388 On the Temperature at which (May, 
the one may still be capable of sliding into the other. This done, 
the larger is to be nearly filled with distilled water; the smaller is 
then to be let down into it in its whole length, so that its thickened 
end may rest on the bottom of the bulb; and its other end is to be 
fixed in the middle of the bore, by wedging in around it some small 
slips of leather or of cork. Lastly, a scale is to be affixed to the 
neck, extending to some distance above and below the level at which 
the surface of the water now stands: and then the instrument will 
be fit for use. 
Two tubes which consist of equal quantities of glass will be 
equally expanded or contracted by the addition or abstraction of 
equal quantities of heat. It is evident, also, from the construction 
of the instrument, that that portion of the inner tube which is con- 
tained within the neck of the outer consists of just as much glass as 
goes to constitute the neck itself; and that portion which is con- 
tained within the bulb, of just as much glass as goes to constitute 
the bulb itself. Hence any variation of temperature which either 
extends to the whole instrument, or, in equal proportions, to its cor- 
responding parts, will occasion an equal variation of bulk, either in 
its two members, in general, or in the corresponding portions of 
them in particular. 
But the space between these two members—that space in which 
the water is contained—will be oppositely affected by similar varia- 
tions of bulk in the two members. Every expansion of the outer 
tube and bulb will tend to enlarge that space, and consequently to 
lower the level of the water in the neck ; whereas every expansion 
of the inner tube will tend to diminish the same space, and to raise 
the level of the water: and the converse will obviously hold good 
with regard to contractions. Hence if the two members of the 
instrument be equally and simultaneously expanded or contracted, 
the space between them will neither be enlarged nor diminished. 
That space, therefore, can suffer no variation of capacity from any 
change of temperature which extends to the whole instrument, or 
in equal proportions to the corresponding parts of it; the level of 
the water in the neck can be affected by no such cause; and conse- 
quently, if it be raised or depressed, this must be occasioned by an 
“actual change of density in the water itself. In short, we here call 
into exercise such a principle of compensation, that the effects of 
temperature on the glass will be completely neutralized, and the 
apparent changes in the volume of the water identified with the 
real ones. 
Such, at least, would be the case if the instrument could be 
constructed with theoretrical perfection. But it must be con- 
fessed that there are some practical difficulties and sources of inac- 
curacy which cannot wholly be avoided. It is only by a rare chance 
that one can expect to light upon two tubes of different widths, 
which are, length for length, exactly of the same weight. Even 
supposing such tubes to be obtained, if the bulb be made very 
1 
