- 106 On the Decomposition and Composition 



currech When the flame of the lamp was immediately 

 Brought to play upon the glass, the combustion was very 

 vivid, so as sometimes to break the tube ; and the alkali ge- 

 nerated partly rose in white fumes, which were deposited 

 upon the slass. 



When the temperature was slowly raised, the bases of the 

 alkalis acted upon the metallic tray and formed alloys, and 

 in this state it was very difficult to combine them with their 

 full proportion of oxygen ; and glass alone could not lae 

 employed on account of its decomposition by the alkaline 

 bases ; and porcelain is so bad a conductor of heat, that it 

 was not possible to raise it to the point required for the pro- 

 cess, without softening the glass. 



In all cases the globules of the alkaline bases were care- 

 fully freed from naphtha before they were introduced ; of 

 course a slight crust of alkali was formed before the com- 

 bustion, but this could not materially affect the result 5 and 

 when such a precaution was not used, an explosion gene- 

 rally took place from the vaporization and decomposition of 

 the film of naphtha surrounding the globule. 



After the combustion, the absorption of gas was ascer- 

 tained, by opening the lower point of the tube under water 

 or mercury. In some cases the purity of the residual air 

 was ascertained, in others the alkali formed in the tray was 

 weighed. 



From several experiments on the synthesis of potash by 

 combustion, I shall select two, which were made with every 

 possible attention to accuracy, and under favourable circum* 

 stances, for a mean result. 



In the first experiment 0*12 grains of the basis were em- 

 ployed. The combustion was made upon platina, and was 

 rapid and complete ; and the basis appeared to be perfectly 

 saturated, as no disengagement of hydrogen took place when 

 the platina tray was thrown into water. The oxygen gas 

 absorbed equalled in volume ]QO grain measures of quick- 

 silver; barometer being at 2Q-G inches, thermometer 62° 

 Fahrenheit; and this reduced to a temperature of 60° Fahr- 

 enheit, and under a pressure equal lo that indicated by 30 



inches. 



