280 K H. Holly— Effect of Magnetic Force on the 



have in the second. In view of the great difference between 

 the E. P. of soft iron and that of tempered steel, the latter 

 being, in some cases at least, about four times as great as the 

 former, the difference in the behavior of these two specimens 

 of cobalt should not, perhaps, excite surprise. 



An experiment was made to determine how much of the 

 effect produced upon the equipotential lines in this more 

 recently used cross by a magnetic field of about 9000 would 

 survive the removal of the cross from the field. It appeared 

 that this permanent effect was rather less than 1 per cent of 

 the temporary effect and in the same direction. 



JViekel. — Thin Cross ; field of varying strength : The speci- 

 men of nickel that gave the quantitative results relating to this 

 metal which I have published in previous papers, was a piece 

 of commercial nickel plating stripped from the surface upon 

 which it had been electrolytically deposited. I had now at 

 hand a thin sheet of rolled nickel quite different in character 

 from this specimen, much softer and more pliable. Little is 

 known concerning the purity of the first specimen. The 

 second, which like the cobalt bar already mentioned was given 

 me by Mr. "Wharton, is probably much the purer of the two, 

 although it shows a trace of iron. The thickness is about 

 Q.gmm p rom this sheet a cross was made, of which the dimen- 

 sions were approximately as follows : 



TMckness Width of Length of 



between Arms. Main part. Arms. Main part. Arms. 



.QQgmm -jT.^mrQ qmm 28 mm >rmm 



With this cross : 



Date. Temp. M. R. P. x 10*. 



Aug. 12. 1885 22-4° / . _ 3595 ) * —9121/ niHH 



Aug. 12,1885 226° r 22 " 5 3533 f 3 ° 64 -9234 \ ~ 9m 



Aug. 13, 1885 22*4° 9152 —5776 



The direct current through the cross was about 0*1 absolute 

 unit. 



It appears that the R. P. of this specimen of nickel is about 

 two-thirds or three-fourths as great as that of the piece pre- 

 viously used, but the uncertainty which exists regarding the 

 thickness of each prevents any accurate conclusion upon this 

 point. The diminution of P. P. when M increases from mod- 

 erate values to higher values is very marked in each. It is 

 somewhat greater in the later specimen than in the earlier, in 

 which there was, according to the experiments of 1881, a fall 

 from about 12500 with M = 3600 to about 8400 with M=9100. 

 This diminution, like that noted with the thin cobalt and in 

 certain cases with iron, indicates a failure of the transverse 



