meter in the meridian of the telescope. After such a sliding of the suspender on the ceiling, which need be performed but very seldom, it is necessary to place on the opposite wall a new mark, to which the telescope may be directed without departure from the meridian. The thread to which the magnet bar is suspended consists of 200 parallel fibres of raw silk, each of which would support thirty grammes without breaking. The weight which this thread has usually to sustain amounts to nearly 2000 grammes, to which, in the measurements of Intensity, two weights of 500 grammes are added when determining the moment of inertia of the magnetic bar. The thread, therefore, never carries more than half the weight with which it would break. It is about two metres long, and has a torsion force, the moment of which amounts, for small deviations, to about the 1000th part of the magnetic force. The thread may be prepared by winding a single fibre twenty-five times round two glass tubes, distant from each other about four times the intended length of the thread; the two ends of the fibre are then tied firmly together, and the twenty-five-fold skein, thus formed, is stretched by drawing the two glass tubes further from each other. A small hook, carrying a weight, is then attached to the skein, midway between the two tubes, which are then raised and brought together, and the two loops are united in one. Thus a hundred-fold thread is prepared, which forms a loop at top and bottom, and which, being again brought together in a similar manner, forms the thread to which the magnet bar is suspended.
9. The stirrup and torsion-circle.—The force of torsion of the thread to which the magnet bar is suspended must not be entirely neglected in absolute measurements of declination and intensity, even though this thread be very long and fine. In order to measure the magnitude of this force, and to diminish its influence, so that the thread in the mean position of the magnet bar may be brought to its natural position when its moment of torsion is zero, it is necessary to be able to turn the thread, at one of its two extremities, round itself, in such manner that the angle of torsion may thereby be measured. In order to have the means of effecting this at hand, the apparatus for this purpose must be at the lower extremity of the thread; but, that the magnet bar may not be turned with it, the stirrup is composed of two parts, an alidade and a circle, which revolve only round a common vertical axis. The alidade supports the magnet bar, and is itself
