RECENT ADVANCES IN SCIENCE 373 



the frequency of auroral displays runs parallel to that repre- 

 senting the solar activity as evidenced by the sun-spots and 

 prominences. The theory has been investigated theoretically 

 by Carl Stormer, and a summary of the results so far obtained is 

 given by him in Terrestrial Magnetism, 22, 23 and 97, 191 7. 



The problem is a difficult one to attack mathematically, 

 and, for simplicity, the assumptions were made that the earth 

 is a uniformly magnetised sphere, that no force other than the 

 earth's magnetic force acts on the particles, and that their 

 velocity is much greater than those of the earth and sun, so 

 that the relative motion of the latter can be neglected. The in- 

 vestigation has involved the tracing, by numerical and graphical 

 methods, of an immense number of different orbits of particles ; 

 it has been continued since 1903, and has included two expedi- 

 tions to Bossekop in 1910 and 191 3, for auroral observations, 

 determination of heights of auroral curtains, etc. 



The theory, combined with the fact that the earth's magnetic 

 axis does not coincide with its axis of rotation, has been found 

 adequate to explain the following auroral phenomena : There 

 are southern and northern limits for the aurora borealis and 

 aurora australis respectively, beyond which aurorse seldom or 

 never occur ; the large majority of auroral displays occur 

 within two well-defined belts ; they are of a sudden and variable 

 character ; they show a tendency to recur about the same time 

 on two successive days and also after about twenty-seven days, 

 the period of the sun's rotation ; aurorse occur at the greatest 

 distance from the earth's magnetic axis at the time of intense 

 magnetic storms. The existence of several auroral curtains at 

 the same time can be attributed to the charged particles 

 possessing several different velocities. 



The theory, therefore, explains satisfactorily most of the 

 phenomena observed. There is one outstanding difficulty — the 

 situation of the belts of maximum auroral frequency. These 

 are about 23 ° from the magnetic axis. The theory gives only 

 6° for /3-rays of radium and 18 for a-rays, so that it would 

 appear that negatively charged particles cannot account for the 

 aurora. It is possible that it is due entirely to the positively 

 charged a-particles. Carl Stormer suggests alternatively that 

 since theory indicates that a large number of particles bend 

 round the earth on the afternoon and night side, the belt so 

 formed may be of sufficient intensity to draw the position of 



