that locality. A considerable number of samples have been collected from other localities in this and in other States, but there has been no opportunity to make a thorough examination of these samples in time for this bulletin. A mechanical analysis has been made of the samples collected from Mr. Donaldson's place and the results certainly seem to explain some of the peculiarities mentioned in the description of these soils.

The mechanical analysis was made by Osborne's "beaker method," and the diameters, given in millimeters, show the range in size of the different grades of sand, silt, and clay into which the soil is divided in the course of the analysis. The most important group in the case of upland soils is the "clay " group, as the grains of clay are so extremely small that there is an enormous number of them. The number of grains in this group is so large that it practically determines the extent of subdivision of the space within the soil in which water is held. It has been found in the upland soils of South Carolina and of Maryland that, under the prevailing climatic conditions, a soil must contain at least 20 per cent of clay to be "strong enough" for a good wheat land. Soils containing less clay than this are excellent for the light-colored tobac coes and for fruit and early truck, but they are rather too light in texture for the economical production of wheat. The very finest wheat lands contain from 30 to 35 per cent of clay. Under the prevailing climatic conditions the soil must contain at least 25 or 30 per cent of clay for a good grass land, and soils containing less clay than this are too light in texture for a permanent sod of grass. Limestone soils, like the Cumberland Valley and the famous "blue grass lands," have from 40 to 50 per cent of clay.

These rice lands can not be compared in this respect with the upland soils, for the large amount of organic matter must greatly modify the simple structure of the soil grains, and besides the soil itself is not called upon to maintain a supply of water for the plant, although it must still provide good drainage and a sufficient circulation of air.

The following tables show the relative amounts of the different grades of sand, silt, and clay in these typical lowland rice lands, calculated on an air-dry basis and also on an organic and water-free basis. The latter results offer a better basis for comparing the structure of these soils with the structure of upland soils, and it will be seen that if the organic matter were removed, as it will be through oxidation in the course of time when the lands are higher and drier than they are at present, they would be exceedingly strong clay soils.

No analyses have been made of rice lands taken from farther up the rivers, but it is probable that such lands, being nearer the up-country where the sediment is derived, would have relatively more sand and silt and less clay than these soils taken from near the mouth of the river and just above the salt water. These soils are considered the very best and finest type of rice lands in the State.

(26) Big cypress field, "clay and alluvial mud."-When this sample was taken the water had been off the field for a short time and the land was being prepared for seeding. The soil was very soft and sticky. It could be readily cut like soft butter. This is considered the very finest type of rice land, and such soil as this can be cropped indefinitely without showing any effect of the continued cropping. It will stand more hard farming and the crop is more uniform and more reliable than on any other type of rice land. It will be seen from the tables that this has relatively more clay in the air-dry sample than the others, and it has only 18.68 per cent of loss on ignition, principally organic matter. When calculated upon an organic and water-free basis this soil contains about 55.22 per cent of clay, and in the normal condition of a well-drained upland soil it would contain about 50 per cent of clay and would have about the same texture as a heavy limestone soil.

(24) Cooter field. The water had been off of this field about twentyfour hours before the sample was taken, and only one tide had been over it then. The soil was black and sticky, and was full of roots

and stubble. The subsoil was a light yellow color; but this could not be sampled lower than 9 inches, as water rose freely into the hole at this depth. This soil is a typical rice soil, and is about as good as that from the "Big Cypress" field. It can be cropped indefinitely as the other can, but it is hardly as reliable in unfavorable seasons. It will be seen from the table that this soil has more clay when the analysis is calculated on the organic and water-free basis, but there is also a greater loss on ignition (24.32 per cent), and there is consequently less clay and more organic matter in the air-dry sample than in the last case.

(25) Sob field. This is a typical swamp bog or peat, and represents a large class of bay lands. As the name implies, it is soggy and shakes for a considerable distance around any pressure on the surface, as in walking. These lands are naturally poor, and are considered the very poorest kind of rice lands. They are exhausted in two or three years, and require rest to produce good crops. They respond readily to commercial fertilizers, and one year's rest with the deposit from the river in the continual ebb and flow of the tide gives splendid crops. It is not known to what the rapid deterioration of these lands is due. It can certainly not be due to an actual loss of plant food, but it must probably be due to some physical change in the structure of the land which interferes with the proper drainage or aëration of the soil, changes which have been brought about by the process of cultivation and which are corrected by the effect of fertilizers or the natural deposits from the river on the texture of the soil. It would be interesting to determine what changes do occur to account for the rapid deterioration of these lands, and this could readily be done. It will be seen from the tables that this soil gave 47.36 per cent of loss on ignition, indicating a very large amount of organic matter. The air-dried sample contained only 31.90 per cent of clay and only 52.15 per cent of total mineral matter. Calculated on an organic- and water-free basis, this soil contains over 61 per cent of clay. The very large amount of organic matter contained in this soil is evidently unfavorable to the growth of rice, and the very presence of so much organic matter probably indicates poor drainage and a very limited supply of air, which would otherwise have oxidized the organic matter, and these are conditions which we have seen to be unfavorable for the proper development of rice.

These three samples represent, it is believed, the most important types of lowland rice lands in South Carolina, representing the finest and poorest types of land in that important locality near the coast where it is considered the very finest rice lands are to be found. These analyses seem to show an important influence of the texture of these soils, and the relative amount of sand, clay, and organic matter which they contain, to the production of rice. If this line of investigation were followed out, and the structure of the rice soils from a number of localities was carefully studied, it would probably have an important practical bearing upon the selection, cultivation, and manuring of rice lands.

UPLAND RICE SOILS.

As stated in another part of this report, upland rice can be grown on any soil adapted to wheat or cotton, where the ordinary climatic conditions are favorable. The lands best adapted to cotton and wheat, under the meteorological conditions which prevail in South Carolina, have from 20 to 40 per cent of clay, and for both of these crops the land must be perfectly well drained and there must be free access of air to the roots of the plant. Rice will thrive with a more limited supply of air in the soil, and for obvious reasons upland rice is preferably grown on low, wet spots, where neither cotton nor wheat could be successfully grown. These areas are usually underlaid with an impervious clay, or for some other reason have insufficient drainage, and the soils are very constantly wet, although there is commonly no water standing over the surface. A limited supply of air favors the accumulation of a large amount of organic matter in the soil.

The following table gives the mechanical analysis of a typical soil of this kind from Lenoir County, N. C.:

It will be seen that there are 24.39 per cent and 34.64 per cent of loss on ignition, respectively, in the soil and subsoil, or about as much, on an average, as in the lowland rice soils of South Carolina. This indicates a large amount of organic matter in this soil, and this accumulation of organic matter indicates a very limited supply of air. The analysis based on the organic and water-free basis shows this land to contain about 40 and 49 per cent of clay, respectively, in the soil and subsoil, which would be regarded as a very strong clay land for other staple crops under better drainage and a more normal amount of organic matter.

The following table gives the mechanical analysis of a soil near Sumter, S. C., very similar to the above, but probably not so good for

This soil is shown to have about 17 per cent of loss on ignition and about 22.88 per cent of clay. A crop of maize had just been gathered from this land when the sample was taken. The crop had been raised on high beds or ridges similar to those used on the Sea Islands and partaking somewhat of the nature of the old Roman beds, and for the same purpose, to secure good drainage and a freer circulation of air in the soil. The ridges were 18 to 20 inches high from the bottom of the alley to the top of the ridge and about 5 feet across at the base of the ridge. The crop was planted on top of this ridge and it was in comparatively well-drained soil, but with an abundant and constant supply of moisture in the saturated subsoil below the ridge, and large elops of corn could be thus produced.