J. L. SmitJi — Determination of the A I'l 



cessfully accomplished, and for ; 

 proved form of crucible. The c 

 cate, is of the following form and 

 slightly conical crucible with rounded bottom and cover (^eitlier 

 with or without the central wire by which to hold it) ; length 

 !'") mm. ; diameter of opening 22 ram. ; diameter of smaller 

 I'U'I just at the turn of the bottom 16 mm. ; weight about 35 

 io 40 grams. These are now made by Messrs! Johnson, Mattley 

 •^ Co., Hatton Garden, London, to whom I have furnislier] all 

 the necessaiy directions. This shape is given it in onl.T tliat 

 the portion of the crucible containing the niixtiir.' mav l>c liraicd 

 strongly, while the upper portion is below a wA heat. 



Manner of heating the crucible. — The ordinary enuiKI*'. it' 



- !. may be heated in the manner commonly "eni{)l<)yr(l ior 



:iision of silicates. If the new form of crucible, however, 



^ iiployed, then the upper partis grasped by a convenient 



'illic clamp in a slightly inclined position, and a moderate 

 --u;t from the table blow-pipe made to play upon it for about 

 25 or 30 minutes. But as gas is to be found in every well- 

 mounted laboratory, Bunsen burners of all dimensions are 

 Tised, and when properly applied, can be made to give all gra- 

 dations of heat ; a simple, cheap, and convenient furnace, with a 

 properly arranged draught, can be made to accomplish all fusions 

 of silicates without the aid of any manual labor, and I there- 

 fore employ such an apparatus. (A description of it is given at 

 the close of this article.) 



Method of analysis. — We have now the pure carbonate of lime, 

 granular sal-ammoniac and the proper crucible. The silicate 

 should be well pulverized in an agate mortar;* and half a 

 gram or one gram" of it is taken. The former amount is most 

 commonly used, it being sufficient and best manipulated in the 

 crucible; a gram, however, may be conveniently employed, 

 f he weighed mineral is placed in a large agate mortar, or better 

 in a glazed porcelain mortar, of half to one pint capacity ; an 

 equal quantity of the granular sal-ammoniac is weighed out 

 \a centigram, more or less is of no consequence), and put 

 into the mortar with the mineral, and the two are rubbed to- 

 gether intimately. After this, add eight parts of the carbon- 



oUn inch, and obtained excellent results. Notwithstanding this, thorough tntura- 



