452 Dr. J. Kerr on a new Relation between 



of the electric field remaining dark. Hardly in any case of 

 strong electric action is there absolutely no perceptible restora- 

 tion of the band at the centre of the field ; but the effect is 

 always comparatively very faint at 0° and at 90°, unless the 

 angular adjustments are very imperfect. 



32. Fifth experiment. — The optical pieces close together, and 

 the flame broadside-on to the polariscope. The light now fills 

 the transverse section of the first Nicol, and extends well beyond 

 the interterminal part of the electric field all round. When the 

 principal sections of the two Nicols are at right angles to each 

 other, there is a good initial extinction between and around the 

 terminals. 



Plane of polarization at 45° to the horizon : the electric 

 action brings out a distinct illumination over a good extent of 

 the field, very faint at the outer parts, but brightening towards 

 the axis of the fields and rising into two patches of comparatively 

 strong light at the terminals. When the compensator is intro- 

 duced and compressed horizontally by a continuously increasing 

 force, it produces extinction first in the outer equatorial parts 

 of the field, then at the central parts ; and then, after a consi- 

 derable increase of compression, it extinguishes or greatly weakens 

 the bright patches at the terminals. 



Plane of polarization horizontal or vertical : the electric action 

 brings out a narrow border of glowing light along the outlines 

 of both wires, and a fainter illumination in other parts of the 

 field, leaving a broad equatorial band very dark. 



33. Summary. — Carbon disulphide is birefringent when di- 

 electrified, acting upon transmitted light as glass extended 

 along the lines of force. The electrostatic force and the bire- 

 fringent power increase together : they also vanish simultane- 

 ously, the optical effect disappearing abruptly and totally at the 

 instant of electric discharge (not through the bisulphide). Some 

 irregularities are observed in the experiments when the eleclric 

 action is intense ; but these are certainly due in great part, if 

 not wholly, to the want of even approximate uniformity through 

 any considerable extent of the electric field. 



34. The optical effect with which we are now concerned can- 

 not be attributed to specks of dirt, however fine; for it im- 

 proves in every respect as the liquid is made cleaner; and 

 when any solid particles are visibly present, it improves for 

 some distance outwards from the axis of the field. Neither can it 

 be attributed to electric discharge ; for the bisulphide is an excel- 

 lent insulator, and any discharge which is known to pass through 

 it in the course of the experiments is very sharply disruptive. 

 The effect appears to be due simply to the action of electrostatic 

 force upon the molecules of the liquid. 



