It is seen from the above analyses that the clay is by far the richest in mineral constituents, of all the ingredients separated in silt analysis, the amount in the clay being more than twice that of all the others combined. Its volatile matter is also the largest. The large amount of soda, however, is probably in part due to the sodium chlorid used in the precipitation of the diffused clay. The following points in regard to the distribution of the different ingredients are instructive:

1. The iron and alumina exist in almost identical relative proportions in each sediment, making it probable that they are in some way definitely correlated.

2. Potash and magnesia also exist in almost the same quantities, and their ratio to each other in all the sediments being almost constant seems to indicate that they occur combined, perhaps in some zeolitic silicate which may be a source of supply to plants.

3. Manganese exists only in the clay, a mere trace being found in the next sediment.

4. The lime appears to have disappeared in the clay, having probably been largely dissolved in the form of carbonate by the large quantity of water used in elutriation. Its increase in the coarser portions may be owing to its existence in a crystallized form not so readily soluble.

5. In a summary of the ingredients, it is seen that there is a loss in potash, magnesia and lime in the sediments as compared with the original soil; and this loss is doubtless due to the solution of these bodies in the water of elutriation.

A noteworthy fact shown in this table is the rapid decrease of acid-soluble matter in the coarser sediments; even what is dissolved from so fine a sediment as 1.0 millimeter hydraulic value, equal to a diameter of 0.04 millimeter, is in this case a negligible quantity. This suggests forcibly the inutility of introducing into chemical soil analysis, grains of as large a size as will pass a sieve of one millimeter aperture. The hydraulic value of these grains would be somewhere between 150 and 200 millimeters per second. While the exact results of the above analysis may not be applicable to all soils, yet the range is so wide that the systematic exclusion from chemical analysis of inert material, by means of preliminary mechanical separation, seems likely to lead to important improvements in the interpretation of the results.

241. Percentage of Silt Classes in Different Soils.—The adaptation of a soil to different crops depends largely on the sizes of the particles composing it and consequently on the relative percentages of the silt classes.

The following table gives the mechanical analysis of some markedly different types of subsoils:^[160]

242. Description of the Soils.—Number one represents the very early truck lands of southern Maryland. It is a light yellow sand, belonging to the Columbia terrace formation. Under an intense system of cultivation and heavy manuring with organic matter, good crops of garden vegetables are produced which mature very early, at least ten days or two weeks before the crops from any other part of the state. Under the prevailing meteorological and cultural conditions this soil maintains about five or six per cent of moisture, while a heavier wheat and grass soil maintains from twelve to twenty per cent. The truck soil is so loose and open in texture that the rain-fall passes through it very readily, and it is undoubtedly owing to this drier soil that the plant is forced to the early maturity which secures it from competition from other parts of the State and insures a good market price.

Number two represents the later truck and fruit lands of southern Maryland. These lands contain rather more clay than those just described; they are somewhat heavier and closer in texture, and are rather more retentive of moisture. This land gives a larger yield per acre than the one just described, and in every way crops make a more vigorous growth and development, but the crop is about a week or ten days later in maturing, and for this reason it brings a lower price in the market. It is much better land than number one for small fruit and peaches. These lands are altogether too light in texture for the profitable production of wheat, and it would cost altogether too much to improve them so that even a moderate yield of wheat could be obtained.

Number three is a tobacco land of southern Maryland. The finest tobacco lands of this locality come between the truck and wheat lands in texture, and contain from ten to twenty per cent of clay. The lighter the texture of the soil and the less clay it contains, the less tobacco it will yield per acre, but the finer the texture of the leaf. The tobacco yields more per acre on the heavier wheat soils, but the leaf is coarse and sappy and cures green and does not take on color. It brings a very low price in the market and does not pay for cultivation. The crop on the lighter lands is of much finer quality; there is a smaller yield per acre but the leaf takes on a fine color in curing, and brings a much better price per pound. Wheat is commonly raised on these tobacco lands to get advantage of the high manuring, and because the rotation is better for the land than where tobacco is grown continuously on the same soil. The finest tobacco lands are, however, too light in texture for the profitable production of wheat. These lands belong to the neocene formation.

