666 CONTRIBUTIONS TO OUll KNOWLEDGE OP SOII.-FERTlLITY,vi., 



multiplication of bacteria occurs in the presence, as in the absence 

 of amoebae. 



In the apparent absence of iDrotozoal activity in these experi- 

 ments, it seemed necessary to confirm some of Russell and Hutchin- 

 son's results. The most telling of their experiments was one in 

 which, as the result of adding a filtered suspension to a toluened 

 soil, the bacteria rose from 66 millions on the 20th, to 166 millions 

 on the sixtieth day. No test was, however, made with the unfil- 

 tered as against the filtered suspension. The experiment was not 

 confirmed, and, as it is possible that the results might have been 

 abnormal, a repetition of a certain portion of it was decided upon. 



An alluvial soil was air-dried, and treated for two days with 5% 

 chloroform. After the evaporation of the solvent, a number of 

 20 grm. portions were weighed out into small bottles, and mois- 

 tened with 4 c.c. of water or extract, a proportional quantity of 

 water being added to the tests which received the gram of air-dried, 

 untreated soil. The amount of water lost bj' evaporation was cal- 

 culated weekly or biweekly from the loss of weight of eight bottles, 

 two from each set, and the loss was made good. The moisture in 

 the soils varied up and down from 19 "6 to 15*4. The extract was 

 made by shaking 100 grm. of soil with 500 c.c. of water for 20 

 miiuites, and filtering half of it through five inches of tightly 

 packed, cotton wool. This removed the larger protozoa, such as 

 C'olpoda cucullus, but the cysts of smaller ciliates wei'e not retained, 

 as was shown by their growth in bean-infusion. The expei'iraent 

 was made in duplicate, one set being incubated at 28°, the other at 



1.5*. 



ExPEKiMKNT iv.,a — The Addition of Soil and its Extracts. 



