ELECTRICAL PHENOMENA IN VISION 597 



the reciprocal relation does break down for very low intensities or very 

 long exposure times. The results here are consistent with those obtained 

 by Wald and Clark (1937) and Crawford (1946) in the human eye and 

 by Riggs (1937) in the eye of the frog. Prolonged exposure to weak 

 illumination is followed by an inordinately long recovery process in all 

 these experiments. This finding is consistent with the concept of a dual 

 restoration reaction as described by Wald and Clark (1937). 



In one other respect the Limulus data are of particular importance 

 for the interpretation of the dark-adaptation process. Hartline and 

 McDonald observed in detail the train of impulses following various 

 intensities of test flash at various times in the dark. Their conclusion 

 was that it was not possible to match such a record in the fully dark- 

 adapted eye with any single record in a partially dark-adapted one, even 

 by the appropriate adjustment of stimulus intensities. In other words, 

 it is true even in this primitive eye that qualitative as well as quanti- 

 tative differences exist in the responses obtained at various stages of the 

 adaptation process. It therefore follows that the "sensitivity" or recip- 

 rocal of threshold intensity is only one of several possible indexes of the 

 course of dark adaptation. 



Light Adaptation. Harthne and McDonald (1947) also studied the 

 adaptation of the Limulus receptor element to fixed levels of illumination. 

 They first adapted the eye for 10 min or more to a given level, then inter- 

 rupted the light and applied a test flash after 1 sec of darkness. The 

 resulting data show that proportionately higher intensities of the test 

 flash must be employed at the higher levels of hght adaptation. At each 

 level, however, the curve relating the number of impulses to the logarithm 

 of the intensity of the test flash is similar to each of the other curves, 

 simply displaced along the intensity axis. In other words, the sensory 

 cell maintains a full range of responsiveness at all levels to which it can 

 adjust itself. 



Another method of studying light adaptation was used by Riggs and 

 Graham (1940, 1945). A test flash was added to the adapting Hght at 

 various times after the adapting light was turned on. The effect of the 

 test flash was to cause a momentary increase in the frequency of nerve 

 impulse discharge. Strangely enough, the receptor unit became increas- 

 ingly responsive to such a test flash during the first minute of light adap- 

 tation; after that the effectiveness of the added flash diminished steadily 

 with increasing exposure to the adapting light. A somewhat similar find- 

 ing in human vision has been reported by Baker (1949). 



Receptor Variability. There appear to be conditions under which rela- 

 tively large fluctuations of sensitivity may take place in a sense cell even 

 when a steady photosensory state may be reasonably assumed. One such 

 condition is that of the refractory period in the sense cell (Riggs, 1940). 

 Immediately following the discharge of any impulse in response to steady 



