246 University of California Publications in Agricultural Sciences [Vol.3 



determinations. Not more than twenty-four hours elapsed between the 

 time the first sample was taken and the time all the samples were at 

 the express office ready for shipment to Berkeley, where they arrived 

 forty-eight hours later. 



At the laboratory, the samples were allowed to air-dry for six days 

 in the original canvas bags, this time for drying being necessary because 

 of the moisture present in the subsoil samples. At the expiration of 

 the six-day period all of the samples were sieved through a two-milli- 

 meter sieve and four one hundred gram portions of each sample 

 weighed into tumblers. One tumbler from each sample was reserved 

 for the determination of residual nitrate ; to a second no nitrogen com- 

 pounds were added, and to the others two-tenths of a gram of am- 

 monium sulfate and one gram of dried blood, respectively, these 

 amounts having been most frequently used in nitrification experiments 

 in this laboratory. All the tumblers were brought to an optimum 

 moisture content by the addition of twenty cubic centimeters of sterile 

 distilled water and placed in the incubator at 28° C for twenty-eight 

 days. During the incubation period, the water lost was replaced at 

 weekly intervals. At the expiration of that period, the soil was dried 

 at a temperature of 100° C and the nitrates determined colorimetri- 

 cally by the phenolclisulfonic acid method as modified by Lipman and 

 Sharp." All results are reported as milligrams of nitrate nitrogen in 

 one hundred grams of soil. 



It was not deemed necessary to make duplicate determinations on 

 all the samples, since from previous results secured in this laboratory, 

 the variation between duplicates, as regards nitrification studies, is 

 small and well within the limits of the error made in the readings, 

 which were never recorded in this study further than to one-tenth of 

 a milligram of nitrate nitrogen. Aside from this fact, it is doubtful 

 if duplicate determinations are of value in experiments of this kind in 

 which the variation between the samples was found to be so large. 



The amount of nitrate nitrogen produced was chosen as the criterion 

 of variability because of the generally accepted idea that nitrate 

 nitrogen is more directly available to plants than other nitrogen com- 

 binations and for that reason more significance may rightly be attached 

 to the amount of nitrate nitrogen found or produced in any given soil. 

 Then, too, small amounts of nitrate may be determined rapidly, with 

 a very fair degree of accuracy, and moreover, are less subject to 

 fluctuations in the duplicate determination as discussed above. The 

 absolute accuracy of the nitrate determination by the phenoldisulfonic 



