WATER SOILS, 457 



Analyses showed the nitrogen content of tlie crops in the 3 cases to be 8o.5, 109.5, and 

 106 kg. per hectare (74.4, 97.5, and 94.4 lbs. i)er acre), a close agreement with the 

 anionnts indicated by the lysimeter measnrenients. The total amount of drainage 

 A'ater obtained from different lysimeters varied from 16.54 to 20.98 liters, yielding 

 from 16.25 to 2.3.5 gm. of solid residue on evaporation. The relative proportions 

 of the different constituents in the dry residue were as follows: Nitri<' nitrogen in 

 the surface soil 45 to 46 per cent, in the subsoil 30 to 37 i)er cent; lime 25 to 32, 

 organic matter 14 to 16, sodium and i)otassium chlorids 2.6 to 4, magnesia 1.8 to 2.4, 

 sulphuric acid 1.7 to 1.9, silica 1 to 1.8, iron oxid and alumina 1.3 to 1.6. The 

 amounts of some of these constituents per liter of drainage water are given as follows: 

 Ammonia 0.2 to 0.5 mg., albuminoid nitrogen 0.8 mg., nitricr aciil very slight traces, 

 phosphoric acid 1 to 3 mg. 



A comparison of the amounts of nitrogen removed from the soil by different crops 

 and returned in the residues was continued as in previous years. The results show 

 that winter wheat required from 50 kg. of nitrogen per hectare (44.5 lbs. per acre) 

 in seasons of poor yield to 85 to 100 kg. (75.7 to 89.1 lbs. ) in seasons of good; winter 

 rye 65 to 76-83 (57.9 to 67.7-73.9 lbs.) respectively. The residues of the former 

 returned from 12 to 20 kg. of nitrogen to the soil, the latter 10 to 15. The previous 

 conclusion that an increase in yield results in a decrease in protein content, 

 especially in the grain, was confirmed. 



Tlie chemistry of the soil as related to crop production, M. Whitney and 

 F. K. Cameron {U. S. Ikpt. A<jr., Bureau uf ,S(iils Bui. 22, pp. 71). — It is stated in 

 this bulletin that the cliemical characteristics of soils have been given little recogni- 

 tion in the classification or mai^ping of soils by the l>ureau of Soils "from the fact 

 that the ordinary methods of determining the plant foods which a soil contains have 

 not given results apparently related in any definite way to the yield of crops." 



Discarding the various methods more commonly employed for the chemical exami- 

 nation of soils, for reasons which are explained, the water extract was adopted as the 

 most satisfactory measure of the available plant food in soils, and methods for the 

 rapid determination of small amounts of mineral matter in the water extracts have 

 been worked out. After trying various means of getting soil solutions, inc4uding the 

 use of a powerful centrifugal machine, "the Bureau finally adopted the convention 

 of taking 100 gm. of the soil sample, stirring it vigorously for 3 minutes with 500 cc. 

 distilled water, allowing it to stand 20 minutes for tlie coarser soil particles to sub- 

 side, and decanting the supernatant solution containing more or less suspended clay 

 and other solid matter." This solution was filtered under pressure through a Cham- 

 berland-Pasteur unglazed porcelain filter, and examined by methods which are 

 briefly outlined as follows: "The bicarbonates are determined by titrating the fil- 

 tered aqueous extract of the soil with a dilute acid solution, using methyl orange as 

 an indicator; the chlorids by titrating with a dilute S(dution of silver nitrate, using 

 carefully prepari'd potassium chromate as an indicator; the sulphates photometric- 

 ally, using a modification of a method proposed l)y Hinds and Jackson; the nitrates 

 by developing tlie color resulting from the addition of phenol-disulphonic acid, mak- 

 ing the solution slightly alkaline with ammonia, and then comparing photometrically 

 with a standard solution of potassium nitrate similarly treated; the jihosphoric acid 

 by comparing the color produced by adding molybdic acid to a solution containing 

 phosphates in the presence of nitric; acid with the color produced in a standard solu- 

 tion of sodium phosphate treated in a similar manner; the silica by the use of the 

 nitric acid and molybdate solution, after allowing the soil solution to stand at least 

 1 hour, reading against the i)hospliatc standard and subtracting the previous reading 

 for phosphate; the potassium from the color ])roduced when an excess of potassium 

 iodid is added to a solution of a platinic salt and photometrically compared with a 

 standard similarly treated, and the calcium and magnesium by an adaptation of th.e 

 well-known Clark soap method, modified by Winkler, Wartha, and others." - 



