804 



LIGHT AND LIFE 



560 



570 



Fig. 6. Variation of ganglion cell response with a small change in wavelength. 

 Same experimental conditions as Fig. 5. Intensity of stimulus, 55 /^w/cm-. 



recording the number of impulses discharged. This procedure has 

 commonly been used in the determination oi luminosity functions and 

 avoids certain difficulties created by the non-linear (quasi-logarithmic) 

 nature of the response mechanism. The minimum intensity flash 

 necessary to evoke the discharge of one impulse was the most com- 

 monly used criterion for threshold response, although the thresholds 

 for two or more impulses were sometimes used where the presence 

 of spontaneous activity tended to introduce an appreciable number of 

 false positive responses. Occasionally, inhibition of pre-existing ac- 

 tivity was used as a criterion for threshold. There seems to be no 

 essential difference in the results obtained with any of these response 

 criteria. 



Fig. 7 shows a threshold intensity versus wavelength plot for a 

 typical ganglion cell. An "on" response was obtained from 400 to 

 610 m/n. At 640 ny no "on" response was obtainable at any in- 

 tensity; but a higher "off threshold could be obtained out to 750 m^u,, 



