the N. W. wind is almost invariably a surface wind, I have, in a
few instances, seen a N. W. set at a considerable elevation, converging
toward a peculiarly stormy state of atmosphere far south of us, about the
period of the spring equinox. And so in one or two instances I think I
have seen light cirro-stratus clouds _above_ the counter-trade, when it
ran very low, setting from the N. E., although the usual and almost
invariable location of the N. E. wind is below the counter-trade and the
stratus clouds of the storm. Aeronauts, too, have found these secondary
currents beneath a serene and cloudless sky. Indeed, the S. E.
counter-trade doubtless often induces a thin secondary current of S. W.
wind between itself and the surface-trade, in the same manner that similar
currents are induced with us, and every where.

A question arises here of considerable interest, which, I confess, I can
not answer to my own satisfaction. It is, whether there be, or not, _an
eastern progression of the body of the atmosphere above the machinery of
distribution_. I have thought there was, and that in set fair weather I
had seen a peculiar kind of cirro-cumulus cloud, in patches, the small
cumuli very distinct and rounded, moving due east, which indicated such a
current. But I am not satisfied, from my own observation, that it is so,
nor is it easy to determine the question. The moisture of evaporation
rarely, if ever, ascends to any considerable elevation, and the upper
strata must be very dry. Hence, condensation, if it takes place, is thin,
and perhaps often undiscernable. Investigations upon mountains prove
little, for the winds of the inferior strata rush up their sides and over
them. It is an open question, and future observation may solve it. The
prevailing opinion seems to be that there is. If the theory of Oersted, in
relation to the circular currents of a magnet, be true, there should be
such a progression produced by opposite secondary currents, unless,
indeed, it be also true that those currents are inoperative at so great a
distance, or their influence barely suffices to retain the attenuated
atmosphere in its place. Perhaps the investigations of Ampere conflict
with it. But it is worth while, I think, for philosophers to inquire
whether the transverse position of the needle upon the wire is not the
effect of the central _longitudinal_ currents, conforming to the circular
currents of the wire, and whether it is not owing to the pr