RECENT WORK L\ AGRICULTURAE SCIENCE. 



CHEMISTRY. 



On the colorimetric determination of small quantities of phosphoric acid 



and silica, F. P. Veitch (Jour. A)ner. Chem. Soc, 25 {1903), No. ^, jrp. 169-184).-— 

 This article reports a study of the conditions affecting the accuracy of the Lepiere « 

 colorimetric method for phosphoric acid as worked out in detail by Woodman and 

 Cay van (E. S. R., 13, p. 319), viz, the influence of turbidity, organic matter, 

 ammonium and iron salts, dissolved silica, etc., and means of overcoming the errors 

 thus introduced. The method proposed by the author for the determination of 

 phosphoric acid and silica in the same soil solution is as follows: 



"The water or extract is tested for iron l)y adding potassium ferrocyanid to the 

 acidified solution. The absence of interfering amounts of iron having been shown, 

 a measured volume of the water or soil extract is freed from suspended matter by 

 filtration or bj- passing through a Chamberland filter (reject the first 100 cc. that 

 ])asses) or by evaporating to dryness and filtration, or in some cases where the water 

 is but slightly turbid the turbidity or color is corrected for by determining its 

 amount in terms of the standard, the reading thus obtained being afterwards sub- 

 tracted from the final readings. Add to the clear extract 5 cc. of nitric acid (sp. gr. 

 1.07) and 4 cc. of molybdate solution. Place in the camera, allow 10 to 30 minutes 

 for development of color, and compare with a standard j^hosphate solution, which 

 may conveniently contain 10 parts per million of phosphorus pentoxid and be con- 

 taine<l in a sliding tube connected b}^ rubber tubing with a side neck tube graduated 

 in cubic centimeters within the camera. (The color of the standard is not affected 

 l)y the rubber tube during one working day, but the standard should be made fresh 

 each day. ) The readings thus obtained (several should be made and the average 

 taken) minus the reading for turbidity, when calculated to a volume of 100 cc, 

 equals P^Os+SiOj in parts per million of solution. 



"Another measured portion of the water or extract is evaporated to dryness twice 

 with a filtration between the evaporations in a porcelain or })latinum dish with 3 cc. 

 nitric acid (s]). gr. 1.07) plus a little magnesium nitrate," heated 2 hours in a 

 water oven, 5 cc. nitric acid (sp. gr. 1.07) added, filtered, washed to about 45 cc, 

 placed in a camera, and compared. If colored, the reading is noted and is finally sub- 

 tracted from the total reading. Add 4 cc. of ammonium molybdate and thoroughly 

 mix. Place in the camera and compare after 2 to 5 minutes. The corrected reading 

 calculated to volume of 100 cc. is P^Os in parts per million of solution. This reading 

 subtracted from the Si.02+P205 reading and the difference multiplied ]jy 0.55 gives the 

 silica. 



""Where the original solution is too nmch colored with organic matter to be accu- 

 rately corrected for by reading the color thus produced against the standard phos- 

 ])hate solution, it is necessary to evaporate with about 0.1 gm. magnesium nitrate & 

 and burn off the organic matter, take up with water +3 nitric acid, evaporate to dry- 



«Bul. Soc. Chim. Paris, 15 (1896), p. 1213. 



'' " In solutions containing sufficient base to form normal phosphates with all the 

 phosphoric acid, the addition of magnesium nitrate appears to be unnecessary," 



24385— No. 0—03 2 833 



