92 
Proceedings of Royal Soeiety of Edinlurgh. [jan. 6, 
these are allowed to cool with free passage to the atmosphere 
through the drying tubes. The bulb is then surrounded with 
broken ice, and the three-way stop-cock t' is opened. The liquid 
of the volume gauge now finds its level ; and, noting the barometer 
roughly (merely to know approximately the pressure), I seal the tail- 
piece at the extremity. The bulb now contains about the quantity of 
air required, and it is now only necessary to remove the tail-piece. 
For this purpose the ice is taken away, and the liquid of the gauge 
is once more drawn back to a considerable extent, thus making a 
partial vacuum to avoid blowing out of the air during sealing. The 
blowpipe flame can then be applied, and the sealing finished off as 
in fig. 2. Finally, the manometer and pressure apparatus are con- 
nected to the volume gauge, and the constant of the instrument 
is obtained by determining the pressure required, including the 
barometric pressure, to bring the liquid of the volume-gauge into 
the marked position, first at the temperature of melting ice, and 
then at the temperature of steam at normal pressure. When read- 
ing the standard barometer, I also, in accordance with a most 
convenient suggestion by Professor Quincke, read at the same time 
my standard aneroid j and this for most purposes, with occasional 
comparison with the standard, is amply sufficient to give the baro- 
metric variations. As in the case of the mercurial thermometer, 
so also in the air thermometer there is sure to be a secular con- 
traction of the bulb ; and, with the large bulbs used for the 
air thermometer it is quite possible that the redetermination of 
the constant of the air thermometer from time to time may be 
necessary. 
Convenient formulm for calculating temperatures from the indica- 
tions of the air-thermometer are easily obtained. Such formulae 
were given by Jolly [Jubelhand von Poggendorf s Annalen), who 
also made fresh determinations of the expansion of air and other 
gases. Some of these formulae are quoted in the Leitfaden der 
Prahtiselien Physik of Kohlrausch ; but curiously enough there is 
nothing said in the description of the air-thermometer by Kohl- 
I have myself carried out, that there is no subsequent perceptible condensation 
of air at the surface of the glass, such as has sometimes been supposed to 
vitiate the readings of the air thermometer. Air only condenses on the surface 
of the glass when there is moisture present — at any rate in such quantity as 
would he perceptible in a case like the present. 
