130 Professor Adams on the Theory of the Thermometer. 
results lead to the conclusion, that air, mercury, and the so- 
lid metals, generally, may be employed without daoger of 
error as thermometric substances. Ur. Ure himself says, that 
rods of pure copper and iron corresponded in their dilata- 
tions pari passu with two of the best mercurial thermome- 
ters, through the extended range between melting ice, and 
540°. Fahrenheit. He also admits that a mercurial, adjust- 
ed by an air thermometer, gives correct indications of tem- 
perature. And as the expansion of air, mercury, &c al 
kcep pace without any sensible discordance, thermometers 
constructed with them, will all be strictly comparable with 
each other. 
In selecting substances for constructing thermometers, 
four appear to be sufficient in every case for which the in- 
strument can be wished. These are air, alcohol, mercury, 
and a combination of platinum and copper. The state of 
the first will remain unchanged, as far as we know, through 
any range of temperature, but the practical management of 
it, except within the limits of a moderate range, is so diffi- 
cult, that it will probably never be entensively employed. 
second substance is best adapted for the measurement 
of temperatures below the point of the congelation of mer- 
cury. The third is extremely well fitted for common use, 
since among other advantages, it remains unchanged in form, 
- through a very long range of temperature. 
In trigonometrical surveying. the base for a series of tri- 
cy. and to do this, the temperature must be constantly 
known. For the French surveys, the celebrated Borda 
prepared a platinum rod, of twelve feet in length for the 
measuring rod, to which was fastened at one end a rod of 
copper, a little shorter than itself, which could slide freely 
ment varies. the rods are unequally expanded, and the grad- 
uation previously adjusted by the points of melting ice and 
boiling water, gives the temperature of the instrument.” 
* Biot, Traite de physique, J. 163. 