Number four is a type of the wheat lands of southern Maryland. These lands represent soil of about the lightest texture upon which wheat can be economically produced under the climatic conditions which there prevail. They contain from eighteen to twenty-five per cent of clay, and are much more retentive of moisture than the best tobacco lands. This type is about the limit of profitable wheat production. These soils will maintain about twelve per cent of water during the dry season. Garden truck is so late in maturing on these lands that there is often a glut in the market when the crop matures, and the crops often do not pay the cost of transportation. The lands are too light in texture for a permanent grass sod. They belong to the neocene formation.

Number five represents the heavier wheat lands of southern Maryland, belonging probably to a different horizon of the neocene formation and containing about thirty per cent of clay. This soil is much more retentive of moisture and produces very much larger crops of wheat than the last sample. It is strong enough and sufficiently retentive of moisture to make good grass lands. It is too close in texture and too retentive of moisture for the production of a high grade of tobacco, or to be profitable for market truck.

Number six is from a heavy limestone soil of lower Helderberg formation. It is a strong and fertile wheat and grass land.

243. Interpretation of Silt Analysis.—The primary conceptions upon which the interpretation of the mechanical analysis is based may be briefly stated as follows:^[161] The circulation of water in the soil is due to gravity, or the weight of water, acting with a constant force to pull the water downward, and also to surface tension, or the contracting power of the free surface of water (water-air surface), which tends to move the water either up or down, or in any direction, according to circumstances. There is a large amount of space between the grains in all soils in which water may be held, ranging from about thirty per cent in light sandy lands to sixty-five or seventy per cent in stiff clay soils. The relative rate of movement of water through a given depth of soil will depend upon how much space there is in the soil; upon how much this space is divided up, i. e., upon how many grains there are per unit volume of soil; upon the arrangement of the grains of sand and clay; and upon how this skeleton structure is filled in and modified with organic matter. It also appears that the ordinary manures and fertilizers change this surface tension, or pulling power of water; that they also change the arrangement of the grains, and consequently the texture or structure of the soil may be changed and the relation of the soil to water, through the effect of the ordinary manures and fertilizers in causing flocculation or the reverse.

244. Number of Particles in a Given Weight of Soil.—The approximate number of particles in the soil can be calculated from the results of the mechanical analysis by the following formula:^[162]

Where a is the weight of each group of particles, d the mean diameter of the particles in the several groups in centimeters, ω is the specific gravity of the soil, and A is the total weight of soil. For the specific gravity of ordinary soils, the constant 2.65 may be used.

In using the formula the per cents are expressed as grams. Thus, if there were twenty per cent of silt, this would be taken as twenty grams, and if the results of the analysis added up ninety-seven per cent the whole weight of soil would be taken as ninety-seven grams. The diameter d is taken as the mean for the extreme diameters taken for any group, for instance, for the silt this would be 0.003 centimeter, which is assumed to be the diameter of the particles in that group. This formula can only give approximate values, as the number of separations in a silt analysis must necessarily be small, amounting usually to not more than eight or ten grades, on account of the time and labor required for closer separations. There is relatively rather a wide range in the diameters of grains within any one of these grades, and absolute values could not be expected without a vast number of separations, so that all the grains in each group would be almost exactly of the same size.

The clay group has relatively the widest limits, which is unfortunate, as this is the most important of all the groups on account of the exceedingly small size of the particles. The figure 0.0001 millimeter is taken as the lowest limit of the diameter of the clay particles. These particles have been heretofore assumed to be ultra-microscopic, but by the use of a microscope of high power with oil-immersion objective and staining fluids, it has been possible to define the clay particles in a turbid liquid which has stood so long as to be only faintly opalescent.

Pending more exact measurements, the figure 0.00255 millimeter has been used as the diameter of the average sized particle in the clay group.

The following table gives the approximate number of grains per gram in the different types of subsoils calculated from the mechanical analysis of the typical soils already given:

245. Estimation of the Surface Area of Soil Particles.—The approximate extent of surface area of the soil grains in one gram of soil can be calculated from the foregoing by the following formula:^[163]

in which d is the mean of the diameters of any group in centimeters, and n is the number of particles in the group.

The following table gives the approximate extent of surface area of the particles in one gram of soil calculated from the preceding table:

246. Logarithmic Constants.—The following logarithmic constants have been used in the calculation of the approximate number of grains per gram and of the surface area, using 2.65 in all cases as the specific gravity of the soil.

