At 8 hours on the 8th, the straight arch ran from S.E. to the zenith, thence 

 branching into four arms covering the sector North to S.W. On the 9th, two 

 small streamers were seen close to the zenith, while on the 10th, the form was 

 similar to that on the 8th, but with only two branches in the sector W. to N.W. 

 Again, on the llth, the display was confined to two small streamers in the north. 

 On the 12th, the straight portion of the arch was very rudimentary, and represented 

 by a streamer in the S.E., the branches being represented by three streamers in 

 the N.W. quadrant. The aurora on the 13th was almost exactly of the same 

 form as on the 12th. 



On the 14th, it was cloudy, but on the 15th, four complete branches were again 

 seen in the N.W. sector, the arch from S.E. to the zenith being somewhat incomplete, 

 but quite recognizable as such. 



Another striking example of this tendency is supplied by the observations at 

 4 hours on the 30th May and succeeding days, a distinct tendency being revealed 

 towards the formation of an aurora in the form of a closed oval-shaped curve. 



4. TREND AND ROTATION OP ARCHES. 



An analysis of all the arches observed during the year at the Cape Adare station 

 is given in Table 9, which shows also the times at which the arches were seen. As in 



TABLE 9. Trend of Arches at Cape Adare, 1911. 



Trend. 



N. to S. 



N.N.W. to S.S.E. 

 N.W. to S.E. 



W.N.W. to E.S.E. 

 W. to E. 



W.S.W. to E.N.E. 

 S.W. to N.E.... 



S.S.W. to N.N.E. 



No. of Observations. 



2 



14 

 33 



10 

 16 



2 

 I 

 1 



Cape Adare Times of Occurrence. 



1.30, 2. 



20, 1, 2, 2, 2, 3, 4, 4, 4, 4, 4, 4, 4.15, 6. 



18.30, 18.55, 19, 19.55, 23.30, 2, 2, 2, 2.40, 3, 3.30, 



3.40, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4.20, 5, 6, 6, G, G, 6, 



fi 8 8 8 



18, 18.25,' 20, 2-1, 3, 3, 3.45, 4, 4, 8.12. 

 15.25, 15.30, 19.45, 20, 21.40, 22.45, 23, 24, 24, 



2, 4.15, 4.15, 6, 6, 8, 8.30. 

 18.30, 6. 

 18.55. 

 20.55. 



the case of the Cape Evans observations, there appears to be a tendency for the chief 

 points of the compass to be favoured at the expense of the subsidiary points. Making 

 due allowance for this tendency, it appears that arches are most likely to form in a 

 direction N.W. to S.E., or more probably, N.W. by W. and S.E. by E. It will be seen 

 that there is a most marked tendency for the arches to avoid the direction which would 

 bring them into the south-western quadrant, which is the quadrant in which auroras 



23 



