400 COMPARISON OF RESISTANCES. 



of which decreases from the centre to the circumference, and the 



mean value of which does not attain the that of the magnets. 



20,000 



It may then be assumed, without appreciable error, that the torsion 

 measures the electromagnetic action; and therefore, for the same 

 velocity and the same field, the conductivity of the liquid. 



It was found that the vessel alone produced no action. In order 

 to allow for variations of the field and for the coefficient of torsion 

 of the wire, each experiment is recommenced by suspending a brass 

 disc in the vessel. The torsion should be the same for the same 

 velocity if the field and the wire had not been modified. 



The authors applied this method to mixtures of sulphuric acid 

 and ^ water. Their results agree well generally with those of Pro- 

 fessor Kohlrausch, especially as regards the position of the maximum 

 and the point of inflection of the curve which represents the con- 

 ductivity of these liquids. 



997. GENERAL RESULTS. As regards resistance, various bodies 

 arrange themselves naturally in three categories : metals, pure metals 

 or alloys, which do not alter in consequence of the passage of the 

 current, and the resistance of which increases with the temperature ; 

 electrolytes, which are the seat of a chemical action correlated to the 

 current ; lastly, simple or compound non-metallic bodies, which have 

 little or no conductivity, like the dielectrics, and the resistance of 

 which diminishes as the temperature rises. The separation of these 

 three classes is by no means absolute, and there are bodies to which 

 it is difficult to assign an exact place. 



998. RESISTANCE OF METALS. Pure silver and copper are the 

 best conductors, but a slight admixture of foreign substances greatly 

 diminishes their conductivity.* The alloys of these metals with 

 each other, or with gold, platinum, nickel, are far worse conductors 

 than that which is the worst conductor. Alloys, on the other hand, 

 which consist of only such metals as lead, tin, zinc, cadmium, etc., 

 have a conductivity which is near the mean. 



Between the temperatures of o and 100 the resistance r of a 

 metal may be expressed as a function of the temperature / by a 

 formula with two coefficients analogous to that for expansions, the 

 coefficient /3 being very small compared with a. 



0. Li CORY. r=r (i 



* Sir W. THOMSON. Proceedings of the Roy. Soc. Lond., Vol. vin., p. 556. 1857. 



