THE AIROU.V BOKEALIS. 245 



somelimcs be found in every degree of longitude, bathing, not only the 

 north pole, but the whole northern hemisphere, as far as to the tropic of 

 Cancer, in its fiery streams. It is worthy of inquiry whether the south 

 pole may not sometimes be enveloped in a similar manner at the same 

 time, and thus the whole globe invested by an aurora, which is visible 

 only in the part lying in its shadov/ or where night exists, and in the 

 regions of greatest activity, which are near the poles. 



8. This leads to the remark that the aurora, contrary to the ordinary 

 opinion, is not a phenomenon developed ly the darkness of the night.^ or 

 the absence of the S7in. A sufficient number of cases are on record 

 which go to prove its existence during day-light. A peculiar brilliancy 

 in the northern sky contrasting it with the rest of the heavens, the bank 

 of what seemed to be vapor or cirrus-cloud lying immovably in the 

 same horizon, the existence of the aurora in all its splendor, with its 

 .streamers, arch and corona as soon as the evening twilight had departed, 

 the melting away of the same into the superior morning twilight, and 

 its existence on several successive nights, and shown by the disturbance 

 of the magnetic needle as having continued during the day, all go to 

 .show that it is not confined to night and is perhaps in no way depen- 

 dent on it. It has, however, variations in intensity and splendor, which 

 it is believed, by many, to have some reference to the hour of the night. 

 It is, for instance, sometimes very active, soon after the departure of 

 twilight, at about 10 o'clock, P. M., at 1 or 2 o'clock, A. M., and about 

 two hours before sunrise ; whilst, during the intermediate periods, the 

 lighting up of the sky is less intense. But as these phases are not en- 

 tirely uniform, they may only prove that the aurora, from some un- 

 known cause, is subject to alternate fits of greater or less splendor. — 

 Neither is it dependent for its display upon the winter, according to the 

 prevailing opinion. A careful comparison of a list of auroras occurring 

 through a number of successive years, will show that as many take place 

 during the summer as the winter months. 



9. The number.) which take place during each year seems not to be 

 uniform. There have never less than two or three been witnessed, and 

 not more than about twenty or thirty. But when we reflect, that in mid- 

 dle latitudes the light is often faint and evanescent, it would require con- 

 stant attention to the heavens during every night to discover all, and then 

 many, may also take place during the day ; so that we may safely infer 

 their number as vastly greater than that just named as the highest, and we 

 may not, perhaps, be far from the truth, when we say that it is probable 

 that the aurora is a constant phenomenon attending our globe, seen only 

 under favorable conditions, and when developed in its greater degrees of 



