THE SWIMBLADDER AS A HYDROSTATIC ORGAN 77 



haemalum and eosin, the pericapillary cytoplasm has a hyaline, homogeneous appearance, which 

 contrasts with the lilac or salmon-pink colour of the other cell contents. The hyaline border varies 

 from 2 to 6/1 in thickness. 



In his plates of the gas-gland of Sternoptyx, Nusbaum-Hilarowicz (1920) shows the pericapillary 

 borders to have a striated appearance, and this has also been reported in other fishes by Vincent and 

 Barnes (1896) and by Bykowski and Nusbaum (1904). Like Fange (1953), I was unable to detect 

 this, but the sections of the giant cells of Vinciguerria did reveal that the interface between the blood- 

 system and the cytoplasm can be the site of intense activity. The borders look as though they are 

 breaking down to form small vacuoles (PI. II, fig. 3) and this process seemed also to be taking place 

 along the intercapillary canals in the cytoplasm. In sections of the gas-gland of Polyipnus laternatus 

 the pericapillary border almost had a striated appearance, but closer examination showed the striation 

 to be a series of elongated ' vacuoles ' aligned normally to the bore of the capillaries. 



These cytoplasmic inclusions cannot be gas-bubbles as they have a different optical appearance, 

 and in Vinciguerria, the hyaline blobs that appear along the capillaries look very like the vacuoles that 

 are lying free in the cytoplasm. It may also be significant that the vacuoles are generally larger near 

 the interface between the gland cells and the swimbladder cavity. 



In a series of papers, Scholander and his colleagues (Scholander and van Dam, 1954; Scholander, 

 1954, 1956; Sundnes, Enns and Scholander, 1958) have argued that the Root effect (the release of 

 oxygen from oxy-haemoglobin by a lowering of the pH of the blood) plays little part in the secretion 

 of this gas against the high oxygen pressures that exist in the swimbladders of deep-sea fishes. (By 

 using marked oxygen (O 18 ) Scholander, van Dam and Enns (1956) showed that the gas must come 

 from the water surrounding the fish (cod) and be transported as oxy-haemoglobin from the gills to 

 the swimbladder). 



In particular, if the lower limit of the blood pH is taken as 6-5, the Root effect is almost non- 

 existent in deep-sea fishes, such as the long-nosed eel {Synaphobranchus-pinnatus), the blue hake 

 (Antimora violacea) and the round-nosed ratfish (Coryphaenoides rupestris), but is well marked in 

 shallower-water species like the tautog (Tautoga onitis). (Scholander and van Dam (1954) determined 

 the oxygen dissociation of the blood at oxygen-tensions from 0-2 to 140 atmospheres and at acidities 

 down to pH 5-6.) Thus in deep-sea fishes the oxygen seems not to be liberated directly from the 

 blood, but to be actively secreted by the gas-gland cells. Furthermore, even in the shallow water 

 toadfish (Opsatius tau), Wittenburg (1958) has shown (in a very neat way) that the gas-gland cells are 

 able to transport oxygen from the blood plasma into the swimbladder. After supplying the fish with 

 carbon monoxide in (presumably) sufficient quantities to immobilize the haemoglobin, oxygen was 

 still secreted into the swimbladder. Wittenburg suggested that this active transport of gas is by way 

 of an iron-haem protein. 



Perhaps the foregoing observations of the gas-gland cells of Vinciguerria also support this idea. 

 Perhaps there are three main phases in the formation of the gas bubbles that are eventually released 

 into the swimbladder. The first consists of an intense interaction between the blood and the peri- 

 capillary cytoplasm, when vacuoles are formed. Fange (1953) has shown this part of the cytoplasm 

 to be devoid of glycogen, while acid-phosphatase activity seems to be concentrated around the peri- 

 pheral zones of the gland cells. The second phase would be the growth of those vacuoles and their 

 transport to the interface between the cell and the lumen of the swimbladder. Some of the energy 

 for this process would be provided by the breakdown of glycogen. The vacuoles may discharge their 

 contents into the mucous fluid covering the gas-gland, after which gas bubbles are formed (third 

 phase) eventually bursting to release their contents (mainly oxygen) into the swimbladder cavity. 

 (Gas-bubbles have never been observed within the gland cells.) The function of the vacuoles would 



