=302. Method of Von Bemmelén.=[195]—According to the view of Von Bemmelén, the soil contains colloidal humus and colloidal silicate, which complicate the determination of water. The colloids retain water in varying quantities, depending upon the following conditions: (1) Upon their composition and state of molecular equilibrium. (2) Upon the pressure of the aqueous vapor of the room. (3) Upon the temperature. At each degree of temperature, the quantity of absorbed water which a colloid can retain in a room saturated with aqueous vapor, is different. The quantity of water which air-dried earth gives off at 100°, has therefore, no special significance unless all conditions are known. In addition to the estimation of the quantity of water which soils, in their natural condition, are capable of taking up and holding, at ordinary temperatures, the estimation of the quantity of water which they can take up in different temperatures in rooms saturated with aqueous vapor should be of interest. It follows, therefore, that there is no special value in data obtained by drying earth at 100° or 110°. For the purpose of comparison, he prefers to select that point at which the soil is dried over sulfuric acid, the point at which the tension of the water vapor in the earth, at a temperature of plus or minus 15°, approaches zero. The water which still remains in the earth under these conditions is characterized as firmly combined water. Von Bemmelén truly observes that only in soils which contain no carbonates and no chlorids and sulfids, can the loss on ignition be regarded as the sum of the humus and water content. By moistening with ammonium carbonate, the correction for lime or carbon dioxid cannot be correctly made as has been the custom up to the present time. In the first place, ignited magnesia, when it has lost its carbon dioxid, does not take this up completely on moistening with ammonium carbonate; in the second place, reactions with the chlorids may take place; and in the third place, the lime which is in the humus will be converted into calcium carbonate. Chlorids on ignition may be volatilized or oxidized. The sulfuric acid formed from the sulfids, on ignition, can expel carbon dioxid; further than this the iron of pyrites takes up oxygen on ignition. All these influences make the numbers obtained from loss on ignition extremely variable.