208 THE SWIM-BLADDER OF FI8HE8 



hepatic portal system. The oxygen which may fill, or nearly 

 fill the swim-bladder cannot be regarded as destined for the 

 purpose of charging with oxygen, or arterialising, the blood, 

 for that is already arterial, and has just been charged with oxygen 

 in the gills. Giinther states (No. 9, p. 142) that the oxj-gen is 

 really sjcreted by the surface of the swim-bladder, and there is 

 every reason to hold that view ; but such a process is therefore 

 secretory, not respiratory in the accepted sense. Again, it must 

 be noted that in such a fish as the haddock (Plate 20, fig. 1. sb) 

 with a swim-bladder having extremely vascular walls, and a 

 rich blood-supply, the organ is closed and the contained gas 

 cannot pass out by a duct, whereas in the carp (Plate 20, fig. 4) 

 with vascular fanlike tufts, in the pike (Estox) with small compact 

 red bodies in which arterial and venous capillaries anastomose 

 and in the sturgeon, salmon, herring, &c., though unprovided 

 with these retia mirabilla, the swim-bladder is not closed, but 

 has in most cases a very capacious opening through the dorsal 

 wall of the oesophagus. The fact, on which Professor Rolleston 

 laid stress (No. 19, p. 424), that in all species, when the branchias 

 are in full activity, the swim-bladder is supplied with the purest 

 arterial blood, lends little support to the supposition that it, in 

 any way, subserves respiration. Of course in Dipnoans, like 

 Lepidosiren and Ceratoclits, the pulmonary function is undoubted ; 

 but the fact that the duct is ventral and not dorsal as in the 

 Teleostei is of the highest importance. What is there to make 

 improbable thj sugg 'stion that these so-called lungs are new 

 structures correlated to the change in the circulation and the 

 more highly differentiated condition of the heart. Wilder, it 

 is true, states that in Amia and Lepidosteits he has found cases 

 intermediate between the dorsal and ventral connection of the 

 swim-bladder ; but even if this variation of the duct be regarded 

 as not wholly exceptional and abnormal, it cannot account for 

 a dorsall3''-placed organ like the swim-bladder becoming a 

 ventrally-placed lung, nor that a pulmonary artery, really a 

 branch of the inferior aortic arch, carrying venous blood, should 



