.RALPH S. LILLIE 135 



of the instrument, the scraped wire becoming more negative, indicat- 

 ing the formation of an anodal area. The range of this excursion 

 is increased by a rapid succession of scrapes, and when a critical point 

 is overpassed a propagated activation wave is started and the whole 

 wire is activated. The reaction thus produced is permanent in acid 

 of specific gravity 1.20 or lower; with higher concentrations it is tem- 

 porary, as already described. The return of the potential to zero 

 after its disturbance by the local destruction of the film evidently 

 indicates the formation of a new film at the scraped area. Any local 

 disruption of the film renders that area anodal, by exposing the under- 

 lying iron, and hence subjects it automatically to the anodal oxidizing 

 influence. This at once reforms the film, unless the free metal is 

 exposed over more than a certain critical area, in which case the 

 active area spreads instead of becoming obliterated and the whole 

 wire becomes active. Slight local disruptions, however, are at once 

 repaired and the continuity of the film is in this manner automatically 

 preserved. Hence the passive state is a stable one in solutions of 

 sufficient oxidizing power. 



It should be noted that this automatic tendency for the film to reform 

 explains why the rapid removal of a considerable area of film is neces- 

 sary for mechanical activation, and also the need for a brief interval 

 between successive scrapes in the summation phenomena just 

 described. It also explains the greater effectiveness of rapidly increas- 

 ing as compared with slowly increasing currents in electrical activa- 

 tion. ^^ Otherwise the film may be reformed as rapidly as it is 

 removed, and the local effect remain insufficient to start a wave of 

 activation. , 



We infer therefore that the condition of equilibrium which the 

 passive metal reaches eventually when immersed in nitric acid is one 

 in which a thin continuous layer of oxygen compound 1 molecule in 

 thickness covers its entire surface. There is nothing to prevent a 

 thicker layer from being thinned by solution until it reaches this 

 limit, but any further removal is prevented by the automatic regu- 

 latory reaction just described. In this sense the preservation of the 

 passivating film in an oxidizing solution is the expression of a "dy- 



^0 LiUie, R. S., Science, 1918, xlviii, 57. 



