[vanderleck] bacteria OF FROZEN SOILS • 23 



together in the plowed field mentioned in Experiment I. A very 

 satisfactory snow covering was maintained in this part of the field 

 and the temperatures recorded are very representative of the average 

 conditions in this field. Unluckily the thermometers did not register 

 below 30° F. Besides the reading of these thermometers was difficult. 

 Several feet of snow had to be removed before the mercury level was 

 reached and then the glare of the surrounding snow together with the 

 semi gloom of the deep snow pit interfered with accurate reading. 

 After each reading loose snow was carefully replaced round the ther- 

 mometers to the former level. The two-feet soil thermometer was 

 placed in the permanent lawn mentioned in the second experiment 

 and had throughout the winter the heavy snow covering of this 

 field. 



A few times temperature observations were taken in the sample 

 pits, but these results were so unreliable that the readings were 

 ignored. The difference between the soil temperature around 32° F. 

 and an outdoor temperature of— 10°F. was too great. The cold air 

 would fall in the test pit and reduce the reading by 10° to 12° F. 



At a depth of 1 foot the soil was frozen from January 10th to 

 March 25th. a period of 2^ months. The thermometers placed at 

 a depth of 2 feet and 3 feet did not record any approach to the freezing 

 point. 



Conn and all the other investigators had found in all their experi- 

 ments a reduction in the number of bacteria when the soil was not 

 frozen and an increase in the bacterial count during the frozen period. 

 The writer therefore planned to execute most of the analyses and 

 experiments during February and March, during which months the 

 soil would be frozen and the few analyses in November, December 

 and the beginning of January were only taken to keep in touch with 

 soil conditions previous to the period of frost. When these earlier 

 analyses showed the presence of larger numbers of bacteria, much 

 larger than found in summer or mentioned by the other investigators 

 in their work. Experiments II IV and V were undertaken, which were 

 started too late in the year to allow the formation of definite conclu- 

 sions. For next winter an extensive campaign has been formulated to 

 settle definitely some of the new viewpoints opened up. 



The data obtained are arranged in five sets of three tables and 

 one diagram each. Table A gives the total bacterial content per gram 

 of dry soil; Table B compares the number of bacteria appearing on 

 beef peptone media, with those on soil extract gelatine*; and Table 

 C gives the moisture percentages at depth of sampling. 



*The samples were also plated on soil extract agar, but so few bacteria grew on. 

 these plates that the counts were ehminated. 



