302 
PROFESSORS T. W. BRIDGE AND A. C. HADDON 
Sagemehl {loc. cit.) has suggested, the function of these areas is to enable the air- 
bladder more promptly to react to pressure oscillations, or, as we have surmised, more 
readily to allow variations in the volume of the contained gases to modify the bulk 
and specific gravity of the Fish. 
The comparative freedom of the anterior vertebrae from anchylosis in the Cyprinidm 
is obviously related to the fact that the air-bladder, including the anterior chamber, 
lies free in the abdominal cavity, and is scarcely, if at all, invested by skeletal 
elements ; hence, as the bladder is less likely to be aflPected by any compression which 
may result from the flexure of the vertebral column during vigorous locomotor move- 
ments, anchylosis is unnecessary, or at all events does not take place to anything like 
the same extent as in the Siluridse. 
Certain distinctive structural variations in the Siluridse are probably dependent 
upon other modifications of a more direct physiological value, and in any comparison 
with the Siluridse it is interesting to notice the necessary sequence of particular 
anatomical differences between the two when an initial modification is once started in 
one of them. In the Siluridm, the restriction of volumetric variations in the anterior 
chamber to its lateral expansion or contraction is in part brought about by the invest- 
ment of its dorsal surface by the complex and fifth centra and the expanded and 
otherwise modified transverse processes of the fourth and fifth vertebrae, which not 
only prevents the expansion of the chamber dorsally, and to some extent anteriorly, 
but indu-ectly aids in preventing expansion in other directions by affording the 
necessary surface for the direct attachment of the walls of the chamber to the skeleton 
at certam definite points. But to prevent the flexure of the trunk vertebrae in 
oi’dinary locomotion from exerting any influence on an air-bladder so closely related 
to the skeleton, and secondarily on the Weberian ossicles, a more or less extensive 
anchylosis of the anterior vertebrae takes place, as well as a more rigid and intimate 
articulation with the skull than is usually the case in other Ostariophyseae. The 
necessity for a more rigid connection between the skull and the vertebral column has 
probably conditioned further modifications in the anterior vertebrae, and may well 
have been one of the causes which have hastened the further atrophy of those rudi- 
mentary vertebral centra that intervene between the basioccipital and the fourth 
vertebra, and have been partially dismembered in the formation of the different 
Weberian ossicles from certain of their constituent elements. The shortening of this 
section of the vertebral column necessarily involves the tripus taking a position more 
anterior than that which it occupies in the Cyprlnidae and exactly opposite the 
scaphium, with the final result that the intercalarium becomes useless as a sesamoid 
ossicle, and degenerates to the insignificant and variably developed vestige by which 
alone it is represented in the great majority of the Siluridce. 
On the other hand, it is not easy to attach any precise physiological significance to 
the fact that the saccus paravertebral is in the Siluridm has entirely lost that com- 
munication with the cranial cavity which it possesses in the Cyprinida;, nor to certain 
