kneading or other means before a close, finegrained and strong ware can be shaped from it.

The necessity for the crumb structure in all soils except the coarsest, sandy types is found in the fact that the open spaces through which water and air must move and into which roots, root hairs and soil organisms must penetrate in order to place themselves in contact with soil surfaces would be very much too narrow to admit of the necessary amount and rate of movement. An illustration will make this point clear; suppose we lay upon the table three balls, all of the same diameter and so that they touch one another, and then set on top of these a fourth, so that this one touches the other three. The open space left between the four balls may represent a pore space in any soil. The largest root which could pass through such a space without having its cylindrical form changed could have a diameter of only about 16-100ths of the diameter of the balls themselves; or, in the case of the soil particles, only .16 to .2 of the diameter of these

particles. But the mean effective diameter of the particles in our coarsest, sandy soils is such that some 200 of them must be placed in line to span a single inch. So the diameter of the root which could pass through the space between these grains, when all are in touch, could only be of the order of that possessed by a single cotton fiber, or about .001 of an inch. But in the average loam soil the diameter of the mean effective grain, after the dry soil has been broken down by rubbing in a mortar requires some 1,200, instead of 200, to span a linear inch, while with the finest clay soils as many as 6,000 grains are needed.

Small, then, as the pores are in our coarsest, sandy soils, those in the loams and in our finest clay types, if they did not possess the crumb structure, would be but one-sixth and one-thirtieth respectively of that stated for the coarsest, sandy soil. But another principle must be here stated to make more clear the immense importance of the granular structure in soils. It is this: The flow of water and of air through soils and sands in

creases in the ratio of the squares of the effective diameters of the grains. That is to say, in the coarse, sandy soil, in the loams and in the fine clay types, if these did not have the crumb structure, the flow of air and of water through them would be in the order of 900 to 36 to 1, the movement through the sandy soil being 900 times more rapid than through the finest clay. If, for example, two and one-half hours were required to permit the coarse sandy soil to dispose of the excess water falling upon the field during a given rain, by underdrainage, the finest clay soil, without the granular structure, would require some three months to free itself of a like amount by underdrainage. It is therefore clear that the finest clay soils can only be surface drained until after they have acquired the crumb structure to a greater or less extent. More than this, in the properly open soil there are but two and one-half hours between rainfalls during which diffusion can carry the soluble plant food into the water draining away, while in the other condition this loss by drainage is continuous.

The low producing power, or absolute sterility, so invariably associated with puddled soils and with those too close in texture, we believe to be primarily due to a lack of available moisture, notwithstanding the seeming paradox that they are carrying an excess of it. It is a familiar fact that crops wilt and cease to grow in close-textured clayey soils when still carrying 8 to 12 per cent of water, while they may grow luxuriantly in coarse, sandy soils possessing but I to 3 per cent. So, too, we often find desert types of vegetation growing in humid climates on extremely close-grained clayey soils, and more strangely still in peat swamps where the water content is excessively high. To understand these facts it must be remembered that there is a certain thickness of water film which is held so firmly to the soil grain surfaces as to be wholly unavailable to the crop. Portions of this layer cannot be driven off completely even at the temperature of boiling water. When all of the facts shall have been worked out, we believe it will be found that

the thickness of the unavailable water about the surface of the soil grains is essentially the same whether these be large, as in the coarse, sandy types, or very small, as in the finest clays; and if this is the case the absolute amount of unavailable water must increase as the internal surface of the soil becomes greater and as the diameters of the soil particles decrease.

The coarse, sandy soils, with their relatively small internal surface, carry a correspondingly small amount of unavailable water, and hence in them small rainfalls in dry times have a relatively high efficiency. So, too, must soluble plant food and fertilizers, when applied to them, for the same reason, have a relatively high efficiency. But in the finest clay soils, especially if they are not strongly granulated, the amount of unavailable water is very large, and hence it is that heavier rainfall during drouth periods and more liberal applications of fertilizers are required to produce the same relative increase. But it is possible to have the finest clay soils so completely puddled