1888.] J. T. Bottomley on a Practical Air Thermometer. 91 
shown, the results obtained from the instrument are but very 
slightly affected by the initial pressure of the air, and this with 
very wide limits ; and by commencing at common temperatures 
with air of small density, very low pressure, the upper limit of the 
range may be extended without increasing the length of the mano- 
metric tubes. , 
The filling I accomplish in the following way : — The proper 
quantity of liquid is first introduced into the volume gauge, and the 
stop-cock t helps in introducing the liquid and in adjusting the 
quantity. For this and the subsequent operations I use a really 
good Bunsen water aspirator, with a Woulfe’s two-necked bottle 
interposed between the aspirator and the work, and a good length 
of small bore non-collapsible india-rubber tubing. With the india- 
rubber tubing the apparatus to be exhausted can be turned into 
any required position while the exhaustion is being carried on, and 
air-bubbles can be got rid of with ease. 
When the volume gauge has been supplied with liquid, I connect 
the three-way stop-cock t' to the aspirator, and draw the whole of 
the liquid up into the bulb h' and the tube leading up to the 
stop-cock itself. The size of the bulbs and of the tubes is, as has 
been explained, such that when this has been done the bulb h is 
empty as well as the tubes on the left-hand side of the gauge almost 
down to the bend. The three-way stop-cock is then closed, and 
the aspirator disconnected. 
I now, with the help of a temporary three-way stop-cock, connect 
together the tail piece of the bulb shown in fig. 3, the aspirator, 
and a train of drying and purifying tubes (sulphuric acid and 
caustic potash). The arrangement is such that, on turning the tap 
of the three-way stop-cock into position No. 1, the aspirator draws 
the air out of the bulb ; while, on turning it into position No. 2, air 
flows into the bulb passing through the drying tubes. The bulb is 
emptied and refilled many times ; and during this process the 
bulb and all the tubes are heated with a Bunsen flame very 
nearly to the melting point of the glass.* When it is perfectly 
certain that there is nothing but pure dry air in the bulb and tubes, 
* By this process every trace of moisture and condensed air is driven up 
from the walls of the tube ; and, the bulb being filled with perfectly dry air, 
it seems certain, from the experiments of Bunsen and from experiments which 
