On the Action of the Barometer in a Hurricane, ete. 19 
The facts to which I principally refer are set forth in the follow: 
ing diagram. Here it will be perceived that from the morning 
April 27th. 28th, 29th. 30th. May Ist. 
9 3 ¢ 
— ss PME 
50 mS 
10 \ 
29-00 
Diagram representing the barometric fluctuations on the day of the tornado, and 
on the three preceeding days at New Harmony, Indiana. 
The horizontal lines are equal to fluctuations amounting to to of an inch. The 
Vertical lines represent the height of the barometer at the time of sunrise; 9 a. Mm. 
and 9 P.M. on each day ; the darker li I ting tl ise ob ti 
A represents the fluctuation of the barometer between the bours of 3 and 6 P. M., 
on the 30th of April, at which hours the atmosphere was calm and clear; but be- 
ween which the tornado came up, committed its ravages and passed on. 
ous to the coming 
inch in the 
