Abstract
1. A test is proposed of the hypothesis that visual purple is the photosensitive substance concerned in dim vision. It is based on the fact that fish visual purple is different from that of other vertebrates. If the hypothesis is correct, the fish dim-visibility function should be different from that of other vertebrates and should be determined by the absorption spectrum of its visual purple. 2. A new method is described for obtaining the visibility function of fish, in quantitative terms. It depends on the measurement of the least amounts of various spectral energies which will produce a visual orienting response to the displacement of a constant background. 3. Data are presented on thirteen animals. It is shown that the maximum of the visibility function is identical with the maximum of the absorption spectrum of fish visual purple. The shapes of the visibility curves obtained are, however, variable and different from that of the absorption spectrum. 4. The possibility that Lepomis visual purple is different from that of other fish is ruled out by a series of measurements which confirm the results of Koettgen and Abelsdorff on other fish. 5. Reasons are given for the conclusion that there are present in Lepomis special conditions which distort the visibility curve out of true agreement with that predictable from the absorption spectrum of its visual purple. The suggestion is made that the presence of light absorbing, but not light sensitive, pigments is responsible for this distortion. One of these pigments may perhaps be carotin while the second is unspecified.

This publication has 3 references indexed in Scilit: