508 MR. R. II. "WllITEIIOUSE OX THE [-^Pl- ^•' 



may be at first considered a urostyle, indicated liy a rid^e 

 along the doisal side of the hypural, has a conijianion lidge 

 directed ventralvvards ; the loss of the urostyle or its fusion 

 with a hypural bone is an advanced character. A lai-ge hypural 

 Ijoaring four derniotxichia is attached to the penultimate vertebra, 

 while that of the antepenultimate supports only one. An 

 interesting feature in this caudal fin is the ejjaxial fin-r;iy 

 support; this consists of a single epural bono broadly e.\p;inded 

 distally, and attached to the penultimate vertebra.. Such an 

 I'xpanded epui'al is quite unusual in caudal fins, and 1 know of no 

 other instance. 



General remarks on the Jlalacopter)/(/ii. 



The examples chosen illustrate how varied the caudal fin 

 structure niny be, even in a single sub-order of Teleosts. Two 

 extreme types are met with : the Clupeoids, by the extension of 

 notochord beyond the bony axis, by the vestiges of once functional 

 arches in the fonn of opisthural cai-tilages, and by the large number 

 of hypural bones to the last veitebral segment, provide an 

 example of a lowly specialized form of caudal fin ; while Sotopierus, 

 bv the loss of the urostyle and the reduction of the nund)er of 

 bypurals to the last vertebral segment to a single bone, is an 

 example of a highly specialized form. 



S T A R I O P II Y S I. 



Clarias sp. ? (Siluridce.) 



Only a single specimen of this fish has been procurable, but the 

 caudal fin appears to be quite normal. The tail-fin is fully 

 diHerentiated although the dorsal and ventral fins rench nearly to 

 the end of the bod}-. Internally the skeleton is remarkable in 

 ])resenting a very long urostyle, longer than that of any fish 

 examined, extending posteriorly as far as the distal ends of the 

 bypurals ; it does not apjjcar to be actually fused to a hypural. 

 The hypaxial part of the fin is supported by five hypural bones, 

 four t)f whicli are associated with the last centrum and urostyle, 

 and also by one venti-al caudal radial occupying the space between 

 the first and second hypural bones. The hypural of tlie penul- 

 timate vertebra has the appearance of being fornud by the 

 coalescence of a radial and a hannal aich. The epaxial iin-rays are 

 borne by one epural bone belonging to the last vertebra but one, 

 and also by a dorsal caiulal radial between the epural and the 

 urostyle. The spinal cord can be traced to the end of the uro- 

 .stvle. 1'he final centrum possesses the ve.stige of a noin-al arch, 

 tiirough which the spinal cord passes. 



Here then is a generalized type of homocercal eau<]al fin, for in 

 it ai'e representatives of all possil)le endoski'letal siqiports, viz., 

 bvpurals, ventral caudal i-adial, and dorsal caudal radial, and 

 be.-^itles, it posses.scs a well-marked urostyle. 



