in Nickel and Cobalt. 159 



We now return to the consideration of nickel. 

 The original experiment with this metal had been made with 

 a specimen so irregular, that it had not been possible to deter- 

 mine the magnitude of the transverse effect except in the most 

 general way. The direction had been determined beyond ques- 

 tion. The specimen of nickel now employed, and with which 

 the results to be given were obtained, was quite different in 

 appearance and physical condition from the first specimen — 

 though it was obtained in about the same manner, viz. by 

 stripping off a piece of nickel plating from the metal upon 

 which it had been electrolytically deposited. The first spe- 

 cimen was very brittle, the second quite tough. The latter 

 was about -001 centim. thick. As to its purity hardly any 

 thing is known except what is told by its physical characte- 

 ristics. It is probably affected by all the impurities of ordi- 

 nary nickel plating. It contains very likely a little cobalt, 

 and perhaps a trace of iron. I understand, moreover, from 

 Professor Wolcott Gibbs, that nickel plating deposited in the 

 usual manner (i. e. from an ammoniacal solution) is much 

 affected in its physical properties by nitrogen in some way 

 retained by the metal. It would have been desirable, of 

 course, in all cases to work with pure metals ; but such were 

 not at hand, or easily obtainable in the proper form, and it 

 was not thought best to defer the experiments until pare spe- 

 cimens could be obtained*. 



The second specimen of nickel showed an effect of the same 

 sign as the first, and numerically greater than the effect which 

 had been observed in the specimens of iron and cobalt used. 



It now became a matter of great interest to determine 

 whether the transverse effect had really any connexion with 

 the magnetic properties of the metals. It was determined 

 therefore to make a series of experiments, keeping the primary 

 current through the metal as nearly as practicable always of 

 the same strength, but varying within wide limits the inten- 

 sity of the magnetic field. We should in this way ascertain 

 whether the transverse effect was simply proportional to the 

 strength of the magnetic field, or was related to it in some 

 more complicated manner. 



* This may strike some readers as unwise. It lias even been suggested 

 that the difference in behaviour of iron and nickel may be due to impuri- 

 ties in one or the other. This suggestion implies that the transverse 

 effect in these metals is so related to the magnetic properties that, as they 

 resemble each other in one respect, they should also in the other, but at 

 the same time admits that slight impurities, such as would certainly be 

 very far from reversing the magnetic property of either metal, may reverse 

 tlie transverse effect in the same. This does not seem probable, 



N2 



