THE METHODS OF MICROCHEMICAL QUALITATIVE ANALYSIS 305 



rens). That is to say, that while we may obtain proof of the 

 presence of an exceedingly minute amount of chlorine through 

 the separation of crystals of thallous chloroplatinate, approxi- 

 mately one hundred times as much chlorine must be present in 

 order that it may be revealed as potassium chloroplatinate. 



This plan of producing a less soluble salt is, in general, to be 

 preferred to that of causing the separation of a solid phase by 

 forcing back the dissociation, by means of strong acids, salting- 

 out, or other similar processes, since well-formed crystals result 

 in the first case, but abnormal, atypical salts are apt to appear 

 in the other cases. 



EXPERIMENTS. 



Repeat Experiment I lie, page 302, gradually reducing the concentrations until 

 no triple salt separates, then add a fragment of CsCl; the triple nitrite of Cs, Cu 

 and Pb will appear. In a new preparation carry the dilution a little farther, so 

 that the Cs salt does not appear at once. Add a fragment of TINO3. The deli- 

 cacy of the reaction will be approximately tripled. 



VI. The reaction can be hastened and the delicacy of the test 

 increased by exposure to alcohol vapors. 



It was stated under Method V, that it is rarely desirable 

 to employ a reagent that will force back the dissociation; the 

 reasons being that the addition of such a reagent causes a too 

 rapid separation of a solid phase and there is a tendency towards 

 the production of malformed, skeleton or exceedingly tiny crys- 

 tals. When, however, the separation of a solid phase is acceler- 

 ated by the gradual absorption of a vapor in the test drop, thus 

 reducing the solubility by forcing back the dissociation very 

 slowly, it requires only a little care to assure the separation of 

 characteristic, well-formed crystals. 



Alcohol is exceptionally well fitted for use in all cases where a 

 crystalline compound is less soluble in alcohol than in water. 



One of two methods will be found convenient. Place near the 

 test drop a small piece of filter paper. Saturate the paper with 

 a drop or two of alcohol, carefully avoiding the addition of more 

 than the paper will absorb. Cover the drop and paper with a 

 watch glass (Behrens) ; or place a piece of paper at the bottom 



