LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 83 



region. In eyeless Amblystoma Laurens (1917) has shown 

 that the melanophores contract in light and expand in 

 darkness. It may then be suspected that the mechanism 

 by which the changes in the melanophores are controlled 

 is more complicated than we know. The slowness of these 

 changes and the fact that they may not be wholly medi- 

 ated through the eye make it difficult to draw conclusions 

 from them with regard to color vision, and especially diffi- 

 cult to compare such conclusions with those derived from 

 training experiments. In the one case we have a slow 

 reaction to background, while in the other we have, if 

 my results be accepted, instantaneous choice of one of 

 two stimuli which differ only in quality of light. In the 

 latter case we may speak of color vision, or wave-length 

 discrimination. In the former it is doubtful whether 

 either discrimination or vision is involved. The choice by 

 Mast's flounders of a background of the same color as that 

 to which they had been previously adapted affords evidence 

 more nearly comparable to that obtained by training. 



B. Sources of Error 



From the foregoing review of the literature it appears 

 that workers have used many methods of experimental 

 attack on the problem of color vision in fish. All workers 

 except Hess conclude that fish have the capacity to dis- 

 criminate wave-lengths of light. My own results support 

 this conclusion., but before discussing them further I shall 

 take up, on the basis of my own experience and from a 

 study of the literature, the sources of error involved in 

 such work. 



a. Errors may arise from lack of proper environment and 

 care. — The need of keeping the fish in normal physiological 

 conditions has already been referred to. The chemical 

 contents of the water, the size of the containers (Franz, 

 1910, 1911), the tameness or wildness of the fish, are all 

 factors to be considered in judging of the effectiveness of 

 environment in modifying appearance and behavior. 

 Through neglect of these factors Hess seems greatly to 

 have lessened the value of his experiments. He tells us 

 (1913) that while he was experimenting two of his twelve 



