'64 Prof. L. Vegard: Results of Northliglit Investigations 



moving in the same direction. If they move in \ direction 

 which forms an angle with the lines of force, they will move 

 in a screw orbit round the magnetic lines of force, and 

 the number of turns made per unit length measured in the 

 direction of the lines of force becomes greater, and when the 

 angle is 90° the rays will continue to move in a circle. 



If now the rays are moving in a gas they will make 

 collisions with the gas molecules and produce ionization and 

 luminosity. Now if the orbits of the main bulk of the rays 

 are not essentially changed from the collisions resulting in 

 light emission, the luminosity produced by a ray per unit 

 length (measured in the direction of lines of force) must 

 for a given pressure be proportional to the number of times 

 the orbit turns round the lines of force per unit length of 

 the streamer. The increase in the number of turns will 

 liave a similar effect to that of an increase of pressure on 

 rays which are moving along the lines of force, which is the 

 apparent direction of the auroral-ray streamers. 



If, then, we have a bundle of electric rays of a given 

 strength, the luminosity produced per unit length along the 

 streamer will vary with the angle which the ray forms with 

 the lines of force. 



The luminosity will be minimum when the angle is zero 

 and approach a maximum when the angle approaches 90°. 



In order to get a variation of this kind it is necessary that 

 the path of the main bulk of the rays be not essentially 

 altered by the presence of gas. 



Now the main bulk of a bundle of a-rays move through 

 matter in nearly straight lines and are not essentially 

 deflected out of their path by collisions with gas molecules. 



As I have shown experimentally in a previous paper * 

 this result also applies to positive rays in general (Kanal- 

 strahlen). It tvas found from these experiments that a 

 collision resulting in light emission did not alter the momen- 

 tum of the ray by an amount comparable with the momentum 

 it possessed at the moment of collision. 



The same rule will also hold for cathode rays with fairly 

 high velocity when they pass into a gas of very low pressure. 

 In that case we know from experiments in vacuum tubes 

 that a bundle of cathode rays may move through a gas of 

 low pressure without being appreciably spread out from 

 collisions with gas molecules. This is clearly shown by the 

 experiments of Birkeland where bundles of cathode rays 



* L.Vegard, "Die Lichterzeugima' durcli Kanalstrahlen von Stichstoff 

 ■imd Sauerstoff," Ann. d. PJujs. xli. pp. 638-40 (1913) 



