J. Nickles on a New Kind of Electro-magnet. 105 
founded on the well known process of oscillation, the other on 
contact attraction. This latter method I will briefly explain. 
made the experiment more to satisfy my own mind than to 
verify the influence which the length of the arms of an electro- 
magnet exercises on the weight sustained: for it always seemed 
to me that the fact of this influence was evident a priori; be- 
cause by increasing the length of a bar magnet we widen the 
distance between the opposite poles and diminish the neutraliz- 
ing effect which these poles exert between themselves. 
Yet as the contrary opinion has been sustained by the au- 
thorities above cited, I could not give as much weight to this 
reasoning as I should have done had it been experimentally 
proved. 'The fact in question is however put beyond doubt by 
my researches, and without giving full details, I can demonstrate 
it by a simple trial, easily repeated even as a common lecture- 
room experiment. I take a bar of iron surrounded by a helix 
of copper wire, which 1 place in the galvanic circuit; I select 
attracted without resting suspended: at this moment I place on 
the superior pole of the electro-magnet a cylinder of iron and 
immediately the armature suspends itself to the magnet and 
adheres there more or less energetically, to fall back when the 
added cylinder is withdrawn. tintin “hq gle : 
_ This experiment has been varied in many ways and has always 
8iven the results above stated. This addition of a cylinder to 
& Sar magnet obviously lengthens it a corresponding quantity, and 
removes farther off the opposite pole, whose disturbing action 1s 
iS Weakened, ; : 
Suppose a bar sufficiently long provided with a helix at one 
of its extremities; bend this into the form of a horse-shoe, we 
& Species of electro-magnet of which one pole only is cov- 
ered with wire, and which will not less exert a more considrea- 
ble attractive force when its two poles act at the same time on 
an armature; and this effect will evidently be obtained without 
” hecessity of augmenting the electric current. : 
_ But if, for illustration, we suppose a straight electro-magnet slit 
n the line of the axis for a sufficient distance, and then if each part 
of the divided portion be bent back parallel with the other portion, 
last, the middle portion, being the pole wound with wire, we 
‘ean electro-magnet, (see figures, next page) like a double horse- 
Magnet, with a single pole wound (7), and two outer arms 
(m) Which are naked : and when these two are made to act simul- 
