IN THE PRESENCE OF DUST-FREE AIR AND OTHER GASES. 
•293 
spending to the least expansion required to make coloured rings appear. With 
greater expansions the size and brilliancy of the rings rapidly increased. 
We cannot suppose that chlorine at the pressures and temperatures used in these 
experiments obeys Boyle’s law even approximately. It is not allowable, therefore, 
to put P 1 /P 3 = vjvy 
Nitrogen. 
This was prepared by adding ammonium chloride to a nearly saturated solution of 
sodium nitrite, till no more would dissolve, and warming gently. In the first series 
of experiments a concentrated solution of potassium bichromate was added. This was 
omitted in the later experiments. The gas was allowed to pass through two U-tubes 
containing strong caustic potash solution. 
The apparatus for the preparation of the gas is shown in fig. 5. 
Fig. 5. 
The liquid was drawn into the flask through the vertical tube. This was then 
made to dip below the surface of mercury contained in a small test-tube. The 
apparatus was then connected to the water-pump, and the liquid in the flask allowed 
to boil under very low pressure, by warming gently. This was allowed to continue 
for some time. The tap was then closed, and the heating continued till the 
pressure exceeded that of the atmosphere and the nitrogen began to escape through 
the safety-tube. Tj was then again opened and the gas allowed to escape through 
Tg and Tg. The temperature of the flask was never allowed to rise more than was 
sufilcient to give a steady stream of the gas, which gradually ceased after the removal 
of the flame. Finally T 2 was closed and the heating discontinued. The nitrogen 
could then be pumped out when required. 
Results. —The results are practically identical with those obtained with air and 
oxygen. 
The experiments of October 6 were made with the same expansion apparatus as 
