LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 



27 



impossible, it must be concluded that the patches were 

 then of equal brightness for the fish and that they were 

 unable to discriminate them by their quality difference. 

 The first tests were made without the slit, that is, with 

 the red much more intense than the blue. They are pre- 

 sented in this section, while those with shifting values of 

 red appear in the following section. The plate illuminated 

 by the spectral band ^ 4600 to ^ 5245 (blue) was chosen 

 for the positive plate, and the fish were fed before it and 



Fig. 11 

 The First Record of Fish H (see p. 28). 



not before the red. The experiments were made with two 

 horned dace, Md'^ and //J" . These fish had been in the 

 laboratory a year, and the previous spring had been tested 

 with red and blue in the experiment aquarium for some 

 weeks. When they were again introduced into the experi- 

 ment aquarium on the morning of October 13, 1914, they 

 showed absence of fear by coming within two minutes to 

 feed from my fingers. A detailed account of a single ex- 

 periment will serve to illustrate their general character. 



