436 
BR. O’SHAUGHNESSY ON ELECTRO-MAGNETISM. 
inches, are disposed twelve small horse-shoe elec- 
tro>magne(8 of the dimeiisioiis already desciibcd. 
One end of each spiral (see fii. 3) is soldered to a 
disk of copper, c, one inch in diameter placeri 
on the axle conceniric with the wheel, and revolv- 
ing in a little cup of mercury m, connected hy a 
wire with one pole of a galvanic battery. The 
other ends of the spirals are arranged on a small 
ivory disk, so that the ends of fZ/ree dii> in a simi- 
lar cup of nieicuty, one always lising from, as the 
third enters, the surface of the mercury in coit- 
iiexion with the other pole of the battery, as is 
readily seen in the fig. 6. The wires on this, the 
interrupted disk, are further So arranged by gently 
twisting their continuations towards the wheel, that 
the wire of the magnet which is horizonial when 
the wheel is at rest, is perpendicular to and beneath 
the surface of the mercury. A glance at the figures 
3, 4, 5, and 6 will make this description readily 
intelligible. 
Let ns now suppose this wheel mounted on pro- 
per supports, and conductois from a galvanic battery 
brought into contact with the wires leading to the 
mercuiial cisterns belonging to the wheel. If the 
wiieel he made to turn, the smaller copper disk 
revolves in the mercury continually, and hence all 
the w lieel irons are ready to be excited on com- 
pleting the galvanic circuit. I his is effected for 
tivo hy the interrupted disk revolving in the other 
cistern, so that of lire twelve irons, are always 
rendered eltclto-magiiets in succession as the 
wheel turns, as sh.ewn in fig. 7. henever the 
bent iron descends to the angle of 30« above the 
horizon, it becomes magnetic ; its second wire then 
touching thesuifareof the mercui iai cistet n at an 
acute angle. When horizontal it is still magnetic, 
its second wire being then perpendicular to and in 
the cistern. Wlien at an angle of 30® below the 
horizon, the wire then enterges from the meicury, 
and the iron loses all magnetic power. Thus, in 
the revolving wheel, two magnets are always exci- 
ted, their poles being in alternate order, and the 
magnetism of each only exists while the bent iron 
is passing ftomSO'* above to 30® below the hoiizon. 
Tire aliernale order of the poles is ensured by the 
ends of the spirals of each magnet being alternate- 
ly crossed, so that liie successive poles at the same 
side are in contact alternately with the zinc and 
copper ends of the battery , (see figures 3, 4, and 5.) 
The next essential part is 
the external or principal electro- 
magnet. 
It consists of a bar of soft iron, bent so that the 
legs are parallel to each other at the distance of 
two inches. The legs of the bar employed in my 
first expetiment, were but ibtee incbes long, | in 
diameter. It was wound with a double spiial in 
the usual iiiauner. Its sustaining force, when ex- 
cited by a ten-plate 4-inch Wollaston’s baiteiy, 
averaged 30 Ihs. It weighed with its spirals com- 
plete, 11 Troy ounces. 
This electro magnet was placed horizontally "U 
a wooden support- its legs on the same plane 
with, and close to, but not touching, the presenting 
legs of the hot izontal wheel magnet. 1 he ends of 
the spirals from this electro-magnet, were led to 
cups containing mercury, in connexion with the 
poles of a galvanic battery, through a moveable 
system of conductors, devised so that tlte revolution 
of the wheel changes the poles of the external 
magnet every tinre that one of the wheel-magnets 
becomes hot izontal, (see fig. 7) This part of the 
model I shall irext describe ; i term it simply 
THE POLE-CHaNGER; 
and its construction is illustrated iii the figs. 8, 9, 
10, and 11. • . 1 j .1 
The object of this contrivance is, to lead the 
current of electricity alternately to the right and 
leftside orlegrrf the principal external magnet. 
