and Nature of the Cosmic Electric Rays. 83 



with a single elementary charge would have a penetrating 

 power .r=10~ 5 cm. 



With the value a = 10 4 a hydrogen and a helium 

 nucleus would penetrate to heights of 135 and 145 km. 

 respectively. 



Thus, even if we suppose that there are no light gases in 

 the upper strata of the atmosphere that give any appreciable 

 absorption, we find that most of the positive rays with the 

 proper magnetic deflectibility would have a too small pene- 

 trating power to account for the height-distribution found 

 for the lower limit of the auroraa. If lighter gases should 

 be present to any appreciable amount, the assumption of 

 positive rays is not at all consistent with the view here put 

 forward with regard to the explanation of the variations of 

 luminosity distribution, and, as already pointed out in the 

 work of Krogness ;md the author, the electron rays are 

 probably those which will best fit in with the view adopted 

 with regard to the luminosity distribution. The very narrow 

 straight-lined streamers of an aurora are then a result of the 

 action of the magnetic field, and the auroral phenomenon is 

 similar to that which under certain conditions is observed in 

 vacuum tubes, and which we know under the name of mag- 

 netic cathodic rays. It is in this connexion of interest to 

 remember that also the direction of the band forms, when 

 regarded in connexion with the spiral-formed precipi- 

 tation of rays observed by Birkeland in his magnetic 

 "terrella/' seems to favour the assumption of negative 

 charged cosmic rays. 



Still the aurorae show a great variety of forms, and although 

 there are strong indications that most forms are produced by 

 rays of essentially the same type, we cannot at present exclude 

 the possibility that rays with different carriers may be engaged 

 in the production of the aurorae. And if nitrogen should be 

 the predominant gas in the upper strata of the atmosphere, we 

 have still to count with the possibility that positive rays of 

 H and He nuclei with velocities of the order of 10 s cm./sec. 

 may produce an essential part of the auroras, 



Our knowledge about the upper strata of the atmosphere 

 "is too uncertain to enable us to draw definite conclusions in 

 this respect ; but it seems that investigations with regard to 

 the structure and height of the aurorae, together with an 

 increased knowledge as regards the properties of the electric 

 rays and with regard to the auroral spectrum, may enable 

 us to fix more accurately the properties o\' the cosmic rays, 

 and at the same time give us more certainty regarding 



G2 



