on the Electric Arc. 137 



in hydrogen. But if the arcs are carbon-metal, the limit 

 varies as follows : — 



Approximate limit of length 

 Arcs in H. (at 110 volts supply and 15 amp.). 



Carbon-Iron 0*05 inch. 



Carbon-Aluminium O07 „ 



Carbon-Copper 0*13 ,, 



Carbon-Carbon 0*07 ,, 



In coal-gas, Cu-C gave an arc-length of 0'1 in. at 

 15 amperes. 



In air, the arcs may always be drawn out longer than in 

 hydrogen or coal-gas. 



Cu-Cu gave 0'2 in. arc-length with only 2'5 amperes, 

 which cannot be obtained from carbons with 110 volts 

 supply. 



The lowest limits found were — 



for C-Al, 0*3 in. at 10 to 12 amperes, 

 and C-Cu, 0-26 to 0-34 in. „ 5 „ 6 „ 



The following comparisons of maximum arc-length were 

 made, showing the effect of reversing the polarity when 

 •carbon formed one of the electrodes : — 



Arc. Max. length. Current. Voltage across Arc. 



inch. 



Al-C 0-5 10 — 



C-Al 0-3 10 to 12 — 



Cu-C 0-6 to 0-8 7 „ 9-5 70 to 53 



C-Cu 0-26 „ 0-34 5 „ 6 85 „ 75 



Rote. — Cu-C with a length of 0*3 in. at 6 amperes gave only 

 47 volts across the arc. 



Volt- Ampere Characteristic Curves of Arcs 0*05 in. 

 in Length. 



The length of arc, 0*05 in., was chosen as it was the 

 longest which could be steadily maintained for all the arcs 

 under examination. This length is so short that accurate 

 measurements of unstable arcs are extremely difficult, and in 

 such cases it is only fair to consider the curves given as gene- 

 rally approximate. Fig. 5 (PL VII.) gives the volt-ampere 

 characteristics of the arcs in air in which carbon was one of 

 the electrodes. It will be seen that when the carbon is 

 negative, it makes very little difference what the positive is 

 in regard to the general position of the curve, while if carbon 

 is positive a great variation occurs according to the material 



