S/K WILLIAM SIEMENS, F.R.S. 33 



much as, in order to weigh a substance, we must have a unit 

 weight, such as the pound or kilogramme. Several units have from 

 time to time been proposed. Formerly the mile of iron or copper 

 wire, or, in France, the kilometre of iron wire, of a certain section, 

 was taken to represent the unit of resistance, but it is evident from 

 what I have just stated that these are very vague measures. A 

 slight variation in the constitution of the metal, or in its ad- 

 mixtures, would vary the value of a mile of wire between very wide 

 limits. It was at the time when electrical tests assumed a practical 

 and serious shape that my brother, Dr. Werner Siemens, proposed 

 his mercury unit ; that is, the resistance offered by one metre 

 length of mercury enclosed in a tube of one square millimetre 

 sectional area at the freezing-point of water. This unit has been 

 generally used in cable-testing up to this time. But, at present, 

 another unit is proposed, the British Association unit, which is 

 derived from Weber's absolute unit of electrical force. It is difficult 

 to explain correctly what that means ; but the following may serve 

 to convey the general idea : If I have a loop of wire of a certain 

 length, and turn it suddenly, with respect to the magnetic axis of 

 the earth through 90, I develope in this wire a certain current 

 which, if it is made to act upon the needle of a galvanometer 

 produces a certain deflection. The force thus set up by this 

 motion, Weber calls the absolute unit of electrical magnitude ; 

 and the resistance which is set up in the loop of wire of given 

 quality and dimensions he calls the unit of resistance. Now, ten 

 millions of these units of resistance would form the unit of re- 

 sistance proposed by the British Association Committee. This 

 unit would have certain ulterior advantages, being referable to a 

 unit of force ; but its production is attended with considerable diffi- 

 culty, which leaves reasonable room for doubt, whether future de- 

 terminations may not materially modify its value. At present the 

 material unit of a certain amount of mercury, which being a liquid 

 is not subject to differences of temper, contained in a certain tube 

 appears to be more certain, it differs only about 4^ per cent, from 

 the British Association unit, and it has this advantage it is 

 largely used. 



MEASURE OP CHARGE. In addition to the unit of resistance, 

 we have to determine the unit of charge, or unit capacity for 

 static electricity, which may be defined as follows : If I have one 



VOL. III. D 



