A simple experimental method has been described for measuring certain physical constants of soil, using small brass boxes into which soil passing a sieve of 100 meshes to the inch has been packed by hand. The quantities determined are:(1) The weight of unit volume (100 c.c.s.) of air-dry soil, or the apparent specific gravity.(2) Amount of water taken up by unit weight of soil.(3) Pore space.(4) Specific gravity of the soil.(5) The volume expansion of unit volume (100 c.c.) of soil when saturated.The results for one soil only are given, and discussed, to illustrate the method. With the co-operation of the Science Masters Association it is being applied to a number of soils by various schools.The particular soil used was obtained in six depths as follows: 0–6–12″, 12–18″, 18–24″, 2–3′, 3–4′, and the above constants were determined on each depth. It was shown that (1) and (4) varied inversely with the percentage of clay in the soil, while (2), (3), and (5) varied directly with the clay percentage. The effect on the constants of the larger quantities of organic, matter present in the top two layers of soil was, weight for weight, approximately equal to that of the clay, except in the volume expansion results where the effect if any was within experimental error.It is possible that the fraction fine silt II, whose upper limit of diameter is ·005 mm., has similar effects to the clay fraction.

Much of the modern work on the physical properties of soil has been interpreted on a colloidal basis and more recently this hypothesis has been extended by the thermodynamical studies of Wilsdon and also by the investigations of two of us on an indirect measurement of the vapour pressure of capillary systems. There is evidence that the colloidal portion of the soil can be regarded as possessing a reticulate structure, possibly analogous to that shown by Zsigmondy to exist in silica gels. The pore space in soils is therefore an assemblage of voids and irregular capillaries ranging from ultramicroscopic dimensions in the colloidal portions to the macroscopic interstices between adjacent compound particles and the larger mineral fragments. Whereas in studies of evaporation and movement of water the total intersticial space is operative, the vapour pressure of soils at different moisture contents is very largely controlled by the minute pores associated with the colloidal portion and the larger voids have comparatively little influence. Vapour pressure measurements therefore afford a promising line of attack on the physical relations between the colloidal soil material and water, especially when the measurements are made on soils subjected to a variety of preliminary treatments, known to have a considerable effect on other physical properties. The effect of successive wetting and drying, heating and addition of salts are of especial interest in this connection.

The movement of water in soil, and the manner in which water is distributed over the particles and within the interstices, i.e. the dynamical and statical aspects respectively of moisture distribution, are of fundamental importance in soil science. The literature of the subject abounds with experimental determinations, but the difficulties of theoretical treatment are great, and it is only within recent years that any serious attempts have been made in this direction.

1. The results of a six-year cultivation rotation experiment are given. The rotation used was wheat-mangolds-barley and the seed-beds for these were prepared either by ploughing, using a rotary cultivator or a tractor-drawn grubber.2. The yields of these crops were barely influenced by the depth of ploughing, a 4 in. depth giving throughout the six years just about the same yield as an 8 in. depth. The mangold crop was possibly a little larger on the deeper ploughed plots.3. The mean yields of the seed-beds prepared with the tractor drawn grubber or cultivator followed by harrows etc. were lower than the ploughed seed-beds for all the crops, and this was particularly so on those seed-beds prepared by only one grubbing down to 4 in. depth.4. The mean yields on the seed-beds prepared by the rototiller were lower than on the ploughed seed-beds for wheat and mangolds. If the seed-bed was prepared by rototillage to a depth of 8 in. by going over the land twice, the yield of barley was the same as on the ploughed seed-beds, but was definitely less on the seed-bed rototilled only once to 4 in.5. Seed-beds prepared by the rototiller or grubber have only a small residual effect on the crop yields in the following year.6. It is concluded that the primary function of ploughing is weed control, and that it is only advisable to omit ploughing either if the land is already fairly clean or if the crop will be hoed very early on in its development.7. For wheat and mangolds differences in weed infestation of the seed-bed were probably of greater importance than differences in tilth in so far as the crop yield was concerned. The reverse may have been true for barley.

Numerous attempts have been made to devise an experimental method that, applied to a variety or a series of soils, enables them to be placed in an order closely reflecting their field behaviour or their most important physical characteristics. They are called “single-value” determinations as they endeavour to specify the soil by a single number, in distinction to the group of figures obtained, for example, from a mechanical analysis. A number of these methods are discussed in the present paper which contains an account of a detailed investigation on 39 soils of certain single value determinations.The methods selected for study were chosen because (i) they required only simple apparatus, and (ii) they appeared to be related to some distinct soil characteristic.The list of measurements was as follows:Percentage of clay.Moisture content of soil in equilibrium with atmosphere of 50 per cent, relative humidity (the ordinary “air-dry moisture content” which was also determined, is close to this value).