134 EXPEIIIMEXT STATION. [Jan. 



times as many in the sterilized as compared with the unsterilized 

 decoctions. However, in the experiments where sterilized and 

 nnsterilized snbsoil were nsed we fonnd more bacteria in the 

 nnsterilizcd decoctions as compared with the sterilized decoc- 

 tions. This fact proves that sterilization does not in every case 

 result in an increased number of l^acteria in the soil thus treated. 



CONCLUSIOXS. 



1. The development of bacteria may be retard(>d or accelerated 

 in soil decoctions by the use of sterilization. 



2. In decoctions of soil rich in organic matter the develop- 

 ment of bacteria is greatly increased, while in soils deficient in 

 organic matter the development of these organisms is retarded 

 by sterilization. 



3. The stimulating or retarding effects on the development of 

 bacteria of the two types of sterilized soil used by us are similar 

 to those produced upon the growth of crops in these soils. {Cf. 

 Table IV.) 



4. From numerous microscopic examinations made of Am- 

 herst soils we do not find that protozoa are abundant ; neither 

 were they observable in our soil decoctions. 



5. The question of protozoa as a biological factor was elimi- 

 nated in the experiments. The stimulating or retarding effect 

 on the development of bacteria was due to other causes. 



6. Our experiments therefore, made with Amherst soils, do 

 not confirm those of Russell and Hutchinson, who maintain that 

 protozoa influence the number of bacteria in soils, since the devel- 

 opment of bacteria differs in soil decoctions according to the com- 

 position of the soil used ; that is, the number of bacteria which 

 develop in a soil depends upon the chemical and physical con- 

 dition of the soil rather tlian u]>ou the number of protozoa. 



7. These experiments do not necessarily preclude the idea that 

 protozoa might play a much more important role in soils other 

 than those with which we experimented. 



