Sept. ii, 1916 Bacteriological Studies of a Soil 961 



while the average bacterial count for No. 29 was 7,840,000 and for 

 No. 30 was 13,632,000. This accords with Temple's idea (20) that 

 the increase in number is largely due to the fermentable material added 

 and not to the bacteria actually carried in the manure. 



FORMATION OF AMMONIA 



The results presented in Table IV seem to the writers clearly to estab- 

 lish one of two facts, either that the systems of cropping under study 

 exert no appreciable influence on the ammonia-forming power of this 

 soil type or that the methods used for determining such differences are 

 valueless. Since many investigators have been able to detect marked 

 differences with essentially the same methods, the former conclusion 

 might seem most likely. We would call special attention, however, 

 to the results obtained by Given and Willis (6, 7) and Perotti (15), 

 together with their conclusions regarding the existence of differences 

 and the value of methods in vogue for determining this phenomenon. 

 It is true that during 191 3 the plots receiving manure gave slightly 

 higher results than those receiving no such treatment, but it is equally 

 true that the reverse is evident for 1914. Furthermore, there seems 

 to be no correlation between the number of bacteria and the amount 

 of ammonia formed. 



It may be argued, in view of the nitrification data given later, that in 

 the plots receiving manure the ammonia formed was transformed into 

 nitrates. But, as previously mentioned, the accumulation of nitrates 

 in seven days' time in the presence of cottonseed meal is practically nil. 

 In fact, the senior writer (5) has demonstrated that under such condi- 

 tions there is a rapid disappearance of nitrate nitrogen during the first 

 few days. To ascertain this condition, determinations of nitrate nitrogen 

 were made during 191 3 which showed, as may be noted, that the above 

 contention was verified. The greatest quantity of nitrate nitrogen 

 recovered after one week's incubation was 3.5 mgm. from plot 10. It is 

 true that there were large differences in the amount of ammonia in the 

 four-week analyses; but if we add the nitrate and ammonia nitrogen, the 

 differences are slight. From the ammonia figures it is also evident that 

 in no case, even after four weeks' incubation, in the absence of calcium 

 carbonate was nitrification limited by the formation of ammonia. This 

 possibly did not hold true in a few cases when calcium carbonate was 

 added. 



FORMATION OF NITRATE 



In our study of nitrate formation we have obtained the most marked 

 and consistent results of any of our studies. This is graphically shown in 

 figure 1 for the four-week incubated samples without calcium carbonate. 

 Tables IV and V give the tabulated results of experiments conducted on 

 nitrate formation in natural soil from different plots. 



