PHYSIOLOGY OF CHROMATOPHORES OF FISHES 531 



ess, 3. That portion of the scale, 2, which normally lies within 

 the dermal pocket makes a convenient ''handle" whereby the 

 scale may be transferred from one solution to another. With 

 very little practice it was found that a number of scales could 

 thus be raised simultaneously, the epidermis serving to hold them 

 together. By removing these scales to the various salt solu- 

 tions, etc., any changes in the chromatophores could very readily 

 be observed under a low power of the microscope. In removing 

 the scales from the living fish and transferring them to the solu- 

 tions care was taken to handle them only as shown in 2, figure A. 

 In this way any mechanical stimulus or injury was entirely 

 avoided. Furthermore the scales were always placed face up 

 in the solutions, that is, in the same relative position which they 

 occupy on the surface of the fish. This was done to insure a 

 uniform oxygen supply. 



Although there are marked differences in the reactions of 

 scales from different fishes, by using adjacent scales from one 

 individual a satisfactory basis for comparative results was ob- 

 tained. But even in the same individual, scales from 'the 

 median, dorsal region vary in the reaction time of their chro- 

 matophores from those of the more lateral portions of the body. 



In many cases the experiments were started by immersing all 

 the scales in a solution of 0.1 M NaCl which causes a complete 

 and lasting expansion of the melanophores. This was done, 

 for example, in all of the trials with potassium and other salts 

 in which the first specific effect is a contraction of the melano- 

 phores. Whenever this was the case and a comparison was to 

 be made between the effects of the different members of a series 

 of salts (for example, the neutral salts of potassium) a preliminary 

 selection of scales was made. Only such scales were chosen for 

 comparison as showed approximately the same size, shape and 

 darkness (that is, number and expansion of melanophores). 

 This selection was easily made without the use of a lens, the 

 scales resting over a light surface. 



The fish average from 6 to 10 cms. in length. They were 

 kept in salt-water aquaria throughout the period of experimen- 



