246 Dr. C. R. A. Wright on the Determination of 



4933 

 an amount equal to » . 00 = 14*5 per cent. ; for the negative 



term in the expression E — H%J will have been overestimated 

 by that quantity. 



Again, suppose that the water is brought by a secondary 

 physical action into a quasi-vaporous state before it is decom- 

 posed by the current into gaseous hydrogen and oxygen (e. g. 

 on the molecular theory of the constitution of matter, suppose 

 that only detached molecules of water are operated on by the 

 current, and not the water en masse in the liquid state); since 

 about 600 gramme-degrees are required to volatilize a gramme 

 of water at the ordinary temperature (the latent heat of steam 

 at near 15° being about 600), or 5400 per gramme-equivalent, 

 the value of e deduced as above will be too large by about 



'oaTKR =15*8 per cent, because in this case the negative term 



— H%J will be underestimated to that extent. 



Consequently, if both secondary actions be supposed to take 

 place (i. e. if the electrolysis proper takes place between de- 

 tached molecules of water converting them into gaseous hy- 

 drogen and ozone), the value of e will be 15*8 — 14'5, or 1*3 

 per cent, too large. 



Suppose, again, that the primary electrolytic action is to 

 convert liquid water into liquid oxygen and hydrogen, and 

 that the conversion of these liquefied elements into the ordinary 

 gaseous condition is due to secondary physical changes ; it is 

 evident that the value of e found as above is too large by 



q a x 100 per cent., where \ represents the heat that is ab- 

 sorbed in the vaporization of a gramme-equivalent of each of 

 the liquefied gases taken together. 



Yet again, in the electrolysis of copper sulphate, suppose 

 that whilst the ultimate chemical change is expressed by the 

 equation 



CuS0 4 + H 2 = Cu-fO + H 2 S0 4 , 



this change is actually the sum of two changes, of which the 

 first is secondary: — 



(1) CuS0 4 + H 2 = CuO + H 2 S0 4 , 



(2) CuO =Cu + 0; 



it is evident that the value of e deduced as above will be greater 

 than that corresponding to the second change only, by an 

 amount equivalent to the heat absorbed during the first change. 

 . 13. During the last three or four years the writer has made 

 various experiments with the view of determining the E.M.F. 

 measuring the " affinity " between various constituents in dif- 



