56 Mr. S. K. Milner and Prof. A. P. Ohattock on 



Values of the Corrections. 



(1) The expression 



^(T + 0-18) 



is the correction for the variation of heating-coil resistance 

 due to variation of the temperature of the experiment. is 

 measured on the thermopiles J, K for each vessel, and a, has 

 the value 0*00333 given above ; hence if 6 X denotes the actual 

 value of the pile readings on the galvanometer before the 

 increase of thickness the percentage value of this correction 

 becomes — 



E.H. vessel +0-1400, + 0-272(1-18), 

 L.H. „ +0-148 6> 1 + 0'272(T-18). 



(2) When the water thickness is increased, the temperature 

 of the heating-coil rises by 6' — 6, and the flow of heat is altered 

 by the consequent increase of resistance. The percentage 

 correction in the conductivity is given by the third term in 

 the brackets of equation (7), or 



R.H. vessel +0-156 0/, 

 L.H. „ +0-164 0/, 



where 0/ is the galvanometer-reading of the piles J, K for 

 each vessel respectively after the increase of water thickness. 

 The average value of the sum of corrections (1) and (2) 

 amounted to about 2'7 per cent. 



(3) In consequence of the length of time required for the 

 temperatures to become steady, it is impossible in practice 

 always to adjust the temperature of the balancing-coil to 

 equality with that of the heating-coil. A sufficient current 

 to produce an approximate equalization of temperature was 

 therefore sent through the balancing-coil at the beginning of 

 tin experiment ; and this was not usually altered afterwards. 

 By measuring the temperature-differences d and d' before 

 and after the increase of thickness the heat flowing upwards 

 can be calculated in terms of the upward conductance (k r per 

 square centimetre). 



To determine k' the water layer in each vessel was replaced 

 by a layer of cotton-wool about half an inch thick, and a 

 balance obtained on the bridge-wire with a small heating- 

 current c, the corresponding values of and d being also 

 measured. A current was then passed through the balancing- 

 coil ; and when the temperatures had again become steady, 

 fresh values 6' and d f were obtained. 



