442 Mr. R. S. Willows on the Variation of ilie 



this interval showed that it remained steady at the same 

 temperature, and showed no tendency to increase. 



As in the case of the 4'8 per cent, and 9*5 per cent, speci- 

 mens, the variation of the resistance with temperature is 

 different from that indicated by fig. 3 if the cycle is taken 

 immediately after the spiral has been heated several times. 



In fig. 6 A gives the curve obtained from an amalgam con- 

 taining 33'3 per cent, zinc, the cycle being taken just after 

 the spiral has been heated three times ; curve B was obtained 

 when the spiral had been allowed to stand for some weeks. 



If the resistance of the amalgam at 15° be represented by 

 R , and if the resistance at any other temperature be assumed 

 to be given by the simple formula H = R {l-\- at), then a 

 reference to curve A in each of the figs. 1, 2, and 6 shows 

 that the consequence of adding more zinc is to increase the 

 value of a for temperatures between 15° and 40°, for in 

 fig. 1 a is negative, in fig. 2 negative but numerically less 

 than in fig. 1, in fig. 6 it is positive. 



It is seen from fig. 6 that the effect of repeated heatings 

 is (a) to make the more or less sudden change at the tem- 

 perature corresponding to P to disappear, (b) to decrease the 

 temperature at which the sudden fall in resistance occurs, 

 (c) to decrease the amount of this fall. 



When this fall in resistance has taken place the curves A 

 and B are nearly coincident over the remainder of the cycle. 



Fig. 6. — Resistance of au amalgam containing 33 '3 per cent, of zinc. 



c __, 



f=- 





- 



u--~ "'// 





~A ~\ 





\ y 's 





I 





>^c*\ 



^ 



^zs^^^ 



Temperature 



,s « 35 <f5 ot> 65 75 85 95 



Experiments were next made to test the reversibility of 

 different parts of the curves. In fig. 6 C shows the curve 

 obtained when the specimen was heated to 34° and then 

 cooled. The resistance quickly takes its value when the 

 bath is cooled, and the final resistance is nearly the same as 

 if a whole cycle had been taken. 



In fig. 3 D shows the result when the highest temperature 

 reached is somewhat lower than that at which the fall in 

 resistance occurs. During the cooling the resistance scarcely 

 undergoes variation. 



