1888.] Magnetisation in the Dimensions of Rings and Rods. 407 



iron, observations being made of the changes which occurred in their 

 diameters under the influence of various magnetising forces obtained 

 by passing currents of electricity through coils of wire encircling the 

 rings. To remove the third objection the rings were demagnetised 

 before every observation, by a modification of the method described 

 by Professor Ewing in the ' Phil. Trans,' vol. 176, p. 537. And lastly, 

 the battery employed was increased from seven Grove's cells to thirty. 



After an explanation of the precautions taken to guard against the 

 effects of current heating, an account is given of some experiments 

 with three rings arranged in slightly different ways, and the results 

 are compared with those of an experiment made under similar con- 

 ditions with a straight rod. It was found that in their general cha- 

 racter the phenomena of elongation and retraction were just the same 

 in both cases, and were in close agreement with those of the former 

 paper. The differences in mere details were not greater than would pro- 

 bably be found to occur in different specimens of iron of the same form. 



Being satisfied that these curious effects of magnetism were prac- 

 tically independent of the form of the iron, and having regard to the 

 fact that it was much easier to obtain intense fields with straight than 

 with circular solenoids, the author thought it worth while to make 

 some further experiments with straight rods. The metals used in 

 addition to iron were cobalt, nickel, manganese-steel, and bismuth; 

 and the highest magnetising force reached about 840 C.G.S. units, 

 the maximum in the old experiments having been 290. 



It was found that the retraction of the iron continued to increase 

 with higher forces until it was finally as much as 45 ten-millionths of 

 the length of the rod, when there were indications that a limit was 

 being approached. The retraction of the nickel reached 113 ten- 

 millionths, when it also was evidently hot far from its limit. 



The behaviour of the cobalt rod was exceedingly curious and inte- 

 resting. ~No evidence of any change of length appeared until the 

 magnetising force exceeded 30 or 40 units. Then the length of the 

 rod began to diminish, and continued diminishing until the force was 

 about 400, when the retraction amounted to 50 ten-millionths. Bat 

 beyond this point the rod gradually became longer again, and the 

 retraction with the highest force of 800 units was only three-fifths of 

 its maximum amount. It was ascertained that the maximum retraction 

 did not coincide with a maximum of magnetisation, as might have 

 been suspected to be the case. It is suggested that iron and nickel 

 might possibly behave in a similar manner under sufficiently high 

 magnetising forces.* 



* It is also suggested that some speeimens of cobalt and nickel might, like iron, 

 begin with a small preliminary elongation, thus accounting for Professor Barrett's 

 observation that cobalt undergoes elongation when magnetised (' Nature,' vol. 26, 

 p. 5S5). 



2 G 2 



