JQ ACTION OF HEAT MODIFIEO 



from it ; here the substance was firm, and was highly marked 

 iu the fracture with crystalline facettes. Wherever the car- 

 bonate touched the tube, the two substances exhibited, in their 

 mixture, much greater proofs of fusion than could be found in 

 the pure carbonate. At one place, a stream of this compound 

 had penetrated a rent in the inner tube, which it had filled 

 completely, constituting a real vein, like those of the mineral 

 kingdom : which is still distinctly to be seen in the specimen* 

 It had then spread itself upon the outside of the inner tube, to 

 the extent of half an inch in diameter, and had enveloped the 

 fragment of porcelain already mentioned* When pieces of 

 the compound were thrown into nitric acid, some effervesced, 

 and some not. 

 rrjcpprimene I repeated this experiment on the same day, with two graitt$ 



repeated with Qf water. The fumace being previously hot : I continued the 

 than before, fire during one half-hour with the muffile open, and another 

 with a cover upon it. I then let the barrel dewn by means of 

 the pulley. The appearance of a large longitudinal rent, made 

 me at first conceive that the experiment was lost, and the bar^ 

 rel destroyed : The barrel was visibly swelled, and in swelling 

 had burst the crust of smooth oxide with which it was sur- 

 rounded ; at the same time, no exudation of metal had happen- 

 ed, and all was sound. The metals were thrown out with 

 more suddenness and violence than in any former experiment, 

 ^)Ut the rod remained in its place, being secured by a cord. 

 The upper pyrometer gave 27^ the lower 23®. The contents 

 of the inner tube had lost 1.5 per cent. The upper end of the 

 kittle lump of chalk, was rounded and glazed by fusion ; and 

 the letter which I have been in the habit of cutting on these 

 small pieces, in order to trace the degree of action upon them, 

 was thus quite obliterated. On the lower end of the same 

 lump, the letter is still visible. Both the lump and the ram- 

 med chalk were in a good semitransparent state, shining a 

 little in the fracture, but with no good facettes, and no where^ 

 appearing to have acted on the tub^. The last circumstanc« 

 is of consequence, since it seems to shew, that this very re- 

 markf^ble action of heat, under compression, was performed 

 without the assistance of the substance of the tube, by which, 

 in many other experiments, a considerable additional fusibit 

 lity ha? been communicated to the carbonate. 



These 



