( 162 ) 



After this, I infcrt the point of the tube reprefented at fg. 27^ 

 wherein the Gunpowder was fired into the aperture of the glafs, 

 fig. 23, at W, fo that the bent part of the t\ibe R, may prefs the 

 bottom of the glafs at T, and continue the preflure until the tube 

 breaks about Q, whereupon the compreffed air in tlie cavity of the 

 glafs, jig. 27, ruflies violently into the glafs, fg. 28 ; but as it can- 

 not efcape through the aperture W, it places itfelf about V, by 

 which means the fame fpace of water is expelled at W, as of air 

 is forced from the glafs, fg. 27, into the glafs, fg. 28, and u]M)n 

 weighing this lail mentioned glafs, when filled with water, and 

 again when part of the water, as before mentioned, has been driven 

 o\it of it, I can make an cxaifl computation of the quantity of com- 

 prefied air driven out of the glafs, fg. 27. 



Again, I take another glafs of the Ihape reprefented ntfg. 27,. 

 withthis difference only, that the tube at Q. is flraight, and the 

 aperture left unclofed, and then infert this into the glals, fg. '28, 

 fo as not quite to touch the bottom ; then, upon liring tlie grain of 

 Gunpowder, the condenfed air rufhes with great force into the 

 glafs 28 ; after which firing, and the heat of the flame elcaping out 

 of the glafs 27, inflantly, as I may fay, the water makes its -way 

 through the orifice Q, into the glafs 27, in greater quantity than 

 one could imagine. 



In thefe experiments, by the firing of the Gunpowder and the 

 rufliing in of the water, the glaties are often broken ; therefore, in 

 making thefe obfervations, it will be neceflary to repeat the experi- 

 ments many times. 



-*- 





