Configuration dependence of scotopic spatial summation.

1. Psychophysical experiments were done under conditions of complete dark adaptation, 7 degrees in the human temporal retina, using blue-green flashes brief enough to be within the range of complete temporal summation.2. Frequency-of-seeing experiments were done for two 3.4' squares separated by 20' or 23' and for each square separately. The results were consistent with the hypothesis that the squares acted independently in determining the absolute threshold. There was no observable summation occurring between them.3. The absolute threshold (50% seeing) for two 3.4' squares (the centre-to-centre separation varied from 3.4' to 56') was measured relative to the absolute threshold for just one such square. It required more photons at the cornea at threshold for two spots than for one such spot, except when they were actually touching.4. The same subject, in the same retinal region and under otherwise identical test conditions, showed complete spatial summation at absolute threshold for circular spots up to 30' in diameter.5. Similar experiments were done on other subjects using 3.4' squares and lines of width 3.4'. All subjects showed complete summation at absolute threshold for lines 34' long. All showed lack of spatial summation for two squares unless the two squares touched to form a rectangle.6. The absolute threshold for N small spots was compared to the absolute threshold for 1 spot, where N varied from 2 to 25. Again complete spatial summation broke down even when all the spots lay within the Ricco area.7. It was shown that the results are inconsistent with the hypothesis of a linear output and a linear summation of excitations.8. Various hypotheses are examined and it is shown that no simple hypothesis can explain all the results. The conclusion is that spatial summation at absolute threshold is configuration dependent.