F.. H. Bigelow—The Earth a Magnetic Shell. 93 
hence the purely magnetic periodic impressed forces are not 
entirely eliminated. The first curve is the European magnetic 
curve, 1878-1889, reproduced; the second is the numerical 
mean of the curves 3 to 15 inclusive as they stand ; the third 
is the relative number of West Indian hurricanes for 20 years. 
(Astron. and Astrophysics, June, 1894). The short period 
synchronism is evident, and in view of the action of the earth’s 
shell on the magnetic field, very interesting. It is thus con- 
cluded that the sun emits a polar radiation, as well as an elec- 
tro-magnetic, which at the distance of the earth continuously 
acts upon the terrestrial 7 corel and meteorological systems. 
European 
Magnetic 
Curve. 
Mean 
Meteorological 
Curve. 
West Indian 
Hurricane 
Curve. 
Magnetic and Meteorological Curves. 
The external polar field is concentrated in two belts, Fig. 3 
(1) the auroral belt, A, (2) the tropical belt T ; with a strongly 
depressed zone at the poles, P, another in the mid-latitudes M, 
and a third at the equator E. Unfortunately our per manent 
magnetic observatories are placed in the two depressions M 
and EK, the American and European being all outside the 
aurora belt A, the Asian being all, except Zi-ka-wei, near the 
magnetic equator. The work in the southern hemisphere is 
nearly suspended. The auroral belt was surveyed satisfactorily 
for only one year, 1882-83, and the tropical belts T T have 
been quite overlooked in planting fixed observatories. It is 
hoped that these defects may soon be remedied. 
The auroral belt receives its concentration from the fact that 
the shell turns aside the magnetic rays from the pole to a 
broad belt in the latitudes indicated. This belt isa rude oval 
embracing the magnetic and the geographic poles. (See 
Petermann’s Mittheilungen, vol. 20, 1874, ix.) The maxi- 
