PHYSIOLOGY OF CHROMATOPHORES OF FISHES 571 



The preservation of the irritability of the chromatophores in 

 sea-water is in keeping with Loeb's view of a physiologically 

 balanced solution (Loeb '03, p. 409). Upon dilution of the sea- 

 water the specific NaCl effect appears in an expansion of the 

 melanophores. Loeb has recently found ('12) that the toxicity 

 of ions varies with the concentration of the solution. Thus, for 

 example in small concentrations the protective action of Ca++ is 

 greater than that of Na+ but from a definite concentration on, 

 Na+ becomes less harmful than Ca++. The antagonism of the 

 constituents of sea-water may vary in a similar way which would 

 account for the appearance of the Na expansion in dilute sea- 

 water. 



Loeb ('10) found that a solution of KCl of the same concentra- 

 tion as in sea-water, killed Fundulus embryos in two days though 

 they lived indefinitely in a NaCl solution of the sea-water con- 

 centration. Toxic KCl solutions were rendered harmless by the 

 addition of NaCl in definite quantity. This antagonism was 

 between the cations Na+ and K+ and not between K+ and Cl"~. 



More recently ('11) Loeb has determined quantitatively the 



KCl 

 'Entgiftungskoeficient' of AT~pi for several concentrations and 



has found it to be approximately 1/17. He states (p. 469) that 

 some investigators (W. Koch and A. P. Matthews) still claim 

 that the antagonism is always between ions of opposite charges. 

 In this case Loeb considers the antagonism to be between the 

 cations. 



Koch ('09, p. 434) states that Loeb overlooked the importance 

 of the anion action when he first discovered the Na-Ca antago- 

 nism. The foregoing experiments upon the chromatophores in 

 salt solutions seem to furnish evidence in favor of both views. 

 The neutral series of K salts showed two constant reactions; (1) 

 a contraction and (2) a cytolytic degeneration. The time of the 

 first contraction was the same for isionic solutions of the five 

 neutral salts (p. 547). The time required for the beginning of 

 the degeneration varied with the anion. Since the number of 

 K+ ions in all five solutions was identical this result suggested 

 a specific cation (1) and anion (2) effect. It furthermore sug- 



