These later Rothamsted results fit in very well with those obtained ten years ago (1903) by True and Gies in their experiments on the physiological action of some of the heavy metals in mixed solutions. Plants of _Lupinus albus_ were tested for 24–48 hours with different solutions in which the roots were immersed. Given the same strength of the same poison, the addition of different salts yielded varying results. For instance, with copper chloride as the toxic agent, the addition of magnesium chloride did not affect the toxicity, calcium chloride decreased it, while sodium chloride slightly increased the poisonous action. Calcium sulphate with copper sulphate enabled a plant to withstand four times as much copper as when the latter was used in pure solution. Calcium salts in conjunction with those of copper proved generally to accelerate but not to increase growth, but with silver salts they did not cause any improvement. Perhaps this amelioration is in inverse proportion to the activity of the heavy metals. With a complex mixture consisting of five salts--copper sulphate and salts of sodium, magnesium, calcium and potassium, all except calcium being present in concentrations strong enough to interfere with growth if used alone--it was shown that “as a result of their presence together, not only is there no addition of poisonous effects, but a neutralisation of toxicity to such degree as to permit in the mixed solutions a growth-rate equal to or greater than that seen in the check culture.” If the concentration of the copper salts was increased, the other salts remaining the same, the poisonous activity of the copper became greater than could be neutralised by the other salts. If the copper remained the same and the other salts were diminished by half (i.e. below toxic concentration) the neutralising action of the added salts was markedly less, and the growth rate _never exceeded_ that of the control. This was apparently due to the action of the unneutralised copper. The indications are that the conspicuously effective part of the molecule is the cation or metal, and that the anion plays little or no part in causing the toxicity; in such great dilutions the metals act as free ions. The hypothesis is put forward that interior physiological modifications are responsible for the observed differences in growth rate, the cell processes being so affected as to bring about different results on cellular growth; in other words, the growth rate represents the physiological sum of oppositely acting stimuli or of antagonistic protoplasmic changes where mixtures of salts occur. This is really an extension of Heald’s idea that the toxic effect of a poison is due partly to changes in the turgescence of the cell, a sudden decrease causing retardation or inhibition of growth, and partly to a direct action on the protoplasm, which differs in different plants with the same salt. Heald (1896) went so far as to suggest that the poisonous action is a