TAMILY, I— SILUEID^. 



439 



In the genus Arius and some allied forms tlie males appear to carry tlie ova in tlieir moutlis perhaps untd 

 the young are pi-oduced.* 



Many of these fishes are credited with causing poisonous -wounds, and we frequently find such cases 

 admitted into hospitals. They may be divided into two classes, (I) those in which the wounds are of 

 a distinctly venomous description, (2) those in which their jagged spines occasion intense inflammation often of 

 a dangerous character. 



The respiration of these fishes is effected in two ways, and it may be appropriate here to refer to the 

 amphibious fishes of India, as the Labyriiithici and Opldocephalldc'. 



Eespii-ation in fishes is carried out normally (1) by their using the air which is in solution in the water 

 to oxygenate the blood at their gills, (2) by taking in atmospheric air direct which is employed at a special 

 or^an, where it oxygenates the blood, which can be returned for use into the general circulation without its 

 going through the gills, thus constituting true amphibious fishes. f 



No doubt we observe that fishes which normally oxygenate their blood solely at their gills do rise to 

 the surface in very hot weather, when the water is foul or insufiiciently charged with air and take in 

 air by the mouth. 



Likewise we find that those which mainlj^ take in atmospheric air direct by the mouth may, to a certain 

 extent, be able to use their gills. 



If fishes having these two different modes of respiration are placed in a globe of water, across which a 

 diaphi-affm of net is inserted below the surface so as to prevent their obtaining access to the atmosphere, those 

 of the "first class which oxygenate their blood at the gills are unafl'ected, whereas those which take in 

 air direct, or amphibious fishes, die, due to blood poisoning. 



If we seek for the reason, we find that the amphibious fishes have accessory breathing organs, those of 

 the Ophiocephalidfe and Lahi/rinthici have been referred to(pp. 362, 3C9). Amongst the Siluridm we also find 

 amphibious forms : thus the Clarias has a dendritic apparatus attached to the branchias (Plate CXII, tig. 6a), and 

 the Saccohranchus has a long airj or respiratory sac passing backwards amongst the muscles of the back from 

 behind the gill cavity proper. The woodcut (after Hyrtl) explains how this is effected. The transverse 



processes of the first vertebra are dilated to enclose the air-vessel (a) in a bony capsule, from each of these 

 air-vessels there passes a duct which goes to the pharynx, this air-vessel has nothing to do with amphibious 

 respiration, it is the swim-bladder partially enclosed in bone. 



The respiratory air-sac (b) passes down amongst the muscles of the back from behind the gills. The 

 blood from the heart goes up the bulbus arteriosus (c) and divides into branches on either side, one of 

 which goes to each respiratory air sac, the anterior on the right, the posterior on the left side. The one on 

 the right (d) goes along the upper wall of the sac, whilst on the left [e) one traverses the lower wall, 

 giving off numerous lateral branches. Prom this respiratory air-sac the purified blood is returned by a 

 vessel (J) which conveys it direct to the aorta. § 



* See rcm.irks under the head of genus Arius. 



t Sec paper on this subject, Proc. Zool. Soc. 1858, p. 274. 



J This is not the air-vessel or swim-bladder seeu in the majority of Teleosteous fishes, as that also exists more or less enclosed 



in bone. 



§ See paper in Journal Linn. Soc. read January ISth, 1877. 



