Before taking up the demonstration that the earth is simply a
giant loadstone, Gilbert demonstrated in an ingenious way that
every loadstone, of whatever size, has definite and fixed poles.
He did this by placing the stone in a metal lathe and converting
it into a sphere, and upon this sphere demonstrated how the poles
can be found. To this round loadstone he gave the name of
terrella--that is, little earth.

"To find, then, poles answering to the earth," he says, "take in
your hand the round stone, and lay on it a needle or a piece of
iron wire: the ends of the wire move round their middle point,
and suddenly come to a standstill. Now, with ochre or with chalk,
mark where the wire lies still and sticks. Then move the middle
or centre of the wire to another spot, and so to a third and
fourth, always marking the stone along the length of the wire
where it stands still; the lines so marked will exhibit meridian
circles, or circles like meridians, on the stone or terrella; and
manifestly they will all come together at the poles of the stone.
The circle being continued in this way, the poles appear, both
the north and the south, and betwixt these, midway, we may draw a
large circle for an equator, as is done by the astronomer in the
heavens and on his spheres, and by the geographer on the
terrestrial globe."[6]

Gilbert had tried the familiar experiment of placing the
loadstone on a float in water, and observed that the poles always
revolved until they pointed north and south, which he explained
as due to the earth's magnetic attraction. In this same
connection he noticed that a piece of wrought iron mounted on a
cork float was attracted by other metals to a slight degree, and