98 ESSAY V. 



blowing must be immediately discontinued, in such a manner 

 that the brown flame shall not touch the glass. When it is 

 afterwards taken out with the forceps, it appears perfectly 

 colourless. 



(4) This destruction of the colour does not seem to 

 happen suddenly and at once, but by degrees ; for when I 

 now and then discontinued the blowing, before the true mark 

 had appeared. I found the glass generally lighter than it was 

 before, and this more or less according as I had blown 

 longer or shorter time. 



(5) After putting this colourless glass again upon the 

 charcoal, and melting it by the brown flame of the candle, it 

 again assumed its former colour, though I kept it melted for 

 a long time in the brown flame. 



This change of colour I have effected several times with 

 the same glass, always with the same success ; but I cannot 

 say whether it may be repeated often. Such an experiment 

 would be too difficult to make. There is some room to 

 doubt of this ; at least if the phenomenon in the following 

 experiment was not owing to some accident. 



(6) Having already twice discharged the colour, I forced 

 the blue flame with violence against the glass, in order to 

 make the colour of the manganese the sooner disappear. I 

 observed some little vesicles to rise in the glass, which after- 

 wards burst, and at the same time dispersed a number of 

 very small glass globules around. I had now continued the 

 blowing for a whole hour, and had been blowing very 

 violently, I therefore was obliged, after a short blast, to 

 desist for two or three minutes. 



I found the glass somewhat lighter than it was before ; 

 but on continuing the blowing with equal force, I found it 

 impossible to expel the colour from the glass. After I had 

 given over blowing, the mass appeared to be somewhat 



