Chemical Composition of various kinds of Glass. 441 



specimens, yet it is evident that the better-insulating glasses 

 were least affected, and that, generally speaking, the glasses 

 display sensibility to the action of carbonic acid in the same 

 order in which they evince readiness to combine with water. 



The increase of weight obtained after drying at ordinary tem- 

 peratures is, in all the specimens, very nearly three times as great 

 as after drying at 160°. Knowing, then, that it is the silicated 

 alkali in the glass that is first attacked by the water, we can ex- 

 plain the phenomena observed by the following formula for the 

 decomposition of glass in presence of a large amount of moist 

 carbonic acid : — 



NaO,2Si0 3 + 2C0 2 + 3HO=(NaO,C0 2 + HO,C0 2 ) 



+ 2Si0 3 ,HO; 



of which last products CO 2 + 3 HO, or about two-thirds of the 

 weight of the matter absorbed, disappear at the higher tempe- 

 rature. 



Pelouze* has shown that water extracts from glass powder a 

 soluble alkaline hydrosilicate. I convinced myself that even in 

 humid air it is not only hygroscopic in the ordinary meaning of 

 the word, but that it binds apart of the water chemically. The 

 glass No. 5, reduced to an extremely fine powder, after having 

 stood eight days over lime-water, and then thirteen days over 

 lime and chloride of calcium, still showed 6 per cent., after drying 

 at 100° C. 2*7 per cent., and after drying at 160° 1*5 per cent, 

 increase of the original weight f. Alkaline silicate, as is 

 known, retains water at a temperature of 150°. By testing it 

 for carbonic acid I found only 0*4 per cent. Dry carbonic acid 

 had, as one might have expected, no effect at all upon the glass 

 powder ; the carbonic acid can but be absorbed in proportion as it 

 decomposes the hydrosilicates formed by the action of the aqueous 

 vapour. 



It was now a matter of interest to subject the glass specimens, 

 which I had received ticketed simply as either "good" or "un- 

 serviceable," to a more accurate test of their insulating-power, 

 and see whether in that respect they followed strictly the order 

 which their chemical qualities gave reason to expect. This in- 

 vestigation I executed in the physical cabinet belonging to the 

 Royal Academy of Sciences, to the Superintendent of which 

 (Professor Edlund) I am obliged for much valuable advice re- 

 ceived on the occasion. 



For the purpose, it was necessary to arrange so that the 

 experiments might be independent of the form and size of the 



* Comptes Rendus, vol. xliii. p. 117; vol. lx. p. 985. 

 t When the powder was placed over caustic ammonia I obtained almost 

 identically the same increase of weight as over water. 



