BY COMPKESSION. 597 



Qirunk, fiill retaining a cylindrical form, though bent by 



partial adhefions. Its furlace bore fcarcely any remains ot 



the impreffion taken by tlie powder, on ran;n)ing it into the 



tube: it had, to the naked eye, the rpughnefs and femitranf- 



parency of the pith of a rufli ftripped of its outer Ikin. By 



the lens, this fame furface was (een to be glazed all over, 



though irregularly, (liewiiig here and there fome air-holes. 



In fradture, it was femitranfparent, more vitreous than cryf- 



talline, though having a few^ facetted : the niafs, was feeraingly 



formed of a congeries of parN, in theaifelves quite tranfparent : 



and, at the thin edges, imal! pieces were vifible of perfeft 



tranfparency. Thefe nnift iiave been produced in the fire; 



for the fparhad been ground with water; and palTed through 



fieves, the fame with the fineft of thofe ufed at Etruria, as 



defcribed by Mr. Wedgwood, in his paper on the conflruClion 



of his pyrometer. 



With the fame barrel I obtained many interefling refults, I" ^^efe experi- 



/, ,. r , n • r • . msaii the efcape 



giving as Itrong proofs of lulion as in any lormer experiments ; ^^ carbonic acii 



with this remarkable difFt-rence, that, in thefe lalf, the fub- appears to have- 

 Hance was compadl, with little or no trace of frothing. In 

 tlie gun-barrels where fufion had taken place, there had al- 

 ways been a If-fs of 4 or 5 per cent. conne6ted, probably, with 

 the frothing. In thefe experiments, for a reafon foon to be 

 fiated, the circuraftance of weight could not be oblerved ; but 

 appearances led me to fuppofe, that here the lols had been 

 Imall, if any. 



On the 6th of April, I made another experiment with the '^'radual failure 

 fc]uare barrel, vvhofe tiiicknefs was now much reduced by fuc- bored from the 

 celTive fcales, produced by oxidation, and in which a fmallfolid. 

 rent began to appear externally, wliich did not, however, pe- 

 netrate to the bore. The heat rofe high, a pyrometer on the 

 breech of the barrel giving 37'^. On removing the metals, 

 the cradle was found to be fixed, and was broken in the at- 

 tempts made to withdraw it. The rent was much widened 

 externally : but it was evident, that the barrel had not been 

 laid open, for part of the carbonate was in a (tate of faline 

 marble ; another was hard and white, without any (aline grains, 

 and fcarcely elfervefced in acid. It was probably quicklime, 

 formed by internal calcination, but in a flate that has not oc- 

 curred in any other experiment, 



Tlie 



