Aug. 1909.] Soil Bacteriology. 235 



A COMPARISON OF THE GRAVIMETRIC AND VOLUMETRIC METHODS. 



Tables I to IV give the comparative results obtained by 

 gravimetric and volumetric methods. The samples were taken 

 by means of the sampler described above. One cubic centi- 

 meter of soil was taken each time, placed in a sterile test-tube 

 and weighed on an analytic balance. In every case duplicate 

 samples were obtained and analyzed. In the following tables 

 the weight of each sample is given in the first column, the 

 number of bacteria per cubic centimeter is given in the second, 

 and the number of bacteria per gram is given in the third 

 column. 



A glance at tables I to IV will show a variation in weights 

 of samples varying in table I from 1.19 to 1.4 grams per cubic 

 centimeter, from 0.965 to 1.14 grams in table II, from 1.021 to 

 1.232 grams in table III, and from 1.01 grams to 1.035 grams 

 in table IV. The comparison also shows remarkable differ- 

 ences in weight between many samples and their duplicates, 

 which were taken side by side. For instance, in table IV the 

 fourth inch gives 1.01 grams while the duplicate gives 1.034. 

 In the sandy soil (table II) the difference becomes more 

 marked, showing in the twelfth inch a difference of 0.113 of 

 a gram. 



In tables I to IV the number of bacteria per cubic centimeter 

 and per gram shows, in duplicate samples of soil, variations 

 which are somewhat proportionate to the variations in weight 

 of samples. These variations in quantitative findings repre- 

 sent the inaccuracy which is due to the volumetric method of 

 taking soil samples. It may be assumed that such variations 

 or inaccuracies will always occur from soil samples which are 

 measured in terms of cubic centimeters. However, in general 

 comparative soil bacteriological work these variations in num- 

 ber of bacteria, due to the volumetric method of procedure, 

 may not be greater than those variations which might occur 

 as a result of differences in specific gravities of different sam- 

 ples of soil, should the soil samples be taken by the gravi- 

 metric method. For instance, two types of soil, Marshall silt 

 loam and sandy loam, were used in this work. The volume 

 weight of the silt loam soil is 1.11 and the volume weight of 

 the sandy loam is 1.35. Their weights stand in the ratio of 

 1:1.21; that is, an acre-foot of the sandy loam weighs 1.21 

 times as much as an acre-foot of the silt loam. Therefore, 



