of the scales do not affect the principle upon which thermometers are
constructed. When we know, for instance, that Fahrenheit's scale is
divided into 212 degrees, in which 32 deg. corresponds with the freezing
point, and 212 deg. with the point of boiling water: and that Reaumur's is
divided only into 80 degrees, in which 0 deg. denotes the freezing point,
and 80 deg. that of boiling water, it is easy to compare the two scales
together, and reduce the one into the other. But, for greater
convenience, thermometers are sometimes constructed with both these
scales, one on either side of the tube; so that the correspondence of
the different degrees of the two scales is thus instantly seen. Here
is one of these scales, (PLATE II. Fig. 1.) by which you can at
once perceive that each degree of Reaumur's corresponds to 2-1/4 of
Fahrenheit's division. But I believe the French have, of late, given
the preference to what they call the centigrade scale, in which the
space between the freezing and the boiling point is divided into 100
degrees.

[Illustration: Plate II. Vol. I. p. 42.

Fig. 1. Thermometer.
Fahrenheit's Scale.
Reaumur's Scale.
Boiling point of Water
Freezing point of Water

Fig. 2. Differential Thermometer.]

CAROLINE.

That seems to me the most reasonable division, and I cannot guess
why the freezing point is called 32 deg., or what advantage is derived
from it.

MRS. B.

There really is no advantage in it; and it originated in a mistaken
opinion of the instrument-maker, Fahrenheit, who first constructed these
thermometers. He mixed snow and salt together, and produced by that
means a degree of cold which he concluded was the greatest possible, and
therefore made his scale begin from that point. Between that and boiling
water he made 212 degrees, and the freezing point was found to be at
32 deg.

EMILY.

Are spirit of wine, and mercury, the only liquids used in the
construction of thermometers?

MRS. B.

I believe they are the only liquids now in use, though some others, such
as linseed oil, would make tolerable thermometers: but for experiments
in which a very quick and delicate test of the changes of temperature is
required, air is the fluid sometimes employed. The bulb of air
thermometers is filled with common air only, and its expansion and
contraction are indicated by a small drop of any coloured liquor, which
is suspended within the tube, and moves up and down, according as th