20 



THE ROYAL SOCIETY OF CANADA 

 Experiments III and IV. 



In these experiments we have the comparison between a cropped 

 and a fallowed soil. In both experiments the bacterial content was 

 low and it is hard to find any definite relations between frost, moisture 

 and bacterial numbers, although no data obtained in these experiments 

 contradict the main conclusions arrived at the year before and men- 

 tioned in the Introduction. 



Experiment V. 



In this experiment we have a soil containing so much water that 

 it needed a very heavy frost to congeal it completely. The frost 

 penetration was accordingly much less than in the other experiments 

 and conditions apparently favourable for the well known increase in 

 acterial numbers. However, on the five and ten inch levels the 

 trend was downward and only on the fifteen inch level was there a 

 slight increase. The moisture and bacteria curves run more or less 

 parallel. However, some of the samples had to be thawed and drained, 

 namely those taken on December 21st, January 9th, January 22nd. 

 These three dates represent the depression in the moisture curve. 

 It was not possible to estimate the influence of this water drainage on 

 the bacterial content of the samples. 



Rink Experiment. 

 Average number of bacteria (in mHlions) per gram of soil at different depths. 



Rink Experiment. 



