462 CLAKKE AND DENTON [OHAP. 10 



cited by Beer, 1897) and the eyes of deep-sea animals, particularly those in the 

 "twilight zone" are generally very large with respect to the size of the animal. 



Now the human eye is very sensitive to small flashing lights and the distances 

 at which man can detect luminescent animals will probably give us a good 

 indication of the kinds of distances over which deep-sea animals can see one 

 another. Nicol (1958) has made estimates based on the known energies emitted 

 by luminescent animals and the known sensitivity of the human eye under the 

 best conditions. He gives, for example, the maximum possible distance at 

 which the human eye could just detect the ctenophore, Beroe, as 120 m, if the 

 sea-water had a transmission of 99%/m, and 41 m if the sea-water had a trans- 

 mission of 90%/m. Observations of an artificial source of light lowered into the 

 clear waters of the Brownson Deep (transmission 97%) at night have shown, 

 however, that in more natural and difficult conditions of observation, a small 

 light about 1000 times brighter than Beroe could only just be seen at 56 m by 

 a man under water and wearing a face mask using foveal vision (Clarke, un- 

 published observations). 



The distance at which a light source can be distinguished depends not only 

 on the intensity of the flux reaching the eye, the region of the retina on which 

 the image falls and the state of adaptation of the eye, but also on the degree of 

 diffusion of the image by scattering and the degree of interference of other 

 light flux present, particularly that from the surface. In the test at sea just 

 described, the intensity of upward scattered light was somewhat greater than 

 10~4 [jtW/cm^ at a depth of 1 m. At greater depths interference from scattered 

 surface light would be correspondingly less (and discrimination would thus be 

 easier), but at any given depth an eye looking upward would receive about 

 100 X more intense ambient flux than a downward directed eye (down to the 

 depths at which perceptible light penetrates). Therefore, under actual conditions 

 and particularly near the surface, the distances at which animals can distinguish 

 the flashes of other animals may be considerably less than those calculated by 

 Nicol for men under the best conditions. Further observations on the visibility 

 of small light sources are greatly needed. 



The eyes of fishes and cephalopods differ from those of land vertebrates in 

 that the cornea plays practically no role in the formation of the image (in the 

 human eye, for example, the principal refracting surface is the curved air-hquid 

 interface of the cornea). In the fish or cephalopod eye the image is formed by a 

 spherical lens which is so beautifully contrived that the retinal image is almost 

 free of both spherical and chromatic aberrations. This seems to be achieved by 

 continuously varying the refractive index of the lens from 1.53 (that of almost 

 pure protein) in the centre of the lens to approximately that of sea-water at its 

 periphery (Pumphrey, 1961; Fletcher, Murphy and Young, 1954). The lens in 

 the fish eye is not generally stopped down by the iris, which usually merely 

 prevents light from passing by the side of the lens. The fish eye has, therefore, 

 a very large aperture (//0.8) and, since the lens can form very good 

 retinal images, it is well adapted for detecting very weak and small sources of 

 light. The lens shows one other adaptation to depth ; in the surface forms where 



