. BY COMPRESSION. 339 



tequ«nce of which, the fufible metal was driven through the 

 iron as water was driven through filver * b)- mechanical per- 

 cuflion in the Florentine CKperiment. It occurred to me, that Remedy. A 

 this might be prevented by confining along with the fufible ^-^^ waTleft'in° 

 meial a fraall quantity of air, which, by yielding a little to the barrel, 

 the expanfion of the liquid, would fave the barrel. This re- 

 medy was found to anfwer completely, and was applied, in 

 all the experiments made at this time f. 



1 now propofed, in order to keep the carbonate clean, to The carbonate 

 inclofe it in a fmallvelTel; and to obviate the difficulty of ^ (^nall feparie 

 removing the refult at the conclufion of the experiment, I vefTel. 

 further propofed to connect that velfel with an iron ramrod, 

 longer than the barrel, by which it could be introduced or 

 withdrawn at pleafure. 



* EJays of Natural Exjieriments made in the Academic del Cimento^ 

 tranflated by Waller, London, 1684, page 117. The fame in 

 Muffchenbroek's Latin Tranflation, Ludg. Bat. 1731, p. QiZ. 



f I found it a matter of much difficulty to afcertain the proper 

 quantity of air which ought to be thus inciofed. When the quantity 

 was too great, the refult was injured by diminution of elafticity, as 

 I (hall have occafion fully to (h€w hereafter. When too fmall, oi* 

 when, by any accident, the whole of this included air was allowed 

 toefcape, the barrel was deftroyed. 



I hoped to afcertain the bulk of air neceffary to give liberty to 

 the expanfion of the liquid metal, by meafuring the aftual quantity 

 expelled by known heats from an open barrel filled with it. But 

 I was furprifed to find, that the quantity thus difcharged, exceeded 

 in bulk that of the aif which, in the fame heats, I had confined 

 along with the carbonate and fufible metal in many fuccefsful ex- 

 periments. As the expanfion of the liquid does not feem capable 

 of fenfible diminution by an oppofing force, this fa6l can only be 

 accounted for by a diftenfion of the barrel. In thefe experiments, 

 then, the expanfive force of the carbonic acid, of the included air, 

 and of the fufible metal, afted in combination againll the barrel, 

 and were yielded to in part by the diileiifion of the barrel, and by 

 the sondenfation of the included air. My objefl was to increafc 

 the force of this mutual ailion, by dirainifning the quantity of air, 

 and by other devices to be mentioned hereafter. Where fo many 

 foices^were concerned, the laws of whofe variations were unknown, 

 much precifion could not be expected, nor is it wonderful, that in 

 attempting to carry the compreHlng force to the utmoft, I fliould 

 have dettroyed barrels innumerable. 



A finaii 



