3 02 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



Fig. 136. 



in taking them thus was to investigate the conditions on both sides of one of the above-mentioned three 

 screens, the one whose position answered to about 6 p. m. 



Photographs i, 2, 3 and 5 show very distinct precipitation on the Z>-side of the screen, and shadows 

 on the a-side, indicating that the precipitation on the terrella has a tangential motion from west to east 

 in the "auroral zone". Photographs 3, 4 and 5 also show, however, a slighter precipitation at the very 

 bottom of the a-side of the screen. 



We obtain a clearer understanding of this twofold phenomenon from photographs 6, 7 and 8. We 

 here see quite distinctly, although not nearly so distinctly as in the experiment itself, that the broad 

 band of light consists of two bands, one more northerly that moves from west to east, and one more 

 southerly that moves from east to west. The northern band of light breaks off just to the east of the 

 screen, while the southern band breaks off just to the west of the screen, in both cases because of the 

 shadow cast by the screen. 



These circumstances seem to give us the key to the apparent enigma of the simultaneous occurrence 

 of a negative polar storm in Spitsbergen, and a positive polar storm at Kaafjord and Matotchkin Schar. 



