ON THE ANATOMY OF FISHES. 
161 
rim has attached to it from before backwards, first, the dorsal edges of the fibres 
derived from the corresponding antero-lateral wall of the anterior chamber and 
traversing one of the thin oval areas, and, secondly, the fibres primarily derived from 
the remaining portion of the lateral wall, as well as those converging from the 
outer wall of the lateral compartment of the same side. The thin outer rim of 
the crescentic process does not receive the insertion of any of these fibres, as the 
latter simply cross its ventral surface on their way to the thickened inner rim, 
which is consequently the only portion of the tripus that can be seen imbedded in 
the dorsal wall of the bladder after the removal of the tunica interna (fig. 60, 
tr.c.). The unusual length of the crescentic process seems to be associated with 
the relatively large size of the anterior chamber, and also with the fact that 
the process not only receives the insertion of the fibres derived from the antero- 
lateral and lateral walls of the anterior chamber, but also from the outer wall of 
the contiguous lateral compartment. The absence of any marked curvature in 
the process may also have conditioned its somewhat exceptional length. At the 
junction of its anterior and crescentic processes a thin, flexible, and highly elastic, 
triangular lamina of bone is given off from the dorsal surface of the tripus and 
at right angles to it (fig. 63, tr.ar.). The surfaces of the lamina look almost directly 
backwards and forwards respectively, and the inner margin, from its jun-ction with 
the body of the tripus to the pointed dorsal externity of the lamina, is confluent with 
the neural arch of the complex vertebra along a line which is slightly oblique from 
below upwards and forwards. This lamina undoubtedly represents the articular process 
of other Siluroids, from which, however, it differs in being directly continuous with the 
neural arch instead of articulating with the lateral surface of the centrum at the bottom 
of a deep pit-like socket. Consequently the movements of the tripus in Auclienvpterus 
will not depend on a normal articulation, but solely on the flexibility and elasticity of 
its peculiar articular process. Then again, in most Siluridse with normal bladders, it is 
the radial fibres of the tripus which cause that ossicle to move inwards towards the 
complex centrum when the lateral distension of the anterior chamber has ceased to 
separate the two, but in Auchenipterus the function of the radial fibres, which, from 
their relatively feeble development and nearly straight antero-posterior course, is 
possibly less efficient than in other Siluroids, must be powerfully supplemented by 
the elasticity of the articular process. As in all probability the tripus is to be 
regarded as the modified transverse process of the third vertebra it will at once be 
obvious that it presents a striking resemblance to the corresponding process of tlie 
fourth vertebra. In each case there is a flexible and elastic root directly continuous 
with the neural arch of its vertebra and in distal continuity with a movable plate 
of bone — the body of the tripus in the one case and the oval plate of the “ elastic- 
spring ” apparatus in the other. The fact may also be recalled that as regards the 
thinness of its root and the curious obliquity of its origin from the neural arch the 
MDCCCXCIII. — B. Y 
