E. T. NEWTON ON FISHES* TAILS. 85 



each of these hypural plates, excepting No. 2, is broad and hollowed, 

 but apparently not for the passage of vessels. Between the end of 

 the notochord and the neural spine of the last imperfect vertebra 

 there are three bones (ep.), which appear like neural spines, but 

 are not connected with any of the surrounding bones ; they cor- 

 respond with those in the Gasterosteus tail, called ejriuralsby Prof. 

 Huxley. 



The tail of the adult sprat differs from that above described — 

 first, in the enlargement of the splints and protective processes of 

 the fifth vertebra, which now completely cover the notochord at its 

 sides ; secondly, in the covering up or obliteration of the bony ver- 

 tebral segments behind the fifth vertebra ; and thirdly, in the 

 hypural plate No. 2 becoming connected by bony tissue with the fifth 

 vertebra. 



A comparison of the sprat's tail with those of the stickleback 

 and flounder shows that it differs from them in three most im- 

 portant particulars, namely, in the first place the notochord is 

 not ossified or enclosed in a long urostyle, but is only protected 

 at the sides by splint-like bones, beyond which it projects for a 

 considerable distance ; secondly, the upturned portion of the 

 notochord is segmented in an early stage ; and thirdly, in place of 

 the two hypural jjlates, the sprat has seven plates which bear the 

 tail fin-rays. 



The tail of the salmon described by Kolliker* and also by 

 Bruchf agrees essentially with that of the sprat in all the three 

 points just mentioned. The greatest divergence from the sprat 

 form of tail which I have yet seen among the Teleostean fishes is 

 that of the cod and that of the eel. 



In the specimen of an adult cod-fish tail which is now before me, 

 there are about 50 fin rays, and of these about 23 are placed above 

 the end of the notochord, and 27 below, so that the displacement 

 of the end of the notochord is only upwards for the space of 

 two fin rays, and at first the tail appears to be diphycercal ; 

 besides this, according to my reckoning, there is only one hypural 

 plate. Prof. Alex. Agassiz speaks of two hypurals in the embryo 

 cod ; but this difference is not one of facts, but of interpretation. The 

 hsemal arches become gradually enlarged, and assume the form of 



* " Ueber das Ende der Wirbelsaiile der Ganoiden," &c, Leipzig, 1860. 

 t " Vergleichende Osteologie des Rlieinlachses " (Salino salar, L.). Mainz. 

 1861. 



Journ. Q. M. C, Series II., No. 2. H 



