Evidence lias already been given to show that the correlation between magnetic 

 disturbance and auroral display is better, the more intense the auroral display (both 

 in high latitudes), and that this correlation is not so much a correlation at particular 

 hours as a correlation common to the whole disturbed period. 



Whether we agree with Vegard in assuming that the solar particles causing auroras 

 are charged particles due to radioactive disintegration in the sun, or prefer the theory* 

 revived by Lindemann and ascribe them to clouds of particles projected from the sun, 

 is not of immediate moment. 



The latter theory, as Lindemann points out, is capable of giving an explanation 

 on the quantitative side of the energy involved in a large magnetic storm. On the 

 qualitative side, there is little to choose between the two, but it should prove possible 

 to discriminate between the rival theories if further information were available. 



For example, information might without too great difficulty be obtained on the 

 following points : 



(i) The period between the instant when the sunspot faces the earth and the 

 initiation of the corresponding magnetic storm ; the variation of this period 

 (the time taken for the charged particles to reach the earth) with the 

 magnitude and activity of the sunspot and with the duration of the storm. 



(ii) The height of the aurora at different (low) latitudes and its dependence upon 

 the activity of the sunspot to which it owes its origin. 



(iii) The relation (if any) between the height and intensity of the aurora and its 

 spectrum. 



In the process of obtaining this information, it seems not unreasonable to hope 

 that a clear indication would be given of the cause of the earth's permanent 

 magnetism. 



Lindemann' s theory that the charged particles causing both magnetic storms and 

 auroras are due to clouds of ionised particles projected from the sun (such as are 

 observed in the prominences), with a velocity due to light pressure on the neutral 

 atoms, is given support by the work of Megh Nad Saha.j The latter points out that, 

 in the chromosphere, ionisation is complete above a certain level, being evidenced by 

 I lie fact that only the "enhanced" lines, which are due to ionised atoms, appear 

 a hove this level. A rough calculation is made of the height in the chromosphere at 

 which this takes place for selected elements using data on ionisation potential, by 

 implication of the " Keaction-isobar " to the process of ionisation. The calculations 

 give figures in rough agreement with the observed heights for several spectral lines, if 

 certain assumptions regarding temperatures and pressures in the chromosphere are made. 

 II, now, selective light pressure acting in opposition to gravity is the cause of the 

 presence of these ionised elements at such high levels, and the velocities imparted in 



* Ijiinleiuami, ' Philosophical Magazine,' December. 1919. 

 t ' Phil. Ma-.,' October, 1920. 



40 



