Relation to Factor Limiting Bacterial Activity in Soil. 447 



untreated soil (see footnote, p. 454). The Untreated + 1870 curve is, on the 

 whole, very even, showing no marked fluctuations up or down, and it is very 



232 284 



DAYS 



Fig. 4. 



interesting to note that during the first 232 days the bacterial counts are 

 below the counts for the untreated soil. This seems to indicate that some 

 factor was introduced with the 1870 soil which, for a time, checked the 

 rapid growth of bacteria and prevented their increasing to the numbers 

 attained by the bacteria in the untreated soil. It is conceivable that this 

 was owing to the action of the protozoa added with the 1870 soil, but 

 the fact that after 232 days the bacterial numbers for the untreated soil 

 came down below the level of the mixed sample, and that the two curves 

 during the last 160 days are practically identical lends no support to this 

 idea. 



Whatever the influence was which during the first 232 days checked the 

 growth of the bacteria in the mixed soil, I am inclined to regard it as con- 

 nected with some other property of the 1870 soil than the presence of 

 protozoa in it. To be more explicit : The 1870 soil was bottled in a com- 

 paratively moist condition and received no drying about 1881 as did the 1846 

 soil along with many others. This drying seems to have effected a very 

 important change in the 1846 soil and, taken in conjunction with the pro- 

 longed period of storage has produced in it a condition comparable with 

 partial sterilisation. At any rate, the 1846 soil along with other dried and 

 stored soils which I have examined gave very high bacterial counts when 

 moistened, whereas the 1870 and other soils which have been stored in almost 

 the same condition as they were in when taken from the field give low 

 bacterial counts and indicate the presence of the limiting factor. The 

 following experiments illustrate my point. 



Bacterial Counts in other Samples of Old Stored Soils. 



Three soils were taken for this piece of work, two of them from bottles of 

 Broadbalk soil stored since 1856 and 1865, and one from a bottle of Geescroft 



