I wish to communicate to you the result of my exp,eriments in making the spirit lamp nearly it not quite safe. I have made my common lamps, as I conceive, free from the danger of exploding, simply by filling the body of the lamp with coarse clean sponge in such a manner as to leave no spaces occupied by the fluid as a liquid free to run. You will perceive that when the lamp or rather the sponge is filled with fluid, that if by accident it should be upset, the fluifl cannot escape from the sponge in consequence of being held by capillary attraction, therefore, as far as the danger of setting fire by spilling is concerned, the sponge prevents it. Again, the danger of explosions does not consist in the fluid being explosive, but the vapor which arises from it, which vapor occupying that (upper) portion of the lamp having no liquid in it, issues out from the screw aperture, and if a flame be within a few inches, it takes fire and explodes, carrying fire to whatever fluid may be left in the lamp. Now, the sponge which fills every portion of the lamp, leaves no space (comparatively) unoccupied, to hold this vapor, hence the amount of vapor Which can possibiy be in a lamp at any time cannot do much damage, as I have frequently shown to my friends by taking off the screw and bringing it in contact with a flame, the fluid held by the sponge would simply take fire and continue to burn until exhausted or blown out. It may be objected to on the ground that the lamp would have to be made much krgeitjgual to ma\!:e np the difrence ofthe' space otSftjjffed- ""he ' spongeon thi point I would say that I tried the experiment with two lamps of the same capacity, and found that there was but ten minutes difference in the time of one hour, the sponge burning one hour, and the plain lamp one hour and ten minutes, so that this is no proper objection. I have now used lamps arranged in this way for a year or more, and would not use the fluid in any other manner. I have nothing to make by publishing the subject, save the satisfaction of doing some little good in saving life, 'c.
This article was originally published with the title "To Prevent Lamp Explosions" in Scientific American 8, 13, 99 (December 1852)