MECHANICS AND USEFUL ARTS. 57 



found its way into the boiler at all, as was proved by actual expgri- 

 ment with it closed, when the first puff of steam generated forbade 

 the entry of more water. This opening, however, could have in no 

 way assisted to prevent the bursting of the boiler, if the views 

 generally entertained with regard to the explosive effect of dash- 

 ing water on red-hot plates were correct; the action of water un- 

 der these circumstances is supposed to be as irresistible and in- 

 stantaneous as that of gunpowder. 



The second boiler, also of copper, weighed 44 pounds, and 

 measured 11| inches in height, Ilk inches in width, by 10 inches 

 in depth at the bottom and 8 inches at the top, having a flue tube 

 running through it 6 inches in diameter, so that it had an internal 

 capacity of about five-eighths of a cubic foot. This was treated just 

 as the first had been, and, when nearly half of the boiler was red- 

 hot, water was suddenly turned into it through a 1-inch pipe, con- 

 nected to the boiler at one end, and to a tank affording ahead of 6 to 

 8 feet in height at the other. Jt was supposed that an increased 

 diameter of pipe would give a more sudden injection of water, 

 and therefore prove more favorable to an instantaneous injection 

 of steam. There was, however, no explosion. The boiler re- 

 mained perfectly still, the only effect produced being the escape 

 of a jet of steam through a hole in the top of the boiler, left open, 

 for the purpose. 



In the third experiment a cast-iron boiler was used, which 

 would possibly prove more favorable to explosion, both on ac- 

 count of the brittleness of the material and from its greater weight 

 of metal, which would afford increased capacity for heat, and thus 

 for the more rapid generation of steam. The boiler weighed 85 

 pounds, and measured 15| inches in length, 10 inches in height, by 

 11 inches in depth at the bottom, and 8| inches at the top, having 

 an internal capacity of less than a cubic foot, while the bottom was 

 arched, which increased the heating surface. This was heated 

 till the greater part became red-hot, when water was let on as in 

 the second experiment ; but there was no orifice left open on the 

 top of the boiler as before, a safety-valve, loaded to a pressure of 

 about 35 pounds on the square inch, being substituted. On let- 

 ting in the water, no result whatever was apparent ; the safety- 

 valve did not blow, and the boiler neither cracked nor trembled, 

 but the feed-pipe got hot some 15 feet from the boiler, as if the 

 steam was beating back and forbidding further entry of the water. 

 Finding no result could be produced with the safety-valve, shut or 

 open, this was removed, and an orifice 1| inches in diameter left 

 open instead. On turning the water on again, a jet of steam es- 

 caped from the orifice as before, and, shortly after, the boiler 

 cracked on one side from the top to the bottom with a sharp re- 

 port. This was due simply to the contraction of the metal, and 

 the rupture did not spread, neither did the boiler stir from its seat. 

 This experiment was repeated, with the opening in the top reduced 

 to three-eighths of an inch, with the same result. The capacity of 

 these boilers was such that a pressure of about 150 pounds on the 

 square inch would have been generated within them by the evap- 

 oration of a quarter of a pint of water in the two larger ones, and 



