average yield of similar unfertilized land; and during the next thirty years—1882 to 1911—the corresponding average yields were 38 bushels an acre on the fertilized land, and 11.7 bushels where no plant food was applied. These statements are not mere opinions, but determined facts whose accuracy stands unquestioned.

On another field at Rothamsted, England, the average yield of barley for the same sixty years was 43 bushels an acre where nitrogen, phosphorus and calcium were regularly applied, 42.6 where all five elements—including potassium and magnesium—were added, but only 14.3 on unfertilized land.

On still another Rothamsted experiment field, where a four-year crop rotation of turnips, barley, clover (or beans) and wheat has been practiced since 1848, the yield of turnips in 1908 was 717 pounds an acre on unfertilized land and 35,168 pounds where the five important elements of plant food had been regularly applied once every four years—for the turnips only—since 1848. In 1909 the barley yielded 33.4 bushels an acre on the fertilized land, but only 10 bushels where no plant food was applied. The yield of clover in 1910 was 8590 pounds an acre on the land fertilized for turnips, but only 1949 on the unfertilized land. The wheat following the clover with no other fertilizer produced 24.5 bushels an acre in 1911, but 38 bushels where plant food is always applied for turnips grown three years before.

These are the established facts from the longest accurate record, and thus the most trustworthy data the world affords; and when one hears promulgated the very pleasing doctrine that the rotation of crops will maintain the fertility of the soil it is time to remember that "to err is human."

Fertility in Normal Soils

Of the four important mineral elements, potassium is by far the most abundant in common soils. Thus, as an average of ten residual soils from ten different geological formations in the eastern part of United States, two million pounds of subsurface soil were found to contain:

Even the depleted, and to some extent abandoned, gently undulating upland "Leonardtown loam," which was farmed for generations and which, according to the surveys of the Federal Bureau of Soils, covers 41 per cent of St. Mary's County, Maryland, and more than 45,000 acres of Prince George's County—still contains in two million pounds of surface soil—corresponding to the plowed soil of an acre about 6-2/3 inches deep:

The brown silt loam prairie soil of the early Wisconsin glaciation is the most common type of the greatest soil area in the Illinois Corn Belt. Two million pounds of this surface soil contain as an average:

The older gray silt loam prairie, the most extensive soil of Southern Illinois, contains in two million pounds of soil:

These data represent averages involving hundreds of soil analyses, and they emphasize the fact that normal soils are rich in potassium and poor in phosphorus. This is to be expected, for most soils are made from the earth's crust, and normal soils should bear some relation in composition to the average of the earth's crust, which contains in two million pounds 49,200 pounds of potassium and 2,200 pounds of phosphorus, as shown by the weighted averages of analyses involving about two thousand samples of representative rocks, reported by the United States Geological Survey.

Measuring Fertility Losses

The plant food required for one acre of wheat yielding 50 bushels, one acre each of corn and oats yielding 100 bushels, and one acre of clover yielding four tons, includes for the total crops:

If only the grain, including a yield of 4 bushels an acre of clover seed, is considered, the straw, stalks and hay being returned to the soil—either directly or in farm fertilizer—then the loss per acre from four years of cropping as above would be as follows:

The average annual loss by leaching from good soils in humid sections is known by the results of many analyses to be about as follows per acre:

The average annual loss of magnesium in drainage water from good soils is probably 30 pounds or more an acre, but the data thus far secured are inconclusive with respect to that element.

A careful consideration of the trustworthy data clearly reveals the fact that potassium is very abundant in normal soils, while phosphorus is relatively very deficient; and, all things considered, calcium—and probably magnesium—is of much greater significance than potassium, from the standpoint of the maintenance of usable plant food in the soil. It should be noted, too, that certain crops which are exceedingly important for economic systems of permanent agriculture require very large amounts of calcium as plant food. Thus a four-ton crop of clover hay takes about 120 pounds of calcium from the soil, or the same amount as of potassium; while such a crop of alfalfa requires about 145 pounds of calcium, but only 96 pounds of potassium. When it is known that the abandoned "Leonardtown loam" still contains in two million pounds of surface soil 18,500 pounds of potassium and only 1000 pounds of total calcium, the significance of these chemical and mathematical data must be apparent.

The Liberation of Fertility

Probably there has never been a greater waste of time and effort in the name of science than in the endeavor to determine the "available" plant food in soils. The almost universal assumption has been that the plant food in the soil exists in two distinct conditions, "available" and "unavailable," and that the determination of the "available" plant food would reveal both the crop-producing power of the soil and the fundamental fertilizer requirements for the improvement of the soil for crop production.

After ascertaining the total stock of plant food in the plowed soil, the next important question is not how much is "available," but rather how much can be made available during the crop season, year after year. In other words we must make plant food available by practical methods of liberation, by converting it from insoluble compounds into soluble and usable forms; for plant food must be in solution before the plant can take it from the soil. For the present, space is taken only to emphasize the value of decaying organic manures in the important matter of making plant food available; and attention is also called to the fact that the decomposition of the organic matter of the soil—including both fresh materials and old humus—is hastened by tillage and by underdrainage, which permit the oxygen of the air to enter the soil more freely, oxygen being a most active agent in nitrification and other decomposition processes of organic matter, as well as in the more common combustion of wood, coal, and so forth.