The reaction of the soil may be determined with a large number of indicators among which, for convenience, sensitive litmus paper, both red and blue, stand in the front rank. A sample of the soil, from fifteen to thirty grams, is mixed with water to a paste and allowed to settle. The litmus paper is then dipped into the supernatant liquid. =293. Determination of Water in Soil.=—The following problems are presented: (a) _The Determination of Water in Fresh Samples taken in situ._—The content of water in this case varies with the date and amount of rain-fall, the capacity of the soil for holding water, the temperature and degree of saturation of the atmosphere, and many other conditions, all of which should be noted at the time the samples are taken. (b) _The Determination of Water in Air-Dried Samples._—In this case the soil is allowed to remain in thin layers, and exposed to the air until it ceases to lose weight. The quantity of water left is dependent on the capacity of the soil to hold hygroscopic water and to the temperature and degree of saturation of the air. (c) _The Determination of the Total Water by Ignition._—This process not only gives the free and hygroscopic moisture, but also combined water present in the hydrous silicates and otherwise. The estimation is complicated by the presence of carbonates and organic matter. =294. Determination of Water in Fresh Samples.=—This determination requires that the sample, when taken in the field, should be so secured as to be weighed before any loss of moisture can take place. For this purpose it can be sealed up in tubes or bottles and preserved for examination in the laboratory. According to Whitney, the relations of soils to moisture and heat are such prominent factors in the distribution and development of agricultural crops, that the determination of the actual moisture content of soils in the fields should be considered a necessary part of the meteorological observations, and of far more importance, indeed, or having far more meaning to the agriculturist than the simple record of the rain-fall. In order to determine the relation of the soil to moisture, uninfluenced by the varying conditions of cultivation and of the different size of crop, he recommends that a small plot of ground be reserved at each station, adjacent to the soil thermometers, where the samples may be taken for the moisture determinations. No crops should be allowed to grow on this area and the soil is not to be disturbed, except that weeds and grass are carefully removed by hand when necessary. Samples of the soil should be taken every morning at 8 o’clock, by correspondents in the principal soil formations from the different parts of the area under observation, and sent by mail to the laboratory.