THE CORROSION OF IRON AND OTHER METALS 315 



impurities in the iron are more than sufficient to act in con- 

 junction with the oxygen as the depolarising electro-negative 

 element in the current of change. 



As highly purified zinc is scarcely attacked by acids in 

 presence of air, it would seem that in the case of this metal, 

 as in that of highly purified iron, oxygen cannot act as depolar- 

 iser in the absence of an electro-negative component in the 

 circuit to condition change. I am not aware that in experi- 

 ments such as were made by Reynolds and Ramsay with 

 highly purified zinc, referred to in the previous article (p. 643), 

 the attempt was made to exclude air. 



In presence of an electro-negative conductor such as mer- 

 cury, oxygen acts readily as depolariser of the hydrogen 

 displaced by the zinc. Traube's experiment is well known in 

 which a zinc-mercury alloy is shaken with lime water in a flask 

 containing oxygen ; the hydrogen displaced by the zinc is 

 absorbed and an amount of hydrogen peroxide produced which 

 is practically equivalent to the amount of zinc dissolved 

 (cf. Ber. dent. chem-. Ges. : 26, 1471, 1893). 



„ fOH O _ (0\ _ HO 



Zn + Ca loH + 6 = z Ho/ Ca+ H6 



In view of these conclusions, it is desirable to ascertain 

 whether acids act on highly purified copper in presence of oxygen 

 alone or whether, as in the case of iron and zinc, a relatively 

 electro-negative conductor be required to condition change. 



The question here raised is one of considerable practical 

 importance ; attention is called to it, however, chiefly in order 

 to emphasise the need of taking into account the part played by 

 the electro-negative component of the circuit in all cases in 

 which the corrosion of metals is concerned. It cannot be too 

 forcibly insisted that action only takes place when the presence 

 of an electro-negative component is assured and that action 

 takes place always and only in an electrolytic circuit. 



It is remarkable and a striking illustration of the truth of my 

 contention (p. 311), that Friend does not deal with the subject 

 of "galvanic action" until the fourteenth chapter of his book 

 is reached and that his treatment of the subject is vague and 

 unsatisfactory. Throughout the earlier chapters he distinctly 

 implies that iron will dissolve in acid even if pure ; and that 



