PHYSIOLOGY OF CHROMATOPHORES OF FISHES 551 



able as regards their initial contraction the difference in time re- 

 quired for the secondary degeneration was less obvious. Exper- 

 iments of the following type were carried out with a view to 

 clearing up this question. 



August 30, 1912 



12.25 P.M. Scales from a 7 cm. female were placed in 8 cc. of 0.1 M KCl 

 and NH4CI. The melanophores contracted in the KCl 

 sooner than in the NH4CI as in previous trials. 

 1.50 P.M. Both sets showed expanded melanophores. 

 3.00 P.M. Degeneration had begun in the KCl. Great expansion but 

 no degeneration appeared in the melanophores of the 

 NH4CI set at this time. 



From these experiments it eventually appeared that the alka- 

 line cations fall into a regular series as regards (1) their power to 

 contract the melanophores and (2) their susceptibility to degen- 

 eration. This series is 



K>NH4>Li>Na 



KCl contracts the melanophores in the shortest time and in this 

 solution they likewise degenerate most rapidly. NaCl expands 

 the melanophores and they may remain irritable in such a solu- 

 tion for more than twenty-four hours. 



Recent experiments with the additional chlorides of caesium 

 and rubidium have shown clearly that the Rb+ cation is practi- 

 cally identical with K+ in its initial effect upon the melanophores 

 although its tune of contraction is shorter than NH4+. Similarly 

 NH4+ and Cs+ are almost indistinguishable although clearly 

 separated from Li+. In the hope of separating the RbCl from 

 the KCl effect, experiments were carried out at reduced temper- 

 atures. Sunilar trials were made with NH4CI and CsCl solu- 

 tions. None of these attempts succeeded in establishing beyond 

 a doubt the positions of the cations Rb+ and Cs+ in the alka- 

 line series. 



Thus, as regards their contraction effect upon the melano- 

 phores the alkaline cations follow the order. 



(A) K & Rb>Cs & NH4>Li>Na 



