The sample of soil, twenty-five to thirty grams, after passing a two millimeters mesh sieve, and long boiling, is washed through a series of sieves of the following diameters of mesh respectively; viz., one millimeter, 0.5 millimeter, 0.25 millimeter, and 0.1 millimeter. The part which passes the finest sieve is placed in a Knop’s cylinder, the cylinder filled with water one decimeter above upper tube and well shaken. The cylinder is allowed to rest for five minutes when the upper cock is opened and the water drawn off. After five minutes more the next tube is opened and so on with equal intervals for the three upper tubes. The operation is repeated with fresh water until the water drawn off is clear. Finally the lowest tube is opened and all the water poured off of the sandy residue. The space between each tube is one decimeter. The dust remaining is dried and weighed and the weight of material carried over as silt determined by difference.

201. Siphon Silt Cylinder.—Instead of the tubulated cylinder one furnished with a siphon can be employed^[136] (Fig. 26).

It should be about forty centimeters high and six centimeters internal diameter. The cylinder first receives twenty-five to thirty grams of the well boiled fine earth and then water until there is but a small space between it and the stopper when the latter is inserted.

The cylinder is marked exactly 200 millimeters below the surface of the water with a narrow strip of paper at a, stoppered, inverted and well shaken. The cylinder being again placed in normal position the soil particles under the influence of gravity tend to sink with greater or less rapidity according to their size. The siphon a b c is filled with water, the cock at c closed and the opened end a placed in the cylinder A just at the mark 200 millimeters below the surface of the water, and the water thus transferred to B when desired. If the suspended matter is allowed to stand for 100 seconds the particles of more than two millimeters hydraulic value will have fallen below the open end of the siphon. If allowed to stand 1,000 seconds the silt value of the particles will be 0.2 millimeter per second. Whatever the number of seconds may be, the operation is continued until the water removed is practically clear. The open end of the siphon a should be bent upwards so that no disturbing current may bring the particles below the line into the liquid discharged into B.

While the results obtained by this method are satisfactory as compared with other similar processes it cannot be highly recommended because of the time and trouble required to get a complete separation and by reason of the difficulty of collecting the separated silt.

202. Wolff’s Method.^[137]—As modified by Wolff the Knop process is conducted as follows: Fifty grams of fine earth are boiled with water and then the entire mixture is passed through three sieves with openings of one millimeter, 0.5 millimeter, and 0.25 millimeter in diameter, respectively. The finest part is mixed with water to a height of eighteen centimeters in a bottle twenty centimeters high and having a capacity of one liter and thoroughly agitated, after which it is left to rest, and finally the turbid liquid is drawn off with a siphon, the bottle refilled with water, agitated, and left to rest, and the process repeated as long as the water carries any suspended matter after a definite time.

Wolff proposes for the first three periods of rest one hour, for the second three, a half an hour, for the third three, a quarter of an hour, and for the fourth three, five minutes.

203. Moore’s Modification of Knop’s Method.^[138]—The sample of soil is first passed through a sieve having round perforations three millimeters in diameter. The weight of the particles remaining on the sieve is then determined, and likewise that of the portion passing through, which is known as fine earth. The last named portion constitutes the material for all subsequent operations of mechanical and chemical analysis.

Thirty grams of the fine earth are boiled rapidly with water until the lumps are disintegrated and clayey portions separated from the sand. The material is then successively washed through perforated metal sieves, the holes of which are respectively 1, 0.5, and 0.25 millimeter in diameter. The portions retained on the sieves are severally dried, ignited and weighed, and the finest portion, or that passing through the 0.25 millimeter sieve, is then submitted to the following process of separation:

The sediment and water passing through the 0.25 millimeter sieve are placed in a glass cylinder fifty centimeters long and thirty-seven millimeters in internal diameter. The cylinder is closed at the bottom and is provided with a lateral tube inserted six centimeters above the bottom. Three other lateral tubes are inserted at intervals of ten centimeters above the first tube, and a ring is etched into the cylinder ten centimeters above the uppermost tube. The lateral tubes are closed with rubber tubes compressed by spring clips. The sediment being placed in the cylinder, water is added to the mark or ring, the cylinder closed with a rubber stopper, and vigorously shaken until the contents are thoroughly mixed. It is then placed upright, the stopper removed, and after standing undisturbed for five minutes the clip on the uppermost tube is opened and the water allowed to flow into a beaker. After five minutes further standing, the second clip is opened and the water drawn off into the same beaker; in the same manner the water is drawn off from the other tubes at intervals of five minutes until the level of the lowest tube is reached. The cylinder is then refilled with water to the mark, thoroughly shaken after inserting the stopper and the water again drawn off at intervals of five minutes, as before; the operation being repeated until the water drawn off is almost free from turbidity. The sediment remaining in the cylinder from this process of washing by subsidence is termed by Knop, fine sand, the material flowing off in suspension in the wash waters, dust, and the process of separation by Knop’s original method ends here.

