88 Dr. Matthiessen on the Electric Conducting Power oj 



wires were dipped. The whole tube was then brought into a 

 trough, which was filled with water or a solution of chloride of 

 calcium, according to the temperature desired. The liquid in 

 the trough, which was kept coutiuually agitated, was gradually 

 heated by lamps placed underneath it, observations being made 

 from time to time after the method used in the former expe- 

 riments. The temperature was read off from a thermometer 

 in the trough. The distance between the two platinum wires 

 was treated as the length of the wire, and the diameter was mea- 

 sured by weighing a quantity of quicksilver which occupied a 

 known length in the tube. From these data and the observed 

 resistances the conducting power was calculated as above. 



In these results an interesting fact was observable, viz. that 

 at some distance from the point of fusion, as well in the liquid 

 as in the solid state, the decrements in the conducting power 

 witn the increase of temperature were almost in proportion, but 

 near the point of fusion the decrease in the conducting power 

 became much more rapid; with sodium this change appears to 

 be very sudden, whereas with potassium it seems gradual. This 

 difference in these metals corresponds with their different beha- 

 viour in fusion ; namely, potassium does not become suddenly 

 liquid like sodium, but first passes through a semifluid state*. 



The values obtained can be expressed by the following for- 

 mulas, in which X, = the conducting power compared with silver 

 at 0° C. = 100, and t the temperature in Centigrade degrees. 



Potassium. 

 For t between 0° and 46°-8, 



\= 30-14. - 0-0819^ + 0-000235^2 . 

 for t between 4G°-8 and 56°-8, 



X = 668-26-40-40.2/ + 0-83801^2_o.o058152<3; 

 for t between 56°-8 and 100°, 



X=: 13-35-0-03393^. 



Sodium. 

 For t between 0° and 95''-4, 



X=32-54-0-1172/ + 0-000127i2; 

 for t between 96°-l and 120°, 



X=23-38-0-07222^. 



Potassium melted between the temperatures 46°-8 and 56°, 

 and sodium between 95°-4 and 96°-l. 



Tlie curves (fig. 7) show the variation of the conducting power 

 for different temperatures as calculated with these formulas. The 



* See Reginault, PoggendorfF's Annalen, vol. xcviii. p. 411 j Phil. Mag. 

 vol. .\ii. p. 489. 



