THE AMERICAN STORM OF DECEMBER 20, 1836. 5 



direction and velocity of its motion from day to day. But unfortunately few if 

 any, observations have been obtained from points north of the centre of the storm 

 At Fort Snelhng near latitude 45° on the Mississippi river, the entire oscillation 

 of the barometer from the lowest point on the 20th to the highest on the 21st was 

 .67 inch. At Natchez the oscillation was .7-1 inch, and at New Orleans (U inch 

 It may then be inferred that along the valley of the Mississippi river, the centre 

 of the area of low barometer was near St. Louis. But tlie greatest depression of 

 the barometer anywhere observed was at Quebec. Assuming St. Louis and Quebec 

 to lie npou the central path of the area of low barometer, the direction of its pro- 

 gress was from S. 60° W. to N. 60° E. There can be little doubt that the barometric 

 wave travelled towards some point north of cast, and not towards the south of east, 

 as has been inferred by Messrs. Hare and Espy. From the morning of Dec. 20th 

 to the morning of December 21st, the centre of the storm advanced in the direction 

 already assigned 1050 miles, being an average velocity of -44 miles per hour. 



Observations of the Thermometer. 



The following table presents a summary of the observations of tlie thermometer 

 at most of the stations. Column first shows the names of the stations ; columns 

 second and third their latitude and longitude ; and colurfin fourth the mean tern- 

 perature at each station for the month of December. The mean temperature for 

 the given hours of observation was obtained by applying to the mean temperature 

 of the month a correction for the hour of the day as determined by hourly observa- 

 tions made at Montreal and Philadelphia. This result was subtracted from the 

 observed temperature, and the remainder indicated the number of degrees by which 

 the observed temperature differed from the mean temperature for that hour of 

 December. The fluctuations of the thermometer were required for 8 A. M. and 

 8 P. M. of each day; and whenever these hours did not correspond to the actual 

 hours of observation at any station, the fluctuations for these hours were interpo- 

 lated by a comparison of the preceding and subsequent observations. The results 

 are inserted in columns five to twelve, which accordingly show how much the ther- 

 mometer was above or below its mean height for the dates given at the top of the 

 table : — 