In order to remedy the imperfect separation into definite particles secured by the above method, Moore proposes the following device:

The fine sand from the first series of subsidences is placed in a separate vessel, the washings are allowed to remain undisturbed for twelve hours, the turbid liquid decanted and the sediment returned to the cylinder. Water is then added to the mark, the whole shaken, and the liquid drawn off at intervals of five minutes, as in the first series. The sediment from this operation is placed in a separate beaker, the washings returned to the cylinder, and again allowed to subside as before; the sediment from this second subsidence is added to that from the preceding operation, and the washings again returned to the cylinder, the operation being repeated as long as any sediment can be obtained from renewed treatment of the washings; the final washings are then placed in a separate vessel for subsequent microscopic measurements.

The collective sediments from the last series of operations are then returned to the cylinder and allowed to subside with fresh additions of water, as in the case of the first series; the fine sand thus obtained being added to that from the first series, and the washings being collected in a large beaker. The latter are left at rest for twelve hours, and the sediment returned to the cylinder and treated as before until no further separation can be effected. The fine sand resulting from all of these operations is then dried, ignited and weighed; the weight of the portion removed by the washing being determined by difference, as it is, owing to its excessively slow rate of subsidence, found impracticable to collect it for direct weighing. The size of the particles of fine sand is then determined by micrometric measurement. Similar measurements are made on the material obtained by long subsidence from the washings from the foregoing operations. The average diameter of the largest particles should not exceed 0.01 millimeter.

204. Statement of Results.—The results of the analyses on three soils from the localities indicated in the table, and the method of stating them, are given in the following table:

205. Method of Bennigsen.—The silt flasks recommended by Bennigsen^[139] are shown in Fig. 27.

The glass flask b carries a long cylindrical neck a the upper part of which is graduated in cubic centimeters. Ten grams of the fine soil are shaken with water in the flask, the neck of which is closed with a rubber stopper. The flask is then inverted bringing the soil and water into the neck. The flask is hung up and sedimentation is assisted by imparting a pendulous motion to the neck for ten minutes. After an hour the soil particles have separated into a coarse layer below and a fine layer above. The relative volumes of the two layers are then read off in cubic centimeters. While this method may be useful in helping to form a speedy judgment concerning the character of a soil it can lay no claim to being an accurate method of silt separation.

206. Method of Gasparin.—The method of Gasparin only gives a very primitive separation of the various components of the earth according to their fineness. It is conducted as follows:

Ten grams of sifted earth are put into a beaker, water is added and strongly agitated; after five minutes the water is decanted into another vessel, the first vessel is filled anew with water, agitated, decanted, and this process is repeated until the liquid remains perfectly clear. Only two portions are weighed, i. e., the pebbles which remain in the sieve and the coarse sand which remains in the beaker; while the argillaceous portion drawn off with the water is determined by the difference.

207. The Italian Method.—The following modification of Gasparin’s process is practiced by the Italian chemists:^[140]

Twenty grams of earth are passed through a sieve having openings of one millimeter in diameter, then the sifted part is mixed with 100 cubic centimeters of water in a 200 cubic centimeters beaker and left to rest for some hours, then strongly agitated and after ten seconds the turbid liquid is poured into another vessel of half a liter capacity. This manipulation is repeated until the liquid is clear.

The decanted liquid is thoroughly agitated, then left to stand until the movement shall be completely arrested, after which the supernatant liquid is poured into another vessel holding two liters. To the residuum is added more water; it is agitated, decanted, and this process is repeated until the water is no longer turbid.

208. Method of Osborne.—In the foregoing paragraphs the methods of silt separation by subsidence as practiced in different countries have been outlined. The good points of the various methods are combined in the process as carried out by Osborne.^[141] The details of this method will be given with sufficient minuteness to make its practice possible by all analysts.

Selecting the Sample.—Several pounds of air-dried, fine earth are secured by passing the soil through a sieve, the holes of which are three millimeters in diameter.

