of the Providence Tornado. 41 
fence e was overthrown northward toward f, and the fence f 
in the contrary direction towards e ; both directions being 
transverse to the line of the axis, which passes between them. 
Such cases have been adduced as supporting a directly in- 
ward course of the wind in the body of the tornado; or as 
indicating two bodies of opposing wind meeting on a central 
line ; but I draw a different conclusion. 
In this figure a repre- Vig. 1. Providence Tornado* . 
sents a wooden dwelling- 
house of two stories with 
chimney at its centre; 6a 
dwelling added to @ and 
extending to the rear; a 
lighter building about 16 
feet by 30, attached to the 
rear of b: g was a large 
wooden barn: a long 
building or shed extending 
from the barn to the car- 
riage-house 7. The width 
of the visible track was 
here about five hundred 
feet, and the course of the 
centre or axis of the tor- 
nado appeared to have 
passed somewhat diagonally 
over the three first-named buildings. 
The house a withstood the shock, receiving some damage; the chimney 
top of 6 was thrown on the roof of a, perforating the same, while 4 was 
unroofed and greatly injured, and a Jong timber or sill from the shed h 
broke endwise into the upper part of the house 6 from a north-westerly 
direction. The building c was turned more than twenty feet to the deft 
about, as regards the axis of the whirlwind, against the top of the prostrated 
pear-tree d, and was there overturned upon it. There were twenty-one 
persons in a@ and 4, including a school of children, none of whom were 
seriously injured. 
The barn g and the shed h were destroyed, and the materials swept off 
toward the first-named buildings. A corn-house, standing on the same 
side with the barn, is stated in the Providence papers to have been blown 
over to the west, but I can find no notes of my own respecting the direction 
of its fall. 
Let fig. 2 represent, horizontally, the directions of such 
centre blowing winds in the body of the tornado, and let it be 
supposed as passing over the area of fig. 1, without revolving, 
so as the course of the centre will coincide with the arrow 
which indicates the course of the axis on that figure. It may 
thus be seen that on this hypothesis the wind must strike 
the fences e, f, either parallel to their length, or but little ob- 
* On these plans the large dot at the end of the several short lines 
shows the original position of the root of the tree; the pointed end of the 
line shows the direction of its top. 
