ir. A. H. RUSHTON 707 



used the statistics o£ quantum fluctuations and nicasined the fre- 

 quency of seeing a flash as a function of its log intensity. The two 

 methods gave reasonable agreement, and the authors concluded that 

 about 6 quanta must be absorbed in order to see, and moreover that 

 each of these quanta was absorbed in a separate rod cell, since the 

 probability of a double-hit upon one rod was substantially less than 

 the probability Avitli which subjects actually saw. 



Contemporary work by van der Velden (12) and subsequent results 

 by many authors, though diverging as to the actual minimal number 

 of rods involved, all agree that a rod can be excited by one quantum. 



Our first question, then, is: "How can a single quantum of light 

 absorbed anywhere upon the outer segment of a rod set up some 

 kind of excitation which travels down the cell to its fine termination 

 0.1 mm away, and there unite with similar excitations of a few other 

 active rods to set up a nerve impulse?" All this must occur within 

 the latent period for seeing — less than 1 second. 



(b) Adaptation 



It is a familiar experience that the eye can adapt rapidly to light 

 and that it takes a much longer time to adapt to darkness. One feature 

 of the change is the transfer of function from rods at low luminance 

 to cones at high luminance. But, confining our attention to rod 

 vision alone, it is a well-known fact that the visual threshold may 

 fall 100-fold or more during the course of dark adaptation. Thus 

 somewhere in the mechanism of vision between the light and its 

 perception is a stage whose sensitivity varies very greatly. Where is 

 this stage? 



It is certainly not chiefly in the brain, for records from the ganglion 

 cells in the excised frog's eye show the same thing. Moreover, a 

 similar effect is seen in the human electroretinogram (ERG) , which 

 is the earliest result of photolysis that can be recorded. 



The records of Fig. 1 (Fuortes and Rushton, unpub.) are taken 

 from the lateral eye of the arthropod Limulus, and they are obtained 

 by leading from a micropipette inserted into one nerve cell in an 

 ommatidium. Fig. lA shows the effect of light adaptation. First the 

 ommatidium is exposed to a weak flash which depolarizes the cell 

 and generates a train of impulses. Then a bright flash 4,000 times 

 as strong is given, and finally the first flash is repeated. The depolari- 

 zation and impulse train is greatly diminished following the bright 

 light. To what extent is the diminution due to the intense previous 

 activity which the bright light had induced in the recorded cell? Fig. 



