238 ON THE PERIODICITY OF THE AURORA BOREALIS. 



curve the ordinates of which represent the various richness of the years in solar spots. 

 The most conspicuous maxima of the aurora-curve are in the years 1730, 1769, 1779, 

 1788, 1839 and 1848 : of the spot-curve, they are in the years 1769, 1779, 1787, 1837 

 and 1848. The principal minima of the aurora-curve are in the years 1758 and 1809- 

 13: the most decided minima of solar-spots are in the years 1755 and 1810. Fritz 

 thinks it probable that the secular period embraces, on the average, five of the eleven- 

 year periods or 55.55 years. 



Wolf 1 draws a secular curve through the summits of the undulations expressing 

 the relative numbers of solar spots, and its maxima appear in the years 1779 and 1837, 

 and its minima in 1816 and 1860. He draws another secular curve through the lowest 

 points of the undulations and the maxima occur in 1775 and 1833-4, and the minima 

 in 1810 and 1856. The long swell is equal to five or six of the short waves. Car- 

 rington 2 explains the smaller elevation of the summits of the secular curve of the 

 solar spots in the latter part of the last century, when compared with the summits of 

 the present century, by improvements in telescopes or by an increase in the number 

 of observers. 



The eleven-year period of the spots rides also on the back of the secular curve of the 

 aurora, these subordinate maxima and minima of the two phenomena coinciding. The 

 years 1723, 1733, and 1745, which are known approximately to be spot minima, are 

 also deficient in auroras. Such exceptions as appear in the interval between 1813 

 and 1820 Fritz explains by the incompleteness of the catalogues, particularly that of 

 the aurora. The interval between the two grand maxima of 1788 and 1839 is fifty- 

 one years ; between those of 1730 and 1788 is fifty-eight years. The average of the 

 two intervals is fifty-four and one-half years. If the year 1848 is substituted for the 

 year 1839 as that of the last great maximum, the first interval will be increased to 

 sixty years and the mean of the first and second to fifty-nine years. Although Wolf con- 

 siders the coincidences just mentioned between the spots and the aurora as proving the 

 cosmical origin of the latter, he objects to the theory of Mairan and Olmsted, which would 

 associate it with the zodiacal light and meteors, that such years as 533, 763, 1096, 1766, 

 1799, 1833-4, which swarmed with shooting-stars, are in no way remarkable for auroras. 

 The secular period of the meteors is only 33-4 years. However, five of these periods 

 contain about the same number of years as three of the secular periods of the aurora. 



Wolf draws a secular curve for the aurora which he calls the European curve. 



1 Vierteljahrsschrift der Naturforsch. Gessellsch. in Zurich. VI. 190, 1. 

 3 Monthly Notices of the London Astronomical Society, XX. 282. 



