18 THE ROYAL SOCIETY OF CANADA 



of frost had taken place. The location of this plot was next 

 to the cleared space mentioned in Experiment II. 

 5. Quantitative determination of bacteria from soil from the top of 

 a hill in a very exposed position. Very little snow could accu- 

 mulate on this spot, thus allowing the frost to penetrate to great 

 depth. A crop of garden peas was harvested from this field. 

 In each experiment samples were taken on the 4", 8", 12" and 

 16" levels and sometimes also on 20", 24", and 26" levels in order to 

 reach below the frost line. 



Conn took all his samples at a depth of 3-6 inches, and Brown 

 and Smith sampled at a depth of nearly 8 inches. 



All bacteria found in the samples have been classified in three 

 groups suggested by Conn, but in this paper this interesting phase of 

 the investigation will be omitted. 



The different spots were selected to afïord a sufificient range of 

 conditions. 



In Experiment I we have a rich arable soil, which was manured 

 and plowed in the fall. In Experiment III we have in the black muck 

 an arable soil, which was plowed but not manured, and contained 

 much humus, and tops from the root crop grown on it during the 

 season. In the well drained cultivated soil which was kept fallow for 

 two seasons we have a rich arable soil but devoid of raw material for 

 bacterial decomposition. In Experiment V we have a rich arable 

 soil full of the roots of the last crop of peas and plowed in the fall. 

 Being on the top of a hill, this plot was in a very exposed position 

 and was frozen to a great depth throughout the winter, giving a sufifi- 

 cient contrast with the well protected snow covered plots mentioned 

 above. 



The plot on the land planted to timothy for the second year 

 mentioned in Experiment III supplies a connecting link between the 

 plots of cultivated arable soil and the plots marked out on a permanent 

 lawn. This timothy plot contained much raw material in the form of 

 roots etc., which was available for bacterial decomposition. 



Finally we have the plots on a permanent lawn. In Experiment 

 II the lawn was most of the time protected by a heavy layer of snow, 

 which prevented or retarded frost penetration. In comparison we 

 have the plot mentioned in Experiment III, situated on an outdoor 

 rink, the soil of which was frozen to great depth throughout the winter. 

 Excep4: for this exposed condition the rink soil was of the same quality 

 and contained the same amount of raw material available for bacterial 

 decomposition as the soil under the permanent lawn mentioned in 

 Experiment II. 



