AND PREVENTING ITS EXPLOSION. . 52° 
and in some places is not more than four feet. On the other 
hand, the occasional falling of the roof forms dome-like cavi- 
ties above this height; it is in these that the inflammable air 
chiefly accumulates, and in the lower passages or workings, 
where they are open, the draught of air must in general pre- 
vent it from being collected. Hence the limits to the eleva- 
tion of the lamp, and of course tothe operation of the principle 
on which the method is founded, are less than they at first ap- 
pear; though even. the height of three, or three feet and a 
half from the floor, while it is probably best adapted to the ne- 
cessary illumination, will give the requisite security. 
A. lamp with oil is more convenient than a candle, as requi- 
ring no adjustment. with regard to the wick; and by the com- 
mon. contrivance of a. plate with a screw-on the aperture, the 
oil is prevented from being spilt, on any occasional inclination 
of the lamp. The usual time of a. miner’s. work is six hours ; 
the lamp, of the size just. now mentioned, with fresh oil and. 
wick, burns seven hours. The miner, therefore, may take it. 
with him newly trimmed, and the lantern need. never be open- 
ed in the mine, by which any risk from the communication of | 
its flame to the surrounding air may be avoided. If it were ne- 
cessary that it should burn for a longer period, it might, with- 
out any inconvenience, be made of a larger size ; and the wick 
might be made sa.as.to admit of being raised by. a contrivance. 
similar to that of Arcann’s. All the joinings of the case or. 
lanthorn, it is obvious, ought to be as close as possible. 
By employing a metallic lantern, with a lens of very thick 
glass in front, the risk from breaking, which is incurred when 
a glass case is used, is avoided. This construction has other: 
advantages. It affords a great deal of light in the most fayour- 
able manner; the illumination. being directed with less loss on 
the space where it is required.. Where the situation admits of 
i, 
