ACTION OF WATER ON IRON. 265 



tina is formed, which, on subsequent digestion in nitric acid, 

 becomes powerfully explosive ; boron here apparently playing 

 the part of the carbon in Mr. Faraday's compounds. 



30. The inflammable powders produced by Magnus, by reduc- 

 tion of the difficultly fusible metals by hydrogen, also connect 

 themselves with the subject. To complete our knowledge of all 

 these remarkable substances, with reference to the immediate 

 subject of this report, will need a careful and extensive series of 

 experiments. 



31. There have been but few observations made as to the 

 variations in composition of cast and wrought iron as regards 

 their acceleration or retardation of the action of solvent agents 

 upon it. It is not known at this moment with certainty what 

 properties'" should be chosen, in either cast or wrought iron, 

 that its corrosion may be the least possible, under given cir- 

 cumstances, when used in construction. 



32. Faraday found the alloys of most of the metals he tried 

 with steel much less acted on by moist air than steel unalloyed; 

 but he also discovered the remarkable circumstance, that a very 

 minute quantity of an alloying metal produced an increased ac- 

 tion of sulphuric acid on steel, within certain limits ; thus, ^i^ 

 of platina greatly increased the action of the acid on the steel 

 with which it was alloj^ed ; with from ^^^ to ^i^ it was power- 

 ful; with 10 per cent, of platina there was a feeble action; 

 with 50 per cent, of platina the action was the same as with 

 unalloyed steel ; and an alloy of 90 platina and 20 steel was not 

 touched by the acid. In these cases even acids of very weak 

 combining power, as oxalic, tartaric, and acetic, rapidly dis- 

 solved the steel. Of three possible modes of accounting for 

 this suggested by Sir H. Davy, Mr. Faraday justly chooses 

 that which supposes the platina in part forming a definite alloy, 

 and the remainder diffused through the substance of the steel ; 

 thus forming an indefinite number of voltaic elements. On the 

 first action of the acid some of the particles of platina are de- 

 nuded, and being strongly negative with respect to the rest of 

 the compound, aid its solution. Upon this ground the action of 

 the excess of platina, in reducing the action, is obviously de- 

 pendent upon the whole alloy becoming definite again. Solu- 

 tions of chloride of sodium did not act more rapidly on these 

 alloys than on steel alone. 



33. It has been long observed how little liable to tarnish or 

 rust native and meteoric iron are, which contain often as much 

 as 9*5 per cent, of nickel, and variable proportions of chrome 

 and cobalt. The following tables embody perhaps the whole of 



