VISUAL ACUITY 



37 



colors of illumination, red (670 mju), and 

 blue (490 m^u). Their results, for one of the 

 two subjects they used, are shown in Fig. 26. 

 Notice in particular the data at the low 

 luminance values, which show visual acuity 

 to be much better with blue light than with 

 red light. Exactly opposite results have 

 been reported by Luckiesh and Taylor 

 (see Fig. 27). Since the data by Luckiesh 

 and Taylor do not cover as great a range as 



24 



20 



16 



12 



08 



4 



-6-4-2 2 4 



BACKGROUND LUMINANCE IN LOG xuL. 



Fig. 26. Visual acuity with red and blue back- 

 grounds of various luminances. (Data from 

 Shlaer et al., 82) 



those of Shlaer et al., the scale in Fig. 27 

 is shghtly expanded. Further, in reworking 

 both sets of data to make them comparable, 

 the writer had to make certain assumptions 

 which may or may not be valid. In any 

 case, the data within Figs. 22 and 23 are 

 strictly comparable, even though the data of 

 Fig. 22 may not be strictly comparable to 

 those in Fig. 23. 



Notice particularly that Luckiesh and 

 Taylor find visual acuity under red light at 

 low luminances to be far superior to visual 

 acuity under blue light. The resolution of 



this apparently complete contradiction of 

 results probably lies in the way the two 

 groups of experimenters carried out their 

 photometric measurements at the low lumi- 

 nance levels. Shlaer et al. do not say 

 explicitly how they made these measure- 

 ments, but the implication is that they used a 

 technique similar to that used by Blanchard 

 (3) about which we commented earlier. 

 They equated luminances at high levels and 

 computed the luminances for the rod levels 

 by means of the densities of the neutral 

 filters they used. In fact, in one place in 



r) 04 



BACKGROUND LUMINANCE IN LOG m L 



Fig. 27. Visual acuity with red and blue back- 

 grounds of various luminances. (Data from 

 Luckiesh and Taylor, 59) 



their paper they say, "The intensity scale is, 

 by definition, that of the cones." This is 

 another way of saying that at —2.5 log mL, 

 for example, the blue light probably was 

 much brighter than the red light. Luckiesh 

 and Taylor, on the other hand, probably 

 used a low-brightness photometer in meas- 

 uring luminance values. This is another 

 way of saying that the luminances at each 

 level were adjusted until they appeared 

 equally bright. 



We cannot be sure, of course, that this 

 explanation accounts for the entire dis- 

 crepancy between the results of these two 

 experiments, but it undoubtedly accounts 

 for most of it. If this analysis is correct, 



