paid to the direction and force of the
wind preceding the barometric maximum--and the same phaenomena succeeding
it, and particular notice should be taken of the time when, and amount
of any change either in the direction or force of the wind. It is by
such observations as these, carried on with great care and made at every
accessible portion of the oceanic surface, that we may be able to
ascertain the continuity of these atmospheric waves, to determine
somewhat respecting their length, to show the character of their
connexion with the rotatory storm, and to deduce the direction and rate
of their progress.

V.--SEASONS FOR EXTRA OBSERVATIONS.

In reference to certain desiderata that have presented themselves in the
course of my researches on this subject (see Report of the British
Association for the Advancement of Science, 1846, p. 163), the _phases_
of the larger barometric undulations, and the _types_ of the various
seasons of the year, demand particular attention and call for extra
observations at certain seasons: of these, three only have yet been
ascertained--the type for the middle of November--the annual depression
on or about the 28th of November--and the annual elevation on or about
the 25th of December. The enunciation of the first is as under: "That
during fourteen days in November, more or less equally disposed about
the middle of the month, the oscillations of the barometer exhibit a
remarkably symmetrical character, that is to say, the fall succeeding
the transit of the maximum or the highest reading is to a great extent
similar to the preceding rise. This rise and fall is not continuous or
unbroken; in some cases it consists of _five_, in others of _three_
distinct elevations. The complete rise and fall has been termed the
great symmetrical barometric wave of November. At its setting in the
barometer is generally low, sometimes below twenty-nine inches. This
depression is generally succeeded by _two_ well-marked undulations,
varying from one to two days in duration. The central undulation, which
also forms the apex of the great wave, is of larger extent, occupying
from three to five days; when this has passed, two smaller undulations
corresponding to those at the commencement of the wave make their
appearance, and at the close of the last the wave terminates." With but
slight exceptions, the observations of eight successive years have
confirmed the general correctness of this type. On two occasions th