88 DISCOVERY REPORTS 



take but one instance, it is surely significant that most deep-scattering layers are detected between 

 depths of 300 and 600 m., the levels between which Leavitt (1938) found minimum quantities of 

 zooplankton. It seems most likely that the fishes take what food they can get during the day, but have 

 a proper meal at night. Like the vertically migrating plankton animals, the up-and-down movements 

 of bathypelagic fishes must be governed by the daily rhythm of submarine illumination, a rhythm that 

 can readily be followed by their highly sensitive eyes. As Denton and Warren (1957) have shown, 

 their pure-rod retinae contain visual gold, a pigment with enhanced sensitivity for the deeper pene- 

 trating, blue rays from the sun. 



But not every species of bathypelagic fish is able to take direct advantage of the near-surface con- 

 centration of food. Adults of the deeper-living pelagic fishes (Lyomeri, ceratioid angler fishes, 

 Cyclothone spp. etc.) are not found near the surface at night. This is also true of many species living 

 in the upper 1000 m. Fishes with tubular eyes (hatchet fishes, Opisthoproctus, Winteria, Scopelarchus, 

 Evermannella, Gigantura, etc.), are rarely, if ever, taken in nets fished close to the surface. It is thus 

 evident that there are different degrees of vertical migration. These may now be summarized. 



A. Bathypelagic fishes migrating upward to the surface mixed layer. 



These are able to cross the thermocline and have already been mentioned in the first part of this 

 ection (p. 85-86). Certain species of the stomiatoid families, Astronesthidae, Chauliodontidae, Melano- 

 stomiatidae, Stomiatidae, Idiacanthidae and Malacosteidae. Numerous species of Myctophidae. 



Of the Astronesthidae, it appears that only species belonging to the genus Astronesthes cross the 

 upper thermocline. It is also likely that some of the myctophids (e.g. Lampadena spp.) do not reach 

 the surface-layer. The position regarding the smaller gonostomatids (e.g. Vinciguerria spp., Mauro- 

 licus spp., Ichthyococcus) is less certain. While they live for the most part in the upper 500 m. and 

 evidently undertake vertical migrations (Grey, 1955), the larger individuals are not often taken near 

 the surface at night. Perhaps they migrate to levels close below the thermocline. 



B. Partial migrators of the upper 1000 m. 



Cyclothone braueri (and probably C. signata), Sternoptychidae, various Myctophidae, Melamphaes 

 spp. Evidence for the migrations of hatchet fishes has already been given (p. 85), while Grey 

 (1955) has noted that Cyclothone braueri and Melamphaes spp. are taken at higher levels by night 

 than by day. 



C. Deeper partial migrators 



In the western North Atlantic diurnally migrating deep-scattering layers can be detected down to 

 1000 m. (Moore, 1958). This may well be near the threshold of light for the highly sensitive eyes of 

 deep-water fishes (see above). In discussing this problem, Clarke (1958) wrote as follows: 'A depth of 

 about 900 to 1000 m. would appear to be the shallowest level in clear water at which day and night 

 changes in illumination would be below the threshold of perception and at which daylight would 

 never be sufficiently strong for the inhabitants to be seen by their predators or by their prey.' Diurnal 

 migrations geared to submarine sunlight would thus seem to be confined to the upper 1000 m. 

 of the subtropical and tropical ocean. In temperate waters the threshold must be nearer 

 the surface. 



But certain of the deeper-living fishes seem to undertake vertical migrations. Off Bermuda, Grey 

 (1955) found that the daytime catches of Cyclothone microdon came from nets fished between 800 and 

 2000 m., while at night the upper level rose to 250 m. The populations of Cyclothone pallida, Melam- 

 phaes microps, M. opisthopterus and M. robustus also occupied higher levels at night. It may be that 



