204 OX THE PERIODICITY OF THE AURORA BOREALIS. 



The foregoing catalogue of auroras may be employed in the discussion : I. of the 

 daily periodicity ; II. of the yearly periodicity ; and III. of the secular periodicity of the 

 aurora. 



I. The record in most cases is silent as to the hour of the day at which the auroral 

 display reached its height. I have not mentioned it even when it is known, as I had 

 in view the yearly and secular periodicities, especially the last. Olmsted l says that the 

 maximum of brilliancy arrives between 10 and 11 o'clock, p. m, and that inferior exhi- 

 bitions terminate before midnight. Quetelet 2 agrees with him that none occur in the 

 morning. Bravais recognized the same law as holding good within the polar circle, 

 where daylight cannot interfere with the visibility of the aurora. 3 Broun has pub- 

 lished a table of the number of auroras observed at Makerstoun during the different 

 hours of the night, for each month separately, and also for the whole year. The maxi- 

 mum hour is about 9 o'clock p. m. at all seasons, irrespective therefore of twilight. 4 On 

 the other hand, the remark of Simpson, on page 124 of this memoir, is deserving of 

 notice. 



II. Mairan deduced from his catalogue of 1441 auroras that the numbers of those 

 which appeared when the earth was in perihelion and aphelion were to each other in 

 the ratio of 7 to IoVtj or °f ^ t° %~s%%> or 0I> 9 to 4|2 1, according as the perihelion and 

 aphelion comprised each a period of two months, four months, or six months. The 

 number of auroras seen when the earth was in the ascending node of the sun's equator 

 is to the number seen when the earth was in the descending node as 9 to 2 : each node 

 embracing a period of two months. More even had been observed when the earth was 

 at the ascending node of the sun's equator than when it was in perihelion, in the ratio 

 of 7 to §3%\. The number at the descending node was to the number at the aphe- 

 lion as 9 to S^y. Then, again, Mairan divides the orbit of the earth into two parts by 

 the line of apsides, and he finds the number of auroras in that part which contains the 

 ascending signs and the number of those in the other half in the ratio of 9 to 7f£f. The 

 numbers at the two equinoxes are almost the same, if each equinox is made to embrace 

 two months. The number of auroras is also nearly the same in the neighborhood of those 

 two places of the earth's orbit which are characterized by being 90° from the nodes of 

 the sun's equator. All these various comparisons were made by Mairan in order to 

 sustain his theory of the cosmical origin of the aurora, and of its dependence on the 

 sun's atmosphere. Mairan did not eompare directly the equinoxes with the solstices, in 



1 Smithsonian Contributions to Knowledge, VIII. 34. 2 Annales de 1'Observ. Roj-ale de Brux. XI. 42. 



8 Thompson's Meteorology, p. 362. Observ. Magnet. Comm. Nord. 4 Edinburgh Philosophical Transactions, XIX. Pt. 2. 



