Length by Magnetization in Nickel and Iron. 135 



changes of angle can be advantageously used in various 

 other researches *. 



The magnetizing current was supplied by Bunsen cells, 

 and its strength was measured by a Thomson Graded Gal- 

 vanometer. The galvanometer was gauged by means of a 

 deciampere balance. The current was always changed con- 

 tinuously by means of a liquid slide included in the circuit. 



A rough plan of the arrangement of the different parts is 

 given in fig. 4. 



The hysteresis accompanying the change of length in nickel 

 is not so complicated as in iron. I will first describe a few 

 experiments made with the former metal. A nickel wire, 

 19 "4 centim. long and 2*04 millim. thick, was carefully 

 annealed by placing it due magnetic east and west in a por- 

 celain tube, and heating it red-hot in a charcoal fire. The wire 

 was then placed in the V-groove of the apparatus, and inserted 

 in the solenoid. The strength of the current was gradually 

 increased, and, at convenient intervals, the corresponding 

 readings of the micrometer and galvanometer were noted, 

 till the magnetic field was 10*2 C.G.S. units. The contrac- 

 tion of the wire was at first very slow, but when the field 

 was about 8 the rate of change was greatly increased. 

 As the field w r as diminished the wire tended to return to 

 its former state, but not by its former course. There was 

 lagging, so that the wire, for the same strength of field, was 

 more contracted than when the field was on the increase. 

 In fact, when the field was diminished to zero, the wire 

 still remained contracted 38*2 x 10~ 8 of its original length. 

 When the current was reversed, the wire continued in its 

 tendency to recover its former length until the reversed field 

 became equal to 5. There the recovery stopped, and the 

 wire began once more to contract, and the contracted length 

 in field — lOl was nearly the same as that in field -f-10'2. 

 On decrease of current, the same succession of changes took 

 place. The changes are graphically represented in fig. 5 

 (PI. I.), where the course is in the order of the letters of the 

 alphabet. The curve c d efg h of a cycle from the highest 

 field and back to it is nearly symmetrical with respect to the 

 line of zero field, forming complete loops on both sides. 

 The measurements are given in Table I. at the end of the 

 paper. 



On reannealing and experimenting betw y een fields + 30, 



* I found lately that it is more advantageous to replace the prism p 

 and the mirror m by a small rectangular prism attached at m. The 

 positions of telescopes and collimators must be suitably changed. 



