58 Mr. S. R. Milner and Prof. A. P. Chattock on 



water thickness the balancing-coil was generally arranged to 

 he a little higher in temperature than the heating-coil (i. e. 

 c?! is negative, and the two parts of the expression are of 

 contrary sign). 



(4) The expression 



°^r/3C 2 R 



in equation (7) gives the effect due to the alteration of 

 thermal resistance of the glass and paraffin bottom of the 

 heating-coil when the temperature of the coil rises. To 

 obtain the values of r and ft the water was replaced by 

 mercury (for the sake of good thermal contact) and the upper 

 vessel containing the heating-coil let down to rest on the 

 bottom of the lower one (B). With water flowing through 

 the false bottom of this readings were taken on the pile 

 JK (fig. 1) with widely altered currents through the heating- 

 coil. An estimate of the values of r and ft could thus be 

 obtained, due regard being had to the temperature-coefficient 

 of the heating-coil, and to the fact that the temperature 

 alteration of the glass and paraffin is only half that of the 

 coil. The mean results were 



r =218 CG.s. 



/3 = 034 per cent, increase per degree rise of temperature. 



Introducing these values the percentage correction in k 

 becomes 



0-0047 8, 



amounting in the actual experiments to about 0'8 per cent. 

 As a matter of fact the value of ft determined in this way 

 does not represent the change in r due to a change of 1° in 

 the temperature of both glass and paraffin, but is somewhat 

 nearer to the value for paraffin alone, as the temperature of 

 the latter is changed in the above measurements more than 

 that of the glass. The smallness of the total correction, how- 

 ever, renders a more detailed consideration of this point 

 unnecessary. 



(5) Since the temperature of the sink and the flow of heat 

 remain constant when the water thickness is increased, the 

 fresh layer of water introduced, the conductance of which is 

 that measured in the experiments, must be taken as that 

 immediately below the bottom of the heating-coil vessel. Its 

 mean temperature is therefore 



