44 Messrs. 0. W. Waidner and F. Mallory on 



error in the system of electric units employed. This, indeed, 

 in the light of the enormous and painstaking work which 

 serves as the basis of these units, is not probable, but, 

 considering the difficulties encountered in the standards of 

 electromotive force and current, it is not altogether impossible. 

 The enormous experimental evidence which served as the 

 basis of the international ohm, together with the recent 

 confirmation of the accuracy of this unit, renders it almost 

 certain that the difference need not be looked for in this 

 direction. 



At the Toronto meeting of the British Association, before 

 which an abstract of this paper was read, the Committee 

 on Electric (Standards received an appropriation for the 

 redetermination of the electrochemical equivalent of silver 

 and the absolute electromotive force of the Clark cell, and 

 their results are awaited with much interest. In the mean- 

 while the results of a determination of these quantities, made 

 by Dr. Kahle at the Reichsanstalt, have been published 

 (Zeitschr.f. Instk. xviii. pp. 229, 267, 1898 ; Wied. Ann. lix. 

 p. 532, 1896 ; Wied. Ann. lxvii. p. 1, 1899). His final value 

 of the E.M.F. of the German H standard form of Clark cell 

 was 



E 15 o = 1-4325 volts. 



By making use of a previous comparison between the 

 Cambridge standard and the German H standard (B. A. 

 Report, 1892), the Cambridge standard, when reduced in 

 accordance with Kahle's value, becomes 



Cambridge i 5 o = 1*4329 volts. 



The value of the Cambridge standard (upon which is based 

 Griffiths' values of the capacity for heat of water) as deter- 

 mined by Glazebrook and Skinner (Phil. Trans. A, 1892) 

 was 



Cambridge 15 o = 1*4342 volts. 



As has already been shown by Dr. F. A. Wolff (Johns 

 Hopkins Univ. Circular, June 1898), when the values of the 

 E.M.F. of the Clark cell, as found by Kahle, are applied to 

 Griffiths' values of the capacity for heat of water, as well as 

 those of Schuster and Gannon, they are brought into very 

 fair agreement with those of Rowland as corrected by the 

 results of our comparisons (differing by about 1 part in 

 1400). This is shown by the following table : — 



