DIURNAL VARIATION OF TERRESTRIAL MAGNETISM. 
183 
1887. Data supplied by H. A. Wilson # show that in the positive column of a 
vacuum tube the conductivity reaches the value 10 -13 and the kathode glow is even 
more highly conducting. The same author has experimented with air ionised in 
contact with hot platinum, and the data supplied by his diagrams!' allow us to fix the 
conductivity of such air as about 4x 10 -17 at a temperature of 1080°. When the air 
was charged with the spray from a 1 per cent, solution of a potassium salt, the 
conductivity rose to l'4x 10~ 13 , the temperature being 1200 1 The conductivity of a 
Bunsen burner has been measured by Gold and found to he 8x 10 -15 . In view of 
these figures, which all apply of course to atmospheric pressure, we ought not, I 
think, to reject the value of 10~ 13 as an impossible one for the conductivity of air at 
high altitudes, hut it is necessary to inquire into causes which produce so strong an 
ionisation. 
The increased intensity of the magnetic variation during the summer months 
suggests directly that we are dealing with a solar action. This action may he simply 
an effect of radiation or it may be due to an injection of ions into the atmosphere. 
The former hypothesis is the one which presents itself as the most natural one, 
though the coronal streamers lend some countenance to the second view, which has 
often been put forward and sometimes even pressed in support of wildly speculative 
theories. 
Ultra-violet radiation is known to ionise air in contact with metallic surfaces, but 
the evidence is somewhat conflicting' as to the effect of radiation on the air itself. 
Unless the air is absolutely free of dust, the observed action may he due to the 
illumination of the dust and not of the air. Dust-free air is so transparent to 
luminous radiation that it would not be surprising if the ionising effect would 
disappear, as some experimenters believe it to do, when proper precautions are taken. 
On the other hand, Dr. V. Schumann has shown that air has a very strong absorbing 
power for wave-lengths which are sufficiently short. Such short wave-lengths are 
supplied by several metallic sparks, and are freely transmitted through hydrogen. 
Nevertheless it seems difficult to believe that, even if emitted by the hottest portion 
of the sun’s envelope, they are not absorbed again by the surrounding cooler layers. 
We are not, therefore, at present in a position to assert that sufficiently short wave¬ 
lengths can enter the atmosphere and be absorbed in the outer layer, thereby causing 
ionisation, but we know so little about the conditions of the uppermost layers that 
we may reasonably retain the view that the powerful ionisation of the air, which we 
must consider to be an established fact, is a direct effect of solar radiation. 
If we turn to the possibility of a direct injection of ions by the sun into our 
atmosphere, we have to deal with the alternatives of supposing that ions of both kinds 
are introduced or only those of one kind. The second alternative must be rejected at 
once, because a simple calculation shows that the outward force due to the volume 
* ‘Phil. Mag.,’ 1900, p. 512. 
f ‘ Phil, Trans.,’ vol. 202, p. 243 (1904). 