Sifting.—Thirty grams of the above fine earth are stirred with 300 to 400 cubic centimeters of water and then thrown successively upon sieves with circular holes of 1, 0.5, and 0.25 millimeter diameter respectively. By means of successive additions of water and the use of a camel’s hair brush, all the fine material is made to pass through the sieves and these at the last are agitated under water in a shallow dish in such a way that the soil is immersed. The finest sieve should be well wet with water on its lower surface just before using. The finest particles which render the water turbid are easily washed through. The turbid water is kept separated from the clear water which comes off with the last portions that pass the sieves. The turbid water usually does not amount to more than one liter.

Elutriation.—The elutriation should be carried on so as to secure three grades of silt; the diameters of the particles ranging in the first grades from 0.25 to 0.05 millimeter, in the second grade from 0.05 to 0.01 millimeter, and in the third grade from 0.01 millimeter to the impalpable powder. The term sand is applied to the first grade, silt to the second, and dust to the third. After the turbid liquid from the sifting has stood a short time it is decanted from the sediment and after standing until a slight deposit is formed, is again decanted and the sediment examined with a microscope. If sand be present, the subsidence of the turbid liquid is continued until no more sand is deposited. As the sand subsides rapidly there is no difficulty in altogether freeing the liquid first decanted from this grade of particles. The sediment thus obtained contains all the sand, a part of the dust and much silt. As only dust and the finest silt render the water turbid the sediment is stirred a few times with a fresh quantity of water and decanted after standing long enough to let all the sand settle. When the water decanted is free from turbidity, the last portions of the soil passing through the sieve with clear water are added to the sediment and the decantations continued so as to remove most of the silt. When no more silt can be easily removed from the sediment without decanting sand, the decantations are made into a different vessel and the subsidences so timed as to remove as much of the silt as possible. By using a little care, at least three-quarters of the sand are thus obtained free from silt. The rest of the sand is mixed with the greater part of the silt which has been decanted into the second vessel. The size of the smallest particles in this vessel is determined with the microscope, to make sure that its contents are free from dust as they usually will be if, after settling for a few moments, they leave the water free from turbidity.

The soil is thus separated into three portions, one containing sand, one sand and silt, and the other silt, dust, and clay. The sand and silt are separated from each other by repeating the subsidences and decantations in the manner just described.

In this way there is removed from the sediment, on the one hand, a portion of silt free from sand and dust, and on the other hand a portion of sand free from silt. Thus is obtained a second intermediate portion consisting of sand and silt, but less in amount than the first and containing particles of diameters much more nearly approaching 0.05 millimeter. By repeating this process a few times, this intermediate portion will be reduced to particles whose diameters are very near 0.05 millimeter and which may be divided between sand and silt, according to judgment. The amount of this is usually very small. As soon as portions are separated, which the microscope shows to be pure sand or pure silt, they are added to the chief portions of these grades already obtained.

The same process is applied to the separation of silt from dust. When all the silt has been removed from the dust and clay, the turbid water containing the dust and clay is set aside and allowed to settle in a cylindrical vessel for twenty-four hours. The vessel is filled to a height of 200 millimeters. According to Hilgard, the separation of the dust from clay during a subsidence of twenty-four hours, will give results of sufficient accuracy, although the clay then remaining suspended will not be entirely free from measurable fine particles up to 0.001 or 0.002 millimeter diameter.

Small beakers and small quantities of distilled water are used at first for the decantations, as thus the duration of subsidence is less and more decantations can be made in a given time than when larger quantities of water are employed. Beakers of about 100 cubic centimeters capacity are convenient for the coarser grades, but it is necessary to use larger vessels for the fine sediments from which turbid water accumulates that cannot be thrown away, as may be done with the clear water, from which the coarse sediments settle out completely in a short time.

It is best to keep the amount of water as small as possible in working out the dust since loss is incurred in using too large quantities.

It is also necessary in most cases to subject the various fractions obtained during elutriation, to careful kneading with a soft rubber pestle so that the fine lumps of clay may be broken up and caused to remain suspended in the water. This treatment with the pestle should be done in such a way as to avoid as far as possible all grinding of the particles, the object being merely to pulverize the minute aggregations of clay and extremely fine particles which always form on drying a sample of soil after removing it from the ground.

Measurement of the Particles.—To determine the size of particles in suspension, a small glass tube is applied to the surface of the liquid in such a way as to take up a single drop which is transferred to a glass slide. This drop will contain the smallest particles in the liquid.

