ACTION OP WATER ON IRON. 281 



" Insulation, or free communication with the substance of 

 one of the two metals, having scarcely any influence on these 

 sorts of reactions, as I have assured myself, and which also 

 conforms to the pi'operties (proprietes) of electric currents, I 

 have thought that these variations might result from magnetism, 

 acquired by the iron placed vertically ; and, to try if in fact it 

 had any influence on the decomposition of the salt, I plunged, 

 in this position (viz. vertically), a bar of soft iron in a tube of 

 glass containing a neutral solution of sulphate of copper, and 

 corked it. As scarcely any disengagement of gas agitated the 

 liquid, I saw that decomposition began towards the two ex- 

 tremities of the bar, advancing progressively towards its middle 

 point ; the two extremities and this point acting meanwhile on 

 a magnetic needle, as the two poles and the neutral point of a 

 magnet, and the poles changing by reversal in the usual manner. 

 The efi"ect of this magnetization appeared then to augment che- 

 mical action, and hence to diminish the quantity of copper de- 

 posited on the platina." 



Are we to conclude from this that magnetism modifies che- 

 mical action, or that chemical action is capable, under certain 

 conditions, of conferring magnetism on iron ? 



The subject is obviously in need of much further experiment, 

 and is one of interest in a general, as also in our particular view. 



68. It will be now necessary to state the nature and extent of 

 the experiments upon the large scale which have been instituted 

 at the desire of the Association, and aided by its funds. On the 

 very first consideration, it appeared important that those expe- 

 riments in which time was an element should be first of all put 

 in progress ; and of these the most important seemed to be, 



1st. To obtain experimentally a set of numerical results of 

 the relative rates of corrosion of all the difi'erent most im- 

 portant makes of British cast iron, exposed under the same 

 circumstances to sea water, and unprotected, except from 

 mechanical abrasion ; and a like set for fresh water, va- 

 rying the conditions in both cases as far as might occur in 

 practice. 



With this view coiiiplete sets of authenticated specimens of 

 cast iron from most of the principal iron-works in Britain have 

 been written for and obtained. A very considerable delay ne- 

 cessarily occurred in procuring these, and it was not until 

 about two months since that it was found practicable to com- 

 plete the collection, which numbers between eighty and ninety 

 specimens. 



These specimens were then all fused, at the works with 

 which the writer is connected, separately in crucibles, to avoid 



