With sea soils, Von Bemmelén has weighed the soil after the elementary analysis and estimated, in addition to the carbon dioxid, both chlorin and sulfuric acid therein. The comparison of these estimations with those of CO₂, Cl, SO₃ and S, made in the original soil, gave the necessary corrections; _viz._, for the increase in the weight through oxidation of sulfur and iron, and for the decrease in weight through the volatilization of sodium chlorid, sulfur, and carbon dioxid. A trace of chlorin was evolved as ferric chlorid, nevertheless, the molecular weight of sodium chlorid, 58.5, is scarcely different from the equivalent quantity of ferric chlorid 54.1. For this reason the estimation of loss of water, on ignition, of sea soils is less exact than that of soils which are free from carbonates and sulfids and which, as is usually the case with tillable soils, contain only small quantities of chlorids and sulfates. _The Strongly Combined Water._—Water which, at a temperature of plus or minus 15°, in a dry room, still remains in the soil, is chiefly combined according to Von Bemmelén with the colloidal bodies therein. Its estimation, presents, naturally, difficulties and is not capable of any great exactness. The quantity of strongly combined water, on the one hand is determined from the difference between the loss on ignition and the quantity of humus present, calculated from the content of carbon; on the other hand, from the difference between the water obtained by elementary analysis and the water which corresponds to the calculated quantity of humus. If the hydrogen content of humus is correctly taken and no appreciable error is introduced through the factor 1.724, both of these differences must agree. On the other hand the hydrogen content of the humus can be computed from the difference between the water found and the calculated content of the firmly combined water. The hydrogen content of humus bodies, dried at 100°, varies between four and five per cent. Eggertz has found the content from 4.3 to 6.6 per cent of hydrogen in thirteen soils which he first treated with dilute hydrochloric acid then extracted with ammonia or potash lye and precipitated this alkaline extract with acid. The method of applying these principles to soil analysis is indicated in the following scheme: A volcanic earth from Deli gave, on elementary analysis: Per cent. Carbon 2.94 Water 14.78 Nitrogen 0.28 Loss on ignition 17.54 FIRST CALCULATION. Per cent. Loss on ignition 17.54 Humus = carbon, 2.94 × 1.724 = 5.07 Difference = firmly combined water 12.47