66 J. Whitmore — Method of Increasing 



several cells, the difference of potential at the terminals of 

 each of the three cells of a series was measured by means of 

 Thomson's quadrant electrometer. It was thus observed that 

 the potential fell regularly along the series for the lower elec- 

 tromotive forces. For example, when the total difference of 

 potential was 0'9 volt, at the terminals of the first cell it was - 3, 

 and between the first and second 06 volt. The distribution of 

 the potential among the different cells was less uniform after 

 a total electromotive force of 1*2 volts' was exceeded. This 

 was possibly due to electrolytic conduction which occurred in 

 the apparatus. Moreover the caliber of each of the capillary 

 tubes was not exactly uniform throughout the entire length of 

 the tube. Thus, these measurements clearly indicate, that 

 with precisely similar cells, the fall of potential in each cell of 

 the series is the same. 



It was next sought to put the apparatus in a convenient 

 form. Since it seemed desirable to dispense with the use of 

 the cathetometer in reading . the deflections of the mercury 

 columns in order that the reading might be obtained more 

 quickly, experiments were made with cells in which a small 

 electromotive force produced a large deflection of the mercury 

 columns. This is notably the case in the electrometer used by 

 Pratt,* in which the capillary tube is placed nearly horizontal 

 and its inclination, and hence also the sensitiveness of the 

 instrument, can be varied at will. Thus the mercury column 

 can be caused to move over a space of a centimeter for a dif- 

 ference of potential of 0*1 volt, and hence a deflection corre- 

 sponding to a thousandth of a volt can be easily read directly 

 on a scale placed behind the moving column. 



Accordingly three cells of this form were joined in series 

 and a difference of potential was maintained between the ter- 

 minals of the series. It was found, however, that very often, 

 when the electromotive force was applied, the mercury col- 

 umns of the three cells, instead of moving together in one 

 direction until the positions of equilibrium were reached, began 

 a remarkable seesawing, the mercury column of one cell ad- 

 vancing while those of the other cells retreated. This balanc- 

 ing continued through a considerable interval of time, since, 

 as the positions of equilibrium were approached the vibrations 

 were very slow. Moreover a given electromotive force at 

 different times caused the mercury columns of the several cells 

 to arrange themselves differently, so that no measurement of 

 electromotive force could be made by observing a single cell 

 of the series. However, when simultaneous readings of the 

 three cells were made, and the curves drawn, though these 

 curves were irregular in form, yet the mean of the three curves 

 was of the normal type. 



* This Journal, vol. xxxv, 1889. 



