1884.] On the Electro- Chemical Equivalent of Silver. 145 



brevity the H-cell, iu which the solid zinc is replaced by a fluid 

 amalgam. The amalgam and the pure mercury, forming the metallic 

 electrodes, are placed at the bottom of two small test-tubes standing 

 vertically. The electric connexion is made by platinum wires sealed 

 through the bottoms, and the communication between the two 

 vertical tubes is through a lateral horizontal branch sealed into them. 

 The cell is filled with sulphate of zinc solution above the level of the 

 horizontal branch, and can then be closed with corks. This form of 

 cell lends itself conveniently to experiment, as by withdrawing the 

 corks it is easy to observe the effect of stirring or of various 

 additions. 



A large number of cells have been compared for more than 

 eight months, and have behaved very satisfactorily. The results are 

 sometimes anomalous for the first two or three weeks, but the values 

 finally attained are in our experience extremely close together. 



The comparisons are made by the method of compensation. The 

 difference of electromotive force of the cells to be compared is com- 

 pensated by a known fraction of the electromotive force of other cells, 

 the value of which is then expressed in terms of one of the Clarks. 

 There would be no difficulty in obtaining still greater sensitiveness, 

 but it is useless to take readings closer than to TOS-QO- 



The Clark cells possess the immense practical advantage (as com- 

 pared, for instance, with Daniell's) of standing always ready for use, 

 but the objection is sometimes expressed that they polarise greatly on 

 the passage of the smallest currents. Our experience has been in the 

 opposite direction, and has shown that moderate short-circuiting is 

 actually advantageous in the case of cells newly set up. When old cells, 

 which have reached their permanent condition, are allowed to make 

 10 1 00 ampere for a quarter of an hour, the disturbance thus occasioned 

 passes off in about half an hour to within a few ten-thousandths, and 

 on the next day there is no indication of any residual effect. 



The absolute determinations of the E.M.F. of Clark's cells were 

 made by compensation with the difference of potentials at the 

 terminals of a known resistance traversed by a known current. The 

 details of the method, which offers no special difficulty, are given in 

 the paper. Of thirteen values, found on different days between 

 October, 1883, and April, 1884, with the current measuring apparatus, 

 the highest is T4552 and the lowest 1'4531. This number expresses 

 the E.M.F. of a certain cell at 15 in terms of B.A. volts. To get the 

 E.M.F. in true volts, the mean number (1'4542) must be multiplied by 

 the number expressing the B.A. unit of resistance in absolute measure. 

 If 1 B.A =-9867 ohm, E.M.F. of Clark=l'435 volt. 



The value of the H-cells would be a few ten-thousandths higher. 



Apparatus capable of giving original determinations of the inten- 

 sities of currents not being generally available, we have shown with 



VOL. XXXVII. L 



