LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 73 



which he allowed to fall through the water while a gray 

 back ground was placed against the outside of the aquar- 

 ium. He also fastened pieces of yellow meat and pieces 

 of colored wool of similar shape upon one side of a glass 

 plate which could be lowered into the water, so that the 

 colored bits were visible to the fish against the same gray 

 background but the glass surface which separated them 

 from these pieces was invisible. The fish went with equal 

 frequency to all the bits of color as the glass was sunk 

 into the water. Next a glass was used to cut off a small 

 space along one side of the aquarium. For one week yellow 

 food w^as allowed to fall on the side of this glass plate where 

 the fish might reach it. Then for two weeks, while yellow 

 food was given on the same side, where it could be seized, 

 blue food was allowed to fall at the same time on the re- 

 verse side of the plate, where it was visible but inaccessible 

 to the fish. After the three weeks of feeding of yellow 

 meat, with or without the inaccessible blue, bits of yellow,, 

 blue, gray, red, and green were dropped into the aquarium 

 behind the glass plate, the fish swam at them freely and 

 showed no preference for yellow. Again Hess insists that 

 fish are color blind. 



In this work of Hess, in which he used training methods, 

 he allowed the food masses to fall through the water and 

 found that the fish rushed to a mass of any color, though 

 they had previously been fed upon yellow-colored food 

 only. 



Hess has here adapted the well known experiment of 

 Mobius (1873) with pike, afterward confirmed by Triplet 

 (1901) with perch. Mobius found that when a pike in 

 an aquarium was separated from minnows by a glass plate, 

 the pike at first dashed at the small fish and received a 

 bump on the nose whenever it did so. After a time the 

 pike stopped rushing at the minnows, and when the glass 

 plate had been removed, the minnows were not at first 

 seized by the pike. A severe bump had become associated 

 with trying-to-seize-minnows. Had Hess, while feeding 

 his fish on food of quite other form and appearance, first 

 succeeded by punishment in establishing an immunity 

 for his blue food masses so that they were not seized when 



