over which they pass appear as surface-winds, the winds in fact being
disposed in alternate beds or layers, S.W., N.E., as in fig. 3.

[Illustration: Fig. 3.]

The Professor also suggests that these parallel and oppositely directed
winds are shifting, _i. e._ they gradually change their position with a
lateral motion in the direction of the large arrow cutting them
transversely.

In the course of the author's researches on atmospheric waves he had an
opportunity of testing the correctness of Professor Dove's suggestion,
and in addition ascertained that there existed another set of oppositely
directed winds at right angles to those supposed to exist by the
Professor. These currents were N.W. and S.E. with a lateral motion
towards the N.E. He also carefully discussed the barometric phænomena
with relation to both these sets of currents, and arrived at the
following conclusions. The details will be found in the author's third
report, presented to the British Association for the Advancement of
Science (Reports, 1846, pp. 132 to 162). During the period under
examination the author found the barometer generally to rise with N.E.
and N.W. winds, and fall with S.W. and S.E. winds, and that the
phænomena might be thus illustrated:--Let the strata _a a a' a', b' b' b
b_, fig. 3, represent two parallel aërial currents or winds, _a a a' a'_
from S.W. or S.E., and _b' b' b b_ from N.E. or N.W. and conceive them
both to advance from the N.W. in the first instance and from the S.W. in
the second, in the direction of the large arrow. Now conceive the
barometer to commence rising just as the edge _b b_ passes any line of
country, and to continue rising until the edge _b' b'_ arrives at that
line, when the maximum is attained. It will be remarked that this rise