SILICATES. 289 



the substance to be analyzed ; the maximum should 

 correspond to the pure silica which requires 110 to 

 112 per cent. 



When an analysis is to be made the quantity of 

 silica is only known approximately by experiments 

 with the blowpipe. With this uncertainty it is better 

 to use too much lime than too little, but a large excess 

 must not be used, for most silicates contain bases some- 

 what votatile, as potash and soda, which if set free will 

 cause loss. 



The silicate is ground, passed through a silk sieve ; 

 it is not necessary to carry this sifting very far, at least 

 if the silicate is not very hard or with very great diffi- 

 culty decomposed, in which case it would be better to 

 pulverize it in a small steel crusher than to employ an 

 agate mortar. 



When the steel crusher or mortar is used, it is 

 necessary to digest the powder obtained in nitric acid, 

 wash with water, and ignite gently to bring the ma- 

 terial to its original purity; when this is done the 

 substance is placed in the crucible and weighed, and 

 the proper amount of carbonate of lime added. 



The mixture being weighed, it should be mixed as 

 thoroughly as possible with a little strip of platinum. 

 All the dust adhering to the platinum should be brush- 

 ed into the crucible with a small feather, then the 

 feather passed around the interior of the crucible in 

 such a manner as to bring all together at the bottom, 

 and at the same time passed between the crucible and 

 the powder, so as to detect the mixture. 



During this time the powder has absorbed a little 

 moisture ; the crucible is placed for a moment over 

 the small lamp* and heated to such a temperature that 



* Ordinary gas lamp without bellows. The lamp with turpen- 

 tine and bellows is termed the larger, and the gas lamp with only 

 bellows attached, the smaller. 



25 



