324 



PROCEEDINGS OF SECTION B. 



allowed to fall lightly through a speciai coarse sieve (20-hole) without 

 rubbing. The coarse powder is then agated, the value of this during 

 the analysis being well worth the trouble. 



Silica (SiOj). — One half gramme is always used, as, unless specially 

 large platinum crucibles are available, the silica is dangerous to handle, 

 and the precipitate unwieldy. 



After fusion the crucible should not be suddenly chilled in cold 

 water whilst hot, as this soon induces cracking in the platinum, especially 

 if it is of any thickness. 



Various methods may be adopted for removing the melt. A useful 

 way is to allow it to cool round a platinum rod and then detach by 

 short, quick heating ; this will remove the bulk. The main object is 

 to get the fusion out of the crucible as quickly as possible and with the 

 least chance of mechanical loss or of injury to the crucible. The 

 particular means adopted may be left to individual choice. The melt 

 is now placed in a 5in. porcelain basin, well covered, and dissolved in 

 HCl, after being moistened with water. The portion remaining in the 

 crucible is carefully dissolved out and added to the bulk. After 

 evaporation to dryness on the water bath the mass is twice moistened 

 with HCl (few drops only), taken up with hot water and filtered. 

 Heating the dried mass at 120° C. for half an hour is unnecessary. 

 Its effect is generally to increase the impurity in the SiOj, as indicated 

 in the residue after treatment with hydrofluoric acid. The dried mass 

 is treated with 10 ccs. HCl and 100 ccs. hot water, and warmed till 

 dissolved. It is then filtered and washed free from chlorides and the 

 silica weighed till constant after prolonged blasting. 



Hillebrand insists on a second evaporation of the filtrate, as he says 

 that the silica contained therein is not certain of recovery in the ultimate 

 treatment ofxthe alumina, iron precipitate. A number of experiments 

 were made in endeavoring to satisfy this point, and comparative 

 analyses were performed, using each method. 



Table I. 

 In each case A = siii;rle evaporation. 

 " B = double evaporation. 



