﻿38 Dr. Ashworth on Magnetic Temperature Coefficients 



annealed, magnetized, and fixed inside a stout copper tube 

 " end on " to the npper pair of poles of the astatic system. The 

 temperature, which could be raised to 800° C. by a row of 

 broad Bnnsen burners, yielding a continuous sheet of flame, 

 was observed by means of a therm o-juncti on of platinum and 

 platinum-iridium, constructed by the Cambridge Scientific 

 Instrument Co. The constant of the junction was verified 

 and found correct. Thin wires of iron containing 0*4 per 

 cenf. of impurity, and of nickel with less than 2 per cent, of 

 impurity, were also tested with this apparatus, and the results 

 agreed closely with those of the first series. 



The general course of an experiment was to heat and cool 

 the magnet several times until a cyclic condition was reached, 

 and then to take a careful set of simultaneous readings of 

 magnetic intensity and temperature as the tube was slowly 

 and continuously heated to the highest temperature required, 

 and then allowed to cool. After this the magnets were re- 

 magnetized and the experiment repeated for the sake of 

 confirmation. The arrangements were such that the obser- 

 vations of temperature covered a range of 45 cm. for 800° C. 

 and the readings of variation of magnetic intensity extended 

 over 7 -5 to 20*0 cm. 



3. The results which have been arrived at are exhibited 

 in Table I. and in fig. 1 (p. 40). In the first column of the 

 table is given the temperature in centigrade degrees, and in 

 the second column this temperature, converted into absolute 

 degrees, is expressed as a fraction of the absolute critical 

 temperature. In the third column is given the magnetic 

 intensity calculated as a fraction of the intensity under the 

 given conditions at absolute zero. The figures in the second 

 and third columns have been carried to the third decimal 

 place so that the second may be correctly assessed. 



From the rate of change of intensity at ordinary tempe- 

 ratures the maximum intensity at absolute zero has been 

 estimated, but there are experiments by P. Weiss and 

 H. Kamerliugh Onnes * which show that the change of 

 intensity from 0° C. to the temperature of liquid hydrogen 

 under very strong forces is 5*5 per cent, for nickel and 2 per 

 cent, for iron. I have adopted these values for nickel and 

 iron, which are nearly the estimated values, and 1*5 per cent, 

 for cobalt. 



The critical temperature is taken to be that point at which 



the magnetic intensity is varying most rapidly with change 



of temperature. This I determined for nickel to be 388° C, 



and for iron 785° C, and I have adopted 1090° C. as the 



* Journal de Physique, ser. 4, vol. ix. p. 555. 



