Charles Grafton Page. 15 
Of similar interest are thor phenomena of which we extract 
the following account from a paper by him in the American 
Polytechnic Journal, vol. i, p. 305, on the subject of the diffi- 
culties encountered in the great electro-magnet engines, from 
the destruction of metallic contact pieces by the induction 
sparks. ‘‘In the attempt to remedy this defect, our first ef- 
forts were directed to the entire suppression of the combustion, 
The current was first broken under oil, naphtha, and other hy- 
drocarbons, with the expectation that the absence of oxygen 
would give a favorable result. To our surprise, however, we 
found that when the current was broken between two metallic 
surfaces entirely covered with oil, the consumption of the metal 
was much greater than when the break occurred in the air be- 
tween clean metallic surfaces. The hydrocarbons were rapidl 
decomposed by the intense heat, and a black powder formed 
which appeared to be a carburet of the metal, though it was 
not chemically examined. is curious result we explain ag 
follows :— The hydrocarbons are non-conductors of electricity 
and heat, and when the metallic surfaces are separated, the 
secondary spark is concentrated, and its action upon the metal 
becomes topical and intense. The action is analogous to that 
in the familiar experiment of perforating a glass bottle with 
the electric spark ; a very feeble spark passing through oil, be- 
ing sufficient to make a hole through a thick glass bottle. "When 
the secondary current is one of great power, the oil is thrown 
about with considerable force. To illustrate this, we broke the | 
cireuit under oil contained in a small cavity in a block of wood, 
having first driven by a hammer a small bullet into the cavity 
over the oil. The spark from a large magnet drove the bullet 
out with some force. 
“We next tried breaking the circuit in the following manner: 
—We procured two thick plates of glass, ground carefully to- 
ough each of these plates we drilled a hole one quarter inch 
in diameter, and when the plates were so arranged that the 
= holes corresponded with each other, they were filled with mer- 
_ eury in the following manner :—The under side of the hole in 
the entire absence of air. Here occurred a iant and unex- 
pected phenomenon, Whenever the break was me, the spark 