To obtain a sample of the coarsest particles the liquid is allowed to stand long enough to form a very slight sediment and a portion of this sediment is collected with a glass tube.

To determine the diameter of the particles in a sediment it is stirred vigorously with a little water and the pipette at once applied to the surface of the water. On decanting the greater part of the sediment, the large particles remain at the bottom of the beaker and may be easily examined.

Time.—The time required to make the separations, above described, is about two hours for each, so that an analysis including the sittings, is made in five or six hours, exclusive of the time necessary for collecting the dust and separating the clay, for which a subsidence of twenty-four hours is allowed.

Weighing the Sediments.—The sediments are prepared for weighing by allowing them to subside completely, decanting the clear water as far as possible, rinsing them into a weighed platinum dish and igniting. The dish is cooled in a desiccator.

Effect of Boiling.—The analyses show a very decided increase in the particles smaller than 0.01 millimeter diameter at the expense of coarser particles as the result of boiling. The surfaces of the coarser particles are seen to be polished and of a lighter color than those not boiled. The surfaces of the unboiled particles are coated with a film of fine material probably cemented to them by clay. When these coarse particles which have not been boiled, are violently stirred with water for a short time, no fine particles are detached from them; and a careful examination under the microscope fails to reveal in any of the sediments more than an occasional grain exceeding the 0.05 millimeter limit by so much as 0.01 millimeter, or the 0.01 limit by as much as 0.005 millimeter. It would, therefore, appear that these small particles thus set free by long boiling are really a part of the larger ones and should be treated as such in a mechanical analysis of these soils.

209. The French Method.—The Schloesing method^[142] as practiced by the French agricultural chemists^[143] differs essentially from those already described in attempting to first free the silt from carbonates and organic matter. It is conducted as follows:

One kilogram of the soil previously dried in the air, is taken and passed through a sieve of which the meshes are five millimeters. The agglomerated particles of earth are broken up by the hand. The pebbles are also taken out and weighed. The pebbles are then treated with hydrochloric acid until all effervescence is over. The insoluble part is dried and again weighed. The difference in weight gives the quantity of calcium carbonate contained on the external surface of the pebbles. The earth which passes the sieve of five millimeters mesh is next passed through a sieve having ten meshes to the centimeter. The masses on the sieve are broken up with the hand or with a pestle, in such a manner as to separate the fine agglomerated particles. The material which remains upon the sieve after being dried at 100°, is weighed. This gives the coarse sand. This is treated with hydrochloric acid as were the pebbles before, washed and the residue dried and weighed. The difference in weight gives the quantity of calcium carbonate adhering to the surface of the coarse sand.

The mechanical analysis is continued with the matter which has passed the sieve with ten meshes to the centimeter and which consists of the soil, properly so-called. Ten grams of this are taken, dried at 100° until no further loss takes place and the moisture thus determined. Another ten grams are taken and placed in a capsule with a flat bottom, and from nine to ten centimeters in diameter. This is moistened with a small quantity of water in such a way as to make a paste. This paste is rubbed with the finger in fifteen cubic centimeters of water. Ten seconds after the stirring is completed the supernatant liquid is poured into a precipitating jar of about 250 cubic centimeters capacity, taking great care not to allow any particles to pass over which have been deposited during that time. This operation is repeated in the same way waiting about ten seconds each time before decanting, until the decanted liquor is almost perfectly clear. In this way the particles of different fineness are separated. The decanted portions contain the fine sand and clay. The remaining portion contains the sand and particles of medium fineness. This last part is dried, being kept at 100° until it has a constant weight. It is afterwards treated with dilute nitric acid to dissolve the calcium carbonate. When the carbonate is abundant, it is sufficient to determine it by difference which is done by washing the material, drying and weighing. But when the proportion of carbonate is very small and in consequence when its exact determination acquires a greater importance, it is better to determine the lime directly. For this purpose the part soluble in dilute nitric acid is collected, treated with ammonia and acetic acid and precipitated with ammonium oxalate. Details of this operation will be given in another part of this manual.

In regard to the matter which is insoluble in nitric acid, it is composed chiefly of silica or silicates, and sometimes also of vegetable débris. The vegetable matter is determined by the incineration of the material which has been previously dried. The loss of weight gives the proportion of vegetable or organic débris contained in the soil and of combined water.

The portion which has been decanted, the volume of which should not exceed 500 cubic centimeters, is treated with nitric acid until effervescence ceases. It is then left to digest for some time, in order to permit the whole of the carbonate to dissolve. It is next thrown upon a smooth filter about one decimeter in diameter. After filtration it is washed to secure the complete elimination of the soluble lime salts. The lime is determined in the filtered liquid.

