same in the movement of water through soils, when the movement is through the capillary pores.

If water is moving horizontally by capillarity away from under an irrigation furrow, the resistance to be overcome in carrying the water through the second inch is double what was necessary to carry it through the first inch, and hence the progress made in the second inch will be slower than that made in the first inch, and hence, as the distance between furrows is increased the drag on the water becomes so great that ultimately a dry area must be left between furrows, and for the same reason that when ground water is a certain distance below the root zone water cannot be lifted by capillarity into the root zone.

These statements make it clear also that if one were trying to saturate the surface foot of soil by upward capillary movement by means of water carried in tile with their tops laid one foot below the surface, it would be impossible to do so without great loss of water by downward percolation, un

less there were some impervious layer not far below the tile.

TIME REQUIRED FOR LATERAL SPREADING

To measure the time required to saturate the soil by horizontal capillary spreading, we filled a six by six foot tray, eight inches deep, with a clay loam, firmed to the density of field conditions. In one corner of the tray was set on end a section of six-inch drain tile. In the bottom of this, water was maintained at a depth of three-fourths of an inch, automatically, in a manner which permitted weighing periodically the water which entered the soil. It was found that at a distance of one foot from the tile water moved horizontally at the rate of 3.73 pounds per square foot, or .46 inch per day, during 44 days; that eight inches of soil became saturated at one foot from the tile; and that no water passed into the soil beyond a distance of 3.5 feet from the tile, the soil beyond this distance becoming increasingly drier than when the experiment started. It is clear

from these observations that in furrow irrigation water cannot be economically applied unless the furrows are pretty close together.

It is not likely that the drag of 3.5 feet

on the streams of water moving horizontally through this particular soil was sufficient to entirely overcome the surface tension pull. What did happen was that the rate of capillary movement upward to the surface and evaporation from the surface prevented any moisture being carried farther than 3.5 feet from the furrow. Of course, the same limiting conditions must always be effective in a field, even though a good mulch exists, for these do not prevent all evaporation

A SPECIFIC CASE

Let us take a specific case where the furrows are laid say three feet apart, deep and broad enough so that gravity flow directly downward beneath the furrow saturates one foot in width of the soil. In such a case one-third of the field would be irrigated by the directly downward gravity flow and

two-thirds of the field would have its soil moistened by the horizontal capillary flow.

RATE OF PERCOLATION DOWNWARD

In the coarser, sandy soils the rate of gravity flow may be relatively very rapid, many times exceeding the capillary flow in ordinary soils. We have measured the gravity flow downward through five rather coarse sands, and through two soils, one a sandy loam and the other a clay loam. These are the rates of movement through the sands and soils, expressed in inches of water which passed into columns eight to 10 feet long, per day.

From the rates at which water entered the sandy loam and the clay loam it ought to be possible to apply two or three inches

of water to a field and have it all enter the soil during 24 hours, if the movement were directly downward, due to the gravity pull alone.

In Dr. Loughridge's studies, made at Riverside some years ago, it was found, where irrigation was maintained during 72 hours in furrows three to four feet apart, and where 4.5 inches of water were applied, that the average depth of penetration of water was about four feet and the distance the water had spread laterally averaged only about 2.5 feet. Besides this, much of the soil received no water, while very large amounts of water had percolated below the depth in which the roots had developed.

It ought to be possible to apply water to an orchard and much more nearly realize the conditions which follow natural rainfall than was secured here. Let us go over carefully the method by which water enters the soil and then point out how it appears possible to secure a quicker and more uniform distribution of the water.

When water is admitted to the furrows,