62 Prof. L.Vegard: Results of Nortlilight Investigations 



magnetic field of the earth a heterogeneous bundle of rays 

 would be spread out over a considerable area and could not 

 form a thin drapery *. 



If the change of luminosity distribution was wholly to be 

 explained from variations in the properties of the rays, it- 

 should follow that each particular manifestation of a certain 

 form should possess a definite luminosity distribution, because 

 the rays are homogeneous, and to get a variation of lumi- 

 nosity distribution we should have to suppose that the 

 various forms were produced by rays of different properties 

 and that even the same auroral form was produced by very 

 different rays. 



Now the curves (fig. 1, PL II.) which give the height 

 distribution of the lower limit indicate that the aurorse are 

 produced by several groups of rays, but the fact that arcs, 

 drapery-shaped arcs, and draperies give the same typical 

 maxima of the height distribution curve seems to show that 

 the three forms are produced by the same groups of rays. 

 This is also in accordance with the fact that an arc bv 

 gradual changes may take the appearance of a drapery- 

 shaped arc, and the latter may change into the form of a 

 drapery f. Further, we have any transition form between 

 draperies and rays. This change of form is accompanied by 

 great changes of luminosity distribution along the rajs. 



The difference of form and the corresponding difference 

 of luminosity distribution cannot, therefore, be due to a 

 change of the type of rays, because the new types of rays 

 could give no such gradual change of a form, which the 

 whole time may be kept coherent, and which during the 

 changes of form and luminosity distribution may remain 

 very nearly at the same region of the sky. We also many 

 times observe that different parts of the same band-form 

 may show a different luminosity distribution, although we 

 found the height of the lower limit to be the same along the 

 whole band, showing that the rays producing a coherent 

 band are homogeneous. 



Although we must regard it as possible or rather as pro- 

 bable that the various auroral displays or the various units 

 constituting a display may be produced by electric rays 

 which differ with regard to physical properties — at any rate 

 as regards velocity, — still we cannot account for the essential 



* See e.g. L. Vegard, Phil. Mag. 1912. 



t See L. Vegard, " Nordlichtuntersuchungen," Vid. SeJsk. Ski\ 

 Krlstiania, nr. 13, p. 45 (1916). See also the work of L. Vegard aud 

 O. Kvogness pp. 99-100. 