The insoluble portion contains the fine sand, the clay and humus bodies. In order to separate the three elements the precipitate which was received upon the filter, is rubbed with water, the filter is broken and all its contents washed through. The volume of wash water is made up to 200 cubic centimeters; two or three cubic centimeters of ammonia are added and the whole left to digest for two or three hours. The volume of the liquid is then made up to one liter with distilled water, vigorously shaking in such a way as to put all the matter in suspension. It is then left to settle for twenty-four hours. At the end of this time the supernatant liquid is decanted by the aid of a siphon. To the residue are added two cubic centimeters of ammonia and one liter of water. The matter is again brought into suspension and allowed to settle for twenty-four hours. The supernatant liquid is again decanted with a siphon, and added to the liquid previously removed. For ordinary soils two decantations are generally sufficient but when the soils contain a large quantity of clay it is convenient to decant three or four times. By an examination of the supernatant liquid it is easy to tell if the washings have been sufficiently prolonged. The decanted liquors contain the organic matter and that which it is convenient to call clay, which is constituted of very fine particles of sand and colloidal clay which play, in arable soil, a rôle somewhat like that of cement.

These matters are estimated in the following manner: The liquor is first treated with nitric acid and the clay and the humic matters are precipitated together. They are collected upon a smooth filter one decimeter in diameter and washed with water. By means of a washing bottle all the solid matters which have stuck to the sides of the filter are finally collected in the bottom of it. Since the last washings pass the filter very slowly, they can be removed after the complete deposition of the matter they contain, by means of a pipette. When all the liquid is removed the filter is placed upon blotting paper, great care being taken to avoid desiccation, having in view only the elimination of the excess of humidity. The folds in the filter are then carefully smoothed out with the finger. The matter which has collected upon the filter is then removed completely with a washing bottle, placed in a dish and dried at 100° and weighed. After weighing, the mass is incinerated in a muffle in order to destroy the humic bodies. The difference in weight before and after incineration, gives the total weight of the humic bodies and since the diminution in weight comprises not only the weight of the humic bodies, but also the weight of the combined water which is lost during the process of incineration, there should be subtracted from the total loss of weight ten per cent of the weight of the residual mineral matter, which represents the water of composition of the hydrated silicate.

210. Statement of the Analysis.—Schloesing in his original paper^[144] recommends that the analysis be commenced with 1,000 grams of soil. The data of the analysis and the method of arrangement are illustrated by the following example.

The physical examination of the earth having been completed as above, the results can be tabulated as follows: taken, 1,000 grams of dry earth, digested in water, thoroughly worked by hand, sifted, and passed through the meshes of the sieve by a stream of water, the meshes having a diameter of one millimeter.

Humidity of the homogeneous paste, twenty-seven per cent. Then 945 grams of the dry sifted earth correspond to 945
1.00 − .27 = 1294.5 of paste.

The analysis, therefore, should be carried on upon this weight or some aliquot part say 0.01 thereof; viz., 12.945 grams.

Treatment of the Fine Elements.—Treated by nitric acid until a complete decomposition of the calcareous matter is secured, filtered, washed, the residual matter collected upon a filter, and the liquid received in a two-liter flask, a little ammonia added, allowed to digest, the flask filled with distilled water, left for twenty-four hours at repose, and decanted:

Calculating these results to the original quantity of 1,000 grams the following data are obtained:

One thousand grams of dry earth contain:

There are counted as clay all the elements which have remained in suspension in the water after a period of repose of twenty-four hours. In fact, these elements comprise a notable proportion of very fine sand which is not deposited during that time. In order that the liquid should become entirely freed from this sand it would be necessary to wait several weeks and even several months. Such a prolongation of the analysis is evidently inadmissible. The period of twenty-four hours of repose therefore has been adopted. This is merely conventional, in the same way that the period of ten seconds adopted for the precipitation of the gravel is conventional. But this convention is justified by the fact that the substance which is called clay presents, when it has a proper degree of humidity and cohesion, a plasticity entirely analogous to that property of natural clay. Moreover, as has already been said, that which is chiefly important in these analyses is the employment of processes always comparable among themselves in their results and generally followed.

211. The Belgian Method.—The method of estimating the percentage of sand and clay practiced at the Gembloux Station^[145] is essentially that recommended by Schloesing with a few minor modifications.