247. Mineralogical Examination of the Particles of Soil Obtained by Mechanical Analysis.—The principal object of the mechanical analysis of soils as has already been set forth is the separation of the soil into portions, the particles of which have the same hydraulic value. It is evident without illustration that particles of the same hydraulic value do not necessarily have the same size. The rate of flow of a liquid carrying certain definite particles does not imply that these particles are of the same dimensions. Of two particles of the same size and shape, that one which has the lower specific gravity, will be carried off at the lower rate of flow. At the end of the operation, therefore, the several portions of the soil obtained will be found composed of particles of sizes varying within certain limits, and of these particles the larger ones will tend to be composed of minerals of lower specific gravity, and the smaller ones of minerals of higher specific gravity. Of the same mineral substance, the particles which are most irregular, exposing for a given weight the largest surface will be found to pass over at a lower velocity than those of a more nearly spherical shape. The same law holds good for particles falling through a liquid at rest, i. e., the heavier and more spherical particles, weight for weight, will sooner reach the bottom of the containing vessel. To complete the value of a mechanical analysis, it becomes necessary to submit the several portions of soil obtained not only to a chemical but also to a mineralogical examination. Only the outlines of the methods of examining silt separates for mineral constituents can be given here and special works in petrography must be consulted for greater details.^[164]

It is evident that the methods of separation and examination from a mineralogical point of view about to be described can only be applied to silts of the largest size. The finer silts can not be separated into portions of different specific gravities by separating liquids of varying densities on account of the slowness with which they subside, thus tending to adhere to the sides of the separating vessels and to form floccules which are not all composed of the same kind of mineral particles. While, therefore, these processes are more appropriately described in connection with the silts obtained by hydraulic elutriation, they can be applied with greater success to the fine particles passing the different sieves used in the preparation of the soil for analysis or to the finely pulverized soil as a whole.

The minerals which have contributed to soil formation, moreover, are better preserved in the larger silt particles and therefore more easily identified. While the desirability of securing like determinations in the finer silts is not to be denied, in the present state of the art the analyst must be content with the examination of the larger particles.

248. Methods of Investigation.—The chief points to be observed in the examination of the fine particles of soil are the following: (1) the size and shape of the particles; (2) measurement of crystal angles; (3) separation into classes of approximately the same specific gravity; (4) separation by means of the magnet; (5) determination of color and transparency; (6) determination of refractive index; (7) examination with polarized light; (8) examination after coloring; (9) chemical separation. For many of the optical studies above noted, it is first necessary to prepare thin laminae of the mineral particles and properly mount them for examination. For the purposes of this manual only those processes will be described which are essentially connected with a proper understanding of the nature of the soil particles. For the more elaborate methods of research the analyst will consult the standard works on mineralogy and petrography.

249. Microscopical Examination.—The direct examination of the silt particles with the microscope should attend the progress of separation. Unless the particles obtained have the same general appearance, the separation is not properly carried on. Especially is the microscope useful to determine that the value of the silt separation is not impaired by flocculation. Unless flocculation be practically prevented during the separation of the finest particles, many of these will be left as aggregates to be brought over subsequently with particles of far different properties. No special directions are necessary in the use of the microscope. The silt particles are removed with a few drops of water by means of a pipette, a drop of the liquid with the suspended particles is placed on the glass, covered and examined with a convenient magnification. A micrometer scale should be employed in order that the approximate sizes of the particles may be determined. A camera lucida may also be conveniently used for the purpose of delineating the form of particles of peculiar interest.

250. Petrographic Microscope.—Any good microscope furnished with polarizing apparatus may be used for the examination of the silt particles and sections. For directions in manipulating microscopes the reader is referred to works on that subject. A special form of microscope for petrographic work is made by Bausch and Lomb of Rochester. The stand of this instrument is shown in Fig. 37. The base, upright pillars and arm are made of japanned iron. The stage is made in two forms, first, plain revolving, having silvered graduates at right angles and second, a mechanical stage with silvered graduations on the edge with vernier and graduations for the rectangular movements. The mirror bar is adjustable and graduated and the mirror is of large size, plane and concave. The double chambered box in the main tube carries the upper Nicol prism (analyzer). The lower Nicol prism (polarizer) is mounted in a cylindrical box beneath the stage to which it is held by a swinging arm. It is adjustable also up or down and is provided with a compound lens for securing converged polarized light. In revolving the prism a distinct click shows the position of the crossed Nicols.

