indications faithfully recorded. The hours of 3 A.M., 9 A.M., 3 P.M.,
and 9 P.M., are now so generally known as _meteorological hours_, that
nothing should justify a departure from them; and it is the more
essential that these hours should be adopted in the present inquiry,
because the series of observations made at intervals terminated by these
hours can the more readily be used in connexion with those made
contemporaneously on land, and will also serve to carry on
investigations previously instituted, and which have received
considerable illustration by means of observations at the regular
meteorological hours; we therefore recommend their general adoption in
all observations conducted at sea.

It is intended in the sequel to call attention to particular parts of
the earth's surface where it is desirable that additional observations
should be made, in order to furnish data of a more accurate character,
and to mark more distinctly barometric changes than the four daily
readings are capable of effecting. The best means of accomplishing this
for the object in view appears to be the division of the interval of six
hours into two equal portions, and to make the necessary observations
eight times in the course of twenty-four hours. In the particular
localities to which allusion has been made we recommend the following as
the hours of observation:--

A.M. 3, 6, 9, noon. P.M. 3, 6, 9, midnight.

In other localities besides those hereafter to be mentioned, when
opportunities serve, readings at these hours would greatly enhance the
value of the four daily readings.

There are, however, portions of the surface of our planet, and probably