or towards
night and plentiful in the early morning." Huxley in his _Physiography_
makes the addition "without production of mist." The formation of mist
is not necessary for the formation of dew, nor does it necessarily
prevent it. If the deposit of moisture is in the form of ice instead of
water it is called hoarfrost. The researches of Aitken suggest that the
words "by condensation of the vapour in the atmosphere" might be omitted
from the definition. He has given reasons for believing that the large
dewdrops on the leaves of plants, the most characteristic of all the
phenomena of dew, are to be accounted for, in large measure at least, by
the exuding of drops of water from the plant through the pores of the
leaves themselves. The formation of dewdrops in such cases is the
continuation of the irrigation process of the plant for supplying the
leaves with water from the soil. The process is set up in full vigour in
the daytime to maintain tolerable thermal conditions at the surface of
the leaf in the hot sun, and continued after the sun has gone.

On the other hand, the most typical physical experiment illustrating the
formation of dew is the production of a deposit of moisture, in minute
drops, upon the exterior surface of a glass or polished metal vessel by
the cooling of a liquid contained in the vessel. If the liquid is water,
it can be cooled by pieces of ice; if volatile like ether, by bubbling
air through it. No deposit is formed by this process until the
temperature is reduced to a point which, from that circumstance, has
received a special name, although it depends upon the state of the air
round the vessel. So generally accepted is the physical analogy between
the natural formation of dew and its artificial production in the manner
described, that the point below which the temperature of a surface must
be reduced in order to obtain the deposit is known as the "dew-point."

In the view of physicists the dew-point is the temperature at which, _by
being cooled without change of pressure_, the air becomes saturated with
water vapour, not on account of any increase of supply of that compound,
but by the diminution of the capacity of the air for holding it in the
gaseous condition. Thus, when the dew-point temperature has been
determined, the pressure of water vapour in the atmosphere at the time
of the deposit is given by reference to a table of saturation pressures
of water vapour at different temperatures. As it