662 
MR. IT. B. DTXON OX CONDITIONS OF CHEMICAL CHANGE 
No difference was found in the coefficient whether lo - 9 or 24’3 volumes of oxygen 
were used, but the mean value of the coefficient in these four experiments was 
appreciably higher than in the two previous series, which gave 5 for the value of a 
both with 14'2 and 17'7 volumes of oxygen. Apart from the variation of the oxygen, 
the only other difference between these and the previous experiments was a small 
difference in temperature amounting to some five or six degrees. In the first series 
the gases were exploded at a mean temperature of 14°'5, in the second series at a mean 
temperature of 13 0, 3, and in the last four experiments at a mean temperature of 8 0, 1. 
On calculating out the coefficient a from Bunsen’s later experiments, 4 'and omitting 
the first as being below the critical pressure, its mean value is found to be 5’8 for a 
mean temperature of 7 0, 2. 
Table XXIX.— Bunsen’s experiments with dry gases. 
Temperature. 
Oxygen. 
Carbonic oxide. 
Hydrogen. 
Coefficient a. 
”C. 
10-3 
107 
78-6 
21-4 
4-9 
f 4 6 
12-5 
74-9 
257 
5 - 6"| 
| 5-3 
14-5 
70-9 
297 
5-8 1 
I 7-3 
147 
7L8 
28-2 
6-2 y 
87 
H -9 
667 
33-9 
6-0 j 
Lio- 
207 
58-5 
41-5 
5-5J 
Mean 7’2 
Mean 5'8 
Here a high value for the coefficient is also accompanied by a low temperature. It 
appeared possible then that a variation of a few degrees in the temperature of the 
gases before the explosion might materially alter the result. A few experiments 
showed this to be the case. 
Experiments on the influence of initial temperature. 
Portions of a mixture of the dry gases were exploded successively under 1000 
millims. pressure at different temperatures. In the first two experiments the 
eudiometer was cooled down by placing ice water in the jacket, in the other experi¬ 
ments the eudiometer was warmed by hot water. 
The mixture was of the following composition : 
Carbonic oxide.70'8 
Hydrogen.2 9'2 
1000 
Oxvgen added . . 16‘8 
* “ Gasometrisclie Methoden, II te Anflage.” 
