STEEL, NICKEL, AND COBALT TUBES IN THE MAGNETIC FIELD. 477 



Assuming then that the applied magnetic force strains the substance at every stage 

 against the molecular forces which determine the elasticity of the substance, we find 

 that the amounts of work done per unit volume in so straining the two metals in the 

 three fields named are respectively : — 



5*3, 10'8, and 16'7 ergs in iron. 

 327 , 595 , „ 766 „ „ nickel. 



A similar calculation applied to the Iron and Nickel Tubes B II., which are much 

 thicker in the walls than B V., gives for the amounts of work done per unit volume in 

 straining the substance by the magnetic forces 200, 300, and 500, the quantities 



6*5, 17, and 27*2 ergs in iron. 

 385 , 730, „ 1114-6 „ „ nickel. 



From both these examples, so very different in the details of the strains, we find, by 

 taking the ratios of the corresponding pairs, that 60 to 40 times as much work is done in 

 straining nickel as is done in straining iron by means of magnetic forces whose values 

 lie between 200 and 500. 



Applying a similar calculation to the coiled Tubes C II., of which the iron tube is, 

 however, somewhat thinner in the wall than the nickel tube, we find for the amounts 

 of work stored up in unit volume in Fields 60, 180, 300, and 500, 



12-1, 24-1, 32'3, 66 ergs in iron. 

 997, 596 , 818-4, 961 „ „ nickel. 



These quantities show that in a field of 250 the work done in straining nickel 

 against the elastic stresses is about 27 times the corresponding work done in straining 

 the iron. Because of the different thicknesses of the tubes, this comparison has not the 

 same claim to attention as the other two. 



The assumptions underlying the calculations that have just been made are, of course, 

 open to criticism. 



The magnetic force acts directly on the molecular groupings, which break up to form 

 new configurations. The mutual action between every contiguous pair of these new 

 configurations is in all probability the effective cause of the strains produced. We 

 know nothing definite regarding this mutual action, but it is not an altogether 

 unreasonable hypothesis to suppose that the strains are produced against the elastic 

 forces which bind the molecules together. It is, at all events, a fact worthy of note, 

 that the less susceptible material is the one on which the greater amount of molecular 

 work is done. 



I have made no attempt to bring these experimental facts into line with the theories of 

 magneto-striction as developed by Helmholtz, Kirchhoff, J. J. Thomson, and others. 

 Cantone's calculation* of Kirchhoff's constants from the results of experiment on 

 iron and nickel ovoids can hardly be regarded as a test of the applicability of Kirchhoff's 



* Atti d. R. Accad. d. Lincei, vi. 1890. 



