110 ON THE SECULAR PERIODICITY OF THE AURORA BOREALIS. 



things had come. The arch which signalized this aurora had been seen by Dr. Holyoke, 

 who was then ninety-nine years old, only twice before, viz. in 1755 and 1769. The 

 sight of a magnificent aurora was so unusual, that in August, 1827, the bells were 

 rung in Salem to call attention to it. 



Accordingly, the discussion of American observations, both of ordinary and extraor- 

 dinary auroras, substantiates the general conclusion first suggested by European 

 observation ; namely, that there is a secular periodicity to the phenomenon, twenty years 

 or more of abundant exhibitions being separated by intervals, equally long or longer, 

 when these displays, if not wholly wanting, are stripped of their more brilliant charac- 

 teristics. Moreover, these repeated interruptions in the return of auroras are such as 

 no failure of memory, no negligence of observers, and no deficiencies of history can 

 adequately explain. 



Another decline in the frequency and brilliancy of the aurora since 1850 is manifest. 

 This will appear from an examination of the Regent's Reports, made since the publica- 

 tion of Mr. Hough's Results &c. It will also appear, by consulting the American 

 Journal of Science, which contains probably a notice of all the conspicuous auroras in 

 recent years. On September 29, 1851, an aurora was visible in the Southern States. 

 On February 19, 1852, a grand aurora was seen in New Hampshire and Vermont. Splen- 

 auroras are also recorded on June 11 and November 11, by Mr. Z. Thompson, of 

 Burlington, Vt. On August 11, 1853, a slight aurora was visible at New Haven. On 



did 



September 2, 1853 



off Cape Race. On May 24, 1853 



aurora appeared at Perryville, in lat. 37° and long. 39°. Auroras were also seen there 

 on April 22 and June 11, 1852. On April 5, 1853, a splendid aurora was witnessed at 

 Burlington, Vt. On April 29, 1859, an auroral arch was seen at New Haven. But 

 the last decade was redeemed from insignificance in respect to auroral displays by the 

 prolonged and magnificent exhibition between August 28 and September 4, 1859, the 

 geographical extent of which embraced the United States and Australia. Poey* 



aurora has been registered only six times at Cuba, viz., on November 

 13, 1784; on November 14, 1789; in November, 1833; on November 17, 1848; and 

 on August 28 and September 1 - 2, 1859. Mr. Logan says that the aurora has been 

 seen in California eleven times in 11 years, of which four were in 1860. The best 

 displays were on August 28, 1859, and July 4, 1860. The late Professor Olmsted, in 

 his valuable Smithsonian Contribution to Knowledge, « On the recent secular period of 

 the Aurora Borealis," accepted for publication January, 1855, has attempted to approx- 



that the 



American Journal of Science. 



