EXPERIMENT STATION BULLETINS. 



463 



bacteria in the subsoil are i-ctarded by some inhibiting factor that they 

 finally overcome. The Siime soil.s in Series III behave quite differently. 

 The surface soil yields a little less ammonia, the subsoil considerably 

 more than in the first series. The soils of both series are from the 

 same sample which had been kept air-dry during the five weeks which 

 lay between the first and third series. The drying and exposure to 

 air had improved the subsoil to such a degree that after the addition of 

 fertilizer it gave slightly larger ammonia yields than the fertilized sur- 

 face soil. Even if the two different soils were not a mutual chock in 

 themselves, the comparison of the sand cultures and of the peptone solu- 

 tion shows that the differences between Series I and III cannot be laid 

 to the inoculum. 



The soils A and B of Series II and IV do not show as much differ- 

 ence as the Surface and subsoils; in fact, the difference is practically neg- 

 ligible. In Series IV, after the soils had been kept dry for about five 

 weeks, Soil B showed no change, while Soil A had slightly improved. 



In comparing the action of B. mycoides upon peptone in sand and in 

 soils, it would not be permissible to compare the cultures of one certain 

 arbitrary moisture content. It seems fair, however, to compare them 

 at their optimum of ammonia production. This is 25% for the first four 

 soils, 20% for sand and 75% for peat. The maximum values for each 

 soil after 20 days are: 



With surface soil 15.8 cc. t^ NaOH 



" subsoil 7.6 



" soilA. 17.3 



• soils 14.4 



With peat 13.5 cc. 



" sand 11.6 



" solution 6.9 



10 



NH, 



According to these data. Soil A is the best culture medium for B. my- 

 coides, and the subsoil is the poorest. It seems remarkable that the 

 amount of ammonia in the best soil is only twice as large as that in the 

 poorest soil. Solution and subsoil are almost alike in their ammonia 

 formation. 



Fertilizer Effects. — The object of Series III, IV and V was to ascer- 

 tain whether and to what extent the better growth in certain soils could 

 be accounted for by the presence of soluble mineral salts. The "fertilizer" 

 used consisted of 0.2 g. of KHjPO^ and 0.1 g. GaClo per 50 g. of dry soil 

 or 50 cc. of solution respectively. In Series V the monabasic phosphate 

 was substituted by the dibasic phosphate. 



The influence of the fertilizer is favorable in every instance except 

 with quartz sands. This is an apparently peculiar fact which, however, 

 can be explained. In Series III the fertilized sand culture showed in 

 the average 2.0 cc. less ammonia than the unfertilized check, while the 

 treated peptone solution was 2.1 cc. higher than the unfertilized solu- 

 tion. In Spries IV, the minerals decrease the ammonia production in 

 sand 2.7 cc, while the same minerals cause an increase of 0.9 cc. in 

 the corresponding solution. In Series V, the effect of the minerals is 

 not the same upon sand, causing a rise of 0.2 cc. at 15% moisture and 

 of 1.5 cc. at 25%. It is true that in Series V, the monobasic phos- 

 phate was substituted by the dibasic which causes a change of acidity. 

 But the acidity was also in the peptone solution, and yet, there was 

 an increase in the peptone solution due to the same salt that caused 

 a decrease in the sand culture. The difference between the sand and 

 the solution lies in the amount of moisture in which the salts are 



