Gorup-Besanez (1863) mixed 30 grams arsenious acid with 30·7 litres[9] soil, growing two plants on this quantity of earth. Most of his experimental plants (_Polygonum Fagopyrum_, _Pisum sativum_, and _Secale cereale_) developed normally, but _Panicum italicum_ died soon after the plants appeared above the surface, the leaves being very badly coloured. Analyses by Marsh’s test showed no trace of arsenic in 20 grams dry matter from _Secale cereale_, but in 148 grams _Polygonum Fagopyrum_ the presence of arsenic was evident, though the mirror formed was weak. With such a large proportion of arsenious acid in the soil it seems hardly conceivable that the plants were not injured to some extent, and also it is probable that with more careful analyses arsenic would have been detected in those instances in which its presence was denied. Yet it must be remembered that Davy (1859) had treated pea plants in pots with a _saturated_ solution of arsenious acid for a short time and had stated that the plants were uninjured. Thus both Gorup-Besanez and Davy concur in the opinion that _Pisum sativum_ is indifferent to relatively large quantities of arsenious acid when presented in the soil, whereas the Rothamsted experiments show that in water cultures the plant is extremely sensitive even to minute traces of the substance. It is possible that the arsenic in the solution added to the soil enters into combination with other substances, forming insoluble compounds, thus being removed from the sphere of action and rendered unable to affect plant life. If this be so, the apparent immunity of certain plants to arsenious acid is explained. F. C. Phillips (1882), in his experiments on various flowering plants, such as geraniums, coleas and pansies, found that compounds of arsenic in the soil exercised a distinct poisoning influence, tending, when present in large amount, to check the formation of roots, so that the vitality of the plant was so far reduced as to interfere with nutrition and growth, or even to kill it outright. He also stated that traces of arsenic were found in all the plants grown upon the poisoned soil. [9] 30 grams arsenious acid to 30·7 “cubik Decimeter” soil = about ·1%. In this connection it is interesting to note that a certain proportion of arsenic is frequently present in the superphosphate used as manure. In view of the known toxicity of arsenical compounds to plant life the question arose as to whether superphosphate manuring would exercise a detrimental influence on account of its arsenic content. Experiments carried out by Stoklasa (1898), however, indicate that there is not sufficient arsenic in maximum doses of superphosphate to exercise a toxic action in the field. (_d_) _Physiological considerations._