Observations on various Liquids. 239 



Things being thus prepared, the second Nicol is turned to 

 pure extinction, and the machine is set in motion. At the 

 instant of each spark through A B, and then only, there is a 

 clear instantaneous reappearance of the luminous oblong in 

 the polariscope. As the interval A B is shortened while the 

 machine is worked at a constant rate, the optical effect becomes 

 fainter by degrees till it disappears, which it does before the 

 air-interval is reduced to zero. The facts are exactly similar 

 to those observed already in 24. 



52. I have repeated this experiment with several variations; 

 and I shall now mention the principal results. 



(1) As long as the plate- machine is used, there is one 

 thing found to be essential in all forms of the experiment : 

 the discharging train from prime conductor through cell to 

 earth must be interrupted by an air-interval (an insulating 

 shell) A B of some length. The optical effect is observed at 

 the instant of the spark through A B, and at that instant only. 

 The position of the interval A B appears to be of no conse- 

 quence ; it may be before or after the cell, near earth, near 

 cell, or near prime conductor. Mere discharge is not suffi- 

 cient. Spark-discharge through a very short air-interval is, as 

 we have already seen, without sensible effect ; so also is glow- 

 discharge, brush-discharge, and even crackling spark-discharge 

 from the shaft of the insulated ball F upon the observer's 

 knuckle or upon an earth-connected ball. 



(2) When the prime conductor is connected with the knob 

 of a small Leyden jar whose outer coating is connected with 

 earth, the optical effect obtained as in 51 is very much in- 

 tensified. And when the discharges through the liquid are 

 obtained from a powerful Holtz (two 20-inch plates) which is 

 provided with the usual condensers, the effect becomes still 

 more brilliant. These observations agree with that already 

 made (in 51) in connexion with the shortening of the in- 

 terval A B. 



(3) When the two Holtz conductors were connected with 

 the outer balls D and F of the cell, the sparks through the air- 

 interval A B in one of the connecting-wires became extremely 

 dense and strong as A B was lengthened ; and now, for the 

 first time in my observations, the purity of the restoration in 

 the polariscope was in some degree lost, the old object (though 

 seen clearly enough) being displaced and distorted irre- 

 gularly. 



(4) Not even with the torrents of electricity rendered by 

 this powerful machine when worked at the hardest, did the 

 continuous discharge (air-interval zero) give any sensible 

 effect from pure extinction in the polariscope. 



