48 University of Calif ornia Publications in Agricultural Sciences [Vol. 1 



these soils and consequently considerable loss may have been 

 sustained through the volatilization of ammonia. The data on 

 ammonifieation, therefore, should be considered as representing 

 the ammonia accumulated rather than the absolute amounts 

 formed. The relative effects of calcium and magnesium carbon- 

 ates on the loss of ammonia were not investigated. 



A second factor in the loss of nitrogen is that of denitrifica- 

 tion. It was recently shown by VogeP° that calcium carbonate 

 under certain conditions can bring about a considerable loss of 

 nitrogen as nitrates in soils through denitrification. In the pre- 

 vious experiments it was observed that with the use of magnesium 

 carbonate a decided reaction for nitrites could be obtained. Deni- 

 trification, therefore, took place and an actual loss of nitrogen 

 is probably traceable to this cause. From the preceding data 

 (Table VI) it is seen that the use of small amounts of magnesium 

 carbonate not only inhibited nitrification but, as previously men- 

 tioned, also caused a considerable loss of the nitrates already in 

 the soil. We have here, therefore, still further evidence of deni- 

 trification having taken place. With the use of larger amounts 

 of magnesium carbonate, nitrification and denitrification were 

 both inhibited but no considerable loss of the nitrates originally 

 present in the soil took place. 



It seems probable, therefore, that the smaller amounts of mag- 

 nesium carbonate were toxic to the nitrifying bacteria M'hile still 

 allowing the denitrifiers to act, but under the influence of larger 

 amounts of magnesium carbonate both the nitrifying and deni- 

 trifying groups were rendered inactive. 



The striking nature of the results obtained in the previous 

 ammonification and nitrification experiments suggested a study 

 of nitrogen fixation under similar conditions. For this purpose 

 the Anaheim soil was employed since it contains a vigorous nitro- 

 gen fixing flora. Mannite was used in these experiments and 

 the usual method followed. The results obtained proved to be so 

 irregular and discordant that their publication is withheld at this 

 time. In one series a slight decrease in the amount of nitrogen 

 fixed followed the use of magnesium carbonate, while in still 

 another series no effects were observed. 



10 Centbl. Bakt. 34, pp. 540-561. 



