A. A. Blair— Chromium and Aluminium in Steel and Iron. 423 



dryness with a little alcohol ; when thoroughly dry, the Cr 2 O s 

 is 'dissolved in HC1. diluted and filtered from' silica, the Cr 2 O s 

 precipitated by NH 4 HO, filtered with the usual precautions, 

 ignited and weighed as Cr 2 3 ,* calculated to Cr by the factor 

 0-6853. Atomic weight of Cr 52-2. 



The only impurity this precipitate can contain is a little 

 A1 2 3 , partly from the Al in the steel, and partly as an impurity 

 in the Na a CO, and KN0 8 . The best method for separating 

 this impurity, is to add to the solution of alkaline chromate 

 obtained alx.'ve taller fusion and filtration from Fe 2 3 ) an excess 

 of KC10 3 . then a slight excess of HC1, and evaporate to syrupy 

 icy on the water bath, adding a little KC10 3 from time 

 to time, so that there may always be an excess to decompose 

 any HC1. Redissolve in water, add a slight excess of (NH 4 ), 

 C0 3 , and boil off all smell of the latter, filter, wash with hot 

 water, add to solution an excess of HC1 and, after the greater 

 part of the KC10 3 is decomposed, a little alcohol, and precipi- 

 tate the Cr 2 3 as before, f In Genth's method (loc. cit.) of evap- 

 orating the ■• chromate nearly to dryness on 

 the water bath, with an excess of (NH 4 )N0 3 , a minute amount 

 of Cr0 3 is reduced to Cr 2 3 , and precipitated ; it is consequently 

 filtered off with the A1 2 3 , Si0 2 , etc., and lost, unless this resi- 

 due should be treated separately to regain this Cr 2 3 , which very 

 much complicates the process. If in the filtrate so obtained the 

 Cr0 3 is precipitated bv HgN0 3 as Hg 2 0, Hg 2 Cr0 4 , some of the 



chromium salt nearly al\va\> pu-sns int lun.«n upon filtering, 



and can be detected, sometimes in very considerable amounts, 

 in the filtrate. The same thing occurs if this method of precipi- 

 tation is - ::dme chro- 

 mate (obtained by fusion with Na 2 C0 3 and KNO s , an 

 from the Fe 2 3 ) by HNOg. In all cases I have obtained the 

 most satisfactory results by reducing the CrO s to Cr 2 3 , precipi- 

 tating by NH 4 HO, and weighing as Cr 2 3 . 



This method for estimating chromium in steel may appear a 

 lite a number can be made at once ana 1 inrj 

 about three to four days is required for a complete estimation. 

 The accuracy is all that can be desired and the delicacy of the 

 method in detecting and otunatiu^ ynry >uu I am, : 

 than that of any method I have used. I think I can safely say 

 that it is delicate to O005 per cent of Cr ; The amount of 

 Fe 2 3 usually precipitated by the BaC0 3 , with the Cr 2 3 is about 

 Fofte^hinalwork the method can be very much shortened 

 by fusing the precipitate by BaC0 3; with Na^CG, and K^0 3 or 



* For precautions necessary to free Cr,0 3 from alkaline salts, see Genth m 

 Chem. News, vi, 30. Fres. Cbem. AnaL Quant 387. ,. •_ Chem inal 



t Dexter's method, see Rose, Chim. Anal. Quant, p. 520. *Yesemu., on . . 



Quant, p. 372. Pogg. Anal., lxxxix, 142. 



