at sea as well as
on land, only those who have long watched their indications, and
compared them carefully, are really able to conclude more than that the
rising glass[3] USUALLY foretells less wind or rain, a falling barometer
more rain or wind, or both; a high one fine weather, and a low, the
contrary. But useful as these general conclusions are _in most cases_,
they are _sometimes_ erroneous, and then remarks may be rather hastily
made, tending to discourage the inexperienced.

By attention to the following observations (the results of many years'
practice and many persons' experience) any one not accustomed to use a
barometer may do so without difficulty.

The barometer shows whether the air[4] is getting lighter or heavier, or
is remaining in the same state. The quicksilver falls as the air becomes
lighter, rises as it becomes heavier, and remains at rest in the glass
tube while the air is unchanged in weight. Air presses on everything
within about forty miles of the world's surface, like a _much_ lighter
ocean, at the bottom of which we live--not feeling its weight, because
our bodies are full of air, but feeling its currents, the winds. Towards
any place from which the air has been drawn by suction,[5] air presses
with a force or weight of nearly fifteen pounds on a square inch of
surface. Such a pressure holds the limpet to the rock when, by
contracting itself, the fish has made a place without air[6] under its
shell. Another familiar instance is that of the fly which walks on the
ceiling with feet that stick. The barometer tube, emptied of air, and
filled with pure mercury, is turned down into a cup or cistern
containing the same fluid, which, feeling the weight of air, is so
pressed by it as to balance a column of about thirty inches (more or
less) in the tube, where no air presses on the top of the column.

If a long pipe, closed at one end only, were emptied of air, filled with
water, the open end kept in water and the pipe held upright, the water
would rise in it more than thirty feet. In this way water barometers
have been made. A proof of this effect is shown by any well with a
sucking pump--up which, as is commonly known, the water will rise nearly
thirty feet, by what is called suction, which is, in fact, the pressure
of air towards an empty place.

The words on scales of barometers should not be so much regarded for
weather indications, as the rising or falling of the mercury; for, if it
stand at _