22 § 31' KEAGENTS. 



uranium. {Ufl^fijifi^.) — Heat uranic nitrate until a 

 small part of the uranic oxide is reduced, digest the 

 yellowish-red residue with acetic acid, filter the liquid 

 and set the filtrate aside to crystallize ; the crystals are 

 composed of uranic acetate, while uranic nitrate remains 

 in solution. 



The solution of the acetate should not be changed by 

 sulphuretted hydrogen after acidification with hydrochlo- 

 ric acid, and should give a precipitate with ammonic car- 

 bonate that is entirely soluble in an excess of the reagent. 



31. Urea. — Recrystallize it from its solution in alcohol. 



32. Water, distilled.— H^O.— This can be prepared by 

 the analyst himself, if necessary. Dealers in apparatus 

 can supply small stills of copper and worms of block-tin, 

 put together and ready for use. The water must be 

 colorless and tasteless, and it should leave no residue 

 when evaporated in a platinum dish. Ammonic sulphide 

 should give no color to it, nor should basic plumbic ace- 

 tate make it turbid, nor should ammonic oxalate or 

 argentic nitrate make it turbid after long standing. 



33. Zinc. — Zn. — This should give no reaction for arsenic 

 with Marsh's test, and, when dissolved in nitric acid with 

 the aid of heat, it should give no red color with potassic 

 sulphocyanate. Fresenius recommends that before using 

 zinc for reducing ferric to ferrous oxide in the estimation 

 of iron by the permanganate process, it should be tested 

 by the same process. Dissolve a piece of the zinc in di- 

 lute sulphuric acid in the small, long-necked flask, as de- 

 scribed in § 52, b, and, after the flask is filled with water 

 and its contents are cold, add a drop of a very dilute so- 

 lution of potassic permanganate, and at the same time 

 add another drop to the same volume of pure water, and 

 stir both mixtures well. The depth of color communica- 

 ted to both liquids should be precisely the same. 



