Observations on various Liquids. 99 



sible flicker ; and it becomes fainter and fainter on the whole 

 as the length of the spark diminishes. A little before the 

 spark vanishes^ the restorations disappear to sense ; and from 

 this point up to contact of ball and knuckle, and afterwards, 

 the action of the machine is without sensible effect in the 

 polariscope. I have made some additional observations on 

 this form of effect; but, to prevent confusion, I reserve them 

 for a little. 



25. Bromtoluol (C 7 H 7 Br). — In its electro-optic relations, 

 this liquid resembles carbon disulphide and benzol on the one 

 hand, and nitrobenzol on the other, giving the two kinds of 

 effect clearly, though not intensely. Tested in the usual way, 

 it acts as an imperfect insulator. When the connecting-wires 

 are in position, the sparks from the prime conductor are 

 greatly attenuated, and are reduced in length from an inch or 

 more to about an eighth of an inch. 



When bromtoluol is examined electro-optically in the same 

 way as CS 2 (8), the electric action gives a continuous restora- 

 ' tion from pure extinction in the polariscope. The effect is not 

 stron o- ; but it is quite certain, and certainly of the same kind 

 as in CS 2 , being neutralized perfectly by horizontal compres- 

 sion of glass. And, again, when the discharging train from 

 the prime conductor through cell to earth is interrupted at any 

 point by an air-interval, there is a good restoration in the po- 

 lariscope at the instant of each spark, the effect being appa- 

 rently of the same kind as that observed in nitrobenzol, 

 though not nearly so strong. Bromtoluol is an inconvenient 

 liquid to work with, partly from the very irritating, onion-like 

 action of its vapour on the observer's eye, and partly from the 

 difficulty of getting a clean charge — a difficulty that I have not 

 once overcome perfectly. 



26. Trial of other Liquids for the Nitrobenzol Effect. — 

 Carbon disulphide, carbon dichloride, terebene, and most of 

 the other liquids already mentioned were all examined care- 

 fully in the same way as nitrobenzol (24). The optical effects 

 accompanying discharge were either insensible, or of the kind 

 noted in the two following cases. 



Amyiene. — The sparks obtained from the earth-ball were 

 Very short and attenuated, and were accompanied by distinct 

 restorations in the polariscope. The restorations were far from 

 instantaneous, each of them rising sensibly to a maximum in- 

 tensity, and then sensibly falling. The effect improved in 

 strength as the knuckle approached the ball till the length of 

 the spark reached zero ; and then the effect was to sense con- 

 tinuous, and certainly at its greatest intensity. 



Benzol. — The sparks from earth-knob to knuckle were very 



