inoxes as above recommended, it would be desirable to continue the
observations until a complete elevation and depression of the barometer
had been observed at these seasons. This plan is adopted at the Royal
Observatory, Greenwich, and would be attended with this advantage were
it generally so--the progress of the elevation and depression would be
more readily traced and their velocities more accurately determined than
from the four or eight daily readings.

III.--LOCALITIES FOR ADDITIONAL OBSERVATIONS.

In sketching out a system of barometric observation having especial
reference to the acquisition of data from which the _barometric
character_ of certain large areas of the surface of the globe may be
determined--inasmuch as such areas are distinguished from each other, on
the one hand by consisting of extensive spaces of the oceanic surface
unbroken, or scarcely broken, by land; on the other by the proximity of
such oceanic surface to large masses of land, and these masses
presenting two essentially different features, the one consisting of
land particularly characterized as continental, the other as insular,
regard has been accordingly had to such distribution of land and water.

As these instructions have especial reference to observations at sea,
observations on land have not been alluded to; but in order that the
data accumulated may possess that value which is essential for carrying
on the inquiry in reference to atmospheric waves with success, provision
is made to mark out more distinctly the barometric effects of the
junction of large masses of land and water. It is well known that the
oceanic surface, and even the smaller surfaces of inland seas, produce
decided inflexions of the isothermal lines. They exercise an important
influence on temperature. It has also been shown that the neighbourhood
of water has a very considerable influence in increasing the
oscillations of the mercurial column in the barometer, and in the great
systems of European undulations it is well known that these oscillations
increase especially towards the north-west. The converse of this,
however, has not yet been subjected to observation; there has been no
systematic co-operation of observers for the purpose of determining the
barometric affections of large masses of water, such as the central
portion of the basin of the northern Atlantic, the portion of oceanic
surface between the Cape of Good Hope and Cape Horn, the Indian and
Sou