Auroral Observations at New Haven. 169 



If we make a proper allowance for the interference of twilight in sum- 

 mer, we must conclude that Auroras at New Haven are more common in 

 summer than in winter ; that is, they occur least frequently in winter, and 

 most frequently in autumn. The same conclusion has been derived from a 

 comparison of a considei'able number of observations in Canada, and also 

 at the Academies in the State of New York. 



The inequality in the number of auroras upon successive years is much 

 more remarkable. From 1763 to 1794, auroras averaged 7 per year; and 

 during eight of these years the average was 20 per year. From 1795 to 

 1826, a; period of 32 years, only 21 auroras were recorded, being an average 

 of only two in three years; while from 1837 to 1854, auroras averaged 29 

 per year, and there were six years in which at least 40 aui'oras per year were 

 recorded. It is true that during the latter period, a systematic watch for 

 auroras was maintained, and most of the time by tAvo independent observers, 

 so that it is probable that nearly every aurora which could have been seen 

 has been recorded. Moreover, many of these auroras were very faint, such 

 as most meteorological observers would not notice ; and even if they had 

 noticed them, would not think them worthy of being recorded in a Meteor- 

 ological Journal; while from 1795 to 1826, it is presumed that no particu- 

 lar watch for auroras was maintained. Under such circumstances, it is 

 probable that during the latter period a good many small auroras may 

 have occurred which failed to be noticed. But it is impossible to avoid 

 the conclusion that auroras were much more frequent during the 33 years 

 succeeding 1826, than during the preceding 33 years. From 1837 to 1854 

 not a year passed without at least one conspicuous aurora, and generally 

 there were at least three or four sufficiently conspicuous to attract consider- 

 able attention. But from 1804 to 1820 a very faithful meteorological Jour- 

 nal was ke^Dt by Pres. Day, of Yale College, (at that time Professor of Nat- 

 ural Philosophy,) who only recorded 17 auroras, adding in one case the 

 remark, " with streamers," and in a second case, " Bow from E. to W." 

 Yet Prof Day was born in 1773, and was therefore of an age to remember 

 many of the splendid auroras which were recorded by President Stiles. He 

 was therefore perfectly familiar with the appearance of auroras, and was 

 well aware how great interest these phenomena had once excited. He still 

 remembers (1865) several auroras which occurred before he entered college, 

 and, in particular, the grand aurora of Nov. 1789, which occurred when he 

 was Freshman in college ; and he well remembers at present that during 

 the period from 1800 to 1820, auroras were much less frequent, and less 

 splendid than they had been during his boyhood. 



These recollections of President Day are abundantly confirmed by his own 

 written testimony. "When Professor of Natural Philosophy, he was accus- 

 tomed to deliver a course of written lectures on Meteorology to the Senior 

 Class of Yale College; and in his lecture on the Aurora, written in 1811, 

 and still preserved, on page 5 th, after describing that form of the Aurora 

 22 



