THE SMITHSONIAN INSTITUTION. 



207 



^0. 4 — Curve of temperature and pressure for Steuben, Washington county, Maine. Latitude 

 44° 44' JV. ; Longitude 67° 50' W. 

 7 8 9 10 11 12 13 14 15 16 



The continuous line represents the changes of the barometer, and 

 the dotted Kne those of the thermometer. The figures on the left-hand 

 margin of the wood-cut indicate heights of the barometer in intervals 

 of one-tenth of an inch, and those on the right hand side, of the ther- 

 mometer in intervals of five degrees of Fahrenheit's scale. 



The figures along the upper margin indicate the days of the month 

 during which the storm continued, and the letters along the lower the 

 direction of the wind during the same time. 



In order to exhibit more strikingly the relation of the temperature, 

 moisture, and pressure, the dotted line is inverted so as to represent 

 the highest point of the thermometer by the lowest point of the curve. 

 With this arrangement the two curves in the first three figures are 

 nearly parallel, showing that in the middle regions of the United States 

 tlie abnormal depressions of the barometer are attended with a rise in 

 the thermometer and consequently an increase of moisture, while in 

 the eastern States, as illustrated by figure 4, the result is almost en- 

 tirely opposite — the increase of temperature is accompanied with a 

 rise in the barometer. 



If we suppose that a south wind, of 10,000 feet in height, with a tem- 

 perature of 70°, occupies the Mississippi valley, and the upper current 

 from the west flows across its surface, in ordinary circumstances this 

 stratum of air would weigh a little more than nine inches of mercury. 

 Now, if we further suppose a cold northwest wind, having its dew point 

 or morning temperature down to 30°, to displace the warmer current, 

 by crowding it to the east and occupying its place, as ten thousand feet 

 of air at this low temperature weighs l-12th more, or f inch of mer- 

 cury, this denser and colder column will cause a rise in the barometer 

 to that extent. Now, if a decrease of temperature of 40° causes a rise 

 of f inch of mercury, a decrease of 5° would cause a rise of nearty 

 1-lUth inch. This effect would be produced without taking into ac- 

 count the changes which might take place during storms in the upper 

 strata of the atmosphere. 



I have compared the fluctuations of the barometer at 7, a. m., and 

 9, p. M., with the changes in the temperature at these hours from the 

 7th to 16th November last, in the States of Maine, Vermont, New York, 

 Michigan, Wisconsin, North Carolina, Kentucky, Missouri, East Florida, 

 Alabama, Texas, and also at Washington. By making 5° of temper- 



