THE EARTH’S MAGNETIC FIELD IN INTERNATIONAL UNITS. 
447 
placed in series, and the potential circuit included either (1) both coils in series, 
(2) coil No. 2 only, (3) coil No. 3 only, or (4) in one set of measurements a third 
10 ohm coil (No. 1941) was placed in series with the other two, and all three were 
included in the potential circuit. By thus altering the resistance between the 
terminals of which the potential dilference was made equal to the E.M.F. of the 
standard cell included in the potential circuit, as well as by using either one or two 
cells in this circuit, it was possible to get a considerable range of values for the 
current in the coils. 
The procedure adopted in an experiment was for the observer at the galvanometer 
to adjust the carbon resistances till the galvanometer was not deflected, the other 
observer then adjusted the Helmholtz galvanometer so that the magnet was in its 
standard position with regard to the coils. A note of the time, the temperatures of 
the cell and the resistance coils, and the azimuth reading of the Helmholtz galvano¬ 
meter having been made, the current was reversed. Two settings were made with 
the current in this reversed direction, and then a second setting with the current in 
the original direction. These four readings constituted a set of readings such as are 
shown in a single line of the table given on p. 456. 
The Standard Resistance Coils .—Three 10-ohm resistance coils were employed in 
the experiments. Of these one (No. 1941) is a manganine standard coil of the 
German pattern made by Hartmann and Braun. Immediately after the magnetic 
experiments the resistance of this coil was carefully compared with those of the other 
two coils, and then it was sent off to the Beichsanstalt to be compared with their 
standards, when its resistance was found to be the same as when compared some 
years previously. The other two coils (Nos. 2 and 3) were manganine coils of the 
German pattern made by the author in 1897. They were “aged” by being heated 
for several days to 140° C., the silk-covered wire being protected by a coating of 
shellac varnish made with absolute alcohol. All three coils are of such a form that, 
when in use, they can be immersed in oil, the oil being allowed to flow freely over 
the wire. The oil in the bath was kept well stirred, a propeller driven by an electric 
motor being used during the silver depositions, which caused violent motion of the 
oil over the wire of the coils. By comparing the resistance of one of these oils with 
that of a coil made of very much thicker uncovered wire stretched on a frame in an 
oil bath, when traversed by different currents, it was found that the maximum 
current used, 0'2 ampere, produced no error in the value obtained for the resistance 
amounting to 1 in 10,000. When making the magnetic observations the current was 
not so large (‘05 ampere) as in the case of the silver depositions, and it was only kept 
flowing for a comparatively short time, hence a paddle worked by hand was considered 
sufiicient to stir the oil. The temperature of the oil was read by means of a thermo¬ 
meter graduated in tenths of a degree. The errors of this thermometer, as well as 
those of all the other thermometers employed, were determined by comparison with 
