740 SUMMARY OF CURRENT RESEARCHES RELATING TO 



Aluminium. — After long searching a very satisfactory reagent was 

 found in csesium chloride. A platinum wire is dipped into the con- 

 centrated solution, and the test-drop stirred with it, brilliant octa- 

 hedrons of csesium alum rapidly form, varying in size from • 035 to 

 • 090 mm. The presence of iron has no effect. 



Iron and Manganese can be so easily detected by ordinary methods 

 that no special microscopic method is required. 



Sulphur requires to be converted into an alkaline sulphate ; sul- 

 phides are fused with nitre aud sodium carbonate, insoluble sulphates 

 with sodium carbonate. The coarsely powdered fusion is put in a 

 drop of water ; near it is placed a drop containing aluminium chloride, 

 hydrochloric acid, and csesium chloride; on connecting the two drops 

 with a thread of glass the formation of csesium alum shows the 

 presence of sulphur. 



Phosphorus and Arsenic. — These are brought into a soluble form 

 by fusion with sodium carbonate, or with addition of nitre if arsenides 

 may be present. A concentrated solution of ammonium chloride is 

 added to the test-drop, and close by side of this is a drop of water 

 containing a particle of magnesium sulphate (see further under 

 Magnesium). The ammonium magnesium phosphate is not to be 

 distinguished in form from the arsenate ; addition of silver nitrate or 

 of sulphuretted hydrogen affords no satisfactory distinguishing test. 

 If it is required to test for both, the substance is to be fused with 

 cyanide of potassium and carbonate of sodium in a narrow tube, the 

 arsenic sublimes as metal and the residue containing only the phos- 

 phorus is tested as above. The test with ammonium molybdate 

 solution is less satisfactory than that with magnesium sulphate. 



Chlorine cannot be detected by silver nitrate, as the precipitate 

 under the Microscope has no characteristic appearance. Mercurous 

 or lead nitrate are more suitable, but have disadvantages ; thallium 

 sulphate is the best reagent. The test is heated with an excess of 

 sulphuric acid in the platinum spoon and the hydrochloric acid 

 gas evolved collected in a small drop of water hanging to a cover- 

 glass, which is cooled by a larger drop of water on the top, and 

 lies on the platinum spoon. The top drop of water is removed, the 

 glass turned over and laid on a slide, and into the test drop is put a 

 particle of thallium sulphate. The crystals of thallium chloride 

 formed by any of these means are octahedrons with rhombic dodeca- 

 hedrons, with a very strong refractive index, size 0*010-0* 015 mm. 

 The crystals are often grouped together in threes or fours, and then 

 reach to 0*050-0*100 mm. Bromide of thallium is scarcely to be 

 distinguished from the chloride, but the crystals of the iodide are 

 distinguishable by their smallness, the largest rosettes measuring 

 0*020 mm., and by their intense yellow colour in reflected light; 

 the fluoride is more soluble, has a somewhat different form, but 

 appears very transparent and pale compared with the chloride. 



Fluorine. — The test is first fused with soda — and silica if neces- 

 sary — and then after addition of acetic acid evaporated to dryness ,* 

 the residue is moistened with sulphuric acid and gently heated, the 

 platinum spoon being covered with a concave lid of platinum foil, the 



