EXPERIMENTS on WHINSTONE and LAVA. 51 



refults already mentioned. " I had placed in the muffle a long 

 and {lender fragment of this glafs, with its extremities re fling 

 on two fupports of clay, and its middle unfupported. Having 

 then increafed the temperature by flow degrees, I expected to dis- 

 cover the loweft point of emollefcence, by obferving when the 

 fragment funk by its own weight. The muffle having attained 

 a moderate heat, I obferved the glafs to lofe its fhape a little. 

 Wifhing to fee it completely melted, the fame heat was con- 

 tinued, but no further change took place. The heat was then 

 raifed feveral degrees, but without effect. At lafl, being urged 

 flill further, the glafs funk down completely between its fup- 

 ports. The pyrometer being then withdrawn, denoted a tem- 

 perature above 30. 



It occurred to me, that, on this occafion, the glafs, by the 

 firfl application of heat, had foftened, and then had cryftallized, 

 fo as to become hard again ; that, in cryftallizing, it had ac- 

 quired fuch infufibility as to yield to no heat under 30. I im- 

 mediately confirmed this conjecture by the following experi- 

 ment. 



A piece of the fame glafs, placed in a cup of clay, was in- 

 troduced into the muffle, heated to 21. In one minute it be- 

 came quite foft, fo as to yield readily to the preffure of an iron 

 rod. After a fecond minute had elapfed, the fragment, being 

 touched by the rod, was found to be quite hard, though the 

 temperature had remained flationary. The fubftance, thus 

 hardened, had undergone a change throughout ; it had loft 

 the vitreous character ; when broken, it exhibited a fracture 

 like that of porcelain, with little luftre ; and its colour was 

 changed from black to dark brown. Being expofed to heat, it 

 was found to be fufible only at 3 1 ; that is, it was lefs fufible 

 than the glafs by 13 or 14 degrees. 



Numerous and varied experiments have fince proved, in the 

 cleared manner, that, in any temperature, from 2 1 to 28 inclu- 



G 2 five 



