12 KR. BIRKELAND. 



for ordinary zodiacal light, is now easy to understand, for the rays mag- 

 netically drawn in always form a fairly well defined cône, and from his 

 station of observation he was looking right along the northern part of the 

 surface of the cône. 



Conditions for Absorption of Corpuscular Precipitation. 



5. What are now the conditions to be fulfilled if these new solar 

 rays are to come so deep down into our atmosphere that their enormous 

 energy can be transformed into heat by absorption? 



To answer this question we can not make use of the results from 

 mathematical analysis, where the earth is considered as an elementary 

 magnet, because the phenomena we are studying, take place too near the 

 earth's surface. But we can very well use the results from the terrella 

 experiments; they are very instructive in this case. It is obvious that the 

 height above the earth's surface at which the bulk of the corpuscular rays 

 is sweeping past the earth largely depends upon the intensity of the ter- 

 restrial magnetism and in a corresponding manner upon the magnetic 

 Stiffness of the solar cathode- rays. In »A. P.« pp. 591 to 595 we find 

 necessary information about the dimensions of the precipitation-rings on 

 the terrella, these rings corresponding to the auroral zones of the earth. 



From two series of experiments with cathode-rays of 1800 volts and 

 of 2400 volts we find in the first place that the stiffer the rays employed 

 are, and the less the magnetisation of the terrella is, the larger are the 

 polar precipitation rings. It further appears that the more the terrella is 

 magnetic, the narrower or thinner does the band of light in, the ring 

 become, and the smaller are the number of rays that are drawn in towards 

 the terrella in the precipitation ring. There will evidently exist a certain 

 magnetisation which gives the maximum of ray precipitated on the terrella. 



In this connection it is of interest that an aurora that occurs in low 

 altitudes on the earth, must according to our theory and experimental 

 analogies, be due to stifter rays than an aurora that only occurs in the 

 ordinary auroral zone: and that the farther the northern aurora extends 

 towards southern latitudes, the greater will its width be, and we should 

 expect that it will be seen simultaneous!}' in the zenith over a great area 

 of the earth. 



The photographs 2 & 8 of fig. 218 »A. P.« p. 593 are of great utility 

 here, especially when compared with the experiment represented by fig. 219. 

 It appears that for cathode rays of 2400 volts and a magnetic intensity 

 of 1600 C. G. S. units at the pole of the terrella, the precipitation ring has 



