(ill rendering the aurora invisible) and the difficulty of making a correct allowance for 

 the variations in weather conditions during the winter months. 



4. TREND AND ROTATION OF ARCHES. 



Alawson,* in his treatment of the auroral observations of Shackleton's Expedition 

 in 1908, gives a full discussion of the occurrence of arches at that station, and records 

 a definite and close correspondence between the trend of the arch and the sun's 

 azimuth, i.e. a dependence of trend upon local time. 



An analysis of the 59 observations of arches of 70 altitude and over at Cape 

 Evans during the dark period in 1911 does not completely confirm Mawson's results, 

 as may be seen from Table 3. 



TABLE 3. Trend of Arches, Cape Evans, 1911. 



As, in this Table, prominence has obviously been given to the eight chief points of 

 the compass at the expense of such directions as N.N.W., W.N.W., etc., the Table 

 has been re-tabulated below, one-half of the number of observations in the subsidiary 

 directions being added to each of the adjacent chief points of the compass : 



Trend. No. of Observations. 

 N. to S. 19 



N.W. to S.E. 16 



W. to E. 4| 



S.W. to N.E. 



The remarkable point disclosed by these figures is the tendency of such arches to 

 avoid the East- West direction. 



The point to which attention has been drawn by Mawson (loc. c.it.) is the general 

 rotation exhibited by the arches observed by him, this rotation being sometimes 

 quick and sometimes slow. (The rotation of the arch is a necessity if his statement 

 in regard to the relation between local time and trend of arch is correct.) 

 * Loc. cit. 

 f Time of 180th meridian, which is about 54 minutes in advance of local time. 



10 



