Mr, E. H. Griffiths and Miss Marshall on the 

 Table II. 



Temp. 



A9 for difference of 

 1 mean millirn. bridge-wire. 



o 



20 

 30 

 40 

 50 



0009046 

 0-009073 

 009101 

 0-009128 



By experiments repeated this summer we found that no 

 change had taken place in the value of the mean bridge-wire 

 millimetre. A slight change (particulars of which are given 

 in a subsequent section) had, however, occurred in the position 

 of the null-point, i. e. in the reading on the bridge-wire when 

 the temperature of the two thermometers was identical. 



6. Brief Description of the Method of obtaining the Value of 

 the Terms in the Expression 



Total heat = Q E « E + Q 8 f a + Sg-. 

 [ne) 2 X t E 



Wh'e — 



K x xJ 



The ends of the platinum-silver coil (immersed in the oil 

 surrounding the evapora ting-flask and spiral) were kept at a 

 constant potential-difference by means of the arrangements 

 described in Paper J, pp. 382-388. This potential-difference 

 was always some integral multiple (n) of the potential-differ- 

 ence of a Clark cell (<?). During the spring of this year the 

 Clark cells used were again compared with the Cavendish 

 standard (Rj), which has shown no signs of change since its 

 standardization by Lord Rayleigh in 1883 and by Messrs. 

 Grlazebrook and Skinner in 1891. 



The mean value of the whole set of 30 cells differs from R z 

 by 0*00004 volt only ; and although individual cells show 

 larger discrepancies than in previous years, their mean 

 potential-difference at 15° C. may be taken as 1*4342 volts 

 (see Paper W, p. 297). A number of these cells were always 

 placed in parallel arc : thus when n is given as 4. we were 

 really using 12 cells as four files of 3 each. The arrangement 

 for keeping the temperature of these cells at nearly 15° C. has 

 been described in Paper J, p. 385. During the period covered 

 by our experiments, some of the days were extremely hot, and 



