Electrostatic Units in the Electromagnetic Unit. 225 



" the E.M.F. of Thomson's gravity Daniell was measured by- 

 comparing it before and after the above experiment [the elec- 

 tromagnetic valuation] directly with that of the above battery 

 by means of Sir William Thomson's quadrant-electrometer:" 

 the which wording conveyed to me the impression that both 

 cell and battery were compared under the same conditions, 

 viz. with insulated poles, no current flowing. Incidentally 

 referring to Mr. Shida's paper (Phil. Mag. May 1880), I 

 noticed that such a mode of working would introduce an error 

 due to polarization, the E.M.F. of the cell being less when 

 generating a current than when its poles are insulated. Mr. 

 Shida rejoined (Phil. Mag. June 1881) that the quadrant- 

 electrometer comparison was made whilst the cell was gene- 

 rating a current, so that the polarization-error was avoided. 

 I then pointed out (Phil. Mag. July 1881) that, if Mr. Shida 

 made the quadrant-electrometer comparison whilst the cell was 

 generating a current, the electrometer-reading would be less than 

 the true E.M.F. in the ratio of the external resistance to the 

 total resistance in circuit, the difference being in Mr. Shida's 

 experiments about 2 per cent. Mr. Shida now replies, in 

 effect, that this 2-per-cent. error was obviated in his experi- 

 ments by multiplying the electrometer-reading reduced to 



C.G.S. units, V— V, by the factor (where b is the 



internal resistance of the cell, and p the resistance of the rest 

 of the circuit*). That this correction was adopted may also 

 be gathered from Sir W. Thomson's denial of the existence, in 

 Mr. Shida's experiments, of any such error as the omission of 

 this factor would entail. I confess, however, that I am at a 

 loss to understand how pointing out that the omission of the 

 factor would introduce an error can be construed into a "mis- 

 understanding of Ohm's Law." 



After Sir W. Thomson's explicit statement that Mr. Shida's 

 experiments were free from the source of error pointed out, I 

 have to express my regret at having, as it appears, too hastily 

 concluded that Mr. Shida did not apply a particular correction 

 when, in point of fact, he only omitted to mention that he 

 applied it. At the same time it is only due to myself to point 

 out that the numerical values quoted by Mr. Shida as his 

 results are not of such a nature as to lead the reader to put 

 implicit confidence in the rest of the work, and that a priori 

 they seem to suggest that when corrections and precautions 

 were not mentioned they were probably omitted. 



In order to calculate the electromagnetic value of the 



* Mr. Shida uses r instead of p. I have altered the symbol to avoid 

 confusion with the other r used by Mr. Shida in his original paper, referred 

 to later on. 



