458 Mr. L. Schwendler on the General Theory 



reference to strength of birefringent action, the liquids appear to 

 be very unequal — carbon disulphide the strongest, paraffin and 

 kerosene the weakest. Compared with glass, they are much 

 weaker insulators ; but if allowance be made for this difference, 

 I think that, for intensity as well as purity of effects, carbon di- 

 sulphide is far superior to glass. In contrast with glass, all the 

 liquids are characterized by the absence of coercive force, and by 

 the rapidity of variation of birefringent action from point to 

 point of the electric field. The birefringent power is sustained 

 in liquids by the present action of electric force at each instant ; 

 it seems also to be determined at each point simply by direc- 

 tion and intensity of force at the point. 



43. Theory. — After the explanations already advanced in the 

 first paper (24), and again in this paper (34), I think I need do 

 no more than enunciate the three following assumptions : — 



1. The particles of dielectrified bodies tend to arrange them- 

 selves in files along the lines of force. 



2. Changes of molecular arrangement consequent upon rise 

 or fall of electric action are effected slowly and with difficulty in 

 solids, easily and at once in liquids. 



3. The lines of electric force, or the axes of molecular files, 

 are lines of compression in one class of dielectrics (glass &c), 

 and lines of extension in another class (carbon disulphide &c). 



The facts, when thus interpreted, afford a strong confirmation 

 of Faraday's theory of electrostatic induction ; and in whatever 

 way interpreted, they give promise of some new insight into 

 that interesting subject, the molecular mechanism of electric 

 action. 



I cannot conclude without expressing a hope, amounting almost 

 to a belief, that the plate-cell charged with carbon disulphide 

 will develop from the present rude beginning into a valuable 

 physical instrument, a very delicate optical electrometer. 



Glasgow, October 20, 1875. 



LV. On the General Theory of Duplex Telegraphy. 



By Louis Schwendler*. 



[Continued from vol. xlix. p. 126.] 



IN the two preceding investigations* I have given the solu- 

 tion of the first problem for the bridge method. This 

 solution established the general result of the double balance 

 being the best possible arrangement for the bridge method. In 

 the present paper I shall endeavour to find the solution of the 



* Journal of the Asiatic Society of Bengal, vol. xliii. part 2, 1874, pp. 1 

 and 218. Phil. Mag. 1874, vol. xlviii. p. 117; and 1876, vol. xlix. p. 108. 

 Journal Ttlegraphique, vol. ii. p. 580. 



