Chloride of Aluminium and its Analysis. 245 



D. To the solution filtered from the. chloride of silver, (vide B,) 

 and containing an excess of nitrate of silver, pure muriatic acid was 

 added in excess, to precipitate all the silver as a chloride, the same 

 precautions being taken to avoid loss by mechanical enclosure as are 

 detailed in B. The was'iings were continued until there was not 

 the slightest trace of cloudiness visible in testing with nitrate of sil- 

 ver, and the solution and washings, containing nitrate and muriate of 

 alumina, were evaporated to dryness. The evaporation was con- 

 ducted slowly, the liquid not being allowed to boil at any time, and 

 during the latter part of the process, the temperature was carefully 

 regulated, to prevent any loss from the swelling up and breaking of 

 the bubbles, which are then apt to form. The evaporation was con- 

 ducted, in a capsule of one gill capacity on the sand bath, adding 

 the solution and washings as there was room for them. 



The dry mass resulting from the evaporation was removed by an 

 ivory spatula, and placed in a small platinum crucible. A little of 

 the nitrate and muriate of alumina adhered to the capsule, and to 

 obtain the last traces, a little water was added to dissolve it, and this 

 was evaporated in a smaller space. This was detached, as much as 

 could be by the spatula, and the same operation repeated. There 

 was still a visible trace of the nitrate and muriate of alumina, but so 

 small as to be inappreciable by our most delicate balances. 



The muriate in solution, in the two last operations, was not put 

 into the crucible until evaporated to dryness, because it might create 

 a loss by bubbling up during the evaporation; and secondly, there 

 being a little nitric acid present, as well as muriatic acid, there would 

 be some muriate of platinum formed, which, during the subsequent 

 ignition, would leave platinum mixed with the alumina, and thus in- 

 crease its weight. The dry mass in the cnicible was then heated 

 over a lamp to about 500° F. Soon after the heat was applied, a 

 slight transparent coating, like the chloride of aluminium in color, 

 appeared on the under side of the cover of the crucible, but it soon 

 disappeared as the heat was increased. It is possible that the mu- 

 riate of alumina, in presence' of nitric acid, may be decomposed into 

 chloride of aluminium and a higher oxide than that in the muriate.* 

 The crucible was finally ignited. 



* To ascertain il" wlmt is above suggested be true, seme of the liyilrated chloride 

 of aliimiiiinm, dissolved in water, was evaporated lo dryness, and then heated to 

 from 400° to 800° F., biU no chloride of aluminium was evolved. Muriatic acid 

 was abundantly given off and condensed in the cool part of the lube receiver. 



