206 ME. r. DAT ON AMPHIBIOUS AND 



In every fish of the fresh waters of India that I have perso- 

 nally observed respiring atmospheric air direct, or of which I have 

 been able to collect indisputable evidence that they do so, we find 

 an accessory respiratory organ to the gills : — in Glarias an arbo- 

 rescent vascular one ; amongst Ophiocephalidse a somewhat simple 

 cavity above the gill-opening also covered with a vascular mem- 

 brane ; in the Labyrinthici an elaborate one of thin bony laminsB. 

 In AmpMpnous we see that " of all the arches the second alone 

 possesses laminae for the purposes of breathing ; and these consist 

 merely of a few long fibrils attached to the middle of the arch and 

 occupying but a very small extent of its surface ; the third sup- 

 ports in the place of laminae a thick and semitransparent tissue, 

 which in large individuals of the species presents a fringed or 

 denticulated appearance at its edge ; whilst the third and fourth 

 are bare, having only the membrane that fills up the space be- 

 tween the arches reflected over them. The principal organs of 

 respiration are two small bladders, which the animal has the power 

 of filling with air, immediately derived from the atmosphere. They 

 are placed behind the head, one on each side of the neck, above 

 the superior or vertebral extremities of the branchial arches, and 

 are covered over by the common integuments, presenting exter- 

 nally, when distended with air, two protuberances of a round 



shape They present, when separated from their surrounding 



attachments and inflated with air, thin, semitransparent, membra- 

 nous parietes, resembling the posterior portions of the lungs of a 



serpent Of the whole volume of blood contained in the 



branchial artery, one third passes through the gills and respiratory 

 bladder, whilst the other two thirds are conveyed directly from 

 the heart to the aorta without being exposed to the action of 

 the air." — Tayloe, I. c. 



I asked Dr. Hubrecht to be so good as to dissect a Saccoiran- 

 cJius for the purpose of tracing out its circulation. This he did, 

 remarking, however, that Hyrtl (Sitz. d. k. Akad. "Wiss. Math. 

 CI. Bd. xi. Heft 2, 1853, p. 305) had already accomplished this, 

 and that his present dissection went, as a whole, entirely to cor- 

 roborate what Hyrtl had published. From the exceedingly de- 

 veloped bulbus arteriosus (c) of the heart spring the four pairs of 

 branchial arteries. Of these the fourth pair on the left and the 

 first pair on the right* are more developed than their neighbours ; 



* This distribution appears to be an exaggerated form of how the pseudo- 

 branchise ai*e supplied with blood, 



