154 



Mr. T. Andrews. 



[May 3, 



Table B. The cast-steel bars were employed in some of the experi- 

 ments, because after magnetisation in the coil their subsequent in- 

 fluence as permanent magnets of lower strength could be observed as 

 referred to in course of this memoir. 



Table B. — Physical Properties of the Metals. 



Description. 



Contraction of 

 area at fracture 

 per cent. 



Extension 

 per cent. 



Breaking strain 

 per square inch 

 of original 

 area. 



Small iron bars (Wortley 







tons. 





23 



2-5 



45-82 





22 



2-0 



55-42 



Large iron bars (Wortley best 











28 



24-0 



24-46 





24 



20-0 



45-81 



The small iron and steel bars were drawn through a wortle, the 

 large iron and steel bars were rolled rods. 



The chemical reagents employed as electrolytes consisted of solu- 

 tions of bromine, ferric chloride, and chlorine- water, ferrous sulphate, 

 ferric chloride, cupric chloride, cupric sulphate, cupric nitrate, cupric 

 acetate, cupric bromide, nickel chloride, hydrochloric acid, nitric acid, 

 and potassium chlorate. 



In the experiments with the smaller rods a pair of bars in each 

 experiment were immersed in the solution in the U"^ u ^ e ? m circuit 

 also with a delicate galvanometer, and after a suitable time had 

 elapsed in every case for normal galvanic equilibrium to obtain, the 

 bar A in the coil was magnetised, and the magneto-chemical effect 

 recorded. It was found to vary with the nature of the metal and 

 solution employed, and also with the extent of the magnetic saturation 

 of the metals. The strength of the magnetism was practically the 

 same in many of the experiments, and it was generally observed that 

 difference in the strength of the solutions affected the results. In 

 other experiments with a uniform strength of solution, but in which 

 the magnetism of the metals was varied or reduced, the magneto- 

 chemical effect became proportionately altered. The possibility of 

 error from temperature causes arising from any slight internal heat- 

 ing of the coil has been referred to and dealt with in Part I (' Roy. 

 Soc. Proc.,' vol. 42, pp. 462-3). The apparatus, fig. 3, used in some 

 of the present experiments was also conducive to accuracy in this 

 respect. Moreover, the early and extensive development of the 

 magneto- chemical effect, noticed in most cases, especially in the 



