188 Lord Armstrong. On a Multiple Induction 



gauge. A brilliant jet of pale-yellow flame was projected from tlie 

 positive wire as shown, but did not go beyond the candle flame. 

 Nothing was to be seen in the interval between the flame and the 

 negative wire, which, nevertheless, heated as usual. It is quite 

 possible, however, that there might be an arc communication 

 rendered invisible by the intense brilliancy of the jet within the 



candle. 



Fig. 2 shows the negative wire immersed in the flame and the 

 positive clear of it. Here the direction of the jet was reversed, 

 and the negative wire, instead of heating, became covered with a 

 clot of carbon ; but with a thinner wire the negative did heat, and no 

 clot of carbon was formed. 



Fig, 3 shows the effect when both wires were clear of the flame. 

 In this case two jets of brilliant light appeared in the flame with a 

 slight separation and an appearance of conflict between them. 



Fig. 4 shows two candles brought together so as to form one broad 

 flame in which both wires were immersed. In this case a large clot 

 of carbon accumulated on both wires, and neither of them heated. 

 An examination of the carbon showed it to be deposited in very 

 beautiful florescent and fern-like forms. It would have quickly 

 filled up the whole interior of the flame had flakes not fallen off as 

 the mass grew in size. 



Fig. 5 shows the two candles drawn asunder so as to leave a small 

 vacant space between them. Nothing could be seen to pass this 

 dividing space, but the results remained the same as in the previous 

 case. 



The repellent action which is exhibited in all these cases is not easy 

 to explain, seeing that the discharge on both sides seems bounded by 

 the exterior of the candle flame. 



I am at present continuing my experiments, but under the altered 

 conditions of discharge in rarefied air. As these experiments are far 

 from complete, I must reserve them for a future communication. 



I must not, however, close this present paper without referring to 

 an experiment of quite a different type from any of the preceding, and 

 which was made long ago with my large hydro-electric machine. 

 In that instance two wine glasses, filled to the brim with specially- 

 distilled water, were placed in juxtaposition, leaving only a space of 

 about inch between them. A long cotton thread was then coiled 

 up in one glass and the upper end of the thread dipped into the 

 water of the other glass. When the steam was turned on the thread 

 was drawn out of the glass in which it was coiled and conveyed with 

 great rapidity into the other glass, and for a few moments a rope of 

 water remained suspended between the lips of the two glasses. It 

 was only when the machine was at its maximum power that I could 

 do this, and it never reached its highest power when used within the 



