ON THE PERIODICITY OF THE AURORA BOREALIS. 13 



have been known to Mairan, no change has been made, as in the auroras of April 11, 12, 

 and 13, of Dec. 15 and 1G, 1716 ; of April 6, 9, 10, and 11, 1717 ; of Feb. 6 and 10, of 

 Nov. 29, 1720 ; of Feb. 17, 23, and 28, 1721. 1 In other cases there is no doubt that 

 Mairan has occasionally been led into error by advancing dates already expressed in 

 the new style. No copy of Frobes's book could be found in this country, and I did not 

 succeed in obtaining a copy from Europe until the printing of my own catalogue had 

 proceeded too far to allow me to introduce any changes. It is to be understood, there- 

 fore, that the dates here attached to Frobes's auroras are the same as those adopted by 

 Mairan, unless it is specially stated to the contrary. 



Although the catalogue of auroras, which I now present to the public, comprise.* 

 nearly 10,000 independent auroras, and more than 50,000 observations, I do not claim 

 that it has exhausted the supply. On the contrary, I can myself poiut out some of i f s 



deficiencies. It is well known that Musschenbroek observed the aurora durino- 29 



o 



years at Utrecht and Leyden, and collected a catalogue of 720 auroras. In one of his 

 works 2 he has arranged these auroras according to the months in which they ap- 

 peared ; but the individual observations, so far as I am aware, have never been pub- 

 lished. A laborious meteorologist 3 of the last century has stated that he possessed in 

 manuscript the observations of Musschenbroek, and that he intended to publish them 

 with others in a third volume of Memoirs ; which, however, never appeared. .A s 

 Musschenbroek has classified his auroras according to the months of the year, they 

 are available in the study of the annua} changes of the aurora ; it is to be regretted that 

 they cannot also be used in the investigation of those fluctuations which observe longer 

 periods. Of the different countries which have contributed to swell the grand capital 

 of auroral observations, Russia, always devoted to meteorology and the physics of the 

 globe, stands foremost ; an almost uninterrupted series of auroras having been regularly 

 observed and published at St. Petersburg, since the foundation of the Imperial Academy 

 of Sciences in 1726. 



Not finding the published record of auroras for St. Petersburg so complete in recent 

 years as formerly, I wrote to M. Kupffer, Director of the Central Physical Observatory 

 of Russia, requesting to be furnished with full materials. In reply to my inquiries I 

 received from Kupffer a table, w^hich contains, as he says, the auroras observed in Russia 

 between the years 1841 and 1861 inclusive. 4 Kupffer remarks that the aurora borealis 

 always announces its presence by very strong oscillations of the magnetic needle, which 



1 Frobes, Nova et Aiiiqna Luminis atque Amorx Borealis Spectacula, pp. 79, 81, 89, and 90. 



2 Cours de Physique, 24S9. 



3 Cotte's Me'moires snr la Me'te'orologie, I. 366 and II. 169. Also, Jonrn. de Phys. LXX. 168 and LXXJII. 153. 



4 It is incorporated in my Yearly and Monthly Table for St. Petersburg, pp. 186, 1ST of this Memoir. 



