INTENSITY 



363 



performance varied little over a wide range 

 of illumination values for the two higher 

 brightness contrasts. And for 36.5 percent 

 contrast, the improvement is slight above 

 20 f.c. For the 3-minute size of object, per- 

 formance is nearly constant for 20 f.c. and 

 above, with the two higher contrasts ; but it 

 increases significantly through 20 but not 

 beyond 100 f.c. for the lower contrast. 



For the 1-minute size of detail all perform- 

 ance is on a lower level. Performance rises 

 rapidly through 20 foot-candles and there- 

 after the improvement becomes slower and 

 slower. Apparently the optimum practical 

 illumination level for 1-minute size is be- 

 tween 20 and 100 foot-candles. It should be 

 noted that, although lower brightness con- 

 trast brings poorer performance at each il- 

 lumination level, increasing the illumination 

 intensity actuallj^ becomes less effective in 

 improving speed of performance as the 

 brightness contrast becomes smaller. In 

 Part II of the experiment, sizes of test ob- 

 fects were 1.5, 3.0 and 4.5 minutes. The 

 illumination levels were 0.5, 2, 8, 32, 128 and 

 512 f.c. The findings are approximately the 

 same as in Part I. The improvement leveled 

 out at around 32 or 128 f.c. With high con- 

 trast (97 percent) the leveling off occurred at 

 8 and 32 f.c. In 3 out of 13 instances it was 

 between 32 and 128. There were slight im- 

 provements which were insignificant from a 

 practical viewpoint up to 512 f.c. Again 

 brightness contrast was potent in determin- 

 ing efficiency of performance. 



The data on brightness contrast may be 

 summarized as follows: (a) When brightness 

 contrast between object and background is 

 liigh, discrimination of 3- to 6-minute sizes 

 is not significantly improved by illumination 

 above about 20 f.c. (b) With small size 

 (1-minute) object, performance improves sig- 

 nificantly with increased light intensity up to 

 about 30 f.c. or to a level between 20 and 100 

 (or 32 and 128) f.c. The curves suggest 

 that the highest practical illumination is 

 around 50 to 60 f.c. (c) The greater the 

 brightness contrast, the better the visual 



discrimination at any illumination level. 



(d) Excessive illumination will not compen- 

 sate for small object size or for poor contrast. 



(e) Nevertheless, slight improvement is 

 achieved up to 500 f.c. for discriminating 

 smaller object sizes (1 to 3 minutes) and ob- 

 jects with poor brightness contrast. 



Efficiency of Performance 



An analysis of various criteria employed 

 for specifying hght intensities has been made 

 by Tinker (31). Efficiency of performance, 

 although not an ideal criterion, is more satis- 

 factory than certain others, which appear to 

 lack validity. The earlier data on the rela- 

 tion of illumination intensity to efficiency of 

 performance are summarized by Tinker (30) . 



Critical Levels of Illumination 



The critical level of illumination is the in- 

 tensity beyond which there is no further 

 increase in efficiency of performance as the 

 foot-candles become greater. In Tinker's 

 summary (30) the following critical levels of 

 illumination are listed: for reading legible 

 print (about 10 point on good paper) by 

 adults, it is approximately 3 to 4 f.c; for 

 reading and study of children, 4 to 6 f.c; for 

 arithmetical computations, less than 9.6 f.c; 

 for sorting mail, 8 to 10 f.c; for the exacting 

 task of setting 6-point type by hand, 20 to 

 22 f.c; for the very fine discrimination re- 

 quired to thread a needle, 30 f.c; and for 

 reading newspaper print, about 7 f.c. 



Other Studies on Illumination Level 



After an extended period of experimenta- 

 tion, Holway and Jameson (12) arrived at 

 20 f.c. as the amount of light needed on the 

 work surface in offices and reading rooms. 

 An intensive study on effects of illumination 

 level on visual performance has been com- 

 pleted by Simonson and Brozek (27). Six 

 subjects were used. The task was to dis- 

 criminate small letters (3.5 points) as they 

 passed a window. Illumination intensity 

 ranged from 2 to 300 f.c. There was a sig- 

 nificant increase in performance up to 50 f.c. 



