448 Mr. T. Goodey. Investigations on Protozoa in 



soil stored since 1865. The two from Broadbalk were dry when taken from 

 their original bottles, whilst the Geescroft sample was in a comparatively 

 moist condition. 



The water-contents of these soils were not determined immediately after 

 taking them from their respective bottles, because it was not my intention at 

 that time to use them for bacterial counts but merely to ascertain the 

 character of the protozoan fauna. As far as I could judge, I should say that 

 the Geescroft soil contained about 10 per cent, of water, whilst the Broadbalk 

 samples were of about the same degree of dryness and contained about 

 2 per cent, or 3 per cent, of moisture. It was evident from the appearance 

 of the soils that the Broadbalk soils had been taken out and dried along with 

 many other soils in 1881, whereas the Geescroft soil had been left untouched 

 and closely resembled the Barnfield 1870 soil, containing about 10 per cent, of 

 water, which I had used in earlier work. 



I found on examining these soils culturally that the Broadbalk 1856 con- 

 tained no protozoa, the Broadbalk 1865 contained amoebae and flagellates and 

 the Geescroft 1865 also contained amoebae and flagellates. 



A weighed quantity of each of these soils was taken and after making 

 initial counts to determine the bacterial content they were all moistened to 

 20 per cent, water-content. After a period of 148 days the soils were 

 remoistened to bring up the water-content to 20 per cent, again, owing to the 

 gradual loss of moisture by evaporation. 



Bacterial counts by the gelatine-plate method were made at different 

 intervals and the results are set out in the table below. 



Bacteria in Millions per Gramme. 





At 



After 



After 



After 



After 



After 





beginning. 



68 days. 



102 days. 



130. days. 



198 days. 



231 days. 



Bd. 1856 



4 



136 



169 



73 



200 



158 



Bd. 1865 



4 



66 



93 



55 



120 



124 



Gr. 1865 



2-5 



11 -4 



5 



10 -7 



18 -6 



9 



The curves obtained on plotting these results are shown in fig. 5. 



The most noteworthy feature of these curves is the high bacterial content of 

 the Broadbalk soils and the low bacterial content of the Geescroft soil. The 

 former have all the appearance of curves of partially sterilised soils, whilst the 

 latter presents the usual appearance of an untreated poor soil, containing a 

 limiting factor. 



The drop at 130 days in the two Broadbalk soils may be accounted for by 



