120 



ACTION OF HEAT MOfilFIEt) 



part was sinking, had been in part left behind, and in part 

 drawn out into a thin but tapering shape, united by a curved 

 surface (at b and c) to the middle rod. When the platina tub& 

 xvas unwrapt, the thin edges (at c and d) were preserved all 

 round, and in a state of beautiful scniitransparency. (I have 

 attempted to represent the entire specimen, as it stood on it? 

 cone of wax, in fig. 34.), The carbonate, where moulded on 

 the platina, had a clean pearly whiteness, with a saline ap- 

 pearance externally, and in the sun, shone with facettes. Its 

 surface was interrupted by a few scattered air-bubbles, which 

 had lain against the tube. The intervening substance was 

 imusually compact and hard under the knife. The whole 

 surface {e h a c d, ^g. 31.), and the inside of the air-bubbles, 

 had a vitreous lustre. Thus, every thing denoted a state of 

 viscid fluidity, Uke that of honey. 



These last experiments seem to obviate every doubt that re- 

 mained with respect to the fusibility of the purest carbonate, 

 without the assistance of any foreign substance. 



Force of the 

 compreasjoa 



The apparatus 

 consisted of a 

 barrel of iron, 

 into which the 

 materials were 

 put, and a co- 

 ver was applied 

 on the mouth 

 and kept down 

 by a loaded le- 

 ver. 



Measwcment of the Force required to constrain the Carbonic 

 Acid. — Apparatus uit/i the Muzzle of the Barrel vptiards, 

 aiid the xveigkt acting by a long Lever. — -Apparatus xvith the 

 Muzzle dowi-wards. — Apparatus with Weight acting direct- 

 ly on the barrel.— Comparison of various results. 



In order to determine, within certain limits at least, what 

 force ha<l been exerted in tlie foregoing experiments, and what 

 was necessary to ensure their success, I made a number of ex- 

 periments, in a mode nearly allied to that followed by Count 

 Rumford, in measuring the explosive force of gunpowder. 



I began to use the following simple apparatus in June 1S03. 

 I took one of the barrels, made as above described, for the 

 purpose of compression, having a bore of 0,75 of an inch*, 

 and dressed its muzzle to a sharp edge. To this barrel was 

 ftrmly screwed a collar of iron (a a, fig. 36. See quarto 

 Plate IV. in this volume,) placed at a distance of about three 

 inches from the muzzle, having two strong bars {b b) project- 



* This was the size of barrel used in all the followiDg experi- 

 ments, where the fact is not otherwise expressed. 



ing 



