ART OP CtASBMAKlNC. (j5 



ture, anr] tramlucency on the edges, evidently annoqnced a 

 return to the state of glass, as far as it could iu the propor- 

 tions of its present composition. 



From these characters I could not avoid suspecting^, that 

 there must be a correspondent change in the specific 

 gravity. This was fully proved ou trial ; for that of the 

 mass thus brought back to the vitreous state was only 2*625; 

 while that of the same glass, when completely devitrified, 

 was always from 2'77 to 2'801. 



Mr. d'Artigues has rightly observed, that glass when Electricity dc, 

 devitrified, is not so bad a conductor of heat and electricity J)^"u;^^tbe* 

 as before. In fact we have seen, that several pieces of dif- prr«cip!esof 

 f«rent kinds of glass, vi'hen brought to this state, no longer ^i.^.i^^^tj-^^** 

 exhibited any sign of electricity by friction. If it were pos- roent, 

 atble to doubt, that this property depended more on the na- 

 ture of the component parts, than on the manner of their 

 arrangeaient, we should be obliged to return to this princi- 

 ple from the result of the experiment, in which devitrified 

 glasa, restored to its former state by being fused again with- 

 out addition, and having resumed ics original density, 

 fracture, and other characters (except the transparency, 

 which appeared only on the edges), showed no more dispo- 

 sition to become electric by friction than before. 



All the products of devitritication I have hitherto men- I^e^itrifcaUoa 

 tioned concur in showing, that it commences constantly at menceB"***^ 

 the surfaces : and thi^ is a fact of sufficient importance for surf«ca 

 us to inquire into the real cause of such accidents, as may 

 give rise to objections against this principle. 



Are there in reality instances of a devitrification eifected ^ spedmen 

 interiorly; or between two portions of unaltered glass?, A contr^ctjog 

 plate from the gl^ss-house of Premontre, given me by Mr. i'»is: 

 d'Arcet, appears, at first sight, to show the possibility of 

 this. The part completely devitrified forms a very white 

 »tratom, absolutely opake, 5 or 6 mil. [2 or 2*4 lines] 

 thick, between two strata, rather thicker, of green glass, 

 which have retained all their transparency, and exhibit the 

 yitreous fracture very decidedly contrasted with the striated 

 fracture of the devitrified part. 



But, on a careful examination of this piece, we soon per- j,ut jj, fjc^ ^ 

 ceive, that it was not cooled at rest ; and that a portion of P"^oof ofit 



th< 



