218 Dr. D. K. Morris on the Magnetic Properties 



One alteration, however, did take place, particularly during 

 the last annealing of specimen B, when a temperature of 

 1150° was reached and maintained for some time : the 

 resistance of the thermometer-wire rose (on this occasion as 

 much as 1 per cent.) ; though, as stated above, the form of its 

 resistance curve remained throughout unchanged. 



Messrs. Heycock & Neville * have repeatedly noticed such 

 changes, though slight in amount, in their resistance pyro- 

 meters ; and conclude that they were due to strains set up 

 by the permanent expansion of the mica frames on which 

 their wires were supported. It is possible that such increase 

 of resistance may also be due, as Prof. Weber has suggested, 

 to evaporation of the platinum at these high temperatures ; 

 none of the author's platinum wires retained, after prolonged 

 heating, their bright metallic surface f. 



In reducing the results, it has been assumed that this 

 alteration of resistance, which was generally very small, took 

 place entirely during the time that the ring was at its highest 

 temperature. 



(3) In order to satisfy the third condition, it was not only 

 necessary that precise measurements of resistance should be 

 possible (-^q per cent, represents more than \° at 1000°) , but 

 also that the resistance of the thermometer-wire and that of its 

 compensating leads should be measured very nearly simul- 

 taneously. 



To obtain the nett resistance from a single observation, the 

 following arrangement (fig. 1) was employed, which is 

 similar to, though not identical with, that used in Callendar's 

 platinum thermometers. 



The platinum thermometer- wire and leads formed one arm 

 of tho Wheatstone's bridge, and the compensating leads 

 (made exactly equal to those of the thermometer), plus a 

 portion AC of the slide- wire, formed a second arm equal to 



* Heycock and Neville, " On the Determination of High Temperatures 

 by means of Platinum Resistance Pyrometers," Trans. Chem. Soc. 1895, 

 p. 160. 



t Since the above was written, Mr. E. H. Griffiths, F.R.S., has kindly 

 drawn my attention to the fact that his own platinum thermometer wires 

 on mica frames, when heated for considerable periods at temperatures of 

 ] 200°-1400°, do not thus lose their former bright surface. It is there- 

 fore rather improbable that in my case the dull surface should point to 

 any serious evaporation of the platinum. I have in this connexion one 

 or two observations which I had not given as I was in doubt as to their 

 reliability : — The mean sectional area of the platinum thermometer wire, 

 as measured by the specific gravity method, diminished from '001182 

 cm. 2 to "001162 cm. 2 during use in Rings A and B ; the length increased 

 from 3T33 to 3P56 cm. during the heating in Ring B, and during this 

 same heating the weight of the wire did not alter. — I). K. M. 



