146 Effects of Heat modified ly Compression. 



little lube, and the cylinder of chalk, I had put some boras' 

 and sand, with a link- pure borax in the middle, and chalk 

 over it. The metal had not penetrated beyond the borax 

 and sand, by a good fortune peculiar to this experiment; 

 the intrusion of metal in this mode of execution being' ex- 

 tremely troublesome. The button of chalk was found in a 

 state of clean white carbonate, and pretty hard, but without 

 transparency. The little tube was perfectly clean. Its 

 weight, with its contents, seemed to have suffered no 

 change from what it had been when first introduced. At- 

 tending, however, to the balance with scrupulous nicety, a 

 small preponderance did appear oh the side of the weight. 

 This was done away bv the addition- of the hundredth of 

 a grain to the scale in which the carbonate lay, and an ad- 

 dition of another hundredth produced in it a decided pre- 

 ponderance. Perhaps, had the tube, before its introduction, 

 been examined with the same care, as great a difference 

 might have been detected ; and it seems as if there had been 

 no loss, as least not more than one hundredth of a gram, 

 which, ou 10-05 grains, amounts to 0-0912, say 0-1 per 

 cent. The carbonate was loose in the little tube, and fell 

 out by shaking. It had a yellow colour, and compact ap- 

 pearance, with a stony hardness under the knife, and a stony 

 fracture ; but with very slight facettes, and' little or no trans- 

 parency. In some parts of the specimen, a whitish colour 

 seemed to indicate partial calcination. On examining the 

 fracture, I perceived, with the magnifier, a small globule of 

 metal, not visible to the naked eye, quite insulated and 

 single. Possibly the substance may have contained others 

 of the same sort, which may have compensated for a small 

 loss ; but there could not be many such, from the general 

 clean appearance of the whole. In the fracture, I saw here 

 and there small round holes, seeming, though imperfectly, 

 to indicate a beginning of ebullition. 



I made a number c ■!' experiments in the same manner, 

 that is to sav, with the muzzle of the barrel upwards, in 

 some of which I obtained very satisfactory results ; but it 

 was only by chance that the substance escaped the conta- 

 mination of the fusible metal ; which induced me to think 



of 



