68 AST OT ©LASSMAKINO. 



casioned by any thing but the loss of this quantity of 

 matter*. 



To these strong reasons for rejecting the hypothesis either 



«f a simple modification of structure, or of a precipitation 



of a portion of the component parts, let us add the increa&e 



of hardness, and diminution of ^ulk. 



farther argu- AmoDiJ; the products of devitrification, which I have laid 



merit from the , ^ 111 1 t 1 11. 



hardness betore the class, there are several, as I have remarkedj that 



cannot be scratched by rock crystal ; there are some, on 

 which corundum scarcely leaves a mark visible by a lens ;, 

 and Mr. Ciffle's No- 5 scratches rock crystal, as an aqua 

 marina would. 



and density. j-^e (jei-jsity, that glass acquires in this process, is still 



more striking ; though, like the hardness, it is only the ef- 

 fect of a more powerful aggregation. AH the pieces, the 

 specific gravity of which before and after the process I had 

 an opportunity of comparing, gave a difference of l6 or 18 



?'!"'""'!''" <^f thousandths in addition. Mr". d'Arcet had two cubes of 



Bulk IB devi- 



trified bottle bottle glass, of the manufactory ot la Garre, cut, for the 



glass. purpose of ascertaining their bulk, by the scale of Wedg- 



wood's pyrometer, before and after devitrification. One 

 advanced 17% the other 11*; which gave for the first the. 

 proportion of 1000 to 909 ; for the second, 1000 to 952. 

 The glass having been cut from the same piece, and conse- 

 quently being of the same quality, the difference betwe«» 

 the results can be ascribed only to a greater or less portion 

 of the cement adhering to the surface of these cubes ; 

 \»hich, however, was far from compensating for the diminu- 

 tion of the original bulk. This is farther proved by the co- 

 lour, which the pieces dewitrified in Reaumur's cement 

 assume at their surface ; a colour, that often penetrates 

 them to some depth, and can arise only from particles of 

 metallic oxides contained in the sulphate of lime em- 

 ployed. 



* Mr, d'Arcet has soTr.etimes fonud the weight of the cubes of glass, 

 which he. subjected to th« process of devitrification, increased 5 thou- 

 sandths pf a gramme on 2 grammes [3l grs]; but he operated in a 

 c«m<;nt. The same must have been the case with the pieces No. a and 

 4 of Mr. Ciffl6« which, as has been seen, came out with a coating. 

 But this proves nothing against the two experiments related above, 

 made without cement, Vinci ill platjna crucibles. 



£ thmk 



