MancJuster Memoirs, Vol. xlvi. (1901), No. 3. 15 



this is true not only for an elastic bar falling on an anvil, 

 but also for a plug of water in a pipe if it be moved at a 

 velocity V and then suddenly stopped. Thus, in the glass 

 pipes which 1 intend to burst to-night, we have a partial 

 vacuum about 2 feet long on the one side of a plug of 

 of water 6 inches long ; by breaking off the small pro- 

 jecting glass tube at the water end the plug is suddenly 

 exposed to atmospheric pressure, and is propelled towards 

 the other end. The mass per square inch of sectional area 

 of this plug is about y|^. The velocity- acquired while 

 travelling a distance of 2 feet is therefore about 94 feet 

 per second, and the pressure, which in a fluid naturally 

 acts in ever)' direction (hydrostatic) is 



Q4 •I20.0CO.Q4 ^ - . , 



P= E . jTr= - — -;: — - = 0,400 pounds f)er square mch. 



Here JFis the velocity of sound in a C)'linder of water, 

 and E is the volume elasticity- of water in pounds per 

 square inch. 



It will thus be seen that the pressure which can be 

 produced is so excessive, that not only will it fracture the 

 glass tubes, but it will carry the glass splinters to a con- 

 siderable distance, and if the pipe were of steel, one 

 might expect at least a swelling to take place at the 

 closed end. 



We have in this experiment a reproduction of a steam- 

 pipe explosion, illustrated as Fig. 2, taken from the Board 

 of Trade Report No. 594. The valve on the boiler was 

 being opened when the pipe just above the steam hammer 

 exploded. Here, evidently, water had settled above the 

 boiler valve and was shot over as in the experimental 

 glass tube. Similar explosions are detailed in a few other 

 Board of Trade Reports. 



Several explosions have occurred near intermediate 

 stop-valves, probably caused by a plug of v.-ater tra. elling 



