MESSRS. R. T. GLAZEBROOK AND J. M. DODDS ON THE 
284 
A graduated circle is fixed to this stand, and the coils can be turned about a vertical 
axis, and their position read by means of a vernier. This was found useful in adjust¬ 
ing the coils parallel to the magnetic meridian. The galvanometer rested on a stone 
bracket built into the wall of the room. A scale placed approximately north and 
south at a distance of about 259 centims. from the magnet was reflected in the mirror 
and viewed through a telescope. 
The scale rested on a solid wooden support on the floor of the room. The mirror, about 
1*5 centim. square, was a specially good one, selected by a fortunate chance from among 
a number in the laboratory. The divisions of the scale were in millimetres, and after 
practice these could be subdivided by the eye with great accuracy to tenths. The 
scale itself was of paper ; though this material is unsuitable for many purposes because 
of the changes produced by the weather in it, in our experiments these changes are of 
small consequence, for we require only the ratio of the throw produced by the induc¬ 
tion current to the steady deflection produced by the permanent current; and the time 
which elapsed between the measurements was only a few minutes. Any shrinking or 
alteration of the scale will go on very approximately uniformly throughout its length 
and not alter the ratio of two lengths, which were never very unequal, as measured 
by the scale. After use the scale was carefully compared with the standard metre 
at the Cavendish Laboratory and the necessary correction applied to the readings. 
The distance between the mirror and the scale only enters our result in the small 
correction necessary to reduce the scale readings so as to give the ratio of the sine of 
half the throw to the tangent of the deflection. It was unnecessary, therefore, to 
measure it with any great accuracy or to take steps to ensure its remaining the same 
from day to day; so long as it did not change during the half hour occupied by each 
experiment, all the conditions required by us were satisfied. 
The resistance coils. 
The standard coil R has been already referred to ; the means adopted to measure its 
resistance will be described later. 
Its value at a temperature of 14°*6 C. was found in May, 1881, to be IGO'821 ohms. 
The coil V used as a shunt to the galvanometer was made of thick German-silver 
wire. About 450 centims. of wire covered with silk were employed. The extremities 
of this were soldered to two stout copper rods with amalgamated ends, connexion with 
the rest of the apparatus being made by means of mercury cups; the ends of the rods 
were pressed down on to amalgamated pieces of copper at the bottom of the mercury 
cups. 
The value of V was determined by repeated comparison at different temperatures 
with the B.A. unit known as “Flat coil” in Professor Chrystal’s report (Brit. Ass. 
Rep., 1876). The value of the Flat coil in mean B.A. units was taken from 
Professor Fleming’s recent comparison of the B.A. units at the laboratory. We 
have 
