relation to Time and Temperature. 465 



acid at a temperature below the zone of instability so as to induce 

 passivity in the first instance, the test-tube containing these 

 ingredients may be plunged into a heated water-bath, with the 

 assurance that there will be no lapse of passivity until the critical 

 point is reached. The behaviour of sheet iron well illustrates this 

 point. If the metal and the acid are heated separately to a 

 temperature as low as 165° F., and then brought together, violent 

 ebullition as a rule takes place at once. On the other hand, when 

 the alternate method is adopted, passivity even at a temperature 

 of 200° F. is assured. It seems therefore that metallic passivity 

 (especially in the case of sheet iron), when once established, 

 exercises a definite restraining influence upon the premature 

 incidence of a crisis. 



In the chart accompanying this paper an attempt has been 

 made to show from actual experiments the time required for the 

 solution of a metallic unit of the three metals, iron, nickel, and 

 cobalt, at any temperature from ten degrees Fahrenheit to the 

 boiling-point of water. Although in the case of iron and nickel 

 the shortness of the time-unit makes the curves incomplete, their 

 general trend is shown. The practical difficulties of pursuing this 

 investigation more fully than has been done were so numerous 

 that it became futile to extend it with the apparatus at hand. 

 I therefore decided to publish it provisionally, and it must be 

 accepted in this light. 



Finally, a few remarks may be made regarding the apparent 

 anomalies observable in these experiments on metallic passivity. 

 First, it is interesting to observe that all the crises fell — and this 

 statement includes copper — well within the critical temperatures 

 of water, between 32° F. and 212° F. This I ascribe to the action 

 of the fuming nitric acid. Future observations upon other metals 

 may confirm the statement. Secondly, it seems probable that the 

 purity of the metal investigated affects the critical temperature. 

 This contingency must specially affect the three metals I have 

 selected, as they are often alloyed with each other. In the former 

 paper the analyses of the sheet and Mond nickel are given. The 

 rolled cobalt contained approximately Co 95 °/» The chief im- 

 purities are iron, nickel, carbon, etc. As regards the ferrotype 

 plate there is unfortunately no assay at hand. Thirdly, the 

 brittleness of the samples of rolled cobalt and its crystalline 

 fracture so far correspond structurally with the grain nickel. 

 Each has a comparatively low critical point also. 



VOL. XII. PT. vi. 31 



