286 Tornado in N. E. part of Ohio, Feb. 4, 1842. 
Mean of the bottom trees N. 55° W.; top trees N. 62° E.; 
intermediate trees N. 16° E, Here is a Sretisalesliis anifornsiag 
which cannot be ascribed to accident. The bottom tree points 
invariably westerly, and the top tree easterly. ‘To this there is 
no exception. ‘The directions of the intervening trees are inter- 
mediate between the extremes, passing round in the order west, 
north, east. T'o this rule there are only four exceptions marked 
with an asterisk. 'The direction of the third tree in the third 
group differs 30° from that of the fourth; the third tree in the 
sixth group differs 38° from the second; the third tree in the 
eighth group differs 65° from the fourth; the fourth tree in the 
fourteenth group differs 80° from the third. These are the anom- 
alies 30°, 38°, 65°, 80°. How are they to be explained? Do 
they invalidate the inference which naturally follows from the 
_ other bearings, that the wind on this part of the track revolved 
in the direction west, north, east? I think not. For first, the 
exceptions are but four to fifty two, or one to thirteen cases. —Sec- 
ond, it is not to be supposed that the rotation of the wind was 
absolutely uniform like the motion of a clock. In the ease of 
what passes for a rectilinear current, the wind often veers by jerks 
- back and forth from its mean direction through an are varying 
from 50° to 90° in five minutes. The mean variableness a 
Hudson for an interval of five minutes is 42°; so that it coul 
not be deemed surprising if objects prostrated By what is called a 
rectilinear current should often differ in their directions by this 
amount. But thirdly, the direction of the principal roots of a tree 
will somewhat influence its fall; and finally where thousands ¢ 
trees are falling at the same idastait, as in the present ease} in & 
dense forest, their tops must frequently interfere, and a tree be 
turned greatly from the position it would otherwise occupy- The 
force of this last remark will be perceived when it is considered 
that the tornado had entirely passed in seventy five seconds, and 
its most destructive violence probably did not exceed fifteen se 
onds. It should however in candor be admitted that some of the 
ing causes here named would have va influence, from the 
extreme violence of the wind. 
If the motion of the air were merely centri seal, that is, directed 
right a cram the circumference of the tornado to the centté, 
a on the two sides of the track should be pe™ 
2 nd a diagram of the directions of prostrale 
