264 



H. K. HARTLINE, F. RATLIFF AND W. H. MILLER 



If the group is of a fixed size and is illuminated at various intensities, the 

 frequency of any one of the units (r) will vary more slowly with the stimulus 

 intensity than if it were illuminated alone (e), by the factor 1/(1 +(«—!) K} 

 (a shght displacement, due to the constant term, might be noticeable). Figure 

 16 shows that this expectation is borne out to a fair approximation. For large 



40 



30 



20 



JO 



0.25 mm. 



2 mm. 



-3.0 



-2.0 



-i.O 



00 



log I 



Fig. 16. Relations between intensity of light (log /. abscissae) and frequency of 

 discharge (ordinates) of a single ommatidium when illuminated alone (upper curve, 

 0-25 mm spot of light centered on its facet), and when illuminated together with 

 a large number (approx. 40) of neighboring ommatidia surrounding it (lower 

 CLU've, 2 mm spot of light centered on its facet). 



areas of retinal illumination, the activity of the individual receptors is reduced 

 proportionately at each intensity. One of the consequences that this must 

 have is that the range of intensities capable of being covered by the visual 

 receptors is increased before a possible physiological Hmit to the frequency 

 of the receptors is reached. 



The relation between r and n for fixed e is hyperbolic, since (5) may be 

 written 



(r - rO) (« + i - l^l = ^^^ (5a) 



K 



K 



If one of the receptors is considered a "test" receptor, the difference 

 (e — r) in the frequency of its response when illuminated alone and when 

 illuminated with the rest of the n members of the group, measures the in- 

 hibition exerted by the group on each of its members. As a function of n 

 this will increase along a hyperbolic curve from zero at n = 1 to (e — r^) 

 at large values of n (in (5a) write r — r^asie — r") — (e — r)). This is shown 



