186 Dr. C. R. A. Wright on the Determination of 



without throwing them out of circuit or otherwise varying the 

 current passing by altering the total resistance in circuit, 

 which is impracticable in experiments in which it is essential 

 that the current should not sensibly vary during long periods 

 of time, — and, secondly, enables the value of an extremely 

 minute current to be determined with as much accuracy as 

 that of a much stronger one, and even with greater accuracy ; 

 for whilst the current from a given electrometer is diminished 

 by throwing more and more resistance into circuit, the differ- 

 ence of potential between the ends of the resistance thus thrown 

 into circuit is not only not decreased but is even increased. 

 For if <?! be the E.M.F. of the electrometer, Rt the resistance 

 the difference of potential between the ends of which e 2 is mea- 

 sured, and R 2 the resistance of the rest of the circuit; the 

 current passing 



R x Rx + R 2 

 whence « 1 -<- 2 =CR s . 



Hence, as C diminishes e 2 increases, and consequently a larger 

 electrometer-deflection is obtained with a smaller current. 

 Tolerably strong currents develop appreciable amounts of heat 

 in the resistance-wire, thus increasing its resistance and intro- 

 ducing an error in deficiency in the current; so that the 

 method is more particularly applicable for minute currents. 



70. The difference of potential set up between the electrodes 

 of a voltameter by the passage of a current C is E = e + CR, 

 where e is the counter E.M.F. set up, and R the resistance of 

 the voltameter (§ 57). If R do not exceed 100 ohms (100 x 10 9 

 C.Gr.S. units of resistance), and C do not exceed 0-0001 weber 

 (0*00001 C.G.S. current-unit), the value of C Rdoes not exceed 

 100 x 10 9 x 0-00001 = 0-01 x 10 8 = 0'01 volt; whilst, whatever 

 the value of C and R, the E.M.F. equal to the product CR is 

 readily calculable. If now the difference of potential between 

 the electrodes E be read off (in terms of the scale of the quadrant- 

 electrometer) whilst the current is passing, by noticing the 

 deflection of the spot of light from the zero (readings being 

 taken on one side only), and the current be then interrupted, 

 a motion backwards of the spot through a scale-length equi- 

 valent to CR will instantly ensue, and subsequently a con- 

 tinuous motion backwards as the value of e falls ; if, however, 

 CR be equivalent to considerably less than half a scale-division 

 of the electrometer, the first sudden motion backwards will be 

 entirely inappreciable. In the following experiments the elec- 

 trometer was so arranged that 1 scale-division represented 

 about 0'02 volt (readings being taken on one side only of the 



