360 De la Rive on the Aurora Borealis. 
2. The deviations were of two kinds, some of large extent, 
even 50°, others small, varying from 1° to 8° ;—the former not 
common, and changing in direction and intensity, so that no 
law can be discovered; the latter on the contrary subject to a 
simple law, and being very regular when the air is dry and the 
sky serene, but with many anomalies when the weather is cold 
and rainy. 
Mr. Barlow has made numerous observations, and obtained re- 
sults demonstrating the exactness of the principle which I have 
laid down. Four main lines starting from Derby, were used in 
his experiments, two running towards the north and northeast, 
and two towards the south and southwest. The direction of the 
currents perceived on the first two lines, was always contrary to 
that of the currents onthe two others, as ought to be the case, 
on the theory proposed. But the most remarkable fact, is the 
perfect concordance which these observations have proved to ex- 
ist between the movement of the needle of the galvanometer 
placed in the circuit of the telegraph wire and the diurnal varia- 
tions of the magnetic needle. The diurnal movement of the 
needle of the galvanometer is subject to disturbances in intensi- 
ty more or less continued, during storms, and also when the au- 
rora borealis is visible; and so also is this true of the compass 
needle. There is this difference, that the currents acting on the 
latter, circulating beneath the earth’s surface, should not be sub- 
ject to disturbances like those which happen to the telegraph 
wires through the influence of the electrical condition of the at- 
mosphere about them. 
The existence then of electric currents circulating eipiie « the 
e m saa 
the surface of the globe. 
As we have said above, the positive electricity with which the 
atmosphere is charged, especially in the upper regions, is carried 
towards the two poles either by the greater conductibility of the 
upper and most rarified strata of the atmosphere, or by the cur 
rents of air in the upper regions which move from the equator ' 
the two poles. It is consequently through the vapors which are 