251. Form and Dimensions of the Particles.—In order to study the contour of the fine silt particles, it is well to suspend them in a liquid whose refractive index is markedly lower than that of the particles themselves, and for this purpose pure water is commonly used. Care must be taken that not too many particles are found in the drop of water which is to be placed on the object holder and protected with a thin, even glass. The tendency to flocculation in these fine particles will make the study of their form difficult if they are allowed to come too close together. The size of the particles, or linear diameter, is to be determined by means of an eye-micrometer. This consists of a glass plate on which a millimeter scale is engraved with a diamond, or photographed. The millimeter scale is the one usually employed, each millimeter being divided into tenths. On microscopes designed especially for photographic work the micrometer is fastened to the eyepiece, and so adjusted as to read from left to right, or at right angles thereto. Sometimes an eyepiece-micrometer has two scales at right angles so that dimensions may be read in two directions without change. With an eyepiece-micrometer, not the dimensions of the object, but those of its magnified image are read, and the degree of magnification being known, the actual size of the object is easily calculated. The actual measurements may also be obtained by placing in the field of vision, a stage-micrometer and determining directly the relation between that and the eyepiece-scale. If, for example, the stage-micrometer is ruled to 0.01 millimeter, and the eye-micrometer to 0.1 millimeter, and one division of the stage-rule should cover three divisions of the eye-rule, then the one division of the eye-micrometer would correspond to an actual linear distance of 0.0033 millimeter in the object. If the two lines of division in the two micrometers do not fall absolutely together, the calculation may be made as follows: suppose that six divisions, 0.6 millimeter, in the eyepiece correspond to nearly five divisions, 0.25 millimeter, in the stage piece. To get at the exact comparison, take ninety-six divisions of the eye-scale and they will be found to be somewhat longer than eighty-one and somewhat shorter than eighty-two divisions of the stage-scale. It follows therefore that

and, hence, one division of the eye-scale corresponds almost exactly to 0.008489 linear measure.

252. Illustrations of Silt Classes.—In figure 38 are shown the relative sizes and usual forms of a series of silt separates made by the Osborne beaker method. The photomicrographs were made by Dr. G. L. Spencer from specimens furnished by Prof. M. Whitney.

The soil represented by the separates is from a truck farm near Norfolk, Virginia.

The particles represented in each class are not all strictly within the limits of size described. For instance, in the largest size (No. 1) are two particles at least which show a diameter of more than one millimeter. The particles in general, however, are within the limits of the class; viz., one-half to one millimeter, and this general observation is true of all the classes. In the case of the finer particles, especially of clay, the tendency to flocculation could not be overcome in the preparation of the slides for the photographic apparatus. The clay particles are so fine as to present but little more than a haze at 150 diameters of magnification. The particles seen are clearly, in most cases, aggregates of the finer clay particles. The larger particles show the rounded appearance due to attrition and weathering. It would have been more instructive to have had the particles of the different classes all photographed on the same scale, but this is manifestly impossible. The lowest power which shows any of the clay particles to advantage is at least 150 diameters, and with the larger particles such a magnification would have been impracticable.

253. Measurement of Crystal Angles.—The fine silt particles rarely retain sufficient crystalline shape to permit of the measurement of angles and the determination of crystalline form thereby. The rolling and attrition to which the silt particles have been subjected have, in most cases, given to the fragments rounded or irregular forms which render, even in the largest silts, the measurement of angles impossible. For the methods of mounting minute crystals and the measurement of microscopic angles, the analyst is referred to standard works on mineralogy and petrography.

254. Determination of the Refractive Index.—For a study of the theory of refraction, works on optics should be consulted. The general principles of this phenomenon which concern the determination of the refractive power of fine earth particles are as follows: if a transparent solid particle is observed in the microscope imbedded in a medium of approximately the same refractive power and color, its outlines will not be clearly defined, but the imbedded particle will show in all of its extent the highest possible translucency. If, therefore, the form or perimeter of the particle is to be studied with as much definiteness as possible, it should be held in a medium differing as widely from it as possible in refractive power. For minerals, water is usually the best immersion material. On the other hand, when the internal structure of the particles is the object of the examination, it should be imbedded in oil, resin (Canada balsam), etc., or in some of the liquids mentioned below.

