68 CORA D. REEVES 



retina the brightness values of different regions of the gas- 

 spectrum are the same as for the central area of the retina 

 ■of the normal eye in daylight. It should then follow, if 

 Hess' mode of reasoning is sound, that the periphery has 

 color vision, but it is well known that it is color-blind. But 

 if the periphery of the human retina has the same bright- 

 ness values of the spectrum as the normal eye in daylight 

 but is color blind, may not the fish eye have the same 

 brightness values as the human color-blind and yet the fish 

 distinguish colors?' With more knowledge of the color 

 discrimination possible in lower vertebrates, theories of 

 human color vision may be enriched and improved. 



Bauer (1910), noting that the results of Hess were at 

 variance .with those of other writers decided that the fact 

 that the work of Hess had been done with fish in the dark 

 rather than in the light might account for some of these 

 differences. Accordingly he placed above the aquarium 

 in which he kept his fish a light, which furnished a constant 

 moderate light intensity for those experiments in which the 

 fish were light-adapted. The stimulus light reached the 

 narrow, black-lined glass aquarium through a clear rec- 

 tangular area in one of its ends. When two colored glasses, 

 used as filters, were placed so as to divide this opening 

 through the center, the halves of the aquarium were illumin- 

 ated by light of two colors. When different thicknesses 

 of white paper or ground glass were before the halves of 

 this opening, the two parts of the aquarium were illumin- 

 ated by white light of different intensities. When white 

 light was used, Bauer was not able to cause a gathering 

 of the fish in either half by differences in the. intensity of 

 the two sides. But with red and blue filters in place certain 

 light-adapted fish tended to gather in the blue half. With 

 the Charax and Atherina tested, he found, in contradiction 

 to Hess, no brightnesses of these two colored lights at which 

 the fish were equally distributed in both halves. When 

 spectral bands were substituted for the filters, the light- 

 adapted fish went into the blue green or yellow area. When 

 the orange or red (between 620 /^/^ and 630 f^f^) were pre- 

 sented to the fish, they rushed to the unilluminated end 

 of the aquarium as though frightened. When tested with 



