THE ELECTRIC AND LUMINIFEROUS MEDIUM. 
295 
longitudinal expansion of the glass could also be measured microscopically at the 
same time. It was found by Quincke that the coefficient of volume expansion was 
always about three times that of this longitudinal expansion of the glass dielectric, 
just as the above theory indicates for the elastic strain in a bulb of uniform thickness! 
The euoneous deduction was however made from an imperfect theory, that electric 
expansion of solids is uniform in all directions, like expansion by heat, and so in no 
part due to mechanical forces of attraction. 
By using the formula of § 76 along with various known physical constants, a test 
of^ the order of magnitude of Quincke’s determinations may be obtained. Thus 
with a^ striking distance of -4 centim. between brass balls 2 centims. in diameter, the 
expansion in volume of a flint glass condenser of thickness -06 centim. was found to 
be I X 10 - 6 .-“ According to Bailee’s experimentst this striking distance corresponds 
to a difference of potential of 47 c.g.s. The expansion of volume of the bulb due to 
the mechanical force is [2 - K + X/p}/(X + |p).FV 87 r, where, according to 
Eveeett’s experiments! for flint glass, = 25 X lO^o c.g.s. and X = p, and accordincr 
to Hopkinson K is about 7. This gives for the thickness under consideration a 
coefficient of expansion equal to - 0-24 X 10-^ while the observed value was 
0-75X10 6 .§ 'J'he difference between them, in so flir as it does not arise from 
experimental uncertainties, is an intrinsic superficial expansion of the glass, which 
arises directly from the transverse polarization itself, and is not due to the mechanical 
forces caused by it. That there is such intrinsic alteration due to electric excitation, 
IS independently suggested by Quincke’s observation that the values of the elastic 
constants of the material are slightly altered by that cause. 
In the case of a fluid the effect of electric excitation is to diminish the hydrostatic 
pressure; consequently expansion should result when the jflates of the condenser are 
fixed. ^ This agrees with what happens for most fluids. But the fatty oils form an 
exception; thus for them at any rate there is an intrinsic electric contraction super¬ 
posed on the expansion due to diminished pressure. 
^ In both these cases the intrinsic change of volume is of order of magnitude not 
higher than the change due to the mechanical stress. 
The observation of Quincke 1 | that a thin glass-tube condenser, with walls thicker 
on one sfoe than the other, becomes curved (in accordance with the theory above) 
Mdien it IS electrically charged, virtually affords a convenient method for studying 
* Loc. cit. ‘Wied. Ann.,’ 10, p. 190. 
t ‘ Annales de Chimie,’ 1882 : quoted in J. J. Thomson’s ‘ Recent Researclies,’ p 87 
t ‘ Phil. Trans.,’ 1868, p. 369. 
§ According to Quincke’s own determinations (loc. cit. p. 187) the value of K would be about 12 
winch IS a great deal higher than Hopkinson’s results, and would give an expansion - ORO X 10-«. It 
seems possible that these determinations, which involve considerable unitary complexity, may be 
wrong y recorded, as Quincke’s reduction of them sometimes appears to give values for K that arc less 
tlian unity. 
II Loc. cit., p. 394. 
