up download

of it. It should have shrunk to a bougie easily detach- able from the crucible. To avoid partial fusion, it is well to rest the crucible in a little cupel containing sand. The contents of the crucible are then extracted twice with boiling water in porcelain dishes, glass being avoided all through the analysis. To the filtrate ammonium carbonate and a few drops of oxalate are added, and the lime precipitate filtered. The filtrate is then evapo- rated to dryness, ammonium salts are gently ignited off, the residue is extracted with water and a few drops of oxalate, filtered into a weighted platinum basin, evaporated to dryness and gently ignited again, and the residue from this process is taken up with weak hydro- chloric acid, evaporated to dryness, heated in the air-bath, and finally weighed as RC1. The equivalent is found by titration with silver nitrate or by precipitation of potassium platinochloride, and hence the amounts of soda and potash V-are calculated. A blank should be conducted, as this may yield 5-10mgms. Lawrence Smith's method has the advantage over the hydrofluoric acid method that borates are in the water- insoluble residue as calcium borate. This may not be of CERAMIC CHEMISTRY. 19 much consequence in clays ; but in glazes borates may amount to as much as 20 per cent, of the whole, and the accurate separation of alkalies and borates is therefore important.