ON THE PERIODICITY OF THE AURORA BOREALIS. 325 



auroras were observed anywhere, in 1810 and 1813, and none in 1812. It must be 

 remembered that a secular minimum of the aurora approximates to the year 1812. 

 21. Another spot-maximum in 1816. ' As this year comes so close to the secular 

 minimum, we have to deal with a very small number of auroras, in comparing one 

 year with another. There are indications of a slight aurora-maximum in 1817, at 

 Eatisbon, Carlsruhe, Germany, Great Britain, Sweden, and Abo. At Arkhangel three 

 auroras were observed in 1816, seven in 1817, and sixteen in 1819. In Maine there 

 are none from 1809 to 1813 inclusive, but many in the years which cluster around 

 1816, though the largest number was seen as late as 1818. At Salem, also, the aver- 

 age near 1816 much exceeds the average near 1810, the greatest number, however, 

 being as late as in 1819, and none in 1816 and 1817. The observations at Boston and 

 New Haven, so far as they prove anything, would postpone the aurora-maximum to 

 1818 or 1819. 22. Another spot-minimum follows in 1823. No aurora was seen in 

 Europe that year, except in high latitudes. One aurora was seen in Abo, and two in 

 Arkhangel. In Boston not so many auroras were seen in 1822, 1823, and 1824, as 

 in the three years preceding or following. In Maine none were seen in 1822, 1824, 

 1825, and 1826, though two were seen in 1823. In the five years of which 1823 is 

 the central epoch, far fewer auroras were seen in Maine than in the five years which 

 preceded or followed. Farry and Von Wrangel have commented on the rarity and 

 poverty of the aurora, even in arctic latitudes, during the winter of 1822 -23. 1 

 23. Seven years bring about the next spot-maximum in 1830. In Great Britain there 

 was a regular increase in the number of auroras from 1824 to 1830, the minimum 

 being zero, and the maximum being thirty-two. The decrease afterwards was far less 

 regular. A similar remark will apply to Abo, the minimum being one, and the maxi- 

 mum being thirty-four. At Arkhangel the number of auroras increases, but not with 

 perfect regularity, from zero in 1824 to thirteen in 1831, and then the series of 

 observations terminates. In Germany there were many more auroras at the maximum 

 of 1830 than at the minimum of 1823, if the maximum and minimum are understood 

 as covering each five } r ears; but the largest number in a single year belongs to 1831. 

 At Leith there is no indication of a maximum, unless it be placed as early as 1827. 

 In St. Petersburg a maximum of twelve auroras was seen in 1830, although in 

 1829 there were none. In the low latitudes of Europe more auroras were registered 

 in 1831 than are recorded for any other year since 1789 ; but the number was only 

 five. In America the observations made in the States of New York and Maine, also 

 in New Haven and Wilmington, favor the theory of an aurora-maximum in 1830, and 



1 See page 103 of this volume of the Memoirs of the American Academy. 



