METHODS OF MAKING DETERMINATIONS. 



11 



This was established at the time and has been also found in the 

 bacteriological laboratory of this Bureau to be a safe procedure for 

 purposes of comparison. 



A comparison of results obtained by the various methods of extrac- 

 tion is given in Table I. 



Table I. — Nitrites, nitrates, and ammonia found in dry soil treated in various ways. 



Soil treatment. 



Fresh untreated soil, using undistilled water 



Fresh untreated soil, using distilled water 



Fresh untreated soil plus chloroform, using undistilled water 

 Fresh untreated soil plus chloroform, using distilled water. . . 

 Soil dried in oven at 100° C 



Parts per million in dry soil. 



Nitrites 

 (NO a ). 



0. 5.5 

 .55 

 .55 

 .55 

 .55 



Ammonia 

 (NH 3 ). 



1.5 

 .8 

 2.5 

 2.0 

 3.33 



Nitrates 

 (N0 3 ). 



26.7 

 22.2 

 22.2 

 25.0 

 36.4 



The oven-dried sample in this case gave a higher proportion of 

 ammonia and nitrates than the other samples. It is possible, though, 

 that nitrification went on to some extent before the sample was prop- 

 erly dried, as the gasoline drying oven was troublesome, this being 

 its first use. The soil sample was in the oven most of the day before 

 the trouble with the burner was located, and it was warmed most 

 of the day without quickly drying. 



The following set was run through the next day, a soil being used 

 that had been carefully prepared for a winter-wheat seed bed the 

 previous year but which had not been planted: 



Table II.- 



-Nitrites, nitrates, and ammonia found in dry soil from a seed bed treated in 



various ways. 



The ammonia was not read, as the solution contained foreign color- 

 ing matter. These and the previous results give an idea of the limits 

 of experimental error. The standard solutions used for comparison 

 in the colorimeter were made up fresh on the day they were used, 

 and all readings were made within a few hours. The nitrite reagents 

 were kept separate and small portions mixed as needed for use. 



The systematic determinations were begun April 27, and from then 

 until July 31, about ten days after harvest, the determinations were 

 made once a week. 



173 



