616 
imaginary level can be diluted at pleasure, and the changes 
which the lamp undergoes can be easily seen through the 
square of glass in the side of the box. A Stephenson lamp was 
inserted, when the gas was turned on there was first a flicker- 
ing of the flame, which had burnt steadily while the air was 
pure, and at one time it appeared totally extinguished ; a tiny 
blue light, however, remained wavering about within the top 
of the gauze, which gradually increased in size as it was 
drawn through the gauze by which it was enclosed. The 
gauze in a short time got red hot, and after the lapse of one 
minute and thirty- five seconds the gas in the box exploded; 
the lamp was then removed and was found to be still burning 
and uninjured, but generally the light was extinguished 
without exploding the mixture in the box. "With the current 
of air passing through the box at the rate of five miles an 
hour, not an uncommon velocity in a pit, as ascertained by 
the anemometer, the following comparative results were 
obtained : — 
The Davy Lamp 
exploded 
in 5 
seconds. 
The Mousard 
do. 
6 
do. 
The Clanny 
do. 
12 
do. 
The Belgian 
do. 
13 
do. 
The Stephenson 
do. 
25 
do. 
All the above-named lamps were tested under the same 
circumstances. Some further experiments showed slight 
variations in the time, but all tended to two results. In the 
first place it was shown that all the safety lamps were really 
unsafe in a current of atmospheric air and gas when mixed 
to an explosive point. And, secondly, that of the lamps now 
in use the Stephenson is the best and most reliable. 
In a still atmosphere of mixed air and gas all the above 
lamps were extinguished without igniting the surrounding 
atmosphere. 
It will be noticed that the foregoing experiments were 
