22 SEASONAL NITRIFICATION AS INFLUENCED BY CROPS, ETC. 



except the 12-inch layer during that week. The same point is 

 brought out during the period from June 5 to June 12, there being 

 no rainfall at all to wash down the nitrates from the surface into the 

 12, 18, and 24 inch layers. And, finally, no such explanation can 

 be given for the sudden drop in amounts of nitrates during the 

 period from June 12 to June 19, when there was practicalh T no rain. 



DENITRIFICATION . 



At the beginning of the work denitrification was . expected to 

 become an obvious factor in affecting the seasonal changes of nitrates 

 in the soil, and ammonification was also expected to become of 

 measurable importance; accordingly, determinations of nitrites and 

 ammonia were made. Neither of these activities, however, assumed 

 any importance at all during any part of the season so far as the 

 analyses brought out, and, accordingly, toward the end of the season 

 ammonia and nitrite determinations were not regularly made — 

 sufficiently often, however (usually every other week), to know the 

 amounts present of these compounds. 



No conveniences were at hand for making ammonia determina- 

 tions directly on the fresh soil, and this is obviously the best way of 

 determining these compounds. The determination of the water- 

 soluble ammonium compounds could not be expected to give reliable 

 data. That ammonia may have been formed in larger amounts than 

 would be suggested by the quantities found during the season would 

 appear from the work of Stevens and Withers, who state that in 

 some soils "nitrification proceeds as rapidly as ammonification, con- 

 verting the ammonia to nitrate as fast as it is rendered available by 

 the ammonifying organisms." 



The amounts of nitrites found varied during the whole season from 

 nothing to" 1.8 parts per million in the dry soil. There was no regu- 

 larity in even these small amounts, one week traces only or none at 

 all being found and the next week traces to 1.8 parts per million. 

 The same irregularities held true in the individual soil layers. The 

 usual amounts of nitrites were from 0.1 to 0.3 part per million, and 

 one single sample — from the 6-inch layer in the wheat plat, on May 

 22 — had 6 parts per million, the only nitrite determination made 

 that exceeded 1.8 parts per million at any time during the season. 



The determinations likewise showed the presence of but small 

 quantities of ammonia per million parts of dry soil, but in this case 

 there was a slight decrease toward the end of the season. During the 

 earlier part of the season, the amounts found averaged 6 to 8 parts 



a Stevens, F. L., Withers, W. A., et al. Centralblatt fur Bakteriologie, Parasiten- 

 kunde und Infektionskrankheiten, pt. 2, 1909, p. 23; abstracted in Experiment Sta- 

 tion Record, vol. 21, no. 2, 1909, pp. 118-119. 

 173 



