86 LI’. H. Bigelow—The Earth a Magnetic Shell. 
The table gives the stations, their geographical position, the 
equivalent magnetic latitude, the year of the observations com- 
puted, and the group of south and north vectors. It is seen 
that the impressed vectors lie practically in the planes of the 
magnetic meridians in all latitudes, and hence approach the 
earth nearly perpendicular to the ecliptic. In the south group 
the value of a is generally positive and in the north group gen- 
erally negative: that is the vectors pointing south generally 
enter the earth, and those pointing north generally emerge 
from the earth. The earth is near the magnetic equator of the 
sun, and the mechanical pressure is sometimes greater on the 
northern and sometimes on the southern side of the ecliptic at 
the earth. The variation in pressure is shown by a certain line 
of force entering the earth at a definite angle in a given lati- 
tude, depending upon the slope of the horizontal plane at the 
station to the axis of the external field, and the material of the 
earth as compared with the ether. 
Unfortunately the accuracy of this kind of work depends 
upon the vertical force magnetometers, and these are much less 
reliable than those for the horizontal elements. Also the use 
of soft iron deflecting bars in so many of the polar expeditions 
of 1882-83 renders the observations less trustworthy. It would 
be easy enough to point out the probable source of the discrep- 
ancy of certain stations in the instruments used, but it may be 
omitted in this abstract. The angle a for a given day is due to 
many combined forces working together, and shows very con- 
siderable unsteadiness in each group from day to day, the value 
of a here given being the algebraic mean. Now it should be 
remembered that whether the angle ais plus or minus the 
values of s and o will be the same at a station, because an 
external line of force will approach the surface at a given angle 
whether above or below, on account of magnetic refraction 
in the case of the earth, as will be seen from sections that 
follow in this paper. Hence we may obtain from the mean 
angle, the general fact of entry and emergence at the surface, 
and from s and a, the value of the angle itself, cos a = =. 
These computed values are then to be graphically adjusted, so 
as approximately to eliminate local irregularities and obtain a 
mean system. The values of a. a. s. are plotted as ordinates on 
an abscissa equivalent to 180°, at the point corresponding to 
the magnetic latitude of each station, and a curve passed 
through these ordinate points, so as to give equal weights to 
the stations. The following table gives the computed values 
of the angle a and the adjusted values of the vectors for the 
south and the north groups. 
