THE PHYSICS AND BIOLOGY OF VERTICAL MIGRATIONS 91 



is three times the value taken by Kanwisher and Ebling (1957), the gas secretion time for a downward 

 migration of 100 m. would still be 11 hr. At this rate the fish could just restore its buoyancy to the 

 neutral level between the daily descent and ascent. 1 Moreover, it seems unlikely that the gland could 

 produce oxygen fast enough to keep pace with the increasing pressure during the descent. A striking 

 proof of this may well have been given by Hersey and Backus (1954). While following a descending 

 deep-scattering layer they found that the resonant frequency of the reflected sound increased, as 

 though elements in the nature of gas-bubbles were being compressed. As they remark, the bubbles 

 were probably the swimbladders of fishes. 



But if the cells of the gas-gland are able to store oxygen in a combined form, the task of restoring 

 buoyancy would be quickened. As we have seen (pp. 73-75), the gland is highly developed in comparison 

 with the volume of the swimbladder. If, by this means, buoyancy can be regained a few hours after 

 the descent, the remaining hours before the ascent could be given to storing oxygen in the gland. If 

 the cells have not reached saturation level by the time the fish is ready to ascend, it may be that some 

 oxygen could be transported to them during the early part of the climb. Deep-scattering layers are 

 known to rise quite slowly for several hours before sunset (see Clarke and Backus, 1956). During this 

 time the (relative) reduction in pressure could be tolerated without the resorptive part of the swim- 

 bladder being brought into play (see pp. 96-97). Thus, while the gland would not be secreting it could be 

 storing oxygen. These considerations are at least in keeping with present conceptions of gas-pro- 

 duction. Scholander (1954) is inclined to think that oxygen in compound form may be stored in the 

 gland-cells for intermittent use. 



The oxygen secreted into the swimbladder comes from the water passing over the gills. How does 

 a fish obtain enough oxygen if it lives in an oxygen minimum layer during the daytime ? To take an 

 extreme instance, and one already cited (p. 86), the oxygen level is below 0-25 ml./l. between depths 

 of 100 and 1000 m. in the eastern tropical Pacific, north of the Equator. It is here that Beebe and 

 Vander Pyl (1944) investigated the diurnal migrations of myctophids. 



Thus, apart from the upper 100 m., the waters in this part of the ocean would seem to contain 

 insufficient oxygen for active metabolism. (At 20 C. the oxygen consumption of the goldfish, a 

 tolerant species, starts to fall below an oxygen content of 2-5 mg./l. (= 1-75 ml./l.) (Fry and Hart, 

 1948).) Apart from the other bodily processes, the maintenance of an oxygen pressure of 40 atmo- 

 spheres in the swimbladder would be against a diffusion gradient of about 10,000 to 1 (Kanwisher 

 and Ebeling, 1957). This seems a gigantic task. 



Active life for fishes in this oxygen minimum layer would seem to pose a physiological dilemma. 

 They would have to work very hard to get enough oxygen and the more active they were, the more 

 oxygen would be required. To consume 0-5 ml. 2 per hour, a lantern fish or Astronesthes would have 

 to pass at least 3 1. of seawater over its gills (assuming that two-thirds of the dissolved oxygen would 

 be taken up by the blood). A fish with a volume of, say 5 ml. would not be capable of pumping 1 ml. 

 of seawater over its gills in a second, or even half this volume. In trout (volume about 900 ml.) Van 

 Dam (1938) found the maximum ventilation-volume to be about 5 ml. /sec. 



Perhaps the activity of these fishes is largely suspended during the day. It was in the Eastern 

 Tropical Pacific that a single haul of a mid-water trawl yielded over 5000 fishes, most of which were 

 the small lantern fish, Diogenichthys scofieldi (Marshall, 1954). This is a phenomenal catch and one is 

 led to wonder whether it might be due to reduced activity. Lantern fishes are not known to concen- 

 trate in great numbers in dense schools nor are they easy to catch. Daytime observations from a 

 bathyscaphe might well be revealing. 



1 There is, of course, no certain evidence that the vertical migrations of any particular individual are carried out day after 

 day. In an adult fish the extent of these displacements may also vary in time and space. 