We accomplish this hy guiding the cut rent diagotral- 
ly (see fi“. 10) in one instant, and making it return 
diagonairy, so that its course from the battery and 
its return are exemplified by an 8 figure. In the 
second instant, fig. It, tlte crrrtettt proceeds di- 
rectly to the opposite leg of the principal magnet, 
attd returns as directly at the other side ; its course 
t < semblitig the letter O.* 
Twelve circular holes, each half an inch iti diame- 
ter, and half an inch in depth, are cut in a piece 
of wood at one inch distance from each other, as 
shewn in the figure, and fixed wiies are led from 
cavity to cavity ; so that 1, *2, 3 at e at tight attgles 
to, and in metallic conitexion with, each olhei. 
Iii3, 4, 5the same occurs, and so with tlte te- 
maining numheis, as may be seen in the plate (figs. 
10 and 11.) Glass tubes two inches in depth are 
cemeuted into the cavities in the wood, and about 
f of an inch of mercury poured into each tube. 
Now by leading a silk covered wire from 2 to 8, 
attd another from 5 to li, it is obvious we make 
the galvanic current flow from 1 to 2, cross to 8, 
circulate round ihe magnet at 7, telurn from 6 to 
II, and reach the battery again at 12, rendering a 
.a north, and 6 a south pole. If then we remove 
these diagotial wires and insert wites from 3to4 
and from 9 to 10, (see fig. 11,) the current takes 
exactly the different direction : instead of pro- 
ceeding to «, it goes to b, which instead of a south 
becomes a north pole. 
These connecting wires are accordingly so ar- 
ranged with a set of levers made of slendei wo >d, 
that when the lifting of Hie lioiizoiual long lever, 
shewn ill the diawing, causes ilie diagonal wires 
or drawbridge to rise from the mercuiy, the direct 
or lateral wires full in at the same moment; the 
c irreiit instantaneously cliaiiges ; and again, when 
the long horizontal lever escapes the lifter I shall 
presenily desciiiie, the diagonal wiies fall into 
the mercury, and a small square slip of ivory on 
which these wires are fastened strikes the ends of 
the litile lateial levers, and causes their wires to 
leave the meicury and break the lateral conimuni- 
caiioii. 
The lifting and falling of the long horizontal 
lever is effected by six wires placed parallel to the 
axle, half an iiicl) within the circumference of the 
wheel, and immediately across the legs of each 
alternate wheel-magnet ; see fig. 5, tv. w. By this 
arraiigemeiil the poles of the pi iiicipal 
magnet change every time that one of the twelve 
wheel-magiiets becomes horizonial. 
Let us then suppose these seveial parts placed 
on a frame, as shewn in the chief diawing, and the 
wheels and exieriial magnet excited eacli hy a 
Wollaston battery of ten-p'ates. The wheel is at 
rest; communication iliiough the pole-changer 
diagonal : the right side of the external magnet has, 
we will suppose, a norih pole, and the wheel-mag- 
net at the corresponding side at an angle of 30® 
above the horizon has a soiitii pole (and vice vers4 
at tlie opposite side). The external magnet for- 
cibly attracts tlie internal one, which desceinls to 
the horizontal plane. At this inomeni the wire on 
the circumference of the wheel, acting on the long 
lever of the pole-changer, lifts the diagonal wires, 
and permits the lateral to fall; the poles of ihe 
external magnet change: they repel those of the 
internal or wheel-magnei, which they had just be- 
foie attracted, and they attract the next wheel- 
magnet exactly in the same manner, owing to the 
alternating anangement of the wheel-magnets pre- 
viously desciibed. When this, the next wheel.- 
magnet, becomes horizontal, the long lever of Ihe 
changer falls, the diagonal commniiicaiion is restor- 
ed, the lateral interrupted, and the poles of the 
external magnet again changed. Thus the extei- 
nai magnet changes its poles every time an inter* 
nal magnet presents; and as the change makes 
the opposite and horizontal magnets repellent, and 
the external and upper wheel-magnet attractive, 
rapid rotation forthwith ensues. 
The model thus constructed was finished on the 
8th October, and immediately tried: and on the 
fust attempt it more than fulfilled my anticipa- 
tions. The moment the batteries were connected 
