Circular Magnetization on Magnetized Iron Wire. 423 



observations were reduced by means of the correction curve 

 obtained. The heating and magnetizing currents were 

 passed through suitable manganin resistances immersed in 

 oil, so that by observing the deflexion of the galvanometer 

 when connected successively to the terminals of the manganin 

 resistances and the specimen itself, the two currents and the 

 resistance of the specimen could be quickly determined at any 

 time with an accuracy of at least 1 part in 1000. By varying 

 the resistance in series with the galvanometer it was possible 

 to obtain accurately readable deflexions through a very wide 

 range of current and resistance. 



The Magnetizing Solenoid. — The solenoid was wound on a 

 thin brass tube about 70 centim. long, with an external 

 diameter of 2*23 centim. The insulation resistance was very 

 high, special pains having been taken to insulate each layer 

 with paraffin and paper. The winding was tested for uni- 

 formity by measuring each fifty turns during the process. 

 The length of the winding was 60 - 25 centim., containing 

 4079 turns of No. 24 B & S double silk-covered wire in four 

 layers, and having a resistance of 28 ohms at 15° G. This 

 gave a magnetizing field of over 300 C.G.S. with 100 volts 

 on the terminals. In order to dissipate the heat due to the 

 magnetizing current at high fields, an internal water circula- 

 tion was provided through an annular space formed by a 

 second concentric brass tube. 



The Mounting for the Iron Wire Specimen. — As the specimen 

 was enclosed in a vacuum-tube, and its resistance at each 

 temperature determined, a special form of mounting was 

 necessary. 



The iron-wire specimen was 0'127 centim. diameter and 26*1 

 centim. long, or a little over 200 diameters. Its ends were 

 fused to copper wires *040 centim. diameter and 10 centim. 

 long ; the diameter of the copper wire being chosen by trial 

 to give a uniform temperature throughout the whole length 

 of the iron wire when the heating current was passed through 

 the circuit. 



The ends of these copper wires were tin-soldered and 

 riveted to stout copper conductors wdiich were brought out 

 through spiral copper springs to the ends of the containing 

 tube. The two copper springs, whose function was to take 

 up the slack of the heated specimen, had each exactly the 

 same number of turns, and were wound oppositely so that the 

 direct effect (on the magnetometer) of the current circulating- 

 through them would be compensated. The whole was centered 

 and kept in place by brass washers which fitted the thin glass 

 containing tube. The glass tube just fitted the inner brass 



2 G2 



