A new field‐magnetometer, designed by Dr. F. E. Smith, and constructed by the Cambridge Instrument Company Limited, has been taken into use by the Ordnance Survey of Great Britain.This magnetometer is the same in principle as the Schuster‐Smith coil now in use at the Abinger Magnetic Observatory, Surrey, England. It consists of twin coils of wire, 40 cm in diameter and about 20 cm apart, mounted on a rigid metal framework over a heavy base upon a tripod. A current of about 0.1 ampere, measured by balancing a standard cadmium cell against a known portion of the resistance in circuit, serves, when passed through the coils, to neutralize the Earth's magnetic field in the latitudes of the British Isles. The observation for horizontal force consists of a reading of the undisturbed position of a magnet‐system at the centre of the axis of the coils, followed by a reading of its position when the current is producing a field nearly equal to, but not quite in the direction of, the Earth's field, so that an elongated right‐angled triangle of forces exists. The relation H = FC sec θ then exists between the horizontal force, the constant F of the coil, C the current, and θ the angle between the two positions of the telescope. The latter is made to give auto‐collimation of its central wire in the focal plane by means of a small optically worked silvered glass cube forming part of the rigid magnet‐system. In the above‐described two operations the telescope reads first on one face of the cube and then on an adjoining one, so that it only has to be swung through the small angle θ (4° to 12°) whereas the magnet‐system swings through (90° − θ). There is a small temperature‐correction, applied jointly by means of a graph, depending on the temperature‐coefficients of the standard cell, the resistances, and the constant of the coil. Values of H correct to about two gammas can be obtained at intervals of about three or four minutes with the instrument. A very interesting feature of the magnet‐system is that instead of being suspended it is completely immersed in light spirit and tries to float, being held down by a jewel‐cup which comes into contact with a jewel‐point on the system. The upward force is only a few per cent of the weight of the system, and may be adjusted by adding or removing a small rider. Torsion is completely eliminated by this device, and friction is overcome by maintaining a gentle tapping on the container during observations. The apparatus, though designed primarily for the determination of horizontal force, has also been used in the field for determining declination. The results have been consistent, though slightly less accurate in individual readings of declination than those made with a Kew‐pattern magnetometer. Some small modifications to the instrument will be carried out to improve it in this respect. About 30 field‐stations were surveyed in England and Wales during May and June with this instrument.