262 Mr. Griffiths on the 



weeks together,) the clear portion afforded a crystalline salt on 

 evaporation, having the characters of muriate of potash. 



It may be remarked that this solution, when perfectly clear, con- 

 tained no lead, on testing for it by sulphuretted hydrogen ; but 

 upon agitating or diffusing the fine powder of glass through water, 

 holding the gas in solution, it was immediately discoloured, or 

 blackened. 



Flint-glass, although chosen for the above experiments, is not 

 the only variety possessing this remarkable property, crown and 

 plate-glass, white enamel, and what is more remarkable, Newcas- 

 tle green-bottle glass, and tube of the same material, (in the com- 

 position of which there is, comparatively, little alkali,) also 

 Reaumur's porcelain, made from the green bottle glass, possess 

 the power of acting upon vegetable colours as alkalies. 



These experiments, tending to prove that glass is a body of 

 irregular composition, parting readily with its alkali by the action 

 of water, it became a matter of some interest to determine how 

 far certain natural combinations of potash with siliceous matter 

 were equally active to the same tests, especially as in green-bottle 

 glass, which contains little alkali, it is thus rendered evident. 

 No analogous effect could, however, be produced by powders of 

 felspar, basalt, green- stone, granite, obsidian, pumice, and some 

 others, even when boiled with water, a method which never failed 

 to produce it rapidly with glass, although cold water is perfectly 

 sufficient. 



Some interesting conclusions may be drawn from the above 

 experiments, which may tend to explain several well-known phe- 

 nomena, and may, perhaps, become matters of practical utility.] 



In the first place, with regard to the glasses employed in the 

 laboratory, or for domestic uses, it must be evident that water 

 has the power of acting upon and dissolving the alkali at the 

 surface, and leaving an insoluble portion spread as a coating over 

 the interior of the vessel, defending it from further immediate 

 action. 



Where, however, time can be allowed, the effect does not 



