ON THE PERIODICITY OF THE AURORA BOREALIS. 345 



I have failed to discover any connection between the secular changes in the earth's 

 magnetism and the secular periodicity of the aurora. The comparison is an imperfect 

 one, inasmuch as, in the aurora, it pertains exclusively to the frequency of the dis- 

 plays, without any regard to their magnitude, and, in terrestrial magnetism, it only 

 embraces the single element of declination. The secular periodicity of the aurora 

 is mainly the same for the whole earth, both as to the length of the period and the 

 times of maxima and minima. This is not the case with the secular changes in the 

 declination of the magnetic needle. Columbus remarked on September 13, 1492, 

 that " 2\° east of the island of Corvo " (one of the Azores) " the magnetic varia- 

 tion changed and passed from N. E. to N. W." * Although the line of no variation 

 moved through an angle of 80° in longitude, in the course of one hundred and fifty 

 years, on the parallel of Paris, at the Azores it appears to have been stationary for 

 the long interval between 1492 and 1600.f Moreover, the invariability of the decli- 

 nation, for a long period, at Spitzbergen, in New Holland, and among the Antilles, 

 indicates changes in the earth's magnetism which affect different places very un- 

 equally. 



On the eastern continents, the needle reached its extreme westerly position about 

 1800, viz., in 1815 at London, in 1814 at Paris, and in 1791 at the Cape of Good 

 Hope. It was near the extreme easterly limit of its oscillation in 1580 at London, 

 and also at Paris. During this long interval of about two hundred and twenty years, 

 the magnetic meridian of London has swept over an angle of 35|°, and that of 

 Paris over an angle of nearly 34^°. From the report made to the superintendent 

 of the United States Coast Survey, in 1855, by Mr. Charles A. Schott,| it appears that 

 the minimum westerly declination in the United States of America was, on the aver- 

 age of all the stations discussed, in the year 1798. The maximum westerly variation 

 was about the year 1679. Therefore the minimum westerly declination here occurred 

 at nearly the same time as the maximum westerly in Europe. But the range of 

 motion at the three places which entered into Mr. Schott's discussion did not exceed 

 G D . In this particular, a broad distinction is indicated between the two hemispheres. 

 The greatest yearly change occurred about 1741, and again about 1850. If we com- 

 pare with these average values the particular ones which Mr. Schott has obtained for 

 individual places, extending over a wide geographical area, we discover differences, 



» Humboldt's Cosmos, translated by Otte, I. 174; and II. 657. 



f Quetelet thinks that, in the time of Aristotle, it may have been 2-4° east of the meridian of Greenwich. Bull. 

 de l'Acad. Brux., XXI. 303. 



X United States Coast Survey Report, 1855, pp. 315, 337. 



VOL. X. 44 



