962 Transactions of tee American Institute. 



necting two like poles of the two together, so that they neutralize each 

 other, and connecting them by screw-cups III and IY, and taking 

 their resistance as we do that of any conductor. 



47. The following method, too, discovered and first used by the 

 writer, is found to be correct and reliable. The cell to be tested is 

 connected between III and IY, using two or more cells for main 

 battery. The switch being to the left, the deflection caused by the 

 main battery is noted. On turning the switch to the right, the force 

 of the cell is added to that of the main battery and the deflection is 

 increased. Now, reverse the poles of the cell so that its force opposes 

 that of the main, and the deflection is much diminished; add the tan- 

 gent, of the smaller to that of the larger deflection and divide by two. 

 Find the degrees corresponding to this mean tangent. Now turn the 

 switch to the left, and introduce rheostat resistance to bring the needle 

 to the same degree. The amount introduced will show the resistance 



of the cell. 



Constant Multiplier. 



■48. To determine the electro-chemical equivalent of a current by a 

 tangent galvanometer, it is necessary to find a number by which the 

 multiplication of the tangent of its degree of deflection will give the 

 equivalent sought. 



This is done in various ways. That recommended and employed 

 by most authors is by the electrolysis of water in the voltameter, and 

 the production of its elementary • gases, the volume of which, whe» 

 properly corrected for temperature pressure and moisture, is directly 

 proportional to the strength of the current. 



49. The correction may be made by the following 



Formula. 

 -y 1 = volume of gas observed. 

 i) **=. vol. corrected for temp, pressure and moisture. 

 o 1 = barometric pressure observed. 

 b = same corrected for temperature. 

 e = tension of vapor of water at t°C 

 t = the number of degrees above freezing point, 

 ^i^ = Kegnault's co-efficient of expansion for each degree C. 

 V273 h—e 

 ^273TI X- T~ 



Barometric Correction.. 



50. The barometric column is to be corrected for temperature, for 

 which we may employ the following table (51), where the scale is of 

 wood, graduated for millimeters ; or we may multiply the co-eflicient 



