1876.] the Length of Electric Sparks. 169 



means of a double key discharger, or by a rapidly revolving commutator 

 reversing up to 352 times in a second. One terminal in the above 

 determinations consisted of a point of 30°, and the other of a slightly 

 convex surface 0*46 inch in diameter. 



While making these measurements, we noticed in a nearly dark room 

 that when the point was negative a glow, in form like a paraboloid, was 

 seen surrounding it long before the spark passed, and its appearance 

 afforded, by its increasing brilliancy, useful information to guide us as to 

 the more cautious approach of the terminals ; gradually the sugar-loaf -like 

 glow extended to the positive terminal. With 1800 cells the glow was 

 seen when the terminals were 0*0545 inch apart, the spark passing at 

 0*0345 inch ; with 2400 cells the glow began at a distance of 0*0865 inch, 

 the spark passing at 0*0535 inch. Moreover it was noticed that the disk 

 (positive) became covered all over with a peach-like bloom, which became 

 stronger in the centre as the terminals were made to approach each other, 

 giving rise to Newton's iridescent rings. 



In order to study more readily the phenomena accompanying the glow 

 preceding the spark, the whole series of 5640 cells was used, the terminals 

 being a point as before of 30° for one, and sometimes a flat disk 1*1 inch 

 in diameter or a slightly convex one of 0*8 inch in diameter for the 

 other. In all cases a peach-like bloom deposited on the disk, which was 

 connected with the silver (positive) terminal ; and when the flat disk was 

 used the deposit was notably greater at the periphery and the centre 

 than in other portions of it. With this number of cells, 



With the flat disk the glow occurred at 1*073 in., the spark at 0*139 in. 

 With the slightly convex disk „ 1*124 in. „ 0*140 in. * 



To ascertain whether a current really passed when the glow appeared, 

 various vacuum-tubes were interposed in circuit between the battery and 

 one of the terminals ; in all cases they were illuminated even before a 

 glow was perceptible on the negative pole ; their interposition, as was to 

 be expected, shortened the spark and diminished the distance at which 

 the glow was perceptible. For example, with a hydrogen tube, having a 

 capillary portion between two larger tubes, such as is used for spectrum 

 experiments, and offering a resistance of 190,000 ohms, the glow occurred 

 at 0*939 inch, the spark at 0*092 inch. 



A tube of 31 inches between the terminals, and offering a resistance of 

 350,000 ohms, was brilliantly illuminated when interposed between one 

 terminal and the battery ; when the terminals were separated the ex- 

 treme range of the discharge was 1*2 inch, and before any glow was visible 

 at the negative electrode. How much further between the electrodes it 



* Postscript, Jan. 7tb. — At the suggestion of Professor Stokes, who saw the experi- 

 ment repeated, the point was made positive, when a longer spark was obtained, namely 

 0*154 inch and 0*164 inch. 



