GL A

G L A

( 505 )

tougher and Iefs liable to crack from changes ofheat and cold. No ufe however has been madeof it in manufacture, but as a curious chemicalor phyfical phenomenon it deferves furthernotice.

Dr. Lewis made the following very valuableexperiments on this fubftance.' A number ofpieces of common quart bottles were put intocrucibles with white fand poured over them,and put into a proper furnace where they wereheated for many hours, and pieces withdrawnfrom time to time to examine the progrefs ofthe change. The pieces that were taken out,after many hours heating, but below rednefs, didnot appear to have fuffered any change what-ever. In a low red heat the change went onthough very (lowly, but in a ftrong red heatapproaching to whitenefs, juft not fufficient tomelt the glafs, the change went on pretty fall,and in two hours the glafs had affumed the ap-pearance of porcelain, the change beginning ateach furface and fpreading gradually to themiddle.

The glafs firft became blueifh on the furface,and, when held to the light, yellowifh, andwith a very fenfible diminution of its tranfpa-rency. After this it gradually became whiteand opake, and the texture was no longer vitre-ous, but fibrous, and’the fibres difpofed nearlyparallel to each other. By degrees the glafsbecame throughout opake and fibrous, and thecolour of a dun white ; the fibres were arrangedregularly from the fides to the middle, wherethe fibres from the two fides meeting, formed akind of partition, in which occafionally werepretty large cavities. A longer continuance offire induced a further change of texture, thefibres became divided or cut into grains at theouter ends and gradually through their wholelength, and the whole fubftance changed intexture from fibrous to granular, like commonporcelain. By a (till further continuance offire the grains, at firlt fine and gloffy, grewlarger and duller, from being very compactbecame porous, and at laft a friable fubftancelike a flightly cohering mafs of white fand, noteafy to be diftinguifhed from the fand in whichit was imbedded.

Glafs thus changed whilfl it remains in thefibrous ftate, is confiderably tougher and harder,fo as to give abundant fparks with (teel, whichcommon green bottle glafs will hardly do, tocut all common glafs with eafe, and fcarcely tobe fcratched by the file. It will bear alfo to beplunged fuddenly from freezing to boiling* Commerce of th.e Arts.

water, without cracking, and at the fame timeits texture is fo denfe that no acrid liquorswhatever either corrode it or tranfude throughit. The circumltance which the molt preventsits ufe in manufacture is, that though the innertexture is fine and white, the outer is coarfeand dirty-looking. When the heat is fo longcontinued that the texture changes from fibrousto granular, it again becomes foft, no longergives fire with Heel, and lofes its cohefion.

Another important circumftance to be ob-ferved is, that when this fibrous porcellanousglafs is expofed to a very ftrong heat it meltsinto a femi-tranfparent mafs, drawing out infirings, which on breaking are now no longerfibrous, but have returned to the vitreous Jlate,and at the fame time again becomes no harderthan the glafs from which it was originallymade. However it does not melt fo eafily asthe glafs itfelf, and the longer it is cementedthe more difficult of fulion it becomes, fo thatthe granular porcelain requires a much higherheat for melting than the fibrous.

Dr. Lewis repeated the experiments, im-bedding the glafs in a variety of fubltances in-ftead of fand, fuch as bone-affi, charcoal, chalk,&c. but in all the change of texture was thefame, the outer colour alone being affedled.He then heated the glafs by itfelf, being ftuckup with a little luting in the middle of the'crucible, and therefore touching nothing withthe part above the luting. The fame changehowever took place, though more flowly, andwith fome inconvenience from the falling in offome of the pieces when foftened by the heat.The fand therefore has little other effect thanto fupport the glafs during the procefs, andprevent it from lofing its fliape. Another im-portant fadt to the explanation of the caufe isthat the glafs does not fenfibly lofe or gainweight in the whole converfion from the vitre-ous to the fibrous (late.

Mr. Gregory Watt, in his mod valuablepaper on Balalt, u very happily brings this por-cellanous change of glafs as an illuftration ofhis important pofition, namely, that bodies whofefibres have a natural tendency to a cryftallinearrangement, or a polarity, when vitrified by afufficient heat and cooled haftily in the vitreous(late, are able fubfequently to return to theirnatural cryftalline arrangement of fibre whenexpofed to a heat merely fufficiently to foftenthe texture, though not enough for fufion. This,in the inftance of bafalt, he (hews by the fin-gular cryftallizations formed in the cells of

* Phil, Tranf. f»v 1804.

VOL, I.

3 *