278 



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



fiber over a period containing the entire transient dip in frequency, and 

 comparing this with the number discharged in a comparable control period 

 of equal length. The deficit measures the integrated inhibition; it may be 

 correlated with the integrated activity of the inhibiting receptor, that is, with 

 the total number of impulses discharged in the first fiber in response to the 

 flash. When this was done in an experiment in which various flash intensities 

 were used, the relation was found to be a linear one similar to that shown for 

 the steady state (Fig. 12). The slope of the plot may be taken as an inhibitory 

 coefficient and the intercept with the axis of abscissae suggests a threshold for 

 the action of the flash. We are therefore led to hope that the linearity found 

 to hold for the steady-state interaction may also find a useful extension in the 

 analysis of the transient phenomena. Our work on this phase of the problem, 

 however, is still at its beginning. 



Fig. 24. Transient inhibition of the discharge from a steadily illuminated omma- 



tidium (upper trace) by a burst of impulses discharged by a second ommatidium 



nearby (bottom trace) in response to a 001 sec flash of light (signalled by the 



black dot in the white band above the time marks). Time in !^ sec. 



The spatial summation of transient inhibitory effects exerted on a test 

 receptor when brief flashes were applied to two regions of the eye in its 

 neighborhood has been examined briefly. The experiments reveal a point of 

 some interest. Since the bursts of impulses from such regions in response to 

 short flashes of moderate intensity may be completed before the beginning 

 of inhibitory effects they produce (as in Fig. 24), two regions that can be 

 shown to interact under conditions of steady illumination may give no evi- 

 dence of affecting one another when their inhibitory actions on a test receptor 

 are produced by sufficiently short flashes. This is illustrated in Table 2, where 

 the inhibitory effects of two regions, A and B, on a test receptor X are shown 

 for these two conditions. When the frequencies of X were measured during 

 steady illumination of A and B separately and in combination, the combined 

 effect of A and B illuminated together (measured by the decrement produced 

 in X's steady frequency), was less than the sum of their separate effects — the 

 consequence of their mutual inhibition, as we have explained in the earlier 



