THE CORROSION OF IRON AND OTHER METALS 327 



It is fairly obvious that the character and distribution of the 

 electro-negative impurities must be the chief determining factors 

 in corrosion : the sooner the problem is looked at and studied 

 from this point of view, the better it will be. 



Much is said in the Report referred to as to the " solution- 

 pressures " of the steels. The results obtained with sea-water 

 are said to be inconclusive (not sufficiently conclusive to warrant 

 any definite statement). 



It may be questioned whether, in such an inquiry, the con- 

 ception be not one that should be set aside altogether as at 

 most of purely academic interest, if it be not altogether mis- 

 leading. Some account must be taken of the infinite complexity 

 of the results obtained in dissolving materials which dissolve 

 mainly in local circuits ; determinations of the electro-motive 

 forces developed between such materials and a negative electrode 

 cannot well be interpreted in any simple manner — this at least 

 is the conclusion the writer came to years ago after spending 

 much time in determining the electro-motive forces developed 

 by copper alloys. 



It is noteworthy and interesting that Mr. Cobb, who bases 

 his opinion on his own observations, refers to manganous 

 sulphide as a non-conductor of electricity which itself undergoes 

 change but is without direct influence on the corrosion of iron. 

 In the Report of the Commitee referred to, manganous sulphide 

 is spoken of as being powerfully electro-positive and as tending 

 to cause serious electrolytic action when present in steel, thus 

 materially increasing the corrodibility. The probability is that 

 it acts indirectly because sulphuric acid is formed from it sooner 

 or later by oxidation. 



Mr. Cobb's experiments were made by exposing a small slab 

 of iron in opposition to a slab of the material to be tested in the 

 so-called feroxyl indicator which has played so considerable 

 a part in recent discussions on the corrosion of iron. The 

 indicator is made by neutralising a solution of gelatin — so strong 

 that it will just set on cooling — with caustic soda and adding a 

 little potassium ferricyanide and phenolphthalein. When such 

 a solution is electrolysed and the iron is attacked, the region in 

 which the attack takes place is indicated by the appearance of 

 Prussian blue, whilst the region in which alkali is liberated at 

 the electro-negative electrode becomes pink. 



Mr. Cobb describes an experiment in which a slab of iron 



