JoLY — On a Direct Reading Electrolytic Ampere Meter. 561 



generated such, a pressure in the bulb that the quantity escaping" 

 through the orifice is 0-1734 c. cs. per second. The column of 

 mercury then comes to rest. In the case of the instrument upon 

 the table, it takes 20 or 30 seconds for this state of rest to be 

 attained. And the mercury column now stands at some 14 cms. 

 above the zero, where the reading on the vertical scale is I'O ampere. 

 When reading as high as 2 amperes, a somewhat longer time is 

 needed before the final reading is attained. This appears to be 

 due to temperature changes going on in the bulb, and suggests 

 that heavy currents should not be run through so small an instru- 

 ment. Of course, if the current is now varied, the mercury column 

 takes up a new level. Thus when a resistance in the circuit is 

 diminished the column rises quickly ; when it is increased, it falls 

 quickly, and takes its stand at a new level. The reason is obvious. 

 The rate of evolution of the mixed gases is different with each 

 current, and the position of the mercury column is peculiar to each 

 rate of evolution, being such that in each case the pressure produced 

 by it is just able to drive the evolved gas through the orifice at th& 

 same rate at which it is being evolved. 



To construct the scale of the galvanometer it is only necessary 

 to place it in circuit with a galvanometer reading amperes reliably, 

 and passing a number of different currents, Ci, c^, Cs, . . . through, 

 the circuit, to mark upon the scale the points to which the mer- 

 cury rises in the tube. Subsequently we measure the height of 

 these points above the normal level of the mercury, and plot these 

 upon divided paper against the currents. The result is a curve 

 upon which readings of level for any intermediate current 

 strengths may be interpolated. The scale is thus marked out in 

 amperes and tenths of amperes, or closer. It will be found that 



the scale opens out in its upper readings, the equation — = k^ 



approximately describing the curve. As c is proportional to the 

 velocity of eflSlux of the gas at any reading of h centimetres of 

 mercury upon the scale, this approximates to the usual law of the 

 efflux of liquid from an orifice under the head or pressure h. 



The instrument upon the table was constructed for purposes of 

 trial and experiment. It has, however, worked very satisfactorily. 

 So little gas is used in obtaining a reading that the consumption 



