SPECIFIC INDUCTIVE CAPACITY 131 



an alternating current. For water at 17 C. he obtained K = 80 '00, 

 and other observers * by various methods have found nearly the 

 same value. For alcohol he found K to be about 26. 



Experiments of Dewar and Fleming at low tem- 

 peratures.f Dewar and Fleming made experiments on the 

 dielectric constant of ice and other substances from a temperature 

 of 200 C. upwards, using a condenser consisting of two 

 co-axial brass cones, about 15 cm. long, the outer tapering from 

 an inside diameter of 5'1 cm. to 2*6 cm., and the space between 

 the two being 3 mm. The condenser was charged and discharged 

 about 120 times per second by an interrupting tuning-fork and 

 the circuit was arranged so that either the charging or the 

 discharging current alone should go through a galvanometer. 

 The equality of the two was taken to show that conduction was 

 not coming into play. The space between the cones was filled 

 with the substance to be examined, and cooled to the temperature 

 of liquid air, and the galvanometer deflection was observed on 

 charging at given potential or on discharging. Then the substance 

 was melted out and replaced by gaseous air at the same temperature, 

 and the galvanometer deflection was again observed. The ratio 

 of the deflections, after certain corrections, gave the specific 

 inductive capacity of the substance. Observations were also 

 made at higher temperatures by allowing the temperature of the 

 condenser to rise gradually. 



The dielectric constant of pure ice at 200 C. was 2 '43, rising 

 with rise of temperature to 70'8 at 7'5 C., though here 

 conduction had set in and the measurement was not so trust- 

 worthy. A large number of solutions and compounds were thus 

 examined, and the general result was that at 200 C. the 

 dielectric constant was not much greater than the square of the 

 refractive index for exceedingly long waves, as calculated from 

 dispersion formulae. For castor oil, olive oil, and bisulphide of 

 carbon they were nearly coincident. 



Drude's experiments with electric waves. As a 

 type of an entirely different method of research, in which electric- 

 waves are used, we shall take Drude^s experiments on water and 

 on other liquids.^ 



If Uj is the velocity of electric waves in air and U 2 is their 

 velocity in another medium, the refractive index for the waves 

 is /x = U 1 /U 2 . But if A! X 2 be the lengths in the two media of 

 waves starting from a source of the same frequency X 1 /X 2 = Ui/lT 2 

 so that /x = X 1 /X 2 . According to the electro-magnetic theory 

 K = /x 2 = X 1 2 /X 2 2 . To find the ratio X^X^ Drude used a Lecher 

 system in which waves were transmitted between two parallel 

 wires, first through air and then through a trough containing the 



* For a summary see Dewar and Fleming, Pt'oc. Rot/, Soc., Ixi. i\ 2. 



t Proc. Roy. Foe.. Ixi. (several papers). 



% Ann. der plnj gf , 1896, lyiij. p. 1, lix. p, 17, 



