1910.] CAT)n\L FIX OF THE TELE05rOMI. 611 



whicli . Occupies a position exactly in line with tlio in'ostvle ; 

 •A. flifficulty conies about, therefore, in detei-niining whethi'i' this 

 element is hvpaxial oi- epaxiiil. In shape it closely resembles th(! 

 «lorsal caudal radials, but the spinal cord runs alon<^ its dorsal 

 edge ; it may therefore be considei-ed a dorsal caudal radial which 

 lias moved ventiul to the spinal cord ; but its present position 

 below the spinal cord makes it possible also that it is a hypural 

 bone, or perhaps a ventral caudal radial. For the present, 

 however, I have decided to call it a hypural in order to simplify 

 description, but the reader may consign it to any of the three 

 suggested classes he ma,y consider desirable. 



Serranus cabrilla. (Perciformes, Serranidse.) (Plate XLIX. 

 fig. 19.) 



The vertebral column ends in a well-developeil and fi-ee 

 urostyle (?«?•.), which is about the length of an ordinary vertebra. 

 Six hypural bones (%.) of varying size are associated with the 

 last vertebral segment ; the most anterior of these possesses a 

 long spur-like process, running parallel with the urostyle, and 

 inidoubtedly functioning .as an attachment for muscles ; the last 

 hypui-al is in direct line with the axis of the urostyle. Between 

 the distal ends of the hypurals of the second and third last 

 vertebrae is a small cartilage (c.) which shares the sup])oit of the 

 smaller anterior ventral caudal fin-rays. Dorsal to the urostyle 

 and extending beyond it, is a slender splint-like lione (n.a.') 

 protecting the delicate end of the spinal cord ; this probably 

 lepresents one or moi-e neviral arches of vertebra* now absorbetl 

 into the urostyle. Independent of this bone, the la.st vertebra 

 (l.v.) possesses a neural arch of its own ; the neural arch of the 

 penultimate vei-tebra (p.v.) is much reduced, and the ante- 

 penultimate vertebra is provided with an cpural bone (ep.). 

 Three dorsal caudal radials {d.c.7\) support the majority of the 

 epaxial fin-rays. 



This caudal fin, therefore, is a fairly typical form of a com- 

 paratively lowly specialized homocercal type. 



Cepola rubescexs. (Perciformes, Oepolidce.) 



This fish possesses a long tapering form of tail. The lu'ostyle 

 is prominent though slender ; four hypural bones are attached to 

 the last vertebral segment, the last oveidapping the ui-ostyle 

 slightly. All the dermotrichia of the caudal fin are hvpaxial, 

 though epaxially there is a bone, not actually supporting a I'ay, 

 which might be called a radial. Extei'ually the median fin system 

 appears to be a continuous one, but closely examined, theie is to 

 be found a very distinct and separate caudal. AH the skeletal 

 structures in the terminal end of the tail are very delicate and 

 loose, the neural and luemal arches not being actually fused with 

 the elongated centra. The tail-fin is, however, definitely homo- 

 cereal. 



