ox THE ANATO]i[Y OF FISHES. 
255 
apposed in the median line, and occupy two shallow depressions, separated by a faint 
median ridge, on the cranial surface of the basioccipital. The sacculi may have a 
fibrous investment dorsally, but it is certain that they are not roofed over by hori- 
zontal ingrowths from the exoccipitals as is the case with all the other S. abnormales 
that we dissected. It is also due to the absence of such ingrowths that, with the possible 
exception of Callichthys, the walls of the cavum sinus imparis are everywhere fibrous, 
except where it rests on the dorsal surface of the basioccipital. Whether a sinus 
endolymphaticus is present or absent has not yet been determined, but its presence in 
Acanthicus suggests that it may also be present here. The Weberian ossicles are 
very similar to those of other Siluridee abnormales. The scaphium is a small 
concavo-convex ossicle without ascending or condylar processes, and lies altogether 
within the neural canal. The intercalarium is absent. The tripus has but a slightly 
curved crescentic process for insertion into the wall of the corresponding air-sac. No 
mention of claustra is made by Reissnek, nor could we detect any trace of these ossicles. 
From this brief summary it is obvious that the Loricaroid types are in many respects 
widely different from those S. abnormales that came more directly under our notice, 
and exhibit a combination of characters in common which no doubt represents a 
further and more extreme modification of the type of structure exhibited by the 
latter. From this point of view the most important of the Loricaroid features are, 
( 1 .) the absence of a distinct separable centrum to the first vertebra. In all other 
Ostariophyseee this centrum can be readily recognized, although in the Siluridse it may 
be much smaller than any of the adjacent normal centra ; but in the Loricaroid genera 
it has either undergone complete suppression, or has fused behind with the complex 
vertebra. (2.) The mode of formation of the osseous capsules for the air-bladder 
by what have hitherto been described as the dorsal and ventral laminm of the trans- 
verse processes of the fourth vertebra, and by the bones forming the posterior face of 
the skull. It is by no means improbable that this method is partly the direct result 
of the reduction in size and conci’escence of the anterior vertebrae, and the close 
relation of the modified transverse processes to the skull which is thus brought about, 
and partly to the further extension of a modification already indicated in other and 
less aberrant Siluridee. The so-called “dorsal lamina” of the modified transverse 
processes in the Loricaroid forms, is clearly the equivalent of the entire process in a 
normal Siluroid, and like the latter has its origin from the neural arch of its vertebra. 
The dowmgrowth of the lamina to buttress the front w’all of the air-bladder as in 
other abnormal Siluridse is here unnecessaiy, inasmuch as the close relation of the 
bladder to the posterior face of the skull enabled the latter to be utilized in forming 
the anterior wall of the investing bony capsule. The so-called ventral lamina is 
probably to be regarded not as a portion of the transverse process as described by 
Ramsay Wright, but rather as a further development of the peculiar outgrowth 
from the superficial ossification investing the lateral surfaces of the complex centrum, 
which, as a rudiment, is apparent in Glyj)tostcrnum and Bayarius, but is more 
