Jensen (1907) worked with “artificial” soils, under sterile conditions, using finely ground quartz flour for his medium and wheat for a test plant, parallel experiments being carried on with solutions. Every precaution was taken to ensure sterility--the corks were boiled first in water and then in paraffin, the seeds were sterilised in 2% copper sulphate solution for ¾ hour, washed in sterilised water, planted in sterilised sphagnum, the transplanting being done in a sterile chamber into sterilised solutions. The criteria used to determine the toxic and stimulation effects were the total transpiration, average length of sprout, the green weight and dry weight of plants. The results obtained with the different substrata showed that it does not follow that a salt highly toxic in solution is equally so in soil, or that one which holds a relatively high toxic position in soil should occupy the same relative position in solution cultures. For instance, while in soil cultures nickel compounds were the most toxic of all the substances tried, in solution cultures silver compounds were more poisonous than nickel. The range of concentrations, both fatal and accelerating, was found to be much greater in solution than in soil cultures. In the sand cultures the toxicity of the copper sulphate was found to decrease as the ratio of the quartz sand to the poisonous solution increased, provided that a water content suitable for growth was present. Jensen states that the fatal concentration of copper sulphate in solution cultures is approximately 1/10th that of the fatal concentration in his artificial soil. When copper salts are added to soil a complication at once sets in due to the double decomposition which is always likely to occur when any soluble salt is added to soil. The reaction may be graphically expressed as follows, in a much simplified form-- _AB_ + _CD_ = _AC_ + _BD_.