544 
Proceedings of the Royal Society 
must be carried several steps beyond ; also, that while it is easy to 
abridge the lengthy results, it is impossible to extend those which 
have proved too short ; then the work must be re-done from the 
beginning. Hence the great advantage already experienced in this : 
that Brigg’s computations to fourteen places served for the prepara- 
tion of Vlacq’s ten-place table, and that again for those in common 
use, of seven and of five places. 
All Electro-Magnetic Declinometer. By A. Tanakadate, 
Assistant to the Professor of Physics in the University 
of Tokio, Japan. Communicated by Prof. J. A. Ewing, 
University College, Dundee. 
The terrestrial magnetic field will in general be disturbed in the 
neighbourhood of an electric circuit; but if the circuit be a plane 
set at right angles to the terrestrial lines of force, the direction of 
the field will remain unchanged at all points in the plane of the 
circuit. To determine the magnetic meridian, we have only to 
place a plane circuit in such a direction that, when a current in the 
circuit is started and stopped, no change takes place in the position 
of a small magnet hung at a point in the plane of the circuit, and 
free to turn in azimuth. The plane of the circuit will then lie 
magnetically east and west. 
The following method of laying down the magnetic meridian on, 
say, a laboratory table, will be found very convenient and accurate 
in practice : — 
A light rectangular wooden frame is made, about 1 meter long, 
15 cm. high, and 3 cm, wide, and its outer surface is recessed 
slightly, except at the edges, to receive 200 turns of fine insulated 
wire, which are wound regularly round the frame. Both ends of 
the coil are led off from the same point, and close together, in 
order to limit the electro-magnetic influence of the circuit to the 
portion wound on the frame. The circuit is completed at a con- 
siderable distance from the frame through a battery, reversing key, 
and box of resistance coils. 
A small magnetometer, consisting of the mirror (with attached 
magnets) of a Thomson’s dead-beat galvanometer, hung in a wood 
and glass case by a silk fibre 5 cm. long, is placed in the centre of 
this frame, resting upon a little shelf which projects into the middle 
