418 



SCIENCE 



[N. S. Vol. L. No. 1296 



These series were carried out at different 

 times and witli different wires, and the 

 results are not strictly comparable. In gen- 

 eral, however, the rate of the activating pro- 

 cess, in the ease of each salt, is approximately 

 proportional to the concentration. A pro- 

 gressive alteration of the surface-film, at a 

 rate determined by the concentration of Cl- 

 ions, appears to underlie the effect. With the 

 above chlorides the nature of the cation seems 

 indifferent; but with the chlorides of noble 

 metals a retarding influence of the cation is 

 apparent; thus ■m/4 HgCU required 2 to 3 

 seconds for activation and m/8 HgCl^ 3 to 4 

 seconds. A comparison of the three chief 

 halides showed a decreasing velocity of activa- 

 tion in the order, CI, Br, I; in m/50 solution 

 the minimal times for activation were: N'aCl, 

 1 to 2 seconds; ISTaBr, 3 to 4 seconds; ISTal, 

 12 to 14 seconds. 



Other Salts. — Activation occurs much more 

 gradually in solutions of other salts. The 

 following results are typical for the minimal 

 activating exposures in m/2 solutions of 

 sodium salts : 



NaOH 10 to 30 seconds 



NOjCOa 2 to 4 minutes 



Na-jSO, 1 to 10 minutes 



Na-HPOj 3 to 5 minutes 



Na-acetate 80 sec. to 4 min. 



Na-tartrate 2 to 4 minutes 



Na-citrate 5 to 10 minutes 



NaNOa 5 to 20 minutes 



NaClOs 2 to 30 minutes 



NaCN . . 1 5 to 7 minutes 



The figures given represent the extremes of 

 different observations with wires of varying 

 sensitivity. Passivity is preserved longest in 

 solutions of nitrates and chlorates, but all 

 salts with terminal oxygen in the anion 



activate slowly. Why cyanide should act sim- 

 ilarly is difficult to say; the whole subject 

 should be investigated thoroughly. It is note- 

 worthy that in nitrate solutions no definite 

 relation between concentration and velocity 

 of activation was found; in all of the solutions 

 examined activity appeared only after the 

 lajjse of several minutes; and apparently its 

 precise moment of appearance depends largely 

 on casual conditions, especially irregularities 

 in the structure of the metal. The following 

 list gives the minimal times of exposure ob- 

 served in a series of solutions of NaNO,. 



m/2 



m/4 



m/8 



m/16 



m/32 



m/64 



m/128 



m/256 



m/512 



Active after 



10 minutes 



4 



>10 

 10 



In this series the least effective exposure 

 was 4 minutes, and there was no apjjarent 

 relation to concentration. Not infrequently 

 wires have been found passive after 15 or 20 

 minutes in solutions of NaNOg. In all prob- 

 ability the structural conditions at the surface 

 of the metal, which in such experiments are 

 determined by casual conditions beyond con- 

 trol, are a main factor determining the precise 

 rate of activation in any instance. An 

 analogy is offered by the case of rusting, the 

 rate of which depends not so much upon the 

 concentration of oxygen and the conductivity 

 of the solution as upon the special character 

 of the metal; this determines the number of 

 local couples to whose electrolytic action the 

 chemical change is chiefly due. Pure and 

 homogeneous specimens of iron rust very 

 slowly, even under the most favorable condi- 

 tions.^ 



In the above solutions the delay in activa- 

 1 Cf. Wallier, Cederholm and Bent, Jour. Amer. 

 Chem. Soc, 1907, Vol. 29, p. 1251 ; Lambert, Jour. 

 Cliem. Soc, 1910, Vol. 97, Trans., p. 2426, and 

 1912, Vol. 101, p. 2056; Dunstan and Hill, ibid., 

 1912, Vol. 99, p. 1835. 



