36 



SCIENCE 



[N. S. Vol. XXXIX. No. 992 



shall work. An additional supply of nitrogen 

 is usually furnished, either in the form of 

 ammonium-sulphate, or some highly nitro- 

 genous, easily decomposed, organic substance, 

 such as cottonseed meal or tankage. Where 

 soil has been employed as the medium, one of 

 three methods of procedure has usually been 

 followed. One method is to add to a definite 

 weight of soil a given amount of nitrogen, 

 incubate and then determine the quantity of 

 nitrate nitrogen present. The amount found, 

 taking no account of the nitrate nitrogen 

 originally present, is regarded as indicating 

 the comparative ability of different soil sam- 

 ples to form nitrate nitrogen from other forms 

 of nitrogen ; , in other words, its nitrifying 

 power — ability — efficiency. A second method 

 is to make a determination of the amount of 

 nitrate nitrogen present in a corresponding 

 sample at the beginning of the experiment; 

 otherwise proceed as above. The difference 

 between the amount originally present and that 

 found at the final analysis is regarded as the 

 correct factor. By a third method duplicate 

 samples are taken and treated exactly alike 

 with the exception that only one receives an 

 additional supply of nitrogen. After incuba- 

 tion both are analyzed and the difference in 

 the nitrate nitrogen content is regarded as 

 the correct factor for comparison. 



Recently, in determining what Stevens and 

 Withers have termed the nitrifying inocula- 

 ting power of a series of plots that are under 

 study, it became necessary to use a soil rela- 

 tively high in nitrate nitrogen. Cottonseed 

 meal was employed as the source of nitrogen. 

 In a large number of cases, securing the com- 

 parative factor by either of the last two men- 

 tioned methods, it was found to be a minus 

 quantity. That is, with the conditions 

 optimum (so far as is known) for nitrifica- 

 tion, the amount of nitrate nitrogen found 

 after incubating was much less than orig- 

 inally present in the soil. If ammonium sul- 

 phate were substituted for cottonseed meal, as 

 the source of nitrogen, such a condition was 

 never noted. This led to a series of investiga- 

 tions to determine the cause of the minus 

 factors that were secured. There were evi- 



dently losses of large quantities of nitrate 

 nitrogen. Was this due to actual losses of 

 total nitrogen or was the nitrate nitrogen 

 merely transformed? To what extent does it 

 occur? What are the conditions influencing 

 the same? These are some of the questions 

 an effort is being made to answer. Certain 

 grave difficulties have thus far prevented a 

 complete solution of the problems involved. 

 However, it is believed that in the near fu- 

 ture sufficient data, together with that upon 

 which the statements herein presented are 

 made, will be secured to completely settle the 

 question. 



One very interesting fact has been brought 

 to light. This is that at least two of the 

 methods now employed in bacteriological lab- 

 oratories for determining the relative nitri- 

 fying power of soils may give us absolutely 

 no indication of true values. The writer has 

 abundant evidence to show that this is true 

 in the presence of appreciable quantities of 

 nitrate nitrogen when easily decomposed or- 

 ganic substances are employed as the source 

 of nitrogen. Oft-repeated observations show 

 that where from five to twenty-five mg. of 

 NO3 per 100 gr. soil is present and cotton- 

 seed meal added, at the end of seven days ab- 

 solutely no trace of NO3 can be found. If ex- 

 amined three weeks later an abundance will be 

 present. According to the last two methods 

 mentioned above, the five to twenty-five mg. 

 NOs would be subtracted from that found at 

 the final analysis and the difference taken as 

 the factor indicating the amount formed dur- 

 ing incubation. In reality, the amount pres- 

 ent reached zero at one time, hence all found 

 at the final analysis must have been formed 

 during incubation, and represents more 

 nearly the correct factor. 



In a series of experiments designed for the 

 purpose of determining how much NOj would 

 disappear it was found that in five days in the 

 presence of .95 gr. cottonseed meal, as high 

 as thirty-five mg. NO^ per 100 gr. soil had 

 been lost. With 1.9 gr. cottonseed meal the 

 quantity ran up to fifty-five mg. With 

 higher quantities of cottonseed meal still 

 larger amounts of ITO3 were lost. With 



