184 
MR. ARTHUR SCHUSTER ON THE 
electrification of air necessary to account for the required conductivity would be more 
than sufficient to overcome gravitation and to drive out the conducting portions at an 
enormous rate. The injection, on the other hand, of a sufficient number of ions of 
both kinds also presents difficulties on account of the large quantity of new matter 
which would have to accumulate in our air, especially if it is considered that 
recombination at a rapid rate would take place both in the journey from the sun to 
the earth and in the passage through the different layers of the atmosphere. The 
only alternative to ultra-violet radiation seems, therefore, to lie in the injection of ions 
travelling with sufficient rapidity to generate other ions by impact. The air itself, 
according to this view, would supply the raw material for the ionisation, the injected 
corpuscles only acting as fertilisers. There are, of course, other possibilities, such as 
the introduction of radioactive matter, or a spontaneous ionisation which may, if the 
rate of recombination is slow, be very effective at a great height; but that the sun 
undoubtedly plays an important part in the process is shown not only by the summer 
effect, but also by the periodic changes of the magnetic variation, which corresponds 
with the sunspot cycle. I have held for many years and frequently expressed the 
opinion that the relationship can only be explained satisfactorily on the supposition 
that the electric conductivity at times of many sunspots is increased. Whether this 
is a direct influence of the sun, or only an indication that an ionising influence is 
brought into the solar system from outside at times of many sunspots, is a question 
which everyone is likely to answer according to his individual views of the cause of 
sunspot variability. 
That the increase in the number of sunspots coincides with an increased conductivity 
of the upper layers of the atmosphere is also indicated by the eleven years’ period of 
the aurora borealis. The distinguishing feature of the relationship seems to be this, 
that auroral displays extend further into moderate latitudes when the solar activity 
is great. An increase of conductivity is the simplest and most natural way of 
accounting for the effect. The primary cause of the electric discharges which manifest 
themselves in the aurora is still unknown. We may look for it, perhaps, in electro¬ 
static forces which are always present, but causing a visible discharge only when their 
intensity rises abnormally, the course and intensity of discharge being much affected 
by inequalities of conducting power. On the other hand, there are other electromotive 
forces of induction not discussed in the paper, such as those accompanying a general 
drift of the atmosphere from west to east, which may well have something to do with 
the cause of auroral displays. Or again, if interplanetary space contains sufficient 
matter to be conducting, as I believe it must, there will be strong electromotive forces 
acting in the earth’s magnetic field between the conducting powers rotating with the 
earth and those of interplanetary space. 
Outbreaks of magnetic disturbances, affecting sometimes the whole of the earth 
simultaneously, may be explained by sudden local changes of conductivity which may 
extend through restricted or extensive portions of the atmosphere. I have shown in 
