ON THE ANATOMY OF FISHES. 
253 
corresponding retrogressive changes in the auditory organ, to the extent at least of 
the atrophy of tiie sinus endolymphaticus, although the ductus endolymphaticus still 
persists as a channel of communication between the two sacculi. We are also of 
opinion that the sinus is similarly wanting in other Siluridae ahnormales — as, for 
example, in Pinielodus sapo ; but, for reasons already stated, we hesitate to speak 
decisively in this as well as in some other instances about which we entertain a similar 
belief. On the other hand, there are several genera of which we entertain no doubt 
as to the presence of a sinus endolymphaticus, but in which the membranous labyrinth 
appears to be in every respect essentially similar to that of the more normal Siluridm 
{e.g., Clarias, Saccohranchuft, Pimelodus p>ulcher, &c.). It may be affirmed, there- 
fore, that, while in some instances the membranous labyrinth would seem to have 
retained its normal structure, there are certainly other cases in which degeneration of 
the air-bladder has been accompanied by a corresponding simplification of those 
portions of the internal ear which are normally brought into physiological relation 
with it through the medium of the Weberian ossicles. 
O 
In addition to the Siluridae abnormales already mentioned as forming the special 
subjects of our investigations, there are several other genera which, on the authority 
of various writers, must be referred to the same series. 
Thus, on the authority of Day (9), the Indian genera Sisor, Amhhjceps, Pseude- 
cheneis, and Exostoma are Siluroids in which the air-bladder is diminutive and 
bone-encapsuled — a conclusion which is in harmony with their hilly habitat and Loach- 
like form. 
Johannes Muller’s references {loc. cit.) to the air-bladder of the South American 
Siluroid Ageniosus militaris, and of Heterohranchus, warrant a similar conclusion 
with regard to these genera. 
On the authority of Reissner {loc. cit.), the following genera belonging to Gunther’s 
group Hypostomatina, viz., Loricaria, Plecostomus, Calliclithys, and Acanthicus, must 
also be regarded as S. abnormales, and, as the genera Arges and Brontes are but 
the naked allies of these cuirassed LojTcaroid forms, it is extremely probable that they 
must also be added. 
According to Sagemehl {loc. cit.) the Hypostomatous genus Trichomyctcrus also 
possesses a diminutive air-bladder encapsuled by bone, and as Bremophiliis is a near 
ally it may be conjectured that its air-bladder is in a similar condition. 
Finally, it has been proved by Ramsay Wright (44) that in Ilypophthahnus the 
air-bladder is degenerate and bone-encapsuled. 
It follows, therefore, that in addition to the thirteen genera which we have been 
able to revise, there must also be Included twelve others, and with some probability 
the three genera, Arges, Brontes, and Ereinophilns. So that of the 1 10 genera 
enumerated in the British Museum catalogue of the family, certainly twenty-five, and 
probably twenty-eight, are referable to the S. abnormales. In these totals we have 
