214 DETERMINATION OF ALKALIES IN MINERALS. 



for an hour or two, when the excess of the chloride of platinum 

 will attract moisture from the air and afford an opportunity for 

 the chloride of platinum and soda to crystallize regularly. For 

 the most perfect success in very minute quantities of soda too 

 great an excess of chloride of platinum should be avoided. 

 Those engaged in mineral analysis, who will employ this means 

 of detecting the presence of the alkali, will find it of great 

 assistance in facilitating their labors, especially when directed 

 to very minute accuracy in their results, for I have reason to 

 believe it rare to find in minerals any one of the alkalies per- 

 fectly free from one of the others-. 



42. When the chloride of lithium is present it interferes 

 materially with this method of detecting small amounts of 

 soda; for, owing to its very deliquescent nature, it abstracts 

 moisture from the air, and dissolves the double chloride of so- 

 dium and platinum, or prevents altogether its formation into 

 recognizable crystals. These investigations have not added 

 any thing to what is already known concerning the detection 

 of lithia mixed with soda and potash; the plan invariably 

 adopted is to treat the mixed chlorides with a solution of 

 alcohol and ether, and examine the part dissolved by the blow- 

 pipe. Details as to the manner of using the alcohol and ether 

 solution are given under the next head. 



Separation of the Alkalies from each other. 



43. Under this head I have nothing to add to what is already 

 known on the subject. It may be well, however, to mention 

 the manner in which Bammelsberg's method of separating 

 lithia has been employed, as it has not yet been fairly tested 

 in this country. His method, it is well known, is based on the 

 solubility of the chloride of lithium in a mixture of equal parts 

 of absolute alcohol and ether, neither of the other chlorides 

 being dissolved by this menstruum. A number of experiments 

 were made on known quantities of the alkalies, and the results 

 of some of them are as follows: 



(a) Five hundred milligrammes Of chloride of potassium 

 treated with the mixture of ether and alcohol, ten grammes 

 of the latter being used, yielded only three tenths of a milli- 

 gramme to the liquid. 



