chemical change by measurement of the gases evolved. 605 
stored up in the solution was evolved, and the subsequent obser- 
vations of the rate of evolution of gas from the stirred solution, and 
those taken in the earlier part of the experiment, were found to 
agree closely together with the theory of a unimolecular reaction. 
This experiment furnishes the most conclusive evidence that 
the rate of evolution of gas from an unstirred solution is not a 
measure of the rate of reaction, and in conjunction with the 
comparison experiment on the decomposition of hydrogen per- 
oxide described above, establishes the validity of the method of 
investigation put forward in the present paper. 
From the above experiment the following results are obtained 
for the rates of evolution of nitrogen from its supersaturated 
solution and the corresponding amounts of gas stored in the 
solution, the volume of the solution being 37 c.c. 
Rate of evolution 
of gas, 
c.c. per minute 
Volume remaining 
supersaturated, 
c.c. 
Rate of evolution 
of gas, 
c.c. per minute 
Volume remaining 
supersaturated, 
c.c. 
9 
18-5 
2-75 
12-3 
11-5 
21-3 
1-5 
10-3 
11-5 
2T2 
0-95 
8-7 
9-25 
19-7 
0-7 
7-3 
6-75 
17-5 
0-5 
6-1 
4-5 
15-6 
0-35 
5-4 
3-3 
14-2 
0-2 
4-3 
Fig. 11 gives the curve obtained by the graphical representa- 
tion of these numbers. In the absence of very certain data as to 
the solubility of nitrogen at about 60° it is difficult to determine 
accurately the number of times by which the greatest amount of 
gas remaining supersaturated in the solution exceeds the volume 
normally dissolved by that volume of water at the same tempera- 
ture. If the solubility of nitrogen is taken to be '006 at 62°, a figure 
which does not seem too low, then the greatest degree of super- 
saturation obtained is nearly 100. This value has been exceeded 
in some experiments. 
The largely increased rate of evolution of nitrogen for high 
degrees of supersaturation seems to be due to a great extent to 
effervescence of the solution. 
