OF GASES AT HIGH EXHAUSTIONS. 
411 
put down to “ experimental errors.” The discrepancies disappear again at an exhaus¬ 
tion of about 1 millim., and from that point to the highest hitherto reached the 
proportion of Ull85 is fairly well maintained. 
670. The curves plotted from the observations of the repulsive force of radiation 
are given in dots at the left side of diagrams B and C. Repulsion commences at a 
somewhat lower exhaustion, and remains much higher than in air until an exhaustion 
of 22 M is reached, when the two curves become practically identical. 
VISCOSITY OF NITROGEN. 
671. Nitrogen is prepared by heating a solution of nitrite of ammonia. If proper 
precautions are taken to avoid intermixture of air, this process yields it in a very pure 
state. It is collected in the bell receiver shown in fig. 12 (666). To eliminate 
aqueous vapour the gas is passed through phosphoric tubes, and then through the tap 
into the viscosity apparatus, as explained in the case of oxygen. The last traces of 
aqueous vapour are removed from the nitrogen by allowing it to pass in a very slow 
stream through a series of tightly-packed phosphoric anhydride tubes, on its way to 
the viscosity bulb. 
Nitrogen was also prepared by burning phosphorus in a confined space of air, and 
also by the copper turnings and ammonia method; but neither of these processes 
yielded a gas which gave uniformly such concordant results as did the nitrite of 
ammonia process. 
The following are the tabulated results of observations with nitrogen :— 
3 H 
MDCCCLXXXi. 
