In many problems dealing with the use of water by plants, containers of various sizes are adopted, in which the system is sealed against loss of moisture by evaporation from the surface, and against loss by drainage. These containers are usually weighed at intervals, and water added in measured quantities to bring about the moisture conditions desired by the investigator. Generally, these experiments contain several series of containers that are supposed to be moistened uniformly, and kept at moisture contents less than that of the field capacity of the soil used. A report was made on the distribution of moisture in soil following an irrigation in various field experiments (1, 3, 5, 6), in which it was shown that water moves but slowly from soil at field capacity to adjoining dry soil, and that, in the absence of roots, the line of demarcation between the wet and the dry soil does not change during relatively long periods of time. In another paper (2) evidence was presented which showed that the movement of moisture from a wet to a dry soil in containers, takes place in the same way that it does under field conditions. The lack of uniform distribution of moisture by autoirrigators was shown by the accumulation of roots immediately around and below the porous cones. A device for automatically bringing the moisture in a soil in a container up to the field capacity is described by Johnston and Atkins (4). The following results show some of the soil-moisture conditions obtained in experiments with sunflower plants in containers, when the soil was wetted to the field capacity, and when attempts were made to moisten soil at the permanent wilting percentage, to a moisture content less than that of the field capacity. Three soils, a sand, a loam, and a clay adobe, were used. The containers were all brought to a uniform weight for convenience in handling, and a known weight (about 600 grams) of oven dry soil was placed in each. The soil was moistened to its field capacity and dwarf sunflowers planted. When the sunflower plants had reached a height of about 12 cm. and each had 3 pairs of full sized leaves, they were allowed to wilt. The wilting percentages obtained agreed with those obtained in previous trials with these soils. After wilting, various amounts of water were added to the surface of the soil. The plants were then cut off and the container sealed until the following day when the cans were cut longitudinally with a hack saw, and photographed. Soil samples of the top and bottom portions were taken for the determination of the moisture content. The moisture equiva-