On a Quantitative Blowpipe Assay of Chromium. 389 



■ 



on account of tlie excess of water the commercial reagent con- 

 tains, and partly because of its powerful action on platinum, 

 which would soon destroy any ordinary crucible. Carbonated 

 alkalies are much more manageable; their effervescence with 

 some substances is inconvenient at times; but there are cases in 

 which they can not well be dispensed with. In the majority of 

 cases, nitre alone ansAvers perfectly, and it is decidedly the most 

 convenient. The anhydrous potash formed in its decomposition 

 does not act on the platinum perceptibly ; it acts quite as power- 

 fully as the carbonates; wherefore at present I generally confine 

 . myself to its use. 



Supposing chromic iron ore to be under treatment; 100 milli- 

 grams of ore reduced to the finest possible division, are weighed 

 off. The crucible is placed on the triangle of the lamp, and 

 the ore being poured in first, three or four spoonfuls of nitre 

 are added at once. The alcohol himp is provided with its 

 chimney, and so placed that by a simple movement of the sup- 

 port round its pivot, the orifice side of the crucible may be ap- 

 proached to or removed from the flame, which latter is so ad- 

 justed that the luminous cone is slightly truncated by the bottom 

 of the crucible. The heat thus produced is quite sufficient in 

 most cases; the rim of the crucible ought not be much heated, 

 otherwise the flux will effloresce on the outside to such an 

 extent as frequently to reach the triangle. If the latter is of 

 iron, the experiment is lost; if of platinum, it may subsequently 

 be washed off like the crucible itself But with ordinary care, 

 the accident is not likely to happen. In prolonged fusions, so 

 much of the nitre is sometimes volatilized as to make it advisa- 

 ble to add some more. Care must be taken in lifting off the 

 covering dish with the pincette, not to lose the drops of flux 

 "which usually adheres to its lower surface. If the ore has been 

 sufficiently comminuted, its decomposition is usually complete 

 in twenty-five to thirty minutes. 



If the substance under treatment be chrome oclire or ciirome 

 green it is better to take only 50 to 75 milligrams for the assay; 

 the nitre must not be added at once, but by halfspoonfuls, 

 leaving the eflPervcscence to subside each time before adding an- 

 other, and allowing the fused mass to solidify on the surface. 

 The heat must be very carefully regulated, taking as a guide the 

 noise of decomposition. Unless this is attended to, the action is 

 so violent as to render mechanical loss unavoidable. In these 

 cases the completion of the oxydation is easily recognized ; with 

 the chromic iron ore, the only test is the absence of chrome in 

 the washed residue. When the alkaline earths are present, 

 either an alkaline carbonate must be used conjointly with the 

 ^itre, to insure the complete separation of the chromic acid; or 

 else the substance may be treated like those containing lead, tin, 

 bismuth, or cadmium. 