If particles of different refractive powers and the same character of surface be studied in the same medium, they will not all appear equally smooth on the field of the microscope. Some of the surfaces will seem smooth and even, others will appear rough and wrinkled. Those particles whose refractive index is equal to or less than that of the liquid appear smooth, because all the emergent light therefrom can pass at once into the environing medium. On the other hand, the surfaces of those particles which have a higher refractive power than the medium will appear roughened, because, on account of the unavoidable irregularities on the surface, many of the emergent rays of light must strike at the critical angle and so suffer total reflection, and consequently those portions of the surface will be less illuminated, producing the phenomenon of apparent roughness above noted. In the case of any given particle, liquids of increasing refractive power can be successively applied until the change in the appearance of the surface of the particle is noticed. The refractive index of the liquid being known, that of the particle is in this way approximately to be determined.

The following liquids, having the indexes mentioned, are commonly employed:

The solution of potassium and mercuric iodid may also be used for all refractive indexes from 1.733 to 1.334 by proper dilution with water.

The mineral particle may also be imbedded in Canada balsam and over it a drop of a liquid of known refractive power placed. By a few trials one of the liquids will be found having practically the refractive index of the particle under examination.

255. Examination with Polarized Light.—The internal structure of a mineral particle can often be determined by its deportment with polarized light. The theory of polarization is fully set forth in works on optics and will not be discussed here. The principle on which the utility of polarized light in the examination of soil particles rests is found in the information it may give in respect of crystalline structure. The structure of mineral particles which make up the bulk of an ordinary soil is, as a rule, so thoroughly disintegrated that all trace of its original form is lost. Some particles may exist, however, in which there is no determinable element of shape and which yet possess an internal crystalline structure which the microscope with polarized light may be able to reveal.

256. Staining Silt Particles.—The finer silts and clays before microscopic examination should be colored or stained. The methods used in staining bacteria may be employed for the clay particles.

Evaporation to dryness with a solution of magenta will often impart a color to the clay particles which is not removed by subsequent suspension in water. The harder and larger silt particles are not easily stained, especially if they be firm and undecomposed. On the other hand, if the particles be broken and seamed, and well decomposed, the stain will be taken up and held firmly in the capillary fissures. Valuable indications are thus obtained respecting the nature of the silt particles. Particles of mica, chlorite and talc are easily distinguished in this way from the firmer and less decomposed quartz grains.

The staining of the particles after ignition and treatment with acids gives better results than the direct treatment. Particles of carbonate which are stained with difficulty before ignition take the stain easily afterwards on account of the decomposition produced by the loss of carbon dioxid. This is the case also with particles containing water of composition or crystallization.

257. Cleavage of Soil Particles.—A microscopic examination of the cleavage of soil particles may be useful in determining their mineral origin. The course followed by cleavage lines and their mutual position is dependent on the direction in which the separation of the mineral fragment takes place. The character of the microscopic fragments produced by crushing a soil particle is determined primarily by the system of crystallization to which it belongs. Perhaps the most distinguishing cleavage marks in soil particles will be found in fragments of mica and orthoclase. These characteristic forms are shown in Figs. 39 and 40. The first (Fig. 39) shows the pinacoidal cleavage in a fragment of mica. Fig. 40 illustrates the appearance of the cleavage lines in a fragment of orthoclase. Figs. 41 and 42 show the characteristic cleavage lines in fragments of epidote and titanite.

258. Microchemical Examination of Silt.—The methods of quantitative chemical examination of silts will be given in another part of this manual. Certain qualitative and microchemical tests, however, are useful in identifying silt particles. For instance, any soluble iron mineral will be detected, even in minute quantity, by the blue coloration of the solution produced by the addition of potassium ferrocyanid. Manganese will be revealed by fusion with soda and saltpeter on platinum foil, in the oxidizing flame, producing the well-known green coloration due to the sodium manganate formed.

More valuable indications of the character of the fragments examined are obtained by microchemical processes. The best method of decomposing the silt particles for this purpose is by treatment with hydrofluosilicic acid. When the particles are composed of silicates, pure hydrofluoric acid is to be preferred.

