230 



Profs. W. E. Ayrton and John Perry. [Mar. 21, 



found that the conductivity increased with temperature, the increase 

 being very rapid near the melting point, at which place, as we used no 

 auxiliary condenser, the rate of loss of charge was very high. To 

 measure the conductivity of melted paraffin, we employed a very 

 delicate reflecting astatic galvanometer of 20,000 ohms resistance ; 

 one Daniell's cell, through a resistance of 120 megohms, gave a deflec- 

 tion of 131*3 divisions on an ordinary scale, at about a metre distance. 

 The current from a battery having a measured electromotive force 

 47 times that of our standard Daniell passed through the paraffin 

 and the galvanometer. Full precautions were taken to avoid all leak T 

 age except through the paraffin, the deflection of the galvanometer 

 needle due to instrumental leakage alone being quite imperceptible. 

 The thermometers did not indicate truly the temperature of the 

 paraffin, being too high when the apparatus was heating, and too low 

 when it was cooling ; but as the cooling was very slow, the results 

 which are given in following table obtained during cooling, and which, 

 when plotted, give the curve JK (fig. 1, p. 231), may be regarded 

 as being fairly correct, for the observations seem to be all very con- 

 sistent with one another. 



The area of the upper plate was 56*25 square centims., and hence if 

 x is the specific resistance in megohms per cub. centim. of paraffin, 

 and if D is the deflection on the galvanometer scale, 

 Log x=8 -3374-log D. 



x 



Temperature. 



D. 



in megohms. 



90° *6 C. 



167*9 



1 *295 X10 6 



96° *8 



229 *9 



*946 xlO 6 



91° *0 



191 *6 



1 T35xl0 6 



84° *0 



151 *9 



1 *432 x 10 6 



80° *0 



132 *3 



1 -644 X10 6 



76° *0 



115 *9 



1 *876 xlO 6 



70° *2 



94*5 



2*301xl0 6 



65° *1 



78 -0 



2 -788X10 6 



60° *0 



64 *0 



3 *398 xlO 6 



55° *2 



55 *0 



3 *954 XlO 6 



50° -0 



43 *0 



5 *058 xlO 6 



In the above table we have given only every fifth observation. The 

 curve ML has for co-ordinates the temperature and galvanometer 

 deflection as the temperature increased. The points M, L, K, J, cor- 

 respond respectively with the temperatures 90° "6, 94° *7, 93° *1, 48° C. 

 In this and in the other curves we have employed such scales as have 

 seemed to us most convenient. In every case we give the true co- 

 ordinates of some points, and from these the scale may be easily deter- 

 mined. Ordinates measured parallel to the line OX represent tem- 

 perature, the temperature increasing towards O. 



