338 THE ELECTRIC TELEGRAPH. 



magnetic lines, is equal to the product of the intensity of the 

 field, the length of the conductor, and its velocity. 



The way this is put into practice is by the employment 

 of a circular conductor instead of a straight one, which 

 would be impracticable. A small magnet is suspended in 

 its centre, the deflection of which is observed during the 

 time the conductor is rotated on a vertical axis. 



The recent determinations of this really beautiful system 

 of units by those members of the committee who have so 

 praiseworthily worked at it, agree with each other suffi- 

 ciently well to have induced the committee to adopt the 

 result, and resistances expressed in the unit have accordingly 

 been distributed. Of course, in practice, a multiple of the 

 unit, which is exceedingly small, must be called for con- 

 venience the practical unit. The resistance accepted as the 

 practical unit, or the so-called B. A. unit, is 10 7 times the 

 ^^ unit, nearly. This B. A. unit is a little larger than the 

 mercury unit issued by Dr. Siemens, in the proportion of 

 1*0456 to 1. An error has crept into the mercury unit, how- 

 ever, by a wrong value having been assigned to the specific 

 gravity of mercury in the calculated resistances of the normal 

 tubes, so that, allowing for this, one B. A. unit is equivalent 

 to 1'0484 mercury units according to the definition. 



A system has been proposed for reproduction of normal 

 resistances, by Dr. Matthiessen, which consists in making 

 an alloy of gold and silver of certain proportions (two 

 parts of gold, and one of silver), and drawing it into wires. 

 The conducting power of an alloy of these metals, in what- 

 ever proportions the metals are taken, is less than that 

 corresponding to the volumes of its constituents. When 

 a series of such alloys is made, beginning with pure gold, 

 and gradually increasing the percentage of silver until pure 

 silver is reached, it is found that, as the percentage of silver 

 increases to a certain point, the conducting power decreases, 

 after which it increases again until it approaches that of 

 the pure metal. From the point where the curve reaches 

 its minimum, it rises only very gradually, so gradually 

 indeed, that for a diflerence of some few per cent, on each 



