516 M. R. Bansen's Bloii-pipe Experiments. 



elapse until it has quite evaporated, counted bj^ means of a pen- 

 dulum or metronome. This point is so generally marked by a 

 sudden alteration of the flame as to be determinable to half a 

 second. 



Taking the volatility of carbonate of soda as unity, we may 

 compare with it the volatility of equal bulks of other substances. 

 If Iq be the time required for the evaporation of carbonate of 

 soda, and ^j, t,^,t^.,,. that required for the evaporation of other 



substances, we obtain the values -j-, —, -^ as approximate 



M ^2 '3 

 measures for their volatility. 



The following determinations of volatility will illustrate the 

 use to which they may be applied in geological chemistry, more 

 especially if the balance be used, as is necessary in very accurate 

 determinations : — 



Sulphate of soda . . . 0'77 

 Sulphate of lithia . . . 0-89 

 Sulphate of potass ... 1 '21 

 Carbonate of soda . . . I'OO 

 Cai'bonatc of lithia . . . 1*70 

 Carbonate of potass . . 2'30 

 Chloride of sodium . . 6*57 

 Chloride of lithium . . . 8-36 

 Chloride of potassium . . 15 "33 



Boracic acid 08 i 



Borate of soda .... 1*02 

 Phosphoric acid . . . .23-00 

 Bibasic phosphate of soda . 0"12 

 All these substances are completely volatile at 2300° C. 

 Other substances lose a constituent, and leave a non-volatile 

 residue — such as the sulphates of baryta, strontia, and lime, 

 which lose SO^, and become alkaline. The corresponding chlo- 

 I'ides are converted in an analogous manner into basic com- 

 pounds. The silicates rich in alkalies lose considerable quan- 

 tities of potash and soda at 2300° C, at a temperature therefore 

 much less than that of many lava-streams. 



Besides these determinations of volatility, the flame can be 

 used for many valuable blowpi})e experiments, among which is 

 the quantitative determination of soda in the presence of lithia 

 and potassium. 



For the qualitative determination of soda in its volatile salts, 

 it is enough to hold a small bead in the melting-space, and to 

 illuminate with it a crystal of bichromate of potass. When the 

 rays of the soda-flame are incident on this salt, it appears per- 

 fectly colourless, transparent, and with an adamantine lustre. 

 But the following reaction is more delicate, aiul better suited 



] 



