Chemical Affinity in terms of Electromotive Force. 265 



where C is the current, R the resistance, t the time, and H the 

 heat evolved. Moreover the existence of such an error in the 

 B.A. unit would cause a precisely corresponding error in La- 

 timer Clark's valuation of the E.M.F. of his standard cell 

 (Proc. Roy. Soc. 1872, vol. xx. p. 444), this valuation being- 

 based on the determination of the resistance R between two 

 points traversed by a known current C, the potential difference 

 E between the points being equal to the E.M.F. of the cell, 

 so that by Ohm's law E = CR ; and the effect of this would be 

 to cause a precisely similar error in the value of J deduced 



by the formula J= /tt , ;\ • as i* 1 ^ ne writer's experiments. 



36. In confirmation of the supposition that the B.A. unit of 

 resistance is too large, Kohlrausch's experiments may be cited 

 (Phil. Mag. 1874, vol. xlvii. pp. 294 & 342). A careful compari- 

 son of several copies of the B.A. unit with coils standardized by 

 himself in absolute measure led Kohlrausch to the conclusion 

 that the B.A. unit is really 1*0196 earth-quadrant per second, 

 or 1*96 per cent, too large ; for he found that the resistance of 

 certain coils graduated in Siemens mercury units averaged 

 0-9717 x 10 9 C.G.S. units, whilst the B.A.-unit coils averaged 

 1*0493 times the resistance of the Siemens coils, giving there- 

 fore for the B.A. unit the value 



0*9717 x 1*0493 x 10 9 = 1*0196 x 10 9 . 



On the other hand, it is to be observed that Lorenz has 

 made a determination of the value of a Siemens mercury unit 

 with a lower result (Pogg. Ann. 1873, vol. cxlix. p. 251), viz. 

 0*9337 earth-quadrant per second. Lorenz's mercury-tubes 

 were made by himself ; and it does not seem from his paper 

 that any evidence was obtained of the purity of the mercury 

 used other than the fact that its specific gravity at 5°*3 C. was 

 13*586 after digesting for a long time with nitric acid. These 

 mercury-tubes do not seem to have been compared directly 

 with any B.A.-unit coil so as to give the value of the B.A. unit 

 in absolute measure ; on comparing them with a reputed Sie- 

 mens unit, Lorenz found a difference of nearly 1 per cent. 

 (0*943 x 10 9 as against 0*9337 x 10 9 )— the difference between 

 Lorenz's mercury-tube value of a Siemens unit and the value 

 of the latter in B.A. units, as deduced by Kohlrausch, being 



nearly exactly double this amount, or nearly 2 per cent., .. n . Q » 



x 10 9 = 0*9530 x 10 9 as against 0*9337 x 10 9 . V() ^ 6 



37. Yet, again, values for J have been obtained by other 



observers considerably higher than that deduced from Joule's 



