SPECIFIC INDUCTIVE CAPACITY 



127 



was used, consisting of a double cylinder in which an insulated 

 cylinder could hang, the vertical cross-section being as represented in 

 Fig 86. 



The capacity of this condenser was com- 

 pared with that of a sliding condenser, first 

 with air and second with the liquid to be 

 tested as the dielectric. The sliding con- 

 denser was brought to equality in each case. 

 The sliding condenser was graduated, and 

 its capacity for so many divisions of the 

 inner cylinder within the outer was known. 



The hydrocarbon oils agree in giving 

 K = /z 2 nearly, while their values for vegetable 

 and animal* oils differ widely. The Table 

 below is extracted from Hopkinson^s paper 

 (Original Papers, ii. p. 85). The square of 

 the refractive index for infinitely long waves, 

 which is given as well as K, is calculated from the dispersion in the 

 \i>ible spectrum by the formula p = a + b/\, using the sodium 

 and hydrogen lines. 



Fia. 86. 



K 



IVtroleum spirit 1'92 



Petroleum oil (Field's) 2'07. 



Turpentine 2'23 



Castor oil 4'78 



Sperm oil 3'02 



Olive oil 3-16 



1-922 

 2-075 

 2-128 

 2-153 

 2-135 

 2-131 



Boltzmann's experiments on gases. Boltzmann made 

 determination! of K for different gases by a simple method, using 

 a condenser within a closed metallic earth-connected vessel which 

 could be exhausted or filled with any gas. The connections with 

 the plates of the condcn^ r were made by wires passing through 

 the tlit \ >(!, but hermetically sealed in, as represented 



diagram mat ically in Fig. 87. The vc^el uas first exhausted, and 

 one plate, A, was connected to a battery of about 300 Danielfs cells, 

 the other plate, H, being connected to an electrometer and to 

 earth. After A was thus raised to a potential which we will call 

 Vj it wa> insulated, B and the connected pair of quadrants of the 

 electrometer \\cre insulated, and the gas to be experimented on was 

 Admitted The potential of B >f ill remained /ero, for, as there was 

 TIO leakage from A, all the electrification on B was connected with 

 that on A, and then wen then-fore no lines of force between B 

 and the case, and no change of its potential. The potential of A 

 fell by the admi-ion of the gas to, say, V 2 , where V x = KV 2 , K 

 being the dielectric constant of the gas, that of vacuum being 1. 



