THE COMBINING VOLUMES OF HYDROGEN AND OXYGEN. 
419 
rejected. In this experiment, when the tap on the explosion vessel was opened to 
admit the hydrogen residue to the pump, the mercury thread in the capillary beyond 
the tap broke up into a series of globules which settled down and obstructed the 
passage of the gas. To remove them it was necessary to heat the capillary strongly 
with a flame, and since the glass in this region was contaminated with tap-grease, 
some chemical action may have occurred. 
Table V. 
1. 
Hydrogen 
excess. 
2. 
Hydrogen 
residue. 
3. 
Difference. 
4. 
Ratio of 
combining 
volumes. 
5. 
Atomic weight 
of 
hydrogen. 
c.c. 
c.c. 
c.c. 
1 
3-428 
2-575 
0-853 
2-00278 
1 
•00767 
2 
3-461 
2-585 
0-876 
2-00285 
1 
•00771 
o 
o 
3-080 
2-219 
0-861 
2-00280 
1 
•00769 
4 
3-131 
2-242 
0-889 
2-00290 
1 
•00773 
5 
3-078 
2-176 
0-902 
2-00294 
1 
■00775 
7 
3-222 
2-351 
0-871 
2-00284 
1 
•00770 
8 
3 • 855 
2-986 
0-869 
2-00283 
1 
•00770 
9 
4 • 339 
3-429 
0-910 
2-00296 
1 
•00777 
10 
3-160 
2-274 
0-886 
2-00289 
1 
•00773 
11 
3-000 
2-090 
0-910 
2-00296 
1 
•00777 
12 
3-155 
2-268 
0-887 
2-00289 
1 
•00773 
13 
3-141 
2-276 
0-865 
2-00282 
1 
•00769 
Mean .... 
2-00287 
1 
•00772 
±0-00001 
Maximum deviation, 1 part in 11,000. 
The mean is practically identical with the corrected mean of the first four series, so 
that the estimate of the “ contraction ” correction is satisfactorily confirmed. 
Although, in this series, no uniformity of procedure was adopted as regards the 
distribution of the gas measurements over the two days, examination of the results 
reveals no discrepancies attributable to this cause. 
In experiments 1 , 3, 9, 11 and 13, for example, the oxygen was measured on the 
afternoon of the first day, whereas in experiments 2, 4, 5, 8, 10 and 12, it was measured 
on the morning of the second day ; yet the mean result for each group is the same. 
Experiment 7 was completed in a single day. 
Except for the longer interval between iceing and setting, the presence of the small 
subsidiary ice bath on the manometer tube and the juxtaposition of the two final 
measurements constituted the only difference from series 4. The lower results of 
series 5 cannot be attributed to the presence of the little ice bath since the volume 
3 L 2 
