Abstract
1. This paper presents increment threshold data for a fairly wide range of parameters but certain of the relationships are slightly complicated by what may amount to large differences in the observers' criterion. The following paper deals with a very limited range of parameters but the measurements are more precise and the analysis more simple.2. Increment thresholds have been measured for a small, brief duration test viewed against large steady and pulsed backgrounds, using the technique of Aguilar & Stiles (1954) which is intended to isolate rod vision. Under these circumstances it is shown that the test is seen by the rods and that the cones do not disturb the sensitivity of the rods.3. The relative positions of increment threshold curves for steady and pulsed backgrounds are usually consistent with an integration time of tau = ca. 0.1 sec.4. Backgrounds of less than 0.1 sec duration are effectively impulse inputs in the sense that the magnitude and time course of the test threshold disturbance (;impulse function') is dependent only on background energy.5. Functions of the impulse function may be substituted for the classical integration time tau. The exact form of the impulse function and its theoretical treatment is pursued in the next paper.6. The threshold disturbance caused by switching on a background light settles to a steady value within 2 to 3 times tau.

This publication has 13 references indexed in Scilit: