40 Mr. Redfield on the Rotary Action _ 
Mr. Allen says further, “* The progress of the tornado was 
nearly in a straight line, following the direction of the wind, 
with a velocity of perhaps eight or ten miles per hour. Near as 
I was to the exterior edge of the circle of the tornado, I felt 
no extraordinary gust of wind, but noticed that the breeze 
continued to blow uninterruptedly from the same quarter 
from which it prevailed before the tornado occurred. I also 
particularly observed that there was no perceptible increase 
of temperature of the air adjacent to the edge of the whirl- 
wind, which might have caused an ascending current by a 
rarefaction of a portion of the atmosphere.” 
Soliciting a careful attention to the observations of Mr. 
Allen, who is well known for his intelligence and his habits of 
correct observation, I proceed to give some account of my 
own examinations of the traces of this tornado. 
From a point on the rocky “ledge” north of the turnpike 
road and nearly three miles westerly from Providence, to the 
house of John Burr on the Cranston road, a distance of about 
one and a quarter mile, I found the course of the tornado to 
have been 8. 86° E. by compass, over a plain country. The 
magnetic variation being here about 8° westerly, makes the 
true course E. 3° N. From this point to Providence river, a 
distance of about two miles, the course was five degrees more 
northerly. 
I agree with Dr. Hare that the general effects observed on 
this track were “ quite similar” to those of the New Bruns- 
wick tornado; and will give such of my sketches, formerly 
prepared, as will best illustrate this similarity and the general 
effects here mentioned. 
The following is a sketch of some of the effects on the farm 
of Mr. Burr: his house is about one mile and a half from the 
Providence bridge. 
The effects here exhibited appear to me to be due to a pro- 
gressive whirlwind, revolving to the left; for we may notice, 
as in the New Brunswick tornado, « more onward direction 
in the trees prostrated on the right of the axis, d, m,n, 0, &c., 
than on the left side; while the outermost prostrations on the 
right, 2, 0, point still more nearly than the average on this 
side, to the course of the tornado: and on the left side of the 
track we have the tree / in a direction inclined several degrees 
backward from the course of the storm. The value of these 
indications of whirling action T have endeavoured to point out 
in my remarks on the New Brunswick case. [Silliman’s 
Journal, vol. xli. p. 70-75.) 
At the front of the house a, however, were two slatted door- 
yard fences, extending from the house to the road. The 
