1887.] Electrochemical Effects on Magnetising Iron. 



463 



Moreover, the exceptional negative position of the magnetised bar 

 under the same temperature conditions, in the case of sulphuric acid, 

 affords evidence that the effect was not due to these temperature 

 causes. To more clearly demonstrate, however, that the results were 

 mainly due to the influence of magnetisation, and acting on the kind 

 suggestion of Professor G. G. Stokes, it was decided to make a duplex 

 series of observations. I accordingly devised the modified form of 

 apparatus fig. 2, which was intended to eliminate possible sources of 

 error from temperature difference, by keeping the U-tube surrounded 

 by a large volume of cold water during the experiments ; the solution 

 in the respective tubes being thus maintained under equal conditions. 

 In comparing the results obtained with the two forms of apparatus, 

 it should be borne in mind that, when using the apparatus fig. 2, the 

 cold water surrounding the U-tube would have a tendency to retard 

 the increase of the temperature of the solutions naturally arising 

 from chemical combination. Further, the coil in fig. 2 being at one 

 end of the long bar would be calculated to modify the magnetisation 

 of the other end of the metal in the solution, compared with its 

 action in fig. 1, where the coil almost entirely surrounded the bar : 

 hence in fig. 2, the coil was made somewhat larger to overcome this 

 to some extent, and the end of the bar B was shortened in expe- 

 riments with fig. 2, so that this bar would be less liable to be affected 

 magnetically by the external influence of the larger coil. I hope that 

 the confirmatory results obtained in the two sets of observations may 

 be considered as fairly satisfactory. 



VOL. XL 1 1. 



