Chemical Affinity in terms of Electromotive Force. 379 



Observed E.M.F. with zinc-cad mium-sul-"] 



phate cells set up with solutions of | 



•25MS0 4 100 H 2 0, amalgamated-zinc ^ ='360 volt. 



and electro-cadmium plates (§ 121, 



Part VI.) _, 



Observed E.M.F. with zinc-cadmium-^ 



chloride cells set up w^ith solutions off __.oqa 



•25MC1 2 100H 2 O, amalgamated-zinc f "* °° U " 



and electro-cadmium plates j 



The difference betwen '365 and '330 volt, or '035 volt, 

 represents nearly 10 per cent, of the former value, and cor- 

 responds to upwards of 1500 gramme-degrees per gramme- 

 molecule, a quantity almost too large to be likely to be due 

 solely to accumulation of experimental errors. This point 

 will be further discussed in a future paper. 



158. No material difference is made in the relationships 

 between the E.M.F.'s observed (deduced from the assumption 

 that the E.M.F. of Clark's cell = 1-457 volt) and those calcu- 

 lated from the heat- values (by using the factor 4410), when, 

 instead of the numerical values 1'457 and 4410 for these quan- 

 tities (hitherto purposely employed throughout for the sake 

 of uniformity), the more exact values are used deduced from 

 more recent observations on the true value of the B.A. unit 

 of resistance, and especially those of Lord Rayleigh, which 

 indicate that the true value is only '9868 x 10 9 . For the use 

 of this corrected value reconciles the apparent discrepancy 

 between Joule's various valuations of the mechanical equiva- 

 lent of heat (§§ 33, 34, Part I.); and, further, causes all the 

 different estimations of the value of J discussed in §§ 55 and 

 56 (Part 111.) to agree fairly well together, and to become on 

 reduction values close to the mean result of Joule's friction 

 experiments, viz. 41'555 megalergs (§ 34). Combining this 

 value of J with the mean value of the " Faraday coefficient " 

 (the quantity referred to in § 7, Part I., as indicated by %, but 

 which may be gracefully indicated instead b'y F in commemora- 

 tion of Faraday) deducible from the experiments of Kohlrausch 

 and the more recent ones of Mascart, viz. 0001048 (§ 103, 

 footnote, Part V.), the end result is attained that 



JF = 41-555 x 10 6 x -0001048 = 4355. 



Now 4355 = 4410 x '9875; so that whilst the actual measure- 

 ments of E.M.F. are reduced in the ratio '9868, those deduced 

 from heat-valuations must concurrently be reduced in the 

 ratio *9875, or almost exactly to the same extent. If J be 

 taken = 41*5 x 10 6 , the reduction ratio becomes '9862 instead 

 of *9875. Hence it appears that in order to reduce to the 



2C2 



