o94 ‘Intelligence and Miscellaneous Articles. 
denser was much too large to render possible the observation of 
such a difference in position: it was therefore necessary to com- 
pensate this deflection, 2. ¢. by means of an equal opposite action 
upon the galvanometer-needle. 
Of the various possible modes of compensation the author em- 
ployed that of a constant current. A weaker battery (compen- 
sation battery of 4 Daniell cells) was employed to send a current 
through a large resistance, and through the coil of the galvanometer 
in the direction required to produce a deflection opposite to that 
given by the discharges of the condenser. By altering the large 
resistance of the compensation-current, it was easy to arrange that 
the needle should remain in its position of equilibrium under the 
simultaneous action of the two currents. If this adjustment had 
been made for a definite capacity of the condenser, and if then 
only the density of the dielectric medium was varied, a deflection 
of the galvanometer-needle was observed, due to the change in the 
capacity of the condenser produced by the change in density, and 
consequently in the dielectric constant of the medium. ‘The ratio 
of the dielectric constants for two different densities of the medium 
may be calculated from the easily determined deflection produced 
by the compensating current and the observed deflection. In this 
calculation it is only necessary to take account of the circumstance 
that a small part of the condenser-discharge passes through the 
compensation-circuit, and not through the galvanometer. As is 
well known, both Professor Boltzmann and Messrs. Ayrton and 
Perry have employed the electrometer for their investigations on 
this subject. The condenser was charged for a much longer period 
than in the present case; and this constitutes an essential dif- 
ference between the two methods. It isa great advantage of the 
ealvanometric method that good insulation is by no means so 
necessary as with the electrometric method, and thus one of the 
chief difficulties which such measurements offer is avoided. The 
question whether the accuracy of the results is disturbed by 
electricity produced by the friction of the gas admitted to the 
condenser does not here arise. 
Besides the seven gases already examined by Prof. Boltzmann, 
the author has also examined the dielectric constant of five vapours. 
The results obtained with the seven gases and the vapour of carbon 
disulphide agree very well with the determination of Boltzmann, and 
with the electromagnetic theory of light. The dielectric constants 
of the remaining vapours do not satisfy the conditions given by this 
theory. The deflection of the galvanometer-needle was observed by 
means of a telescope, mirror, and scale, and it should be noticed that 
a change of pressure of 690 millim. in atmospheric air caused a 
change in position of 185 scale-divisions. The results are given in 
the following table, in which D denotes the ratio of the two di- 
electric constants which the gas possessed at 0° C., and at two 
pressures differing by 760 millim. 
The refractive indices ” are taken from the determinations of 
Mascart. For the sake of comparison the corresponding results 
obtained by Boltzmann and by Ayrton and Perry are also given. 
