THE COMBINING VOLUMES OF HYDROGEN AND OXYGEN. 
421 
The preparation and purification of the gases and all subsequent precautions were 
the same as in series 5. 
First of all, exactly two volumes of hydrogen and one volume of oxygen were 
exploded and the small residue of oxygen measured. In three other experiments a 
slightly greater proportion of oxygen was taken. The volume of hydrogen was 
614 c.c. in all cases. The oxygen residue was estimated in exactly the same way as 
the hydrogen residue. It would have been more logical to fill the pipette with oxygen 
instead of hydrogen, which was selected because it was more easily procurable. On 
mixing 1 or 2 c.c. of oxygen with 300 c.c. of hydrogen, the resulting volume is not 
strictly additive, but the deviation, though not exactly determinable from existing 
data, is probably very small. 
In these four experiments a new phenomenon was observed. In each case, after the 
final explosion in which oxygen had been present in excess, brown stains were noticed 
on the walls of the explosion vessel. These stains were presumably oxide of mercury 
since they gradually disappeared when mixtures containing an excess of hydrogen 
were detonated in the vessel. If a significant quantity of oxygen was fixed in this 
way, the value of the ratio would be too low. It was a moot point whether it was 
advisable to spark the dried oxygen residue. Coward ( 12 ) has shown that, when the 
pressure is low, traces of uncoinbined hydrogen and oxygen remain after detonating 
electrolytic gas. On the other hand, sparking might lead to the formation of ozone. 
In one case the result would be too high ; in the other too low. The treatment in 
this respect was varied in the different experiments as follows :— 
Example 1. Example 2. Example 3. Example 4. 
0'5 mins., sparking. No sparking. 0'75 mins., sparking. 3 mins., sparking. 
It is significant that the result of experiment 2 was the highest. It should be 
mentioned here that nothing was condensed when the oxygen residue was pumped 
through the liquid-air-cooled spiral. In the following table columns 1 and 2 represent 
the surplus volume of oxygen taken and the residue of oxygen left after explosion, 
respectively. Column 3 gives the difference between the first two columns, repre¬ 
senting the volume by which 307 c.c. of oxygen must be diminished to be chemically 
equivalent to 614 c.c. of hydrogen. 
Without laying any particular stress on these results it may be claimed that they 
do not suggest the presence of a constant error in the earlier experiments. We 
therefore conclude that the ratio of the combining volumes of hydrogen and oxygen 
is represented with a high degree of probability by the figure 
2 , 00288 at 0° C., and 760 mm. pressure, 
which differs from the value of Scott by only 3 parts in 200,000. 
