740 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOIL-FERTILITY, vii. -XI., 



The experiment shows that a soil, from which the toxins had 

 decayed during long storage, became less nutritive or more toxic 

 upon being moistened. At a comparatively low temperature 

 (15°-20°), the change was slower to develop than at a higher 

 temperature (28°), but when formed, it was more persistent. 



Observations made, during these investigations, show that soil 

 has a variable, bacteriotoxic content. Rain washes the toxin 

 out, and the soil becomes non-toxic. When the rain ceases, 

 bacteriotoxins are again formed, and persist, if the soil remains 

 moist. With continued dry weather, and consequent lowering 

 of the soil-moisture, the toxins decay. Experiments in the 

 laboratory confirmed these observations. A soil originally toxic, 

 became non-toxic when washed with water, and, upon incubation, 

 again became toxic. Another soil, originally nutritive, became 

 very much less so upon incubation. 



x. The action of Chloroform upon Blood treated with Vaseline. 



When a soil is treated with a volatile disinfectant, it behaves 

 as if more nourishment had been liberated for the growth of the 

 surviving bacteria. The volatile disinfectants are also fat- 

 solvents, and, after noting their visible action in the soil, I sug- 

 gested that the effect, produced by the disinfectant, was by virtue 

 of its removing the fat or agricere, and so facilitating the decay 

 of the organic matter. Certain experiments* upon the growth 

 of bacteria, in various layers of disinfected soil, bore out this 

 suggestion. In endeavouring to confirm the hypothesis, I have 

 made experiments with the ammoniacal fermentation of blood 

 saturated either with paraffin or vaseline, and treated with chlo- 

 roform. These, however, are rather against the hypothesis, and 

 are here recorded as a contribution to the subject. 



Dried blood was heated for a day, at 56°, with paraffin melting 

 at 43°, and the excess removed. After being ground and sifted, 

 two gram portions were weighed out, mixed with 50 grams of 

 sand, and treated with chloroform. One hundred and thirty 

 grams of dry soil were added, and 30 c.c. of soil-infusion. The 

 tests were incubated at 22°, for 6 days; then water was added, 



* These Proceedings, 1911, pp.696 tt seq. 



