READING PRINTED TEXT 



65 



the range, on the following grounds, among 

 others: 



1. The bhnk rate and allied criteria are 

 questionable. 



2. Luckiesh and Moss present little or no 

 data in any terms for the region between 10 

 and 100 foot-candles. 



3. The recommendations rest in part on 

 estimates from various threshold measures 

 which have not been • validated for actual 

 visual work activities. 



In contrast to the Luckiesh and Moss 

 minimum of 100 foot-candles for the most 

 severe visual tasks, Tinker considers a maxi- 

 mum of 50 to be adequate. For a critique 

 of this controversy, see Bitterman's review 

 mentioned above (6) . 



Brief Reading Periods 



Reading is usually treated as a visual task 

 of intermediate difficulty. Returning to the 

 situation of brief reading periods, we find 

 that Luckiesh and Moss (30, p. 330) recom- 

 mend 10 to 20 foot-candles, while Tinker 

 (54), not differentiating between long and 

 short periods, recommends 10 to 15 foot- 

 candles. Tinker's recommendations are in- 

 tended to include a safety margin. His o^vn 

 experiments (50), covering a range from 0.1 

 to 53.3 foot-candles, showed no improvement 

 above three foot-candles when the subjects 

 were in every case well adapted to the read- 

 ing illumination level. Rose and Rostas 

 (39) found no significant improvement in 

 speed or comprehension from two to 55 foot- 

 candles. Luckiesh and Moss (30, pp. 107, 

 114, 190) report several sets of data, which 

 show, among other things, that reading speed 

 increases above 10 foot-candles, but only 

 at the rate of five percent for the interval 

 between 10 and 100. Tinker's placing of the 

 critical level (the point above which further 

 gains are inconsequential) at three foot- 

 candles when the subjects are adequately 

 adapted seems to fit the central tendency of 

 these several sets of data reasonably well. 



Sustained Reading 



For periods of sustained reading, Luckiesh 

 and Moss recommend 20 to 50 foot-candles 

 (30, p. 330) as against Tinker's 10 to 15 (54). 

 Luckiesh and Moss (30, p. 114) report that, 

 for periods of 30 minutes to an hour, reading 

 speeds increase as illumination increases from 

 one through 10 to 100 foot-candles, the maxi- 

 mum difference being about seven percent. 



A number of experiments have been set up 

 in such a way that illumination could be 

 related to time trends in some aspect of the 

 reading process or its concomitants. Tinker 

 (50) found that clearness of vision was im- 

 paired by two hours of reading at illumina- 

 tions below three foot-candles, but was not 

 affected at that level or above. Luckiesh 

 and Moss (30, p. 360) found that the power 

 of the convergence response of the eyes was 

 impaired 23 percent by one hour of reading 

 under one foot-candle, but only seven per- 

 cent under 100 foot-candles. 



Certain sets of data are in open conflict. 

 Luckiesh and Moss found, for periods of 

 sustained reading, that blink rate increased 

 and heart rate decreased less rapidly under 

 the higher illuminations (30, pp. 107, 353). 

 Bitterman (4) and McFarland, Holway, and 

 Hurvich (31), on the other hand, found in- 

 versions in the bhnk rate-illumination rela- 

 tionship, with large individual differences in 

 trends, and McFarland, Knehr, and Berens 

 (32) found no relation between heart rate and 

 illumination. These several authors inter- 

 pret their results, not as disproving illumi- 

 nation effects, but as invalidating the blink- 

 rate criteria. Bitterman (4) found heart 

 rate changes generally in accord with those 

 of Luckiesh and Moss, but on the grounds of 

 other evidence rejected the interpretation 

 that this indicated greater visual effort at 

 the lower illuminations. 



In looking for points on which there is no 

 conspicuous disagreement among these di- 

 versified data, we find, first an indication of 

 definite impairment below three foot-candles, 

 and second, the fact that, as illumination 



