and the Aurora borealis. 431 



question indicates only approximately the direction of the 

 declination-needle. From theory, then, it follows that the 

 direction of the declination-needle must indicate generally and 

 approximately the summit of the boreal arc, though we are not 

 authorized to maintain that the two must entirely coincide 

 with one another. 



In the localities on the terrestrial surface situated beneath 

 the ring of maximum electric density the electricity descends 

 in the direction indicated by the dipping-needle of the locality; 

 for, as we have demonstrated above (p. 364), the action of the 

 inductive force of the earth is equal to zero in that direction. 



The descending currents cannot be compelled to deviate 

 from the above-mentioned direction, unless, in consequence of 

 an accidental meteorological state of the atmosphere, the elec- 

 trical resistance of the air is greater in that direction than in 

 another, in which case the intensity of the descending current 

 will be most considerable in the direction of least resistance. 

 As a case of this kind may easily occur, we are authorized to 

 maintain, on the ground of the theory, only that the descend- 

 ing currents must be in general parallel to the dipping-needle. 

 Now the current betrays its path through the air by a line of 

 light. If those luminous lines parallel to one another be 

 viewed from the surface of the earth, they will appear to con- 

 verge to a point, by the same laws of perspective to which is 

 due the visual convergence of the rows of trees of a long 

 avenue. As to the point, it will be found in the direction 

 which the dipping-needle at the place of observation indicates 

 in the sky. To this optical phenomenon is due the auroral 

 crown which appears in complete aurorse boreales. The cur- 

 rents in question are formed as soon as the difference of elec- 

 tric tension between the atmosphere and the earth is great 

 enough to surmount the obstacle presented by the resistance 

 of the air. Now, the earth being a good conductor of elec- 

 tricity, and constituting a sphere, its negative charge must be 

 nearly the same everywhere. Doubtless no one will maintain 

 that the positive charge of the air in the southern hemisphere 

 is constantly equal to the same charge in the northern hemi- 

 sphere. The forces which tend to render the air electropositive 

 and the earth electronegative are equal in the two hemispheres; 

 but the result of the activity of those forces depends in part 

 on the meteorological state of the air, which may be different 

 for each hemisphere. Nevertheless, as we have said, there 

 cannot be any great difference in the negative charge of the 

 earth ; and consequently one at least of the causes on which 

 the discharge depends is common to both hemispheres. There- 

 fore nothing verv extraordinarv can be found in the fact that 



