Sept. ii, 1916 Bacteriological Studies of a Soil 957 



EXPERIMENTAL METHODS EMPLOYED 



Practically all the methods here mentioned have been severely criti- 

 cised during the past few years, particularly by Lohnis and Green (11), 

 Allen and Bonazzi (1), and Noyes (14). However, in our laboratory 

 the methods described below have proved, for the object in view, equal 

 or superior to any suggested prior to the beginning of this work. 



Samples for the various analyses were taken with a i^-inch soil auger. 

 Ten to fifteen samples were collected from each plot to the depth of the 

 soil, which was about 10 inches. The cores of soil were taken uniformly 

 all over the plot, avoiding close proximity to the surrounding alleys. 

 They were placed immediately in sterile Mason jars in order to prevent 

 loss of moisture and to avoid contamination, so far as possible. The 

 samples were then brought to the laboratory as soon as possible, where, 

 under aseptic conditions, the soil was passed through a 2-mm. sieve and 

 thoroughly mixed. Samples were immediately taken for quantitative 

 analyses and for moisture determinations. 



For moisture determinations 50 gm. of soil were dried at 1 10° C. for 

 two hours. To determine the water-holding capacity, 50 gm. were 

 placed in a carbon filter containing a perforated porcelain bottom and a 

 measured quantity of water poured on top. The water was permitted 

 to percolate through the soil. The process was repeated two or three 

 times. From the amount of water absorbec plus the quantity lost in 

 drying, the water-holding capacity, expressed in grams of water held per 

 100 gm. of dry soil, was determined. Data obtained by these methods 

 are, of course, not absolute. But the process possesses two essentials: 

 Quickness of manipulation and comparativeness. Since slight dif- 

 ferences in water content, when near the optimum, exercise but little 

 influence upon bacterial activity, it is believed the error introduced is 

 not appreciable. 



Quantitative analyses were made by carefully weighing 1 or 2 gm. of 

 soil, placing it in 98 or 99 c. c. of sterile water and shaking vigorously for 

 one minute. From this suspension dilutions were made in the ordinary 

 way. Finally, 1 c. c. was placed in each of three sterile Petri dishes and 

 thoroughly mixed with 10 c. c. of Temple's agar (20). The dishes were 

 then incubated for one week at room temperature and all colonies counted 

 with the aid of a hand lens. If one of the dishes varied widely from the 

 other two it was discarded. The same was true if dishes were overrun 

 by spreaders or molds. The results are reported in millions of bacteria 

 per gram of dry soil. 



The ammonia- and nitrate-forming experiments were carried out by 

 thoroughly mixing into fresh soil (the equivalent of 100 gm. of dry soil) 

 sterile cottonseed meal containing 60 mgm. of nitrogen. This was 

 placed in a sterile 500 c. c. wide-mouthed bottle and the moisture con- 



