a Dielectric in an Electrostatic Field. 200 



mechanical compression along the lines of force, and others, 

 as resin, a mechanical expansion ; whereas Maxwell's formula 

 in all cases calls for tension alone. 



Let us suppose that the Kerr effect can be explained by a 

 mechanical compression produced by the charged electrodes. 



The following calculation makes clear that this deformation 

 is not sufficient to produce the observed changes of length 

 normal to them : — 



Let V=2 cm. spark =130 C.G.S. units. 



d =0-25 cm. 

 K=5. 

 E = 3xlO u , 



Then F/S = K(Y *~J ^ = 54000 dynes. 



The change of length along the lines of force equals 



?x d/E = 54000x^f- 5 x 10- n = 4-5xl0- 8 cm. 



From Poisson's ratio of 0*25, the change of length normal to 

 the lines of force equals l'l x 10~ 8 cm. My tube was 57 cm. 

 long, giving a total change of length of 6'3 x 10~ 6 mm. This 

 corresponds to less than half a division of the micrometer- 

 screw and hence could not have been observed. 



From experiments of D. B. Brace * on the action of a mag- 

 netic field upon transparent media — " We may also conclude, 

 according to Maxwell's view of the state of polarization and 

 stress in such a medium, that the pressure at right angles to, 

 and the tension along, the lines of force (equal in both cases 

 to H 2 /87r) affect the propagation of light by an amount less 

 than 2*0 x 10 _14 Xfor a C.G S. unit of intensity per centimetre/' 



Note. — In 1895 the writer published an article t " On the 

 Changes in Length of Iron Wires by Magnetization," and, 

 induced by his interpretation of Maxwell's theory, he made a 

 correction for an elongation due to the strains in the medium 

 proportional to B 2 /8tt, which materially altered the interpreta- 

 tion of the results obtained. The paper aroused some criticism 

 which appeared in volume liii. of ' Nature/ Dr. Chree and 

 Dr. E. Taylor Jones argued that the true correction was a 

 contraction. Professor J. A. Ewing wrote : — " For some time 

 I have been aware that the passage referred to in my book on 

 the ' Magnetic Induction in Iron ' requires correction. The 



* Brace, Phil. Mag. xliv. p. 349. 



f Phil. Mag. (1895) ; Phys. Rev. iii. p. 210, 



