from forty to one hundred and eighty rods--the most usual
width being from sixty to ninety rods. Sometimes when still wider, they
have more the character of thunder-gusts, and are brightly luminous.

3d. Two motions are usually visible, one ascending one near the earth and
in the middle, and a gyratory one around the other. The latter is rarely
felt, or its effects observed, near the earth. Occasionally, and at
intervals, objects are thrown obliquely backward by it.

4th. It is composed, at the surface of the earth, of _two lateral
currents_, a northerly and southerly one, varying in direction, but
normally at right angles in most cases, although not always, with its
course of progression, extending from the extreme limits of its track to
the axis; which currents are most distinctly defined toward the center,
and upward. These currents prostrate trees, or elevate and remove every
thing in their way which is detached and movable. There does not seem to
be any current in advance of these lateral ones tending toward the
tornado, save in rare and excepted cases, and then owing to the make of
the ground or the irregular action of the currents; nor any following,
except that made by the curving of the lateral currents toward the center
of the spout as it moves on, and perhaps a tendency of the air to follow
and supply the place of that which has been carried upward and forward,
like that of water following the stern of a vessel. The south current is
always the strongest, and often a little in advance of the other, and
covers the greatest area. The proportion of the two currents to each other
is much the same that the S. E. trades bear to the N. E. This excess in
volume and strength of the southerly current will explain the
irregularities in most cases, and the fact that objects are so often
_taken up and carried from the south to the north side_, and so rarely
from the north and carried south of the axis. These irregularities are
such as attend all violent forces, and something can be found which will
favor almost any theory; but the two lateral currents appear always to be
the principal actors, except, perhaps, when it widens out and assumes more
the character of a straightforward gust. See a collection by Professor
Loomis, American Journal of Science, vol. xliii. p. 278.

The following diagram is a section of the New Haven tornado, from
Professor Olmstead's map accompanying his article in the "American Journal
of Science