DIURNAL VARIATIONS OF TERRESTRIAL MAGNETISM. 
47 
difficulty may be the assumption of a smaller pressure in the conducting layer. 
He shows, indeed, that if the variation of the quantities involved in his calculations 
follow the same laws at low pressures as those actually determined at ordinary 
pressures, the conductivity should theoretically tend to an infinite value with 
increase of altitude. Perhaps the inference to be drawn from this is that whatever 
ultra-violet light is present is absorbed only in some particular layer of the atmosphere 
of appropriate constitution. 
Swann does not discuss the pressures and composition actually existing in the 
upper atmosphere. It appears likely, however, that at about 100 km. height the 
atmosphere contains roughly equal proportions of hydrogen and nitrogen, with only 
about 2 or 3 per cent, of oxygen; the pressure is approximately 3. 10~ 6 atmosphere. 
At 170 km. hydrogen is altogether the preponderant constituent, the only other 
which is at all appreciable being helium (6 per cent.); the pressure is approximately 
6 . 10~ 7 atmosphere. Owing to the lightness of hydrogen, the pressure diminishes 
with height much more slowly than near the base of the stratosphere. Even at 
800 km. height, where hydrogen is the sole constituent (within a small fraction of 
1 per cent.), the pressure is probably 10~ 9 atmosphere.* Perhaps at such high levels 
as these the ultra-violet radiation (X < 1350) of the amount considered by Swann 
might be sufficient to produce the required conductivity ; his calculation related to 
oxygen, however, and how far it would be modified in the case of hydrogen is 
uncertain—I am not aware of the existence of the data necessary to examine this 
point. But it may be doubted whether, in any case, the suggested agency can be 
sustained as a probable cause of the ionization. In the first place, even though the 
solar atmosphere should allow such short-wave radiation to escape, its intensity must 
be much diminished, relatively to the red end of the spectrum, by scattering, and its 
total energy must be much less than that appropriate to a black body spectrum at 
6000° C. Moreover it seems likely, in view of the close connection between solar and 
magnetic activity and the auroral, that the two latter terrestrial phenomena may 
originate in similar regions of the atmosphere. Piecent observations indicate that 
the level of aurorae is generally between 90 and 140 km.t; Swann’s calculation 
seems to preclude ultra-violet radiation as the ionizing agency in the conducting 
layer, if this is indeed situated at the auroral level. 
Whatever the origin and situation of the conducting layer, the main cause of its 
ionization must be in the sun, since the magnetic data of this paper indicate a very 
strongly marked diurnal variation. Hence spontaneous ionization, uninfluenced by 
* For these atmospheric data cf. Jeans’ ‘Dynamical Theory of Gases’(2nd ed.), p. 356. If, however, 
as some authorities believe, there is no appreciable amount of free hydrogen in the atmosphere, the 
pressure will fall off much more rapidly than is described above, and the conclusions would be modified 
accordingly. 
t Stormer, ‘Terrestrial Magnetism,’ XX., p. 159, 1915; Swinne, ‘ Phys. Zeit.,’ 17, p. 529, 1916, has 
discussed 2500 parallax determinations of the aurorae, and finds that 2098 lie between 90 and 130 km., 
and 322 between 1.30 and 200 km. 
