180 Lord Armstrong. On a Multiple Induction 



the area of No. 31 gauge, and from these figures an estimate may be 

 formed of the difference in the development of heat on the negative 

 and positive sides. When I employed two carbon electrodes of equal 

 thickness the greater development of heat on the negative side was 

 still very decided, though the difference was not so conspicuous as in 

 the case of the platinum electrodes. In the arc lamp the superiority 

 of heat is largely on the positive carbon, and it is difficult to account 

 for the contrary result obtained in my experiments. 



I further varied tWse experiments by taking the positive discharge 

 from the surface of acidulated water, in which case the negative 

 electrode was melted by the arc flame springing from the water. I 

 also used for the positive electrode a lump of ice sprinkled with salt 

 to make it conduct, and obtained the same result. A reversal of the 

 current made the water boil at the surface and melted a hole in the 

 ice, while the positive platinum remained unheated as usual. In 

 another experiment I enclosed the two platinum electrodes in glass 

 tubes sealed at the outer end, leaving about ^ inch of the platinum 

 wire (No. 27) projecting beyond the glass. These sealed ends I im- 

 mersed in distilled water, and succeeded in melting .the exposed por- 

 tion of the negative wire while submerged in the water. 



In another case I used an iron wire (No. 20 gauge) on the negative 

 side, retaining the platinum wire for the positive electrode. In this 

 instance the melted globule ran back very quickly to a distance of 

 about \ inch, then stopped and burst into intensely brilliant flame, 

 which showed a strong disposition to cross over to the opposite side, 

 but appeared to be beaten back by the blue flame of the arc. Part, 

 however, did reach the positive electrode, and condensed upon it in 

 the state of an oxide. My attention having been thus directed to an 

 appearance of conflict in the arc, I discarded the iron wire and sub- 

 stituted a platinum wire dipped in a brine of common salt, so as to 

 impart a distinguishing yellow colour to any flame that might issue 

 from the negative side in opposition to the flame emanating from the 

 positive side. This caused the arc to be exhibited under two colours 

 a very decided yellow on the negative side, and the same pale hazy 

 blue as before on the positive side, but the yellow flame was beaten 

 back by the blue flame. Flecks of yellow could, however, be seen to 

 get across occasionally. It was not so easy to produce yellow flame 

 from the positive electrode, because there was not sufficient heat in 

 the positive wire to volatilise the salt ; but by using a separate wire 

 encrusted with salt, and holding it alternately immediately in front 

 of each electrode, I could produce yellow flame on either side and 

 observe the difference of its behaviour in the two cases. When held 

 on the positive side a dense unbroken stream of yellow flame sup- 

 lanted the previous blue, and passed over bodily to the opposite 

 wire; but when held on the negative side the yelloAv flame struggled 



