328 Journal of Agricultural Research voi. xii. no. 6 



was undertaken in which larger aliquots than those recommended in 

 Bulletin 31 were used. The results obtained by the use of larger portions 

 of solution yielded, in general, more reliable figures. If larger quantities 

 of solution were to be used, a great deal of the argument in favor of the 

 colorimetric and turbidimetric procedures disappeared. The small por- 

 tions of solution required for these methods have always been considered 

 one of their great merits. In the soil extracts which contained the 

 largest quantities of plant nutrients, if larger aliquots were taken, it 

 was very clear that the ordinary volumetric and gravimetric procedures 

 could be used. It then became important to learn whether the standard 

 methods of determination could not be applied to the weaker solutions 

 if a carefully standardized technic were adopted. In the work under- 

 taken along this line the dilute and average solutions previously men- 

 tioned were used. As a result, it is believed that a more satisfactory 

 procedure than the colorimetric has been developed for phosphates, 

 calcium, and potash. The methods used are quite well known, but a 

 great deal of work was done to establish the exact conditions for accurate 

 results with the dilute solutions employed. 



The description of the detailed procedures used follows. 



Phosphate. — Evaporate two portions of soil extract of 200 c. c. each in a 200-c. c. 

 porcelain casserole. This size of casserole stands ignition excellently. Add a few 

 drops of hydrochloric acid diluted i to i before the above solution reaches dryness 

 to aid in decomposing soluble silicates. Ignite the dry residues in the casserole 

 over a Meeker burner at a moderate temperature till a grayish white residue is 

 obtained. Cool. Takeup the residue with loc.c. of nitric acid diluted i tog. Cover 

 with a watch glass and digest on the steam bath for approximately 10 minutes to 

 insure complete solution. Filter into a 200-c. c. Erlenmeyer flask. Wash the 

 casserole with two more portions of 10 c. c. each of hot nitric acid diluted i to 9. 

 Wash the dish, and filter with small portions of hot water. Keep the total volume 

 of solution to about 50 c. c. Cool. Add a few drops of methyl orange indicator 

 and neutralize rapidly with concentrated ammonia. Bring just to acid reaction 

 with concentrated nitric acid. Add 2 c. c. of saturated ammonium nitrate. Place 

 the flasks and a quantity of properly acidified ammonium-molybdate solution ^ 

 in a water bath at 50° C. When all the solutions have reached this temperature, 

 add 5 c. c. of molybdate to each determination. Keep at 50° C. for K tour. Remove 

 from the water bath, and filter at once on prepared asbestos felts. Wash with cold 

 distilled water till free from acid by the usual tests. ^ Transfer filter felt to the same 

 flask, using approximately 25 c. c. of distilled water free from carbon dioxid. Add 

 15 c. c. of sodium hydroxid of which i c. c. is equivalent to o.i mgm. of phosphorus- 

 pentoxid and observe carefully whether the solution of the yellow precipitate is 

 complete. Titrate the excess alkali with hydrochloric acid of the same strength, 

 using phenolphthalein as indicator. Calculate results to either elementary phos- 

 phorus or the phosphation, as desired. 



Calcium. — Evaporate 200 c. c. of water extract to dryness in a 200-c. c. casserole. 

 Ignite at a moderate temperature over a Meeker burner till a grayish white ash is 

 obtained. This step is desirable to remove traces of organic matter, even though the 



'Wiley, H. W., ed. ofpiciai, and provisional methods of analysis, association of official 



AGRICULTURAL CHEMISTS, AS COMPILED BY THE COMMITTEE ON REVISION OF METHODS. U. S. Dept. 



Agr. Bur. Chem. Bui. 107 (rev.), p. 4. 1908. Reprinted in 1912. 



