406 COMPARISON OF RESISTANCES. 



an equation from which the value of k is deduced. The formula 

 holds so much the better the smaller is /, but Lorenz has shown 

 that the error may be disregarded with bars 1*5 cm. in diameter 

 when good conductors, whose length / is 2 cm. 



This method was carried out in a very simple manner. The 

 bar was perforated by 9 very fine holes 0-4 mm. in diameter, 

 intended to receive the junctions of thermoelectric couples, and 

 numbered o, i, 2, ... 8. Two couples, one with its junctions at 

 o and i, and the other at 7 and 8, are joined in opposite direc- 

 tions in the same circuit, and give 8 directly. Seven other couples 

 in series have their uneven junctions in the holes i, 2, 3 ... 7, 

 the even junctions being at the same temperature, and give 

 directly. We have thus all the elements for calculating the value 

 of k. 



Electrical conductivities were determined on the same bars at 

 temperatures of o and of 100. 



The results obtained by Lorenz confirm the law of Wiedemann 

 and Franz, and bring out this other remarkable relation that the 



ratio - : is sensibly constant and equal to 1*367. Lorenz 



^100 C Q 

 thinks that we might put generally 



--CT 



c~ 



T being the absolute temperature, and C a constant. In the 

 case of very good conductors, the coefficient k varies very little 

 with the temperature ; the formula expresses then simply the 

 known result that the resistance of pure metals is proportional to 

 the absolute temperature. It is very remarkable that the relation 

 pointed out by Lorenz holds for such alloys as brass and argentan ; 

 the coefficient of variation in electrical conductivity is then less, 

 but that of thermal conductivity far greater, than for pure metals; 

 there is exact compensation. 



1003. RESISTANCE OF ELECTROLYTES. The most divergent 

 numbers are obtained for the resistance of water, whatever pains be 

 taken to purify it. The least traces of dissolved solids, or even 

 gases, considerably increase the conductivity; the greatest resistances 

 correspond to the purest water. Kohlrausch* obtained the value 

 7.io 6 ohms for the specific resistance that is to say, 74 billion times 



* F. KOHLRAUSCH. Bericht der Berl. Akad. Oct., 1884. 