The method of treatment is essentially that of Boricky.^[165] The slide used is protected by a film of Canada balsam, and a few of the silt particles are placed thereon, and fixed in place by slightly warming the balsam. Each particle is then treated with a drop of hydrofluosilicic acid, care being taken not to let the drops flow together. The acid must be pure, leaving no residue on evaporation. The acid should be prepared by the analyst from a mixture of barium fluorid, sulfuric acid and quartz powder, or the commercial article should be purified by distillation before using. The acid should be kept in ceresin or gutta-percha bottles and must be applied with a ceresin or gutta-percha rod. Each particle should be as completely dissolved as possible by the acid, and the rate of solution may be hastened by gentle warming, provided the heat is not great enough to remove the balsam and allow the acid to attack the glass. The bases present in the silt particles crystallize on drying as fluosilicates. In case of a too rapid crystallization, the mass may be dissolved in a drop of water or of very dilute hydrofluosilicic acid, and allowed to evaporate more slowly. Some fragments need more than one treatment with acid to secure complete solution, and particles of mica may even resist repeated applications. In such a case the decomposition may be made in a platinum crucible with hydrofluoric acid, adding afterwards an excess of hydrofluosilicic acid and evaporating to dryness. The crystals may then be dissolved in a little water and a drop of the solution allowed to crystallize on the slide.

259. Special Reactions.—The number of microchemical reactions is very great, but there will be given here only some of the more important for silt identification.

Sodium.—Sodium mineral fragments dissolved in hydrofluosilicic acid and dried give the combinations shown in Fig. 43. With sodium and aluminum the forms shown in Figs. 44 and 45 are obtained. With an increasing amount of lime in the mineral, the crystals tend to become longer. For microscopic work it is not advisable to try to produce the tetrahedral crystals of the double uranium sodium acetate because the commercial uranium acetate often contains sodium and even the pure article will often take up sodium from the bottles.

Potassium.—Fragments containing potash give isotropic clear cubes, or octahedra of low refracting power, or combinations of these forms with each other and with rhombic dodecahedra. These crystals have the composition K₂SiF₆. Their forms are shown^[166] in Figs. 46 and 47. In case much sodium be present, the first crystals obtained may be strongly double refractive rhombohedra, but on dissolving in water and allowing to recrystallize, the normal forms will be obtained. If the crystals be dissolved in hydrochloric or sulfuric acids, and treated with platinum chlorid, the characteristic yellow octahedral crystals of K₂PtCl₆ will be obtained. Ammonium and cesium compounds also give this reaction.

Lithium.—When fragments containing lithium are treated with the solvent mentioned, monoclinic crystals are produced on drying. These crystals dissolved in sulfuric acid and freed from calcium sulfate by treatment with potassium carbonate give aggregates of lithium carbonate resembling a snowflake. At a high temperature lithium solutions treated with sodium phosphate give spindle-shaped crystals of lithium phosphate. The double lithium aluminum silicofluorid is shown in Fig. 48. The ease with which traces of lithium may be detected by the spectroscope renders unnecessary any further description of its microchemical reactions.

Calcium.—Nearly all mineral particles, save quartz grains, contain calcium. When these particles are dissolved by treatment with hydrofluosilicic acid, they form on drying hydrated monoclinic crystals of calcium silicofluorid (CaSiF₆ + 2H₂O). These crystals assume many forms, some of which are shown in Figs. 49 and 50. These crystals are easily decomposed by sulfuric acid, the well-known long prismatic crystals of gypsum taking their place. On treatment of silt particles containing lime with hydrofluoric and sulfuric acids, only a part of the lime passes into solution if the content thereof be large. Where but little lime is present and the sulfuric acid is in large excess, all the lime passes into solution and the characteristic gypsum crystals appear as in Fig. 51.

Magnesium.—Rhombohedral crystals of magnesium silicofluorid separate from the solution of particles containing magnesium in hydrofluosilicic acid. They have the composition MgSiF₆6H₂O and their common forms are shown in Fig. 52. Quite characteristic also are the crystals of struvite (NH₄MgPO₄ + 6H₂O), which are produced in a very dilute solution of the magnesium compound first obtained by carefully adding ammonium hydroxid and chlorid until a faint alkaline reaction is produced, and then placing a drop of dilute sodium phosphate at the edge of the solution. The crystals should be allowed to form slowly in the cold. Their form is shown in Fig. 54.