384 Messrs. Wright and Thompson on the Determination of 



produce two of the third, is given by the equation 



TEL=h l +h 2 --U d . 



In order to obtain the data for the calculation of the various 

 values of H in the case of the solutions examined in the pre- 

 ceding pages, a number of calorimetric determinations were 

 made, substantially in the way described by Julius Thomsen 

 (Thermochemische Untersiichungen, i. p. 18, 1882), using a 

 calorimeter containing 1 litre of fluid in the lower compart- 

 ment. The quantity of salt employed was usually close upon 

 -^ gramme-molecule, the weight being arranged so that the 

 resulting fluid was of the strength MSO 4 400H 2 O, or 

 MCl 2 400H 2 O. The thermometers used were graduated in 

 millimetres, about 11 millimetres representing 1° C; so that 

 readings being made to ±-go nrillim., the thermometer-error 

 of 1° C. As determined by means of' 



was 



— 220 



various 



blank experiments (with water only intermixed), the limit of 

 accuracy of each experiment was within ± 10 gramme-degrees; 

 so that each observation might be erroneous within the limits 

 of not exceeding ±70 gramme-degrees per gramme-molecule; 

 but as several observations were usually made and averaged, 

 the probable error of each average was ordinarily within + 25 

 gramme-degrees. The following numbers were ultimately 

 obtained as means from upwards of fifty observations, indi- 

 cating the heat-evolution, in gramme-degrees per gramme- 

 molecule, on diluting mMS0 4 100 H 2 to -25MS0 4 100 H 2 0, 

 or mMCl 2 100H 2 O to -25MCl 2 100H 2 O. PI. VI. fig. 6 

 represents the values graphically. 





Zinc 



Copper 



Cadmium 



Zinc 



Cadmium 





sulphate. 



sulphate. 



sulphate. 



chloride. 



chloride. 



m=2d 









9000 





20 



.... 







8050 





10 









6100 





7 







1450 







6 







1075 







5 



*375 





825 



4800 



100 



4 



175 





625 



4400 





3 



100 





450 



3800 





2 



50 



50 



275 



2700 





1-5 1 



Less than 



Less than 



200 







1 [ 



errors of 



errors of 



75 



1250 





* J 



observation, observation. 



I 





400 





161. On calculating from these figures the values of h u /i 2 , 

 and A 3 , and hence those of H, for particular cases, it becomes 

 evident that, as a general rule, the E.M.F. corresponding to 

 the value of H falls below the potential-difference actually 

 set up between two plates immersed in solutions of different 



