or three
days before any perceptible change takes place; the more time elapses
before the rain comes, the longer the wet weather is likely to last."

"Then what indicates dry weather?"

"If, after a great deal of wet weather, with the barometer below its
mean height, the mercury begins to rise steadily and slowly, fine
weather will come, though two or three wet days may first elapse; and
the fine weather will be more permanent in proportion to the length of
time that passes before the perceptible change takes place."

"Is this the case at all times of the year?"

"The seasons affect the barometer, it is true. A sudden fall of the
barometer in the autumn or in the spring indicates wind; in the summer
or in hot weather it prognosticates a thunderstorm; in winter, after
frost, a sudden fall of the mercury shows a change of wind or a thaw
with rain; but in a continued frost a rise of mercury indicates
approaching snow."

"It seems, then, that a man must be pretty well versed in the weather to
be able to read the signs."

"That is a correct observation. The instrument in the hands of one who
has had experience with its use is absolutely necessary; it is not a
very satisfying device for those who do not take the time or trouble to
read all the signs, and note all the indications."

As detailed in a preceding chapter, the hardened steel rod for the
compass was brought out for the purpose of securing it in a little case,
so that it might be utilized to give them the true north.

It was a difficult task to find a means of suspending it, for the reason
that they had no tools which would make fine and carefully pivoted
balances, but eventually this was done, and they were gratified to see
the little rod or bar swing around and point north and south.

The work of arranging suitable closets for the various provisions and
providing a miniature kitchen was the next thing in order. This occupied
several days. Instead of taking the bedding in their house, it was
decided that new mattresses should be made up from the barley, of which
there was quite a quantity on hand.

One of the receptacles taken from the cave was a copper jar, which held
five gallons of water. A top was made for this which could be sealed up,
to hold a reserve supply of water. In addition two other vessels were
also provided for the regular supply, and also fitted with covers, so
that they had about ten gallons, an amount which was considered
sufficient.