720 LIGHT AND LIFE 



During the second line a much brighter light was shone for 10 

 seconds, and the slow pulses are seen to be far more numerous and 

 apparently smaller. And it could be argued that the steady potential 

 change upon which these little pulses seem to rest is itself composed 

 of the statistical sum of innumerable pulses. This interpretation is 

 far from proven, but it is the sort of thing which could account for 

 Hartline's results upon the probability of responding. The third 

 line of Fig. 7 shows a return to the steady light of line 1, but now 

 the preparation is light-adapted and the pulses are so small that 

 (except for the giant pulses) they can hardly be seen. 



Unless man is very much less efficient than Limuliis, then, Ave must 

 look for a mechanism of adaptation in which each quantum pro- 

 duces an effect, biu in which the effect diminishes rapidly and ex- 

 ponentially with the fraction of pigment bleached, according to 

 equation (4) . 



I do not think being confronted with this ancient arthropod should 

 shame us into claiming a visual efficiency which we do not possess. 1 

 am not convinced that we see much better near absolute threshold 

 with a light-adapted eye than we do with a dark-adapted eye whose 

 threshold has been equally raised by interposing a neutral density fil- 

 ter. If this is so, we waste quanta after light adaptation, whatever 

 Limulus may do. This question is not an easy one to settle but we 

 badly need an answer. 



Excitation by Light 



We may now state the jjrincipal features which must be explained 

 by any theory relating light absorption to nerve stimulation. 



(a) A quantum absorbed by visual pigment almost anywhere in 

 the rods and cones may produce an effect 0.1 mm away of a kind which 

 will add to similar effects from other receptors and stimulate a nerve, 

 and the effect occurs well within a second of the flash. 



(b) Light of intensity / produces a nerve signal V such that 



gge-«^+l) 



V - Vo = lo 



where F/, is the value of V in the dark, /„ is the absolute dark-adapted 

 threshold, a is a constant which is about 1 for cones and 10 for rods, 

 and X is the fraction of visual pigment bleached at that time. 



(c) There should be another statement relating not just the magni- 

 tude but the information content of the nerve signal to the quantum 

 content of the light flash. But I do not know what this is. 



