When this gas is mixed with atmospheric air, it inflames by a 

 lighted taper ; but unless the quantities are duly proportioned, the in- 

 flammation is too slow to be accompanied with sound. The com- 

 bustion appeared to be most sudden when there were about 7 or 8 

 parts of air to 1 of the gas. When 1 of the gas is mixed with 15 or 

 1 6 of air, the candle burns with an enlarged flame, which does not 

 extend to the whole quantity, and the same phenomenon occurs 

 when the gas constitutes no more than -Vth part of the mixture. 



The author next made trial of the degree of heat necessary for the 

 combustion of this gas, which in fact is difficult to ignite, for even 

 the electric spark will not explode a mixture containing 5 parts of 

 air to 1 of the gas ; but when the air is increased as far as 6 to 1 , or 

 nearer to the most explosive proportion, then it does take fire by the 

 electric spark. Even the most explosive mixture will not take fire 

 from well burned charcoal, nor by iron heated to any degree short 

 of actual combustion. But on the contrary, hydrogen whether sim- 

 ple or sulphuretted, olefiant gas, or carbonic oxide, may each be set 

 on fire very readily by all these means ; and hence the gas procured 

 by distillation from coal is liable to be fired by ignited iron or char- 

 coal, in consequence of the olefiant gas which it contains along with 

 the carburetted hydrogen in the first stage of the process of distil- 

 lation, and on account of the carbonic oxide and hydrogen which it 

 contains in the latter stages of its production. 



Since the fire-damp of coal mines is free from these admixtures, it 

 might be possible to light a mine by means of a charcoal fire, if free 

 from flame, which is necessary for the ignition of the gas. 



The method, however, which the author proposes, is not depend- 

 ent on this property, but upon the supply of air to a lamp employed 

 for that purpose being limited, in such a manner that the lamp will 

 be extinguished by impurity of the air before an explosion can take 

 place. 



If a lighted lantern, in the bottom of which there is only a small 

 aperture just sufficient for its supply, be inserted in a mixture of 

 common air, with about iVth of fire-damp, the flame first appears 

 feeble, next enlarges, and then is uniformly extinguished. In the 

 lantern here proposed for coal mines, a chimney is added of such 

 length as to prevent any communication from the enlarged flame up- 

 wards with the external air ; and the opening below is designed to 

 be at a fixed distance from the flame, by using oil lamps instead of 

 candles. As an additional security in some cases, the author con- 

 ceives it may be useful to add valves at the apertures, which may 

 be made to shut by a thermometric spring when the heat is increased 

 by enlargement of the flame. 



Beside the foregoing, which the author terms the safe lantern, he 

 has constructed two others on the same principle with that proposed 

 by Dr. Clanny, and described in a late volume of our Transactions : 

 but as that of Dr. Clanny has been found objectionable, not only on 

 account of manual labour and constant attention necessary in work- 

 ing the bellows, but also on account of its size, those now proposed 



