452 Dr. Clarke's Observatio?is on Gehlenite. [Dec. 



the mass, which, little by little, was poured into muriatic acid. 

 As soon as all effervescence had ceased, more distilled water 

 was given to it ; and the whole being placed in an evaporating 

 dish, the excess of acid was volatilized, and water again added ; 

 the whole being then thrown upon a filter, the substance which 

 remained upon the filter was washed (until the washings effected 

 no change of colour in litmus paper), and afterwards calcined, 

 It proved to be pure silica, and weighed 29*50 gr. making no 

 allowance for loss. 



B. The silica having been separated, as mentioned in A, an 

 excess of the hydrate of ammonia was poured into the solution, 

 whereby a copious precipitate was caused ; this, being collected 

 on a filter, was carefully washed until the washings did not alter 

 the colour of litmus paper. It consisted of alumina and oxide 

 of iron. 



C. The alumina and oxide of iron, precipitated from their 

 muriates by the process mentioned in B, were taken off the filter 

 in a gelatinous state and heated, according to the method 

 recommended by Thenard,* in a capsule, with a great excess of 

 caustic liquid potass, which dissolved the alumina, but had no 

 action upon the oxide of iron. The capsule being then removed 

 from the fire, and the liquid cooled down to 100° of Fahr. and 

 filtered (the filter being washed until it ceased to exhibit signs of 

 alkalescence), the oxide of iron remained on the filter, and was 

 removed with an ivory knife, and dried, and calcined. It after- 

 wards weighed 12*20 gr. making no allowance for loss. 



D. The washings from B being added to the liquid whence 

 the alumina and oxide of iron were precipitated, and which con- 

 tained all the lime in the state of a muriate, subcarbonate of 

 potass was added to it to effect the decomposition of this 

 muriate. A copious precipitate of subcarbonate of lime, contain- 

 ing a trace of magnesia, ensued ; and this being washed, dried, 

 and exposed to a white heat in a platinum crucible during half 

 an hour, afterwards weighed 27*80 gr. making no allowance for 

 loss. 



E. The next operation related to the obtaining the alumina, 

 which had been dissolved in the alkaline solution mentioned in 

 C. This liquid was first saturated with muriatic acid ; then an 

 excess of the hydrate of ammonia being added, the alumina was 

 all precipitated. After being carefully washed, dried, and cal- 

 cined, it weighed 14*50 gr. making no allowance for loss. 



In all these experiments, no allowance whatever has been 

 made for any loss which, in a slight degree, must be sustained 

 where filters are used ; and filters here were unavoidable, owing 

 to the partial gelatinization of the precipitates, especiallyin that of 

 alumina. I prefer, however, stating the results exactly as I 

 obtained them by weight, leaving the deficiency to go to the 



* Traitc de CWmie, torn. ir. p. 182. Paris, 1816. 



