n found which were consistent with
gyration--for irregularities attend the violent action of all forces, and
particularly the motion of electricity through the atmosphere, as every
one who has seen the zig-zag course of a flash of lightning knows--yet the
evidence of two lateral inward currents, or lines of force, has
predominated over all others. In all cases, where the path is narrow,
those lateral currents are the actors; they constitute the tornado; their
_irregularities_ of action produce the exceptions; but the exceptions are
neither numerous nor uniform, and do not prove either the theory of Mr.
Espy or that of Mr. Redfield. The action is not that of moving air,
merely, but of a power exceeding in force that of powder, which nothing
but electricity or magnetism can exert. As the path widens, the wind
becomes more like the straight-line gust which follows beneath the
ordinary severe thunder-showers. His theory finds no substantial
confirmation or support in the path of the tornado.

Several storms were investigated by Professor Espy, some of them the same
which Mr. Redfield had attempted to show were of a rotary character; one
or two by the Franklin Institute of Philadelphia; one by Professor Loomis,
already alluded to; and recently, two by Lieutenant Porter, from logs
returned to the National Observatory. None of these investigations confirm
the theory of Mr. Redfield. Indeed, Mr. Redfield himself has found it
necessary to resort to suppositions of _modifying causes_ to explain the
evident inconsistencies. It is assumed that the axis, or center,
oscillates, and describes a series of circles; and thus, one class of
difficulties is avoided. Again, it is assumed that simultaneous storms
converge and blend upon the same field, and another class of difficulties
are surmounted. And, again, inasmuch as it is notorious that violent gales
are rarely if ever felt with equal violence around the area of a circle,
but from one or two points only, it is assumed, that the storm winds
ascend, superimpose, and descend again, when they return to the place of
their first violent action, etc. The _simple truth_ requires no such
resort to _modifying hypothesis_.

Still, another objection is, that the changes in the barometer, which
occur before, during, and after storms, do not sustain the claims of Mr.
Redfield or the requirements of his theory.

The barometer sometimes rises before storms. It generally commences
falling about the