1879* 1 Carbons in the Electric Lamp . 529 
centimetre from the end of the carbons. At other times it 
would revolve about the electrodes something like a spiral 
flame in a pyrotechnic display. This leaping and dancing 
of the arc is, of course, fatal to its employment for pro- 
jection. 
In order, if possible to remedy these defects in a lantern 
which I have in almost daily use, I made the following ex- 
periments : — I first took the specific gravity of three speci- 
mens of carbon, obtained from different dealers, one in 
France and two in America. The specific gravity of the 
French carbon was 1*85; of No. 1, American, 1*53; of 
No. 2, American, 1*55. The French carbon is hard, of a 
grayish black colour. The American carbon is soft, easily 
broken up, and has no sign of a metallic lustre. The light 
from the French carbon is quite steady, and displays very 
little of that tendency to flicker so troublesome in the 
American varieties. 
A positive French carbon, which had been used for several 
hours, until consumed nearly to the lamp, burned away at 
the point, but otherwise retained its original shape. This 
carbon was used without any previous preparation. 
A soft carbon, however, of the same size as the preceding, 
became red-hot to a distance of 4 to 6 centimetres from the 
end, and rapidly wasted away; after being in use for half 
an hour it was reduced to a slender, tapering form. 
I first tried the plan so well known in France, but so 
seldom tried here, of coating the carbons with a film of 
copper. The precipitation of the copper should take place 
slowly, and with a current so regulated in quantity and in- 
tensity as to produce no spongy deposits. When the soft 
carbons were thus prepared they worked beautifully for a 
short time. The light was brilliant and steady, while any 
green tint imparted to it by the volatilised copper produced 
no effeCt whatever prejudicial to the purpose in view. But 
as the carbons, little by little, became heated, the copper 
film oxidised, and after half an hour the carbon was again 
reduced to the slender form above described. 
I next tried the expedient of setting a copper wire, 
0*4 m.m. in diameter, into the centre of the carbons. With 
a thin saw I cut a longitudinal groove to the centre of the 
carbon, and after inserting the wire fixed it in place by filling 
the groove with plaster. The upper end of the wire was 
left projecting so that it could be brought into aCtual contact 
with the clamp. I hoped from this contrivance to hold the 
origin of the arc steadily at the end of the carbon, and at 
the same time, by increasing the conducting power, to pre- 
