1894.] on Scientific Uses of Liquid Air. 403 



ture, it had lost 50 per cent, of its original magnetic moment. Such a 

 magnet would apparently diminish in magnetic moment on cooling 

 and heating the first time the action was examined, but a repetition 

 of the process when the action of magnetisation and temperature 

 were strictly reversible might lead to an opposite conclusion. To 

 settle this question a series of experiments on the magnetic moment 

 of small magnets cooled to — 182° were carried out. Small magnets 

 from half an inch to an inch in length were made of watch-spring or 

 steel wire and were either used separately or in bundles • they were 

 fixed rigidly in a block of wood by means of copper staples, and in 

 this condition were easily clamped firmly in the field of a magneto- 

 meter. The cooling was effected by applying a cotton-wool sponge 

 of liquid air. The relative deviations of the magnetometer are 

 proportional to the magnetic moment of the magnet under the 

 respective conditions of + 15° and — 182° C. After the first cooling 

 the magnet is allowed to regain the ordinary temperature, and the 

 operation of cooling and heating is repeated three or four times. 

 The following table gives some of the results, and these may be taken 

 as typical of a large additional number unrecorded. 



Change of the Magnetic Moments op Permanent Magnets at + 15° and 

 - 182° 0. per cent, of the value at the beginning of each cycle, which is 

 always 15°. 



— 182° C. + 15° C. 



(1) Hard steel, 0*5 inches long and 0*4 inches diameter. 



First Cycle +0 -30 



Second „ +33 — 5 



Third „ +36 



(2) Soft steel. 



First Cycle +12 -28 



Second „ +51 



Third „ +51 



(3) Hard steel, 1*03 inches long, 0'4 inches diameter. 



First Cycle -24 -43-4 



Second „ +23 



Third , +23 



(4) Nine steel wires in bundle. 



First Cycle +12-5+3 



Second „ +38 - 2 



Third „ +33 



Tested four days after. 



First Cycle +50 



If the experiment marked (1) is examined we find cooling to 

 — 182°, in the first cycle produced no change of magnetic moment, but 

 that on heating to +15° C. the magnet had lost 30 per cent, of the 

 original strength. In the second cycle cooling increased the magnetic 

 strength of the magnet, in the condition in which it is left after the 

 first cooling by 33 per cent., and heating diminished it by 5 per cent. ; 

 whereas in the third cycle cooling showed 36 per cent, increase and 

 no loss in heating. It was only after three alterations of tempera- 

 tures from +15° to -182° O. that the magnet reached a steady 



