96 DISCOVERY REPORTS 



swimbladder is found in the deep-sea eel Synaphobranchus kaupi (= S. pinnatus) (Scholander and 

 Van Dam, 1954) (depth range: 200-3599 m -) an d i n tne macrourid, Nematonurus armatas (Hagman, 

 1921) (281 to 4600 m.). 'At least forty-six of the more than seventy-six specimens recorded were 

 caught between about 2600 and 3660 m.' (Grey, 1956). 



In view of these findings I have examined the following species, the depth range 1 of which appears 

 after the scientific name: Moridae, Antimora rostrata (403-2904 m.); Macrouridae, Lionurus filicauda 

 (2515-4846 m.); Brotulidae, Aphyonus gelatinosus (2550-4360 m.); Typhlonus nasus (3932-5090 m.); 

 Mixonus laticeps (3200-4575 m.); Bassozetus taenia (4575-5610 m.) and Bassozetus compressus 

 (1920-2744 m.). Except in Aphyonus gelatinosus and Typhlonus nasus, each of these species has a 

 capacious swimbladder extending down the greater part of the body cavity. Drawings of some of 

 them appear in Text-fig. 41. 



Thus a well-developed swimbladder is present in fishes that may range down to depths of about 

 5000 m. This is remarkable in view of the previous discussion concerning the compressibility of 

 gases (p. 84). At a depth of 5000 m. the density of oxygen is about seven-tenths that of 

 seawater. 



A fish with a fully developed swimbladder containing oxygen would require tissues with a density 

 of about 1 -04 to be neutrally buoyant at this depth. But this requirement may well be met. These 

 deeper-living fishes have rather lightly ossified skeletons and the muscles sheathing their flanks are 

 fairly thin. In PI. 3 a radiograph of Lionurus filicauda is shown against one of the poor cod, Trisopterus 

 minutus, a shallow-water fish. It will be seen that the skeleton of the cod is more sharply revealed 

 than that of the macrourid: 2 there is also a striking difference in the size of the otoliths. Denton and 

 Marshall (1958) have shown that the density of fish muscle is relatively high, which is not surprising 

 in view of the heavy protein molecules that are required for contraction. In bathypelagic fishes 

 without a gas-filled swimbladder, reduction of muscle and bone can bring the animals close to neutral 

 buoyancy. 



The swimbladders of these fishes are also remarkable in having extremely long retia mirabilia 

 (see Text-fig. 41). In Lionurus filicauda there are six retia, each having a length of at least 20 mm. The 

 two retia of Bassozetus taenia are about 25 mm. long, while the single one of B. compressus has a 

 length of 17-5 mm. In the deep-sea eel, Synaphobranchus kaupi, the span of the rete is about 10 mm. 

 (Scholander, 1954). 



These findings are surely significant in view of the previous discussion on retial length and the 

 efficiency of gaseous exchange between the arterial and venous capillaries (pp. 70-72). As we saw, the 

 deeper the living-space the longer should be the length of the capillaries. This is strikingly revealed 

 in Table 7. (The data for Antimora viola, blue hake; Sebastes marinus, rose fish; Nezmnia bairdii, 

 common rat-fish; and Synaphobranchus kaupi are taken from Scholander and van Dam (1954) and 

 Scholander (1954).) 



Apart from the general trend, which is quite evident, it is interesting that the deeper living of the 

 two Bassozetus species has the longer retia, and two in place of one. Here, then, is indication that the 

 swimbladders of abyssal fishes may well be functional at great depths. 



But the gas-glands do not appear to be highly developed, which might suggest that their function 

 is simply to keep the swimbladder ' topped up ' with gas. Abyssal fishes are unlikely to move very far 

 from the deep-sea floor, except perhaps during the breeding season, although they could have a 

 relatively wide freedom of movement in the vertical plane. If the upper limit of this zone is marked 

 by a pressure reduction of 20 per cent (Jones, 1952), an abyssal fish with a closed swimbladder and in 



1 Depth-range data have been taken from Grey (1956), Nybelin (1957) and Bruun et al. (1956). 



2 The 'Challenger' fishes were preserved in alcohol so there is no danger of the degree of ossification being reduced. 



