166 
PROFESSORS T. W. BRIDGE AND A. C. HADDON 
Its slightly thickened dorsal extremity is almost in contact with the inner margin of 
the root of the crescentic process of the tripus (tr.c.), with which it is connected by a 
few short radial fibres. The inferior limb of the post-temporal (pt.i.) is very massive, 
and articulates internally with the basioccipital (bo.), and along its dorsal edge with 
the angle which the posterior and opisthotic plates of the exoccipital make with each 
other at the postero-lateral region of tlie skull. 
The air-bladder (fig. 66) is also very similar in shape and relative size to that of 
Anchenipterus. The anterior chamber («.c.) is fairly large, and its lateral walls are in 
close and extensive contact with the superficial skin. The transverse septum {t.s.) is 
attached dorsally to the body of the fifth vertebra, and anteriorly to this to a longi- 
tudinal ridge which traverses each of the lateral surfaces of the complex centrum and 
marks the dorsal limit of the superficial ossifications, and also coincides with the inner 
margin of the corresponding triangular fibrous sheet of the dorsal wall of the anterior 
chamber. The lateral portions of the anterior wall of the bladder are extremely thin 
and translucent where the distal plates of the “ elastic spring” mechanism are applied 
to them, in fact, the latter project into the cavity of the anterior chamber, pushing in 
the thin wall of the bladder before them. At the point of closest contact of the two 
structures the tunica externa is entirely wanting, and only tlie thin inner tunica 
invests the inw'ardly projecting extremities of the two plates.^' The peripheral 
margins of the plates are somewhat firmly attached to the wall of the bladder by 
connective tissue. The posterior extremity of the bladder has a small rudiment of a 
posterior ccecum {p.c.) sub-divided internally by a backward prolongation of the 
longitudinal septum which, anteriorly, separates the two lateral compartments from 
each other. 
The Weberian ossicles are much the same as in Auclienipterus. The crescentic 
process of the tripus (fig. 65, tr.c.) is, however, shorter than the anterior process 
{tr.a.) and without a trace of any inward curvature. From the manner in which the 
fibres of the dorsal wall of the anterior chamber are connected svith it the crescentic 
process appears somewhat triangular in shape wlien viewed from the ventral surface 
after the removal of the tunica interna (fig. 66). The articular process {tr.ar.) is a 
thin flexible lamina of bone, continuous with the arch of the complex vertebra, and in 
correlation with this modification the radial fibres are so feebly developed as to be 
almost non-existent. 
Synod out is schal. 
Johannes Muller (28) has figured and described the “ elastic-spring” apparatus of 
this African Siluroid and its relations to the air-bladder {loc. cit., Plate 3, figs. 1 to 4). 
Sorensen, in his paper “ Om Lydorganer hos Fiske,” also gives an external lateral 
view of the air-bladder of Synodontis with the object of showing the relations of tlie 
* These facts suggest tliat in the mode of formation of the nodular terminal jilates of the mechanism 
Oxyduras resembles Doras, Synodontis, and Auclunipterus, 
