DARK ADAPTATION AND NIGHT VISION 



15 



in. In these experiments all the colors were 

 measured photometrically at a brightness of 

 10 millilamberts which is safely above the 

 Purkinje effect, and the lower intensities 

 calculated from a knowledge of the filter 

 densities. 



Unfortunately, these data cannot be put 

 to practical use because it is impossible to 

 reconvert to any meaningful system of 

 units from the data given by Blanchard. 

 The same is true for many other studies of 

 visual functions for different colors. 



In addition to the specific problem of 

 measurement which vitiates Blanchard 's 

 data, there are some others arising from his 

 experimental conditions. Only a few of 

 these will be mentioned here. Blanchard 

 used a very large adapting field, about 70° in 

 angular size, and a test field about 5° in size. 

 This size of test field, unfortunately, 

 stimulates an area somewhat greater than 

 the fovea, i.e., a region of very uneven 

 sensitivity (see Figs. 2, 3 and 4) . Blanchard 

 himself recognized this objection to his own 

 data, but did not offer a solution to the 

 difficulty. Still another difficulty with 

 Blanchard 's experiment is that it was 

 apparently done with only one subject. 

 Blanchard, as a matter of fact, does not say 

 exphcitly how many subjects were used, 

 but the implication is that only one was 

 used. Further, to be of greatest practical 

 usefulness, this experiment should be ex- 

 tended to measure the instantaneous thresh- 

 old with differently colored test lights, 

 but with the adapting field always white. 

 As we shall see later, the color of the adapting 

 field has a very important effect on the 

 subsequent dark adaptation of the eye and 

 it is reasonable to infer that the instantane- 

 ous threshold would be similarly affected. 

 Finally, Blanchard does not state explicitly 

 how complete light adaptation was. This 

 also is a very important point in experi- 

 mental procedure, because, as we shall 

 also see later, the subsequent sensitivity of 

 the eye is markedly affected by the amount 

 of previous light adaptation. 



The above discussion indicates clearly 

 that this is a problem which needs rein- 

 vestigation. Good basic data on this func- 

 tion would enable the engineer to solve 

 several sorts of practical seeing problems. 

 During the war, for example, radar screens 

 were placed in fast day bombers. A very 

 practical problem at the time was this one: 

 how luminous must a radar screen be so 

 that the operator can see it immediately 

 after scanning the sky? Blanchard 's data 

 provide us with a reasonable basis for 

 guessing, but they are hardly adequate for 

 precise work. 



The Typical Curve of Dark Adaptation. 

 The process of dark adaptation is ordinarily 

 traced by determining the dimmest light a 

 subject can see at various times after the 

 lights are turned out in a dark room. 

 The data in Fig. 9 were obtained by Sloan 

 (84) and are typical for this function. 

 The test light in her experiment was white, 

 1° in diameter, and was situated in the nasal 

 field of view, 15° from a fixation point. 

 The solid line in Fig. 9 represents the average 

 values for 101 subjects; the dotted lines 

 include the thresholds for 95 percent of the 

 subjects. 



It is evident from Fig. 9 that there are 

 two segments to the dark adaptation curve: 

 an initial very rapid decrease in threshold 

 which levels off at about 10 minutes, and 

 another more gradual decrease in threshold 

 which starts at about 10 minutes and 

 continues for some time. The decrease in 

 threshold which occurs during the first 8 or 9 

 minutes is due to an increase in the sensi- 

 tivity of the cones and, to a small extent, to 

 dilation of the pupil which increases the 

 light-gathering power of the eye. The 

 leveling off at about 10 minutes is now 

 known to represent the cone threshold. If 

 colored test lights are used to measure dark 

 adaptation, the colors can usually be 

 recognized up to this point (see Fig. 10). 

 The decrease in threshold which occurs 

 after 10 minutes is due to the increase in 

 sensitivity of the rods. If colored lights 



