ELECTRICAL UNITS OF MEASUREMENT. 113 



which must be accurately measured. Arrange a 

 mechanism to charge it to an accurately known 

 potential of moderate amount (for example, in 

 electrostatic measure, about 10 C.G.S., which is 

 about 3,000 volts), and discharge it through a 

 galvanometer coil at frequent regular intervals 

 (for example, ten times per any convenient unit 

 of time). This will give an intermittent current 

 of known average strength (in the example, io 5 

 electrostatic C.G.S., or about 1/300,000 electro-mag- 

 netic C.G.S., or 1/30,000 of an ampere), which is to 

 be measured in electro-magnetic units by an 

 ordinary galvanometer. The number found by 

 dividing the electrostatic reckoning of the current, 

 by the experimentally found electro-magnetic reck- 

 oning of the same, is " v," in centimetres per the 

 arbitrary unit of time, which the experimenter 

 in search of the mean solar second has used in 

 his electrostatic and electro-magnetic details. The 

 unit of mass which he has chosen, also arbitrarily, 

 disappears from the resulting ratio. 



But there is another exceedingly interesting 

 way a way which, although I do not say it 



I 



