1. The investigation described is a continuation of the researches of certain workers in the British West Indies on the significance of the shrinkage coefficient of clays and soils. A brief outline of these researches and their results is given.

2. Tempany's hypothesis, which explains shrinkage as due to contraction consequent on loss of water by evaporation from the saturated gel-skeleton that ramifies throughout a mass of soil at its point of maximum plasticity, is criticised on the grounds that it does not completely fit the facts. In particular, it fails to account for the abnormally low shrinkage coefficients exhibited by lateritic soils, notably the red upland soil of Barbados. Mason explained this abnormality by assuming that kneading does not entirely destroy aggregation of soil particles, which is especially well marked in the Barbados red soil. The writer demonstrated, however, that soils of similar colloid content, and belonging to one and the same geological type, but exhibiting different degrees of aggregation, possess similar shrinkage coefficients. A different hypothesis to explain soil shrinkage was therefore sought.

3. The hypothesis finally accepted is based on the belief that colloidal gels possess a reticulate structure. At the point of saturation, a hydrogel probably contains water in two phases. The first of these is adsorbed in the walls of the gel; it has been shown by Wilsdon, in the case of soil colloids, to represent the moisture content at the hygroscopic coefficient stage. The second phase fills the vesicles of the gel, and is a crystalline phase.

1. The use of a positive air pressure on closely packed soils decreases the time of recovery and increases the yield of solution obtained by water displacement.

2. Equipment and technique for the recovery of displaced solutions is described.

3. Data from three sandy loams and two silty clay loams are reported.

4. Successive displaced solutions of equal electrical conductivity were obtained from each soil. The concentration thus measured was in each case the same as that of the liquid obtained without the use of displacing water.

5. Concentrations of individual ions in successive portions of displaced solution are uniform in soils at optimum and this also appears to be true at half optimum.

6. There is an inverse proportion between total moisture content of soil and concentration of displaced solution as indicated by measurements of specific resistance.

7. When the solution displaced by water from a given mass of soil was used as displacing agent on another portion of the same soil, the newly displaced solution had the same concentration of electrolytes as the displacing solution. The opinion is expressed that this result indicates that the displaced solution has the same concentration as the solution with which it came in contact in the soil.

8. Typical differences in behaviour of the various solutes in displaced solutions are discussed.

9. The significance and limitations of water extractions of soils are discussed in the light of the results from displaced solutions.

(1) A new and simple method of measuring the shrinkage of moist soil on drying is described, which at the same time gives values for the pore space and specific gravity of the soil.

(2) Diagrams for a number of diverse soils are given, illustrating the character of soil shrinkage.

(3) The shrinkage is shown to take place in two stages in both of which there is a linear relationship to the moisture content.

(4) Tentative explanations of these two stages are advanced, based on the well-known colloidal-coating hypothesis, and confirmatory experiments described.

(5) By means of the method the effect of alternate wetting and drying of soil in producing a good tilth is illustrated.

It is not infrequently assumed that varieties of cultivated plants differ not only in their suitability to different climatic and soil conditions, but in their response to different manures. Since the experimental error of field experiments is often underestimated, this supposition affords a means of explaining discrepancies between the results of manurial experiments conducted with different varieties; in the absence of experimental evidence adequate to prove or disprove the supposed differences between varieties in their response to manures such explanations cannot be definitely set aside, although we very often suspect that the discrepancies are in reality due to the normal errors of field experiments.

(1) The colorimetric method is unsuited to the examination of heavy, alkaline soils owing to the turbidity of the suspension.

(2) Where the nature of the suspension permits of colorimetric determinations being made, they agree with those obtained electrometrically: with the latter method practically identical results are obtained using soil-water mixtures or moderately clear extracts.

(3) No disturbing effect is likely to be introduced by amounts of nitrate up to 500 parts per million of soil.

(4) Owing to the effect on the pH of a soil suspension caused by varying the proportion of water and time of extraction, these conditions should be fixed for routine work. We have found 1 hour's extraction with 5 parts of water satisfactory.

(5) On account of the amphoteric or buffer nature of clay, soil shifts the reaction of acids and alkalis in the direction of neutrality.

(6) The effect of sodium salts on a soil is to displace aluminium and so reduce alkalinity: the residual soil after leaching is found to be more alkaline.

(7) The effect of drying alkaline soil is to cause the pH of the extract to be lower than that obtained from the undried soil. If however the time of extraction is prolonged, the differences disappear almost entirely.

Field experiments with maize at the Bahtim Experimental Station and laboratory tests indicate that the lower efficiency of ammonium sulphate as compared with nitrate of soda is due principally under these special conditions to the liberation of ammonia from the fertiliser in contact with an alkaline soil. The rate of this liberation is a function of the aeration and of the soil reaction.

The maximum water-retaining capacity of a soil may be defined as the quantity of water, measured as a percentage of the mass of oven-dry material, that the soil can retain after it has been thoroughly wetted and then allowed to drain freely. It is generally estimated in the laboratory by Hilgard's method (1893). The air-dried sifted soil is packed by jarring into a small perforated cylindrical brass box which is then placed in a vessel containing water, so that the bottom layer of the soil is in contact with the liquid surface. After standing an hour, the cylinder is removed, surplus water wiped away, and the whole weighed. The cylinder and contents are then dried to constant mass in an air-oven at 110° C. The weighings are corrected for the mass of the cylinder and that of the filter paper employed to cover the sieve. From the results, the maximum water-retaining capacity may directly be calculated.

As has long been realised by soil bacteriologists, the present plating technique used in estimating bacterial numbers in soil suffers from the lack of a really efficient means of disintegrating the soil, so as to produce an even suspension of the organisms in the diluting fluid. As a consequence of this, although we are able, by careful standardisation of the shaking and dilution, to compare the bacterial numbers in two or more samples of the same soil type, the figures obtained are only relative, and do not give an estimate of the total bacterial numbers of the soil.