commencement of hours, and the horizontal tenths of an inch, would be
quite sufficient for this purpose. The _force_ of the wind should be
noted at, or as near to the time of the passage of the maximum as
possible. During the fall of the mercury particular attention should be
paid to the manner in which the wind changes, should any change be
observed; and should the wind continue blowing steadily in _one_
direction, but gradually _increasing_ in force, then such increments of
force should be most carefully noted. During the fall of the barometer,
should the changes of the wind and its increasing force indicate the
neighbourhood of a revolving storm, (independent of the obvious reasons
for avoiding the focus of the storm,) it would contribute as much to
increase our knowledge of these dangerous vortices to keep as near as
possible to their margins as to approach their centres. The recess from
the centre towards the margin of the storm, will probably be rendered
apparent by the _rising_ of the mercury; and so far as the observations
may be considered valuable for elucidating the connexion of atmospheric
waves with rotatory storms (other motives being balanced), it might be
desirable to keep the ship near the margin--provided she is not carried
beyond the influence of the winds which characterize the latter half of
the storm--until the barometer has nearly attained its usual elevation.
By this means some notion might be formed of the general direction of
the line of barometric pressure preceding or succeeding a storm.

Should a gale be observed commencing without its having been preceded by
an unusual elevation of the mercurial column, and consequently no
additional observation have been made; when the force of the wind is
noted in the usual observations at or above 5, then the three-hourly
series should be resorted to, and the same care taken in noting the
direction, changes, and force of the wind as pointed out in the
preceding paragraph.