If the field-water capacity has not been filled, there is always the danger that an unusually dry season or a series of hot winds or other like circumstances may either seriously injure the crop or cause a complete failure. The dry-farmer should keep a surplus of moisture in the soil to be carried over from year to year, just as the wise business man maintains a sufficient working capital for the needs of his business. In fact, it is often safe to advise the prospective dry-farmer to plow his newly cleared or broken land carefully and then to grow no crop on it the first year, so that, when crop production begins, the soil will have stored in it an amount of water sufficient to carry a crop over periods of drouth. Especially in districts of very low rainfall is this practice to be recommended. In the Great Plains area, where the summer rains tempt the farmer to give less attention to the soil-moisture problem than in the dry districts with winter precipitation farther West, it is important that a fallow season be occasionally given the land to prevent the store of soil moisture from becoming dangerously low. To what extent is the rainfall stored in soils? What proportion of the actual amount of water falling upon the soil can be stored in the soil and carried over from season to season? This question naturally arises in view of the conclusion that water penetrates the soil to considerable depths. There is comparatively little available information with which to answer this question, because the great majority of students of soil moisture have concerned themselves wholly with the upper two, three, or four feet of soil. The results of such investigations are practically useless in answering this question. In humid regions it may be very satisfactory to confine soil-moisture investigations to the upper few feet; but in arid regions, where dry-farming is a living question, such a method leads to erroneous or incomplete conclusions. Since the average field capacity of soils for water is about 2.5 inches per foot, it follows that it is possible to store 25 inches of water in 10 feet of soil. This is from two to one and a half times one year's rainfall over the better dry-farming sections. Theoretically, therefore, there is no reason why the rainfall of one season or more could not be stored in the soil. Careful investigations have borne out this theory. Atkinson found, for example, at the Montana Station, that soil, which to a depth of 9 feet contained 7.7 per cent of moisture in the fall contained 11.5 per cent in the spring and, after carrying it through the summer by proper methods of cultivation, 11 per cent.