apparently so. Now as the term homocercal 

 has always been applied to those forms, chiefly Teleosteans, which 

 were externally equal, but had now been shown in many instances 

 not to be so internally, it was decided to retain the name for these, 

 and to call those forms diphycercal which were truly equal above and 

 below, internally and externally, as in the Ceratodus. And fishes' 

 tails are now generally divided into three groups as follows : — 



1st, Diphycercal. Those tails in which the vertebral column or 

 notochord passes directly backwards, without turning upwards, and 

 divides the tail into equal upper and lower portions. 



2nd, Heterocercal. Those which externally are seen to have the 

 end of the body turned upwards, and to a greater or less extent 

 passing into the upper part of the tail, and internally have the 

 vertebral column or notochord passing inwards in the same manner. 



3rd, Homocercal. Those which externally appear to be as much 

 above as below the middle line of the body, but internally show the 

 end of the vertebral column turned upwards, and the larger part of 

 the tail placed below it. 



There is another matter to which I must call attention before pro- 

 ceeding to describe the sprat's tail. The development of the flounder 

 (Pleuronectes) has been worked out by Prof. Alex. Agassiz,* and 



* " Proc. Am. Acad. Sci.," vol. xiii., p. 117. 



