556 B1RKKLAXD. THE NORWEGIAN' AURORA POLARIS KXPKDITIOX, 1902 1903. 



It has not been necessary, however, to reproduce more than a few of these photographs, as other ex- 

 periments that have subsequently been developed from the above, more easily show clear results. 



In the following pages we speak of the north and south sides, i. e. respectively the upper and 

 under sides of the horizontal screen, and the east and west sides of the vertical. We calculate the 

 angle between the vertical screen and the centre line between the terrella and the cathode positive 

 eastwards from o to 360. 



In order to have an unmistakable manner of indicating the angles which we have occasion to 

 mention in the following pages, we shall refer them to the axis about which the terrella ran be 

 rotated -- which, in these experiments, is always vertical - - and a horizontal plane through the centre 

 of the terrella. 



We employ the designations easterly hour-angle and declination to indicate the position of a place. 

 The hour-angle is then calculated in the horizontal plane eastwards from the centre line between the 

 centres of the cathode and of the terrella, to the projection of the place upon the horizontal plane, and 

 the declination is an angle with its vertex in the centre of the terrella, and one side passing through 

 the place in question, and the other through the projection of the place upon the horizontal plane. The 

 northern declination is positive, the southern negative. 



In the eight photographs reproduced in fig. 201, the experiments were made under a pres- 

 sure of about 0.002 mm., with 25 milliamperes through the discharge-tube, and 30 amperes upon the 

 terrella. 



In Nos. i and 2, the easterly hour-angle of the vertical screen was 330, No. i being photographed 

 90 east of the screen, and No. 2 90 west of it. The terrella, it will be noticed, is seen very little from 

 above. In Nos. 3 and 4, the easterly hour-angle of the vertical screen was 30", No. 3 being taken 90 

 east of the screen, and No. 4 90 west of it. In Nos. 5 and 6 the hour-angle of the vertical screen 

 was o, the photographs being taken as before, but from a place with a declination of 25, so that tin 

 terrella is seen from considerably above. 



It should be remarked that the light-figures here seen upon the northern side of the horizontal 

 screen are of course exactly repeated upon the southern side, since the axis of the magnet coincides 

 with the axis of rotation; but they are not visible here. 



In Nos. 7 and 8, the hour-angle of the vertical screen was respectively 300 and 60, and the 

 photographs were taken respectively 120 and 55 east of the screen. 



With regard to the luminous precipitation upon the phosphorescent screens, that upon the vertical 

 screen shows that a very considerable part of the cathode rays are deflected towards the left before 

 they reach the terrella, and then, as we have seen, thrown against the left side surface of the discharge- 

 box, looking from the cathode towards the terrella (fig. 200). 



This phenomenon, in which a large proportion of the rays arc carried past the terrella, and arc 

 nearest to it some way out on its afternoon side in a direction opposite to that of the earth's rotation, 

 must be regarded as very important. We shall frequently return to it in the course of the experiments. 

 It is deflected rays from the sun such as these, that we have previously assumed to be the principal 

 cause of the positive equatorial storms. We shall also return to this ray-phenomenon in discussing the 

 diurnal variation of the terrestrial magnetism and the zodiacal light. 



In photograph No. 5, there are two places, A and B, in which rays descend upon the horizontal 

 screen, and it is these two instances that we shall first consider here. We shall see that the more abundant 

 of the two, which we will call A, and whose eastern boundary is very nearly a straight line, is due to rays 

 which, if not arrested by the screen, would travel round the terrella in a direction from west to east, oscillating 

 alternately above and below the plane of the magnetic equator, and most of them descending at last in "the 

 auroral zones"; but they never seem to come into contact with the terrella to the north of the northern 



