22 HOT SPRINGS, ARKANSAS. 



boric acid and the quicklime were stirred together for about fifteen 

 minutes with a platinum rod to be sure that all boric acid was fixed. 



The volatile contents of the crucible were now evaporated off at a 

 low temperature. It was found necessary to grease the edges of the 

 crucible with vaseline to keep the solution from crawling over. After 

 the contents of the crucible had been evaporated to dryness the cruci- 

 ble was fully dried in the air bath and finally blasted. The increase 

 in weight of the crucible gives the weight of boric anhydride (B 2 3 ) 

 present. 



Iron, aluminum, and manganese. — The residue spoken of previ- 

 ously that was reserved for the determination of iron, aluminum, and 

 manganese was treated with hydrochloric acid and evaporated to dry- 

 ness. It was thoroughly dried at about 120° C, again taken up with 

 water and hydrochloric acid and filtered. The filtrate was evaporated 

 to dryness and dried at 120° C. It was then taken up with hydrochloric 

 acid and water and filtered again. This filtrate was heated to the 

 boiling temperature, and ammonia added, a drop at a time, until it could 

 be very faintly smelled coming off from the solution. The solution 

 was then filtered and the precipitate well washed with hot water, burned 

 and weighed as Fe 2 O s and A1 2 3 in the ordinary manner. 



The ammoniacal filtrate from above was treated with a few drops of 

 bromine, more ammonia was then added, and the whole boiled after 

 stirring up. The vessel was removed from the source of heat, cooled 

 a little, and a little more bromine and ammonia added. This process 

 repeated once or twice precipiated all the manganese as the oxide. 

 The solution was made slightly acid with acetic acid, filtered and 

 washed at once with hot water. The filter and contents were burned 

 and weighed as Mn 3 4 . This is the method by which the iron, 

 aluminum, and manganese were determined in springs 2-i to 46, inclu- 

 sive. In the first 23 springs these three elements were determined in 

 the same portion that was used for the estimation of calcium and 

 magnesium. 



Silica. — In this determination a large quantity of water was evapor- 

 ated to dryness in platinum with the occasional addition of small 

 amounts of hydrochloric acid. After all the water had been evaporated 

 to dryness, the dish and contents were completely dried at 120° C. 

 The residue was taken up with hydrochloric acid and water, heated 

 and filtered, washing the residue thoroughly with hot water. This 

 process took out most of the silica. The filtrate was then evaporated 

 to dryness, dried thoroughly at 120° C, again taken up in hydro- 

 chloric acid solution by heat, and filtered. The filtrate was made to a 

 definite volume, aliquot portions of which were used for the determina- 

 tions of calcium, magnesium, sulphuric acid, potassium, sodium, lith- 

 ium, and phosphoric acid. The two residues were transferred to a 

 crucible, burned and blasted in the ordinary wa}^. and finally weighed 

 as silica. 



Calcium and magnesium. — An aliquot portion of the above filtrate 

 was first treated with ammonia and filtered, then treated with ammonia 

 and bromine water and filtered, and finally treated with ammonium 

 oxalate in the usual manner. This was allowed to stand over night, 

 the liquid filtered off, and the precipitate dissolved in hydrochloric 

 acid and reprecipitated with ammonia and a little extra ammonium 

 oxalate. This was allowed to stand over night and filtered and washed 

 on the same paper previously used. The precipate was dried, trans- 



