90 EFFECTS of HEAT 



I thus expected to confine the carbonic acid; but the at- 

 tempt was attended with confiderable difficulty, and has led 

 to the employment of various devices, which I fhall now fhort- 

 ly enumerate, as they occurred in the courfe of practice. The 

 fimple application of the principle was found infufficient, from 

 two caufes : Firft, The carbonic acid being driven from the 

 breech of the tube, towards the muzzle, among the pores of 

 the pounded filex, efcaped from the comprefling force, by 

 lodging itfelf in cavities which were comparatively cold : 

 Secondly, The glafs of borax, on cooling, was always found to 

 crack very much, fo that its tightnefs could not be depended 

 on, 



To obviate both thefe inconveniences at once, it occurred to 

 me, in addition to the firfl arrangement, to place fome borax 

 (fig. io. C) fo near the breech of the tube, as to undergo heat 

 along with the carbonate (A) > but interpofmg between this 

 borax and the carbonate, a ftratum of filex (B), in order to 

 prevent contamination. I trufted that the borax in a liquid 

 or vifcid ftate, being thruft outwards by the expanfion of the 

 carbonic acid, would prefs againft the filex beyond k (D),. 

 and totally prevent the elaftic fubftances from efcaping out of 

 the tube, or even from wandering into its cold parts. 



In fome refpecls, this plan anfwered to expectation. The 

 glafs of borax, which can never be obtained when cold, with- 

 out innumerable cracks, unites into one continued vifcid mafs 

 in the low eft red-heat ; and as the ftrefs in thefe experiments, 

 begins only with rednefs, the borax being heated at the fame 

 rime with the carbonate, becomes united and impervious, as 

 foon as its action is necefTary. Many good refults were accord- 

 ingly obtained in this way. But I found, in practice, that as the 

 heat rofe, the borax began to enter into too thin fufion, and 

 was often loft among the pores of the filex, the fpace in which 

 it had lain being found empty on breaking the tube. It was 



therefore 



