METHODS OF MAKING DETERMINATIONS. 9 



FIELD HISTORY OF THE PLATS. 



The prairie sod, covered originally with western wheat-grass 

 (Agropyron occidentale) , buffalo grass (Bulbilis dactyloides), and 

 gramma {Bouteloua oligostachya) , with minor quantities of sedges 

 (Carex flifolia and C. stenopTiylla) and other grasses, together with a 

 few vetches, was broken in the spring of 1907 to a depth of abfmt 3 to 

 4 niches, backset (replowed) in the fall of 1907 to a depth of about 5 

 inches, and left open until the spring of 1908, when it was disked and 

 harrowed until a good seed bed was secured. Such was the treat- 

 ment of the spring- wheat plat and the corn plat used in the work. A 

 plat reserved for summer-fallow used in this work was spring-plowed 

 to a depth of about 8 inches, smoothed witli a disk, and harrowed. 

 This plat was cultivated during the summer to keep weeds down and 

 to conserve the moisture. 



The determinations on the corn plat were not begun as early as on 

 the wheat and fallow plats. As this plat was not planted as early as 

 the wheat plat, it was taken for granted that the soil conditions would 

 be the same as those of the fallow plat until the time of planting the 

 corn. 



METHODS OF MAKING DETERMINATIONS. 



The chemical methods used were those described in Bulletin No. 31 

 of the Bureau of Soils. The soil samples were collected in duplicate 

 from 6-inch layers to a depth of 2 feet — that is, in to 6, 6 to 12, 12 to 

 18, and 18 to 24 inch depths — and the two cores from each soil layer 

 made into a composite sample. The samples as bored up with the 

 auger were transferred to air-tight tin cans, taken immediately to the 

 laboratory, weighed, dried in an oven at a temperature of about 100° 

 C, and again weighed. In preparing the solution it was found 

 necessary to adopt this drying method in order to insure thorough 

 mixing of the soil and the water, as it was practically impossible to 

 break down the moist gumbo pieces directly. The ratio of soil to 

 water was 1 to 5 (100 grams of soil to 500 c. c. of water). After 

 drying, the soil particles readily broke down and after thorough 

 stirring in the water would settle out quite clear in about 20 to 30 

 minutes. The supernatant liquid was then poured into Chamberland 

 filters and filtered. Such a procedure of drying the soil could, of 

 course, not be employed were determinations to be made of such soil 

 elements as are derived directly from the soil minerals, as potassium, 

 phosphoric acid, etc., but owing to the origin and nature of the 

 nitrates and nitrites in the soil the solubility is not increased ; neither 

 is it likely that decomposition of a disturbing nature would take place. 

 Experiments were made to compare the two methods of obtaining the 

 24493— Bui. 173—10 — -2 



