410 Electro-Magnetic Experiments. 
Extract of a second letter from J. Saxton, of Philadelphia, to Isaiah Lukens, 
dated London, May 11, 1832. ; 
‘Since my last I have beard of a method of producing a spark 
from a magnet, discovered, I believe, by an Italian. This experi- 
ment [I made at once upon a large horse-shoe magnet which I am 
making for Perkins and his partners. One of your large magnets 
will answer the same purpose. Make a cylinder of soft iron of an 
inch or three fourths of an inch in diameter, and of the usual length 
of the keeper ; place two disks of brass or wood upon this cylinder, 
and at such a distance apart that they will conveniently pass be- 
tween the poles of the magnet; between these, wind, say fifty feet of 
bobbin wire, which may be of iron covered with cotton ; let the ends of 
this coil be bent over the ends of the cylinder and brought down ua- 
til they touch the poles of the magnet, the ends should be of such a 
a length that on bringing the cylinder to the magnet, one of the ends 
will touch when the cylinder is about half an inch from the magnet, 
and the other atone fourth of an inch. ‘The cylinder being thus ar- 
ranged, and in contact with the magnet, on drawing it suddenly away 
a spark will pass between the end of the wire and the pole of the 
magnet.” 
The apparatus alluded to in these letters was, soon after their re- 
ceipt, put together by Messrs. Isaiah Lukens and Benjamin Say ; it 
is figured in the cuts which follow. —__ 
Fig. 1 represents the apparatus of No- 
bili for procuring the spark from a mag- 
net. A,B, is acylinder of soft iron; up- 
on this are two brass rims, between which 
bobbin wire (wrapped wire of iron and 
copper were both successfully used,) is 
wound ; the ends of the wire coil are car- 
ried out at the opposite ends of the cyl- 
inder, and being bent downwards, pass 
through holes in two brass (or wooden) 
pins. . The spring of the wire causes both 
the ends to touch the magnet when the 
keeper is attached; when it is withdrawn 
one end projects beyond the other.* 
This apparatus has been improved by 
C 
* This condition appears by subsequent experiment to be by no means essential. 
