326 MEMOIRS OF THE CARNEGIE MUSEUM. 



3. Urosphen attenuata sp. nov. 

 1911. Urosphen dubia Eastman, Mem. Carnegie Museum, Vol. IV, No. 7, p. 361, 

 PI. XCVI, fig. 2. 



Type. — Nearly complete fish; Carnegie Museum (Cat. No. 4499). 



A small species attaining a length of about 20 cm. having about the same pro- 

 portions as the type of U. dubia, but more vertically compressed, and differing 

 in the conformation of the caudal fin. This is intermediate in character between 

 the caudal fin of U. dubia, which is cuneiform with all of the rays gradually in- 

 creasing in elongation above and below axially, or medianwards, and that of Fistu- 

 lariids in which two axial rays are excessively elongated. The neural and haemal 

 spines of the last vertebral centrum are expanded into fan-shaped laminse medially 



Fig. 2. Tail of Urosphen attenuata Eastman. X 1- C. M. Cat. Foss. Fishes, No. 4499. 



in contact and together forming a urostyle,' which supports in all six slender, greatly 

 elongated and closely apposed caudal fin rays, half the number being epiaxial and 

 half hypaxial (see Fig. 2). 



In addition, a series of ten short rays, increasing gradually in length from the 

 anteriormost onwards until about the fifth, after which all are of uniform length, 

 arise from the dorsal and ventral margins at the posterior extremity of the body, 

 being supported by the neural and haemal spines of the last five vertebrae. The 

 dorsal and anal fins are remote, similar, and opposite, the former with eighteen 

 rays, and the latter with twenty. The trunk and head are vertically much com- 

 pressed, and the small, terminal mouth is provided with minute conical teeth. In 

 the type-specimen the undigested skeletal remains of a small teleost are seen in the 

 forward part of the intestinal tract. The holotype has already been figured in 

 the Memoirs (1911) under the name of U. dubia. 



' Reference may be made here to the writings of R. H. Whitehouse (Proc. Roy. Soc, Vol. 72 B, 1909, 

 p. 139), and C. T. Regan (Ann. Mag. Nat. Hist., (8) Vol. V, 1910, p. 5.31) on the caudal fin of Clupeoids, 

 and the Tcleostean urostyle. 



The last-named author remarks (l. c, p. 533): 



"A comparative study leaves no room for doubt that in many cases the urostyle is merely the 

 result of anchylosis of the uroneurals and that centra take little or no part in its formation, and I do not 

 think that there are any fishes in which a urostyle has been formed simply by anchylosis of posterior 

 centra; but that is a matter which requires further investigation." 



