the movement of
the end of the temporary magnet.

[Illustration: _Fig. 20. MAKING A PERMANENT MAGNET_]

This was repeated over and over a great many times, and finally the hard
steel bar was found to have a charge of magnetism, and for the purpose
of providing a means for holding the magnetism, a C-shaped piece of iron
was put on the bar, as shown in the detached figure.

"Is that the reason," asked Harry, "why a small piece of metal is always
put across the ends of a horseshoe magnet when it is not in use?"

"Yes."

CHAPTER X

STARTING ON THE VOYAGE TO THE WEST

"If I remember correctly, you stated some time ago, Professor, that the
barometer indicated the pressure of the atmosphere, and in that way it
was useful in letting us know what the weather would be. Before we sail,
would it not be well to make one of them? If we had possession of one of
the articles, we might not have been caught in the storm the first time
we took out No. 3."

"That is a good suggestion. I intended to propose that, because with the
barometer and the compass we shall be equipped with two of the most
useful instruments needed."

"I cannot comprehend how the air pressure has anything to do with the
weather. Is the air pressure really greater at one time than at
another?"

"Heated air ascends, does it not?"

"Yes; I can understand that."

"As it ascends it is, therefore, lighter at that point than normally. On
the other hand, moist air is heavier than dry air. These two conditions
would be indicated by the barometric column, would they not?"

"I presume they would; but when the air is moisture laden we don't need
a barometer to tell it is going to rain. We know it and feel it. What I
particularly wanted to know was how the barometer by its actions would
indicate it ahead for any length of time."

"The barometer does not indicate with any degree of accuracy on land;
but on sea it has a much better application. The instrument shows the
present pressure of the atmosphere, and its variations correspond to
atmospherical changes which have already taken place, the effects of
which may follow their cause at a greater or less interval."

"Then how could it be ascertained from the instrument when there would
be a storm or rain?"

"After a continuance of dry weather, if the barometer begins to fall
slowly and steadily, rain will certainly ensue; but if the fine weather
has been of long duration, and the mercury may fall for two