390 ANNUAL OF SCIENTIFIC DISCOVERY. 



derived from meteorological journals, I do not hesitate to conclude that 

 the common opinion of an unusual fall of rain at London, about the 

 time of the autumnal equinox, has been taken up without reason. I 

 now proceed to inquire what foundation there may be for a similar 

 opinion in the United States. And here we encounter a difficulty 

 arising from the want of a continuous register of the fall of rain at a 

 single locality for any long period. In this country, accurate observa- 

 tions extend only over a period of 25 years. And these are only known 

 at one point, which is at Albany. These were taken under the direc- 

 tion of the regents of the University. But the great inequalities 

 of these observations show that the period of observation is too 

 short to elucidate any correct results as yet ; and that, before arriving 

 at any correct conclusion, we must extend these observations over a 

 long series of years. If the first five days of the month of September 

 be taken, we find 20.62 inches to have fallen ; the next five days only 

 8.81 inches ; the third group of five days 13.34 ; the fourth group 

 of five days 13.82; the fifth 17.16; the sixth 13.48 inches. It will 

 thus be seen that more rain fell on the first clays of September than on 

 any other part of the month. From these figures it would be prepos- 

 terous to come to the conclusion that the first week in September 

 would be wet, and that the second would be peculiarly dry. Another 

 fact worthy of notice is the recorded fact that, for 23 successive years, 

 no rain has fallen on the 6th of September in Albany, yet no popular 

 proverb is prevalent in that vicinity noticing the fact. In order to sup- 

 ply, as far as possible, the want of a series of observations sufficiently 

 long, I have had recourse to the journal of Dr. Holyoke, kept at Salem, 

 Mass., from 1786 to 1828, in which it was recorded, each day, whether 

 it was fair, cloudy or rainy, although the amount of rain was not 

 registered. I have taken the sum of the rainy days for each day of 

 the month, and have appended to this the Albany register, making 

 thus a continuous record for 65 years. The greatest number of rainy 

 days on any one day of the month, for the entire period, was 25, on 

 the 5th, and the least number was 12, on the 6th. So far, then, as 

 these observations indicate any influence of the day of the month upon 

 the amount of rain, they lead to the conclusion that the first five days 

 of September are unusually rain} 7 , and the second five days unusually 

 dry. Still, it would be premature to draw definite conclusions from 

 these facts. On the whole, we may conclude that there is as much 

 reliance to be placed on a storm happening in the New England States 

 at the equinox as at the annual meeting of the Quakers ; or, in the 

 language of the poet, 



"If the first of July be rainy weather, 

 'T will rain more or less for forty days together." 



Prof. Guyot said that he was not surprised that there should be doubts 

 as to the prevalence of any storms at the season of the equinoxes in the 

 United States ; that opinions relative to it came, as he supposed, from 

 Europe, in the central portions of which certain atmospheric changes 

 had a regularity which did not prevail in Great Britain, which stood 

 at the extreme verge of the area over which these periodic changes 



