W. A. H. RUSHTON 715 



pigment, the increment threshold was recorded. Superimposed upon 

 the steady bright 60,000 Td background was a small Hashing light 

 whose intensity adjusted so as to be just threshold. At the end 

 of half a minute the 00,000 Td was extinguished and the absolute 

 threshold traced for the next minute or less. 



These results are shown by the irregular tracings of Fig. 3. Starting 

 from darkness, the increment threshold jumps to 3.2 log units and 

 falls back to an increment threshold of 0.6 upon extinction of the 

 background field, the pigment being 20 per cent bleached at this 

 moment. When, on the other hand, we start from the fully bleached 

 state, though the increment threshold is hardly altered, the ab 

 solute threshold falls only to 2.8 units at 0.5 min, wdien the pigment is 

 still 80 per cent bleached. It is therefore plain that the absolute 

 threshold depends very greatly upon the amount of pigment bleached, 

 but the increment threshold very little. That, however, is precisely 

 what would be expected from equation (2) : 



A//A V = I + Id- 



For, by extracting the values of / from the log / tracings of Fig. 3, 

 we find that at 0.5 min., when 20% of the pigment was bleached 

 Ij) = 4, I -}- In = 1000, and thus / is about 1000. On the other hand 

 in the light-adapted trace at this moment when 80% of the pigment 

 was bleached, Ijj — 600, hence / + /» should be 1600 and log 

 (/ _p Ijj) at this point should read 3.2, as it does. Thus equation 

 (2) predicts the 0.2 difference which is observed between the two log 

 increment thresholds in Fig. 3. 



(c) Relation of Absolute Threshold to Visual Pigment 



The foregoing section makes it clear that it is not in the increment 

 threshold but in the absolute threshold that we must look for a re- 

 lation to pigment level. Is there a unique relation between the rod 

 threshold and the level of rhodopsin at the moment, and if so, what 

 is this relation? 



Or on the contrary, do other factors intervene and in particular, 

 does some nervous inhibition regulate the sharing of single nerve 

 paths? For electrophysiological records from animals undoubtedly 

 prove that rods and cones may use the same nerve, and many authors 

 have invoked a rod-cone inhibition. 



Fig. 4 is taken from an experiment by Fuortes and me and shows 

 the time course of regeneration of rhodopsin in a 5° area 15° tem- 

 poral to the fovea on Fuortes' left eye. The white circles give the 



