EUSHA MITCHELL SCIENTIFIC SOCIETY. 21 



in the cracking of the crucibles during the cooling after fu- 

 sion. The crucibles used measured 10.5 c. m. in diameter 

 by 8 c. m. in height and held a charge of 100 grams zircon, 

 400 grams sodium hydroxide and 20 grams sodium fluoride. 

 This is one-half the amount of sodium fluoride recom- 

 mended by Linnemann, but proved sufficient. The sodium 

 fluoride should be dried beforehand. The sodium hydroxide 

 is first thoroughly melted and the fluoride then added. The 

 mass should be brought to a fairly high temperature and 

 then the zircon powder added. A rapid evolution of gas 

 follows the introduction of the powder. The mass should 

 be well stirred by means of a nickel stirrer — a narrow^ strip 

 of sheet nickel fastened to a glass rod answers the purpose 

 and keeps the hands beyond the reach of hot alkali occasion- 

 ally thrown out. If the bubbles threaten to rise over the 

 edge temporary removal of the lamp secures their subsidence. 

 The crucible should not be allowed to cool too far, how- 

 ever. Much seems to depend upon carrying through the 

 reaction rapidly at a high temperature. I have at times 

 doubled and even tripled the length of fusion at a lower 

 temperature without securing the thorough breaking up of 

 the zircon secured at a higher temperature. After the first 

 violent boiling a quieter period follows. The end of the 

 reaction is showm by a thickening of the mass and the 

 rising of large bubbles here and there, also sometimes by a 

 fine spitting or spray. In several instances where weights 

 were kept the undissolved or unattacked portion of the 

 zircon powder amounted to less than five per cent. 



The melted mass was poured out upon pieces of sheet 

 nickel for cooling. After solidifying enough to handle 

 wnth tongs it was broken off" and plunged in a beaker of 

 cold water. Water was also put in the crucible after it 

 had cooled, to dissolve off" the portions adhering to the 

 sides. 



The water separates the sodium silicate from sodium 



