THE INDUCTION OR ELECTRIC STRAIN 



59 



each of two portions, one positively, the other negatively electri- 

 fied, and that the two portions are connected by tubes, or, as more 

 easily drawn, lines, of strain, and that they are pulled towards each 

 other by the forces accompanying the strain. In making this sup- 

 position we are really identifying chemical with electric attraction. 



Fio. 50. 



\\ nm>t al>o have motion of the two parts of the molecule round 

 each other, otherwise the tuo uould lx? pulled into contact and the 

 tuU-s would disappear. But \\r may for our present purpose leave 

 the motion out of account and con vcntionally represent a molecule 

 as in Fig. 51, where we suppose, for dcliniteness, that there are 

 iinccting the two parts. If the dielectric is 



in a neutral condition we nm>t suppose that the molecular axes 



are distributed tonally in nil dhfebons, so that there are equal 



iiiiinhfrs of and negative ends of axes facing in any 



fion, and Unit is no resultant electric action outside, and 



tore no external indication of the molecular charge-*. 



: 



Fio. 52. 



Let us now suppose that the medium is between two conducting 



i ~M, and that these become charged. We may suppose 



that, in tli < li irging, -f and arc moving along the surfaces, each 



pair' iii< < i. (1 }>\ ;i tube of strain aa PQ, and \^- 



think ot MM ii ,i tube as picking out, as it \\eiv. the moli-cuh - 



