8 MEMOIRONTHE 



an iron alembic holding about twenty pounds of fused nitre, on hoisting the 

 alembic off the fire, a jet of the liquefied salt fell accidentally upon some water in 

 a tub, which was unfortunately too near. It also brought to mind that potassium, 

 when thrown upon the surface of water, is, by combustion with the oxygen of that 

 liquid, converted into a fused globule of red-hot oxide, which, in the act of com- 

 bining with water, detonates violently. This detonation struck me as being 

 clearly owing to a sort of double reaction, in which, while one portion of water, 

 by uniting with the oxide of potassium, converts it into hydrate of potash, another 

 portion, uniting with the heat, flies off" explosively as steam. 



10. In a letter to Hayes, immediately after his explanation appeared, I stated these 



before you the facts connected with the subject of the action of saltpetre on substances usually called 

 combustible. 



Saltpetre, or the nitrate of potash, or soda, alone, does not burn, or explode by heat, however intense. 

 It parts with one of its constituents, oxygen, by heat, and it is to the combination of its oxygen with 

 other bodies that it owes its power of burning with them. Wood and other fibrous substances do not 

 burn with saltpetre until they have become partially charred ; they then produce deflagration, or burn 

 with sparks. A large quantity of saltpetre, enclosed in gunny bags, as it is usually stored, after fire was 

 communicated to it, would burn with the bags, emitting much smoke and sparks, precisely as paper 

 which has imbibed saltpetre would. It would not be consumed ; only the small quantity required to 

 burn with the bags would be changed. If an addition of burning wood or charcoal were made to the 

 extent of one-fifth the weight of the saltpetre, an intense and continued deflagration would result, and 

 all the saltpetre would be changed. No explosion would follow from applying fire to mixtures of charcoal, 

 or wood and saltpetre ; the rapid combustion called deflagration would be produced, but, unlike explo- 

 sion, time would be required for the mutual actions ; and where the quantities were large, many hours 

 would be necessary before they would cease. The recent destruction of life and property in New York, 

 the loss of a homeward-bound Indiaman and her cargo, by a similar cause, have created an anxiety which 

 has led to many inquiries respecting the origin of the explosions attending the burning of saltpetre. I 

 need not remind you of a case which occurred at Central Wharf, about ten years since, when the Hart- 

 ford Packet was destroyed. The testimony obtained in the last instance led me to make some experi- 

 ments on the effects produced by dropping water on a burning mixture of saltpetre and charcoal. It was 

 ascertained that a very small weight of water, relatively to the saltpetre, caused explosions, which might 

 be made successive, so long as the materials remained. The quantities of the substances acting, being 

 increased to between one and two hundred pounds, the addition of water, in the form of spray, caused 

 an explosion which destroyed the vessel, and shook all the buildings in the vicinity. The temperature of 

 a burning mixture of saltpetre and charcoal, at the points of contact, is superior to that .of " white hot " 

 iron, and the form is that of a bubbling fluid. Water falling on the mass is instantly conveited into 

 steam, having the elastic force of that used in steam-guns ; exceeding gunpowder in destructive energy. 

 The red-hot particles, dispersed by the sudden action, pass over considerable spaces, and the appearance 

 of flame is produced. 



In cases where water falls on highly heated polished surfaces, such as melted glass, copper, or silver, 

 steam is formed rapidly, but silently ; the water does not touch the hot surface. The spreading of a film 

 or crust over the polished surface, instantly alters its relation to water, and causes steam to form with ex- 

 plosive violence, attended by a loud report. 



I do not hesitate in expressing my belief, that the disastrous efi"eots produced in New York were caused 

 by water or other fluid falling on saltpetre, while burning with the bags investing it. The facts which I 

 have stated may have interest or importance in connection with attempts made to extinguish fir^ in build- 

 ings containing saltpetre. The danger of throwing water on the fire is manifest, while the loss to ;the' 

 owner of the saltpetre would doubtless be greater from water than from fire. 



Respectfully, A. A. Hayes. 



