AND ON PREVENTING ITS EXPLOSION. 53 



extinguishing the flame. But even this is easily obviated, by 

 admitting the air at any height from the floor which may be 

 found requisite. 



Note D. page 42.— To the observations in the text, on the 

 construction of the lamp, a few details may be added. 



The principal circumstance requiring adjustment, is that of 

 the size of the aperture by which the heated air escapes. If it 

 is not sufficiently wide, the flame is faint, and on the move- 

 ment of the lamp becomes unsteady, and is liable to be extin- 

 guished. If it be too wide, there may be some risk of a cur- 

 rent of air entering the lantern by it, especially if the under 

 tube is not sufficiently wide, by which the whole security from 

 the method would be lost. The due size is most easily found, by 

 affixing to the aperture at the top a conical tube, and cutting 

 this down in successive trials, until the diameter is attained at 

 which the flame is steady and bright* It is not easy to give a 

 precise dimension, as the flame is dependent on the breadth 

 and height of the wick, the purity of the oil, and the state of 

 the air ; but I find that in a lantern of the size of the move- 

 able one mentioned in the text, — five inches in height by 

 three in width, with a flat cotton-wick three-tenths of an inch 

 broad, and burning so as to consume about two ounces of oil 

 in six hours, the diameter of the aperture being a very little 

 less than half an inch, admitted of the flame being steady and 

 bright when the combustion was fully established. In the 

 mine it may be required to be a little larger. When the lamp 

 is kindled, it remains for a minute or two more faint, and l£ 

 moved hastily in this state, is liable to be extinguished. One 

 aperture is preferable to two or three smaller apertures, as there 

 is less risk of any counter-current; and to guard also against 

 this in. any movement of the lamp, it is proper to have the 



opening 



