BAXTER. REVISION OF THE ATOMIC WEIGHT OF IRON. 255 



will have no attractive iufluence. Moreover, since the pure metals used 

 in the atomic weight investigations upon iron, cobalt, and nickel in this 

 laboratory were incapable of retaining magnetism, the earth's field will 

 not have even a directive iufluence on these substances, but will only 

 aflfect the polar direction of the weak induced magnetism. 



If any further argument were needed as to the infinitesimal nature of 

 the effect of the earth's magnetism upon the atomic weights in question, 

 it might be found in the analytical results. Compounds are well known 

 to be usually far less magnetic than the metals which they contain * ; 

 hence the magnetic attraction would increase the weight of the metal to 

 a greater extent than that of the compound. But the complete analysis 

 of the cobaltous and nickelous bromides accounted for all the material 

 taken, without any considerable surplus, even although the metal was 

 weighed in the elementary state. Hence, one concludes that the mag- 

 netic attraction must have been unimportant. Still, in order to prove 

 the point directly, the following experiments were carried out. 



The method of determination consisted in weighing iron in a normal 

 fashion in the earth's magnetic field, and then weighing the iron again in 

 a non-magnetic field, formed by neutralizing the earth's magnetism by 

 means of a temporary magnet. 



The vertical component of the earth's magnetism was, of course, the 

 one which was to be eliminated ; but if the magnet is placed in the line 

 of the dipping needle, an intensity which would neutralize one component 

 would also neutralize the other. Hence, the position of neutrality was 

 determined by a horizontal needle or compass. 



Upon the pan of a very sensitive balance was placed a small sensitive 

 magnetic needle. The needle was then brought into an astatic condition 

 by bringing near it one pole of a straight electro-magnet, thirty centi- 

 meters long, which was placed in a position as nearly as possible parallel 

 with the lines of the earth's maijnetic force. This electro-magnet con- 

 sisted of a core of soft iron wire covered by six layers of stout copper 

 wire. The magnetic attraction of the earth and of the electro-magnet 

 neutralized each other when the distance of the electro-maijnet from the 

 needle was eighteen centimeters, and the current passing through it was 

 0.15 ampere. 



The magnet was now removed and the compass replaced by a weigh- 

 ing bottle containing 6.7 grams of pure iron (reduced from the oxide by 

 ignition in a current of hydrogen), which was then carefully counter- 



* For example, see Pliicker, Pogg. Ann., 1848, 1851. 



