OF ARTS AND SCIENCES. 179 



tuted for the hydrogen, and the explosive property remain essentially 

 the same, though of unequal energy, (g) 



"In these facts, a, b, c, d, e,f, g^ lies the explanation of the phe- 

 nomena that have been observed with burning-fluids. 



" The following experiments were made : — 



" I. A current of air was directed into the upper part of a loosely- 

 stoppered laboratory glass spirit-lamp, while burning, causing thereby 

 a mixture of alcohol-vapor and air to rush past the flame. After a 

 moment or two, the jet took fire, and was instantaneously followed by 

 explosion. This result was invariable. 



" 11. After permitting a drop of alcohol, in a large glass flask of 

 small neck, to evaporate for a moment, and applying flame to the 

 mouth, explosion resulted generally, though not invariably. 



" III. Ether similarly treated yielded less uniform results, because, 

 probably, of the greater difficulty of obtaining the proper mixture of 

 ether-vapor and air. 



" IV. A variety of burning-fluid in extensive use, said by the 

 venders not to explode, was subjected to similar experiment, with still 

 less frequent affirmative results. They were, however, sufficient to 

 show that explosions with it are possible. Similar experiments have 

 been made with another variety of burning-fluid, by Dr. Morrill 

 Wyman, with like results. 



" It is, then, conceivable, that, when the proper relative amounts of 

 the vapor of burning-fluid and atmospheric air are mixed together, as 

 they may be in the upper part of a partially filled can or lamp, and a 

 flame is brought suflicienlly near, explosion must result. If the quan- 

 tity of mixed gases be large, the explosion may cause the destruction 

 of the containing vessel, or if that remain entire, it may drive out a 

 portion of the fluid, which, taking fire, may cause more or less injury. 

 The course of safety has been pointed out by the dealers in these 

 articles for illumination. It is to fill the lamps (the tops of which 

 screw on and are not supplied with special air-holes) in the absence 

 ofjlame, by daylight, for example ; in which case no explosion can 

 occur." * 



* " Similar accidents to tliese have taken place in tlie use of the so-called air- 

 tight stoves for burning wood. After the wood has been fired, and the supply of 

 air for some time shut off, on reopening the draft, and sometimes without, occa- 

 sional explosions of great violence have occurred, attended with the blowing out 

 of the stove-door, and in some instances producing still greater injury to the stove. 



