BY U. OREIG-SMITII. 



697 



allowed to evaporate. When all odour had disappeared, the 

 layers were separated into 100 gr. portions. The five layers 

 were tested in the usual manner, by infecting them with a 

 suspension of Bac. prodigiosus, and incubating the portions 

 for 20 hours at 28°. Another set was tested by moistening 

 20 gr. of the soil with 5 c.c. of sterile water, and incubating 

 at 28° for three days. The numbers of the bacteria in the 

 two sets closely follow one another, and, with the exception 

 of the top soil, which is exceptional*, the results agree with 

 the previous experiment. 



Experiment xxi. 



Other experiments, made by simply moistening the soils, 

 confirmed these results. Portions of soil, weighing 100 gr., 

 were placed in beakers, and wetted with chloroform, ether, 

 or carbon bisulphide. "When all odour of the solvent had dis- 

 appeared, 20 gr. portions were removed from the top, middle, 

 and bottom. These were exposed to the air to ensure the 

 volatilisation of the last traces of solvent, and were then 

 moistened with 5 c.c. of water, and incubated at 28° in a 

 moist atmosphere. 



* This exceptional behaviour was confirmed in another experiment with 

 garden-soil, in which 10 cells became 582, 252, and 450 for the top, middle 

 and lower laj-eis respectively. It has also been found in soils exceedingl}' 

 rich in agricere, e.g., sewage-sick soils, that the top layer.s are, after 

 treatment, much more nutritive to Bac. prodigio>m8 than the natural 

 spore-forming soil-bacteria which show most growth in the lower layers, 

 'i'he reason for this difference will probalily be found to be that, in the 

 rich soils, tiie action of the toxins is of more moment than the agriceie, so 

 fai- as tlie growth of the comparatively strong toxin-sensitive Bac. pro- 

 digiosiis is concerned. 



