and Retractions of Rods of Iron and Steel. 



185 



the poles of a strong electromagnet. The heating effects were 

 indicated by a thermo-electrical couple. Cobalt and nickel gave 

 similar results, but less marked; whilst non-magnetic metals 

 were not heated in the same circumstances." I have made 

 many experiments on a tube of iron, weighing two hundred- 

 weight, which confirm these results; the experiments will be 

 given in Part III. of this memoir. 



I here present two Tables of experiments on rod No. 2, of 

 Ulster iron. The successive discussion of these two Tables will 

 give to the reader a clear physical conception of the phenomena, 

 and serve to elucidate the account I have above given of the 

 heat developed on demagnetization. 



Table I. 



No. of 

 experiment. 



Scale, 



Scale, on 



Scale, on 







circuit 

 open. 



closing 

 circuit. 



breaking 

 circuit. 



Elongation. 



Retraction. 



1. 



376 



39-2 



38-0 



1-6 



1-2 



2. 



38-0 



39-2 



380 



1-2 



1-2 



3. 



38-0 



392 



380 



12 



1-2 



4. 



380 



39 2 



380 



1-2 



1-2 



5. 



38-0 



39-2 



38-0 



1-2 



12 



6. 



38-0 



392 



38-1 



1-2 



11 



7. 



381 



39-3 



381 



1-2 



1-2 



8. 



38-2 



39-4 



38-3 



1-2 



11 



9. 



38-3 



39-5 



38-4 



1-2 



11 



10. 



38-4 



39-55 



38-6 



115 



0-95 



11. 



38-6 



39-6 



38-fi 



10 



10 



12. 



38-6 



400 



38-85 



1-4 



115 



13. 



38-85 



401 



390 



1-25 



11 



14. 



390 



40-2 



391 



12 



1-1 



15. 



39-1 



402 



39-2 



11 



10 



16. 



39-2 



40-3 



39-2 



11 



1-1 



17. 



39-2 



40-4 



39-3 



1-2 



11 



In exp. No. 1 we passed the current for the first time round 

 the unmagnetized rod, and observed an elongation of 1*6 of a 

 division of the telescope-scale; immediately after the observa- 

 tion we broke the circuit, which had remained closed about five 

 seconds, and observed a retraction of 1*2 division ; the rod now 

 remained at a constant temperature for three hours, and the 

 scale-reading remained steady at 38'0 — thus showing that the 

 rod had received a permanent elongation of *4 of a division on 

 receiving its charge of residual magnetism. 



On repeating the experiment we find an elongation and retrac- 

 tion of 1*2 division, which is the quantity the rod retracted on 

 the first break-circuit. Experiments 2 to 5 inclusive give the 

 same result; but on the 6th and subsequent break-circuits we 

 observe a retraction less than 1*2; and this effect we attribute 



