f, as is well known, these bodies so
damped be put into a dry atmosphere, as, for instance, one confined over
sulphuric acid, or if they be heated, then they yield up this water again
almost entirely, it not being in direct or permanent combination[A].

[A] I met at Edinburgh with a case, remarkable as to its extent, of
hygrometric action, assisted a little perhaps by very slight solvent
power. Some turf had been well-dried by long exposure in a covered
place to the atmosphere, but being then submitted to the action of a
hydrostatic press, it yielded, _by the mere influence of the
pressure_, 54 per cent. of water.

622. Still better instances of the power I refer to, because they are more
analogous to the cases to be explained, are furnished by the attraction
existing between glass and air, so well known to barometer and thermometer
makers, for here the adhesion or attraction is exerted between a solid and
gases, bodies having very different physical conditions, having no power of
combination with each other, and each retaining, during the time of action,
its physical state unchanged[A]. When mercury is poured into a barometer
tube, a film of air will remain between the metal and glass for months, or,
as far as is known, for years, for it has never been displaced except by
the action of means especially fitted for the purpose. These consist in
boiling the mercury, or in other words, of forming an abundance of vapour,
which coming in contact with every part of the glass and every portion of
surface of the mercury, gradually mingles with, dilutes, and carries off
the air attracted by, and adhering to, those surfaces, replacing it by
other vapour, subject to an equal or perhaps greater attraction, but which
when cooled condenses into the same liquid as that with which the tube is
filled.

[A] Fusinieri and Bellani consider the air as forming solid concrete
films in these cases.--Giornale di Fisica, tom. viii, p. 262. 1825.

623. Extraneous bodies, which, acting as nuclei in crystallizing or
depositing solutions, cause deposition of substances on them, when it does
not occur elsewhere in the liquid, seem to produce their effects by a power
of the same kind, i.e. a power of attraction extending to neighbouring
particles, and causing them to become attached to the nuclei, although it
is not strong enough to make them combine chemically with their substance.

624. It would appear from many cases of nuclei in