THE OSTEOLOGY OF AMIURUS CATUS. 297 



the presence of several haemal arches. Of these there are in all six, 

 the four tying immediately below the terminal filament of the noto- 

 chord being separated from the other two by a distinct interval, cor- 

 responding to the longitudinal axis of the body. The lowest (A) 

 (i.e., the one posterior to that of the last vertebral centrum) is fused 

 with the posterior inferior portion of the last vertebral centrum, and 

 bears at its base a slight lateral ridge. It expands very much towards 

 its extremity, being the broadest of all these fin-bearing arches. 

 The next four (B, C, D & E) arise from the posterior surface of the 

 last centrum, being fused with it. They are triangular in shape, 

 expanding posteriorly, and diminishing in size from below upwards. 

 The last (F) (i.e., that immediately below the notochord) is small, 

 and partly enclosed by the lateral bones enclosing the notochord. It 

 seems to arise from these structures near their base. 



We have thus six haemal arches which are well developed, and 

 specially modified for the purpose of supporting the rays of the 

 caudal fin, the centra and upper arches corresponding to them having 

 become aborted, or perhaps the centra are represented by the last 

 body, several having fused to form it. Lotz 1 has investigated the 

 structure of these vertebrae in Cyprinoids and other fishes, and in 

 the former there appears to be an arrangement very similar to that 

 of Amiurus. The specialization however does not seem to have pro- 

 gressed quite so far. In Barbus the third or second vertebra bears 

 two dorsal arches. The spinous process of the last dorsal arch is 

 similar to that of Amiurus, Lotz naming the free spinous process a 

 ' falsche Dorn,' believing it to be either a part of the true spinous 

 process or a tree tin-bearer. I prefer the former hypothesis. The 

 three lower arches, which have no distinct vertebrae, are fused with 

 the last centrum, as in Amiurus, but the upper four are independent- 

 It would appear from this that the last vertebral centrum really 

 consists of three fused centra, those of the four upper haemal arches 

 having become aborted, the fusion of these arches with the last 

 centrum in Amiurus being secondary. All these lower arches are 

 tipped with cartilage, but there are no intervening cartilaginous 

 pieces as in Barbus. 



Extending back from the posterior superior angles of the last cen- 

 trum on either side of the notochord filament are two bones (NS) 



i Lotz.— Ueber den Bau der Schwanzwirbelsaule der Salmoniden, Cyprinoiden, Percoiden. 

 und Cataphiacten. Zeit. f. wiss. Zool. XIV., 1864 



