Electro-Magnetic Experiments. 4b 
Mr. Lukens, by coiling the ends of the wire into a spiral, thus forming 
a. spring to press the points of the wire against the poles of the mag- 
net; and by carefully insulating the ends in their passage through 
the brass stems. 
From this improved apparatus a a can be obtained on remov- 
ing the keeper with seldom a case of failure 5 frequently two sparks 
appear, one at each end of the wire, and in some experiments of 
Messrs. Lukens and Say a third spark was seen to pass from one of 
the brass rings. I have very frequently obtained. sparks also by ra- 
pidly replacing the keeper. 
To ascertain the nature of this spark, I endeavored to determine 
whether the removal of the keeper which produces the spark was at- 
tended with any electrical effect which would warrant the supposition 
that this was an electric spark. 
The cylindrical keeper, arranged as at first described, was used 
for this purpose. The ends of the wire spoken of as touching the 
poles of the magnet were bent outwards, to remove them from the 
poles, and connected with the wires of a galvanometer. On ap- 
proaching the keeper to, or drawing it from the magnet, an agitation 
of the needle of the galvanometer was produced, demonstrating the 
existence of a current of galvanic ae to which, therefore, the 
spark is probably due. — 
A few turns, say four or five of iron bonnet wire about the keeper 
of a magnet, which supported, by a contact with its whole surface, 
fifty pounds, produced a very sensible vibration in the needle; the 
contact and removal being made at times tending to increase the arc. 
Jn the experiments of MM. Nobili and Antinori, the account of 
which has been received since those which I have described were 
made, it seems that this development of a galyanic current which 
they observed led them first to suspect that the spark might be ob- 
tained. 5 
The following observations were made at the same time with those 
above described, on the direction of the galvanic current. The needle 
of the galvanometer used, was suspended by a fibre of silk; upon the 
stand of the instrument were two brass cups in the same plane with 
the coils of the wire surrounding the needle; in the cup corres- 
ponding to the south pole of the needle, the wire passing from south 
to north above the needle, and from north to south below, was placed ; 
into the north cup was inserted the opposite wire. Thus arranged a 
current of galyanism passing from sai to north will deflect the north 
