SECOND VOYAGE OF DISCOVERY. 



135 



After ten P.M. the stationary light shifted more to the southward, and then^ 1821 ^ 

 gradually disappeared. At ten P.M. on the 18th this phenomenon assumed k^v^j 

 a similar appearance in the S.b.W. quarter. On the evening of the 23d the 

 Aurora Borealis made its appearance in the N.W., vivid coruscations shoot- 

 ing at times across the zenith to the opposite horizon. The gold leaf of the 

 electrometer was not perceptibly affected by it. On the morning of the 24th 

 it was again faintly seen in irregular streams of white light, extending from 

 the western horizon to the zenith: for several hours the same night also 

 this extraordinary phenomenon was visible from the south-east round by 

 south to west, being principally confined to a space about five degrees above 

 the horizon. The magnetic needle, which Avas attentively watched, was 

 not at all affected by any of these phenomena. On the 26th, both in the 

 morning and evening, the Aurora again appeared from south-east to south- 

 west, the brightest part being about ten degrees above the horizon, 

 and with pencils of rays shooting upwards towards the zenith. In almost 

 every instance it is observable that the light, however irregularly disposed 

 in other respects, has a tendency to assume an arch-like form ; but I think 

 a plane bisecting the arch would more generally have coincided with the 

 true than the magnetic meridian, in the phenomena we had here an opportu- 

 nity of observing. This was particularly the case on the morning of the 27th, 

 when at six A.M. the Aurora formed one broad continuous and well-defined 

 arch, its centre passing rather to the southward of the zenith, and its legs 

 appearing to rest upon the horizon at east and west. For several hours on 

 the evening of the 28th it was seen in the south-east, with rays darting ra- 

 pidly up nearly as high as the zenith. There is almost always one stationary 

 patch of light near the horizon, appearing, as it were, the source whence 

 the shifting or variable part of the phenomenon proceeds. It will be seen 

 from about this period how much more frequently the Aurora seemed to 

 issue from the south-eastern quarter than from any other during the rest of 

 the winter. 



On the 1st of December there was a space of many miles in which none of Decemb. 

 the " old" ice was visible. The sea was here for the most part covered with Sat * L 

 a very thin sheet of " young" ice, probably the formation of a single day, 

 since the westerly wind had driven the floes off the land. The whole of this 

 was in motion with the tide, which breaking the thin floes left several spaces 

 of clear water. It was observable that though a considerable frost-smoke 

 arose from the young ice, it was not so dense as that from the clear water, 



