INHIBITORY INTERACTION IN THE RETINA 267 



5.0 



4.0 



3.0 



2.0 



— 

 



mm dia. 



mm dia. 



25 



Frequency (Impulses per sec.) 



Fig. 18. Effect of size of group on the inhibition it exerts on a nearby ommatidium. 

 Decrease in frequency of a receptor B (one of a small group of fixed size) produced 

 by illumination of a nearby group, A, is plotted (ordinate) as a function of the 

 frequency (abscissa) of one of the receptors in A. For the lower plot the diameter 

 of the spot of light on A was I ■ mm. For the upper plot the group A was enlarged 

 by using a spot of light 20 mm in diameter, and the slope of the line (Kba) was 

 increased, as predicted by theory (the change in threshold is not expected 

 according to the theory of idealized group action). 



the coefficient of its inhibitory action on a nearby group of fixed size. It also 

 decreases the coefficient of action of that fixed group on it (not shown). This 

 is as predicted by the definitions (6D), for an ideahzed situation. Exact 

 quantitative prediction cannot be made because of the departure of the actual 

 groups from the ideal uniformity in receptor properties and interactions, re- 

 quired in the derivation of (6) and (6D). 



Especially noteworthy is the case in which one of the groups is reduced to 

 just one ommatidium (a "test receptor"). Then the coefficient of the other 

 group (/; receptors) on it is n times the coefficient of one of that group's 

 receptors acting on the test receptor, while the action of the test receptor 

 back on the receptors comprising the other group is diminished by the factor 

 1/(1 +(«—]) A"}. Here K' is used to designate the internal self-inhibition 

 of the group — for a compact group perhaps of the order of OT to 0-2. Hence 

 for a large group, the test receptor's "back effect" might be substantially 

 diminished compared with the effect it would have on any one of the receptors 

 of the group by itself. 



