80 E. T. NEWTON ON FISHES' TAILS. 



into the upper lobe of the tail ; while the lower lobe has no such 

 extension of the body into it, is often much smaller than the upper 

 part, and looks like a fin placed under the tail. This is termed a 

 heterocercal tail, and occurs in most sharks, in the sturgeon, and in 

 other forms. In the second form the upper and lower halves are 

 alike, and the fleshy termination of the body is as much below as 

 above the middle line. Such tails are called homocercal, and are 

 found in the ordinary bony fishes, such as salmon, cod, sprat, &c. 



Judging from external form alone, therefore, fishes' tails are either 

 homocercal or heterocercal ; but when we come to examine the 

 foundations on which these tails are supported, that is their skeletal 

 structures, they are found to present differences which render it 

 necessary to modify these terms. The tail of a shark, such as a 

 dog-fish, is found to have a series of the vertebra running upwards 

 into the upper lobe and occupying the middle of the fleshy part, 

 while by far the larger portion of the tail-fin is placed below the 

 end of the vertebral column. The internal structure, therefore, in 

 this case agrees with the external, and the term heterocercal is 

 strictly applicable to it. The Lepidosteus, or bony pike, of N. 

 America, has the tail more nearly equal, but still the fleshy 

 portion of the body is seen to be directed upwards, and the vertebral 

 column is found to be directed upwards also (PI. Ill, fig. 4.) So that 

 this tail is really heterocercal, though apparently much less so than 

 the shark's. 



If, now, we take one of the ordinary homocercal fishes, we shall 

 find a very different structure. The stickleback has an externally 

 homocercal tail, the fleshy body terminating in an equally rounded 

 end, on which the fin rays are symmetrically disposed. Internally 

 the vertebral column seems to end in the middle of the tail, fig. 2, 

 and to have attached to it two broad triangular plates, on which the 

 fin-rays are so disposed as to form a tail apparently as much above 

 as below the end of the spinal column. But this appearance is 

 delusive. It had long been known that in the salmon embryo the 

 tail became turned upwards as in the heterocercal fishes, but it was 

 supposed to become equal in the adult. Prof. Huxley, in 1859 (loc. 

 cit.), worked out the development of the stickleback's tail, and has 

 shown us that the apparent homocercality of the adult is a secondary 

 development, and that the tail is really heterocercal. At an early 

 stage the end of the notochord becomes strongly bent upwards, and 

 below it two triangular j^lates appear, upon which the fin rays are 



