THE PERCEPTION OF LIGHT 617 



being eliminated. The owl may be taken as typical of nocturnal birds ; 

 Hecht and Pirenne (1940) taking the minimal observable contraction 

 of the pupil to green light as their criterion, reported an absolute 

 threshold of 1-5 x 10~" ml., their owti threshold under the same 

 conditions being 4-0 x 10"'^ ml. The astonishing visual performance 

 possible in owls in dim illumination was verified by Dice (1945-47), who 

 found that the barn-owl can detect and pounce upon dead mice at a 

 distance of 6 feet under an illumination of 7-3 x 10~' f.c. It is to be 

 remembered that clear starlight has a much higher intensity than this, 

 of about 8 X 10-5 f.c.i 



Among arh}i:hmic animals, adaptations to varying luminosity 

 may also reach a much higher efficiency than in man : it has long been 

 traditional that the wise rider relies upon the horse to pick its own 

 footsteps during the night. In this respect most scientific work has 

 been devoted to the cat. In behavioural experiments involving a choice 

 between darkness and a minimal degree of light, remarkably constant 

 results for the absolute limits of retinal sensitivity have been found 

 —Mead (1942), 1-3 X lO"" ml. ; Bridgeman and Smith (1942), 8-2 x 

 10"^ ml. ; Gunter (1951), 9-92 x 10"^ ml., the average threshold for 

 man being higher — of the order of 5-8 X 10"" ml. 



While the sensitivity to light in most Vertebrates is high, the 

 more highly evolved faculty of the discrimination of variations in 

 INTENSITY, is apparently less efficient, although in many cases the 

 failure may be due to lack of attention in experimental conditions 

 rather than lack of appreciation. Little conclusive work has so far 

 been done on this problem and some of it is contradictory, v. Hess (1909) 

 claimed that the silverside, Atherina, responded to white lights differing 

 only by 1 : 1-23 in brightness ; but the carefully controlled observations 

 of Bauer (1909) on this and other types of Teleosteans, which have 

 since been confirmed, tend to show that they are relatively indifferent 

 to wide variations of intensity of white light (Sgonina. 1933, on the 

 minnow, Phoxinus, and others). Thus, working with the mud-fish, 

 Umbra, and the stickleback, Eucalia, White (1919) found in training 

 experiments that they had difficulty in making any discrimination 

 between pigmentary greys and whites ; this she corroborated at a 

 later date (writing under the name of Hineline, 1927). Similarly, Cora 

 Reeves (1919) substantiated that the dace, Semofilus, could not dis- 

 tinguish differences in intensities greater than 1 : 4 although the sun- 

 fish, Lepomis, showed a better performance (1 : 2), an observation 

 corroborated by Hurst (1953). ^Moreover, goldfish have been taught Goldfish 



to choose one of three intensities, and to choose one light on the basis 

 of its relationshii) with two others \\hen the intensities as well as the 



1 The factor to convert foot-candles into millilamberts is X 3-382. 



At her 



