Electricity and Light. 453 



The character of the effect in the first and second experiments 

 (29, 30) is of some consequence in this connexion. When the 

 liquid is perfectly clean and the initial extinction pure, the light 

 is brought out by electric action with remarkable distinctness 

 and delicacy, and without a trace of unsteadiness or distortion. 

 The whole surface of the visibly restored flame is continuous, and 

 its outline beautifully sharp. These patent characteristics of the 

 effect in the polariscope are strongly suggestive of some regular 

 arrangement of the ultimate molecules of the bisulphide as the 

 immediate cause of the phenomena. And the compensator con- 

 firms the suggestion, while it shows, further, that the new 

 arrangement of the molecules either is in itself, or is at least in 

 effect equivalent to, a positive uniaxal structure with its axis 

 directed along the lines of force. 



Of such a structure in the case of a liquid, I see no probable 

 explanation except that already advanced in the case of glass. 

 The particles of the bisulphide, when electrically polarized, throw 

 themselves into files along the lines of force ; and the system 

 of electric curves thus physically constituted gives the liquid a 

 new structure, which is symmetrical at each point with reference 

 to the line of force through the point. If we adopt Faraday's 

 theory of electrostatic induction, I think that the common ex- 

 periment of the magnetic curves affords rather a proof of the 

 present theory than an illustration. Another illustration of the 

 same kind is afforded by the interterminal chain of solid particles 

 in the present experiments. 



35. I may mention here that I found it very useful in some 

 of the preceding experiments, and still more useful in some of 

 the following, to draw sparks occasionally from the prime con- 

 ductor while attending to the polariscope. The abruptness of 

 the change thus produced in the field of vision adds remarkably 

 to the delicacy of the observation. 



Benzol. 



36. The plate of liquid is very transparent, and an excellent 

 insulator. A steady electric action of the requisite intensity is 

 obtained from the plate machine. RnhmkorfFs coil is not suit- 

 able in this case, as a few of its powerful discharges through the 

 cell are sufficient to contaminate the liquid with a fine deposit of 

 carbon. 



Here again the only difficulty is to get the liquid clean 

 enough. A few specks of organic fibre or of other dust from 

 the air of the room or elsewhere, if they get into the cell, are 

 sure to obscure the principal effect extremely, if they do not 

 quite mask it. When the specks are numerous enough, they 

 form a chain between the terminals, as in the case of carbon 