With the ball of the thumb or with the finger, 100 grams of fine earth are rubbed with water in a porcelain capsule or mortar with a capacity of about 250 cubic centimeters. The suspended particles are poured off with the wash water and the process repeated five or six times, using in all about 200 cubic centimeters of water.

The water containing the sediment is rendered slightly acid (hydrochloric acid) adding the acid in minute particles with constant stirring for about an hour in order to dissolve all the carbonate and to separate the organic acids from the bases with which they are combined.

The liquid is allowed to remain at rest for five or six hours and a part of the liquor decanted to remove any supernatant particles of organic matter which may have passed the sieve in the original preparations of the sample. Filter through a smooth filter about twelve centimeters in diameter, wash until the chlorin has disappeared, and throw the filtrate away.

Break the filter paper over the vessel in which the soil was treated with hydrochloric acid and wash all the contents of the filter into this vessel with as little water as possible (about 100 cubic centimeters), add five cubic centimeters of strong ammonia water, allow to stand for three hours, shaking from time to time and with distilled water make the volume up to 250 cubic centimeters. Stir vigorously with a glass rod or spatula, take this out and wash any adhering particles back, leave at rest for twenty-four hours, siphon the turbid liquid into a two-liter vessel. Make the volume up again to 250 cubic centimeters and treat as above described and repeat the operation until the water becomes clear after standing for twenty-four hours. Usually eight or ten washings are necessary. Wash the residual sand into a weighed dish, evaporate to dryness, ignite and weigh. The weight obtained divided by the weight of the original sample gives the per cent of sand. The sand is separated by sieves of varying fineness into coarse, fine, and pulverulent sand.

Add to the ammoniacal liquor collected in the two-liter flask some powdered potassium chlorid (five grams per liter) to hasten the coagulation and rapid deposit of the clay.

After twenty-four hours siphon the clear liquor, collect the deposited clay in a smaller vessel, allow to remain at rest and decant as much of the clear liquor as possible. Pass through a plain tared filter about nine centimeters in diameter, dry at 150° and weigh the clay.

212. The Italian Method.—Schloesing’s method as carried out by the Italian chemists^[146] is as follows:

A kilo of earth dried in the air is passed through a sieve the threads of which are separated a distance of five millimeters; and with this the small pebbles are separated.

With another sieve having spaces of one millimeter, the coarse sand is separated. The pebbles and sand are dried, weighed, treated with hydrochloric acid and again weighed in order to find the quantity of calcareous matter contained in them. In ten grams of this fine earth the humidity is determined by drying at 100°.

Ten grams are mixed in a capsule with fifteen to twenty cubic centimeters of water and after eight to ten seconds the supernatant liquid is poured into a beaker having a capacity of 250 cubic centimeters. The same operation is repeated until there are contained in the beaker the fine sand and the clay, while the coarser sand remains in the capsule.

This last is then dried and weighed and the quantity of calcium carbonate determined by treating it with diluted nitric acid. By means of calcination the organic matter is determined. The liquid decanted in the beaker, the volume of which must not surpass 200 to 250 cubic centimeters, is treated with nitric acid, filtered after some time, washed and the calcium is directly determined by precipitating the solution with ammonium oxalate.

The part in the filter which contains the fine sand, the clay, and the humus material is mixed with water to a volume of about 200 cubic centimeters; there are then added to it two to three cubic centimeters of ammonia and after two or three hours it is diluted to a liter and strongly agitated.

After twenty-four hours of rest it is decanted and the residuum is treated a second time with diluted ammonia, decanting after twenty-four hours. Ordinarily these two treatments suffice, if, however, the earth is very argillaceous, this operation should be repeated three and even four times.

The clay which is found in the liquid suspended in colloidal form coagulates and is precipitated by adding thirty to forty cubic centimeters of a saturated solution of potassium chlorid, while the humus substance, under the influence of the ammonia remains dissolved.

Sestini found that the method of Schloesing was the only one which indicated exactly the quantity of clay in the soil. He modified this method by reducing the time of rest from twenty-four hours, as proposed by Schloesing, to only twelve hours, a reduction which in his opinion does not in the least impair the exactness of the method.

Sestini also proposes twelve treatments instead of six.

SEPARATION OF THE SOIL PARTICLES BY A LIQUID IN MOTION.

213. General Principles.—The laws, already discussed, applying to the subsidence of a solid particle in a liquid, are equally applicable to the separation of the particle by imparting a motion to the liquid at a given rate. If a solid particle subside in a given liquid at the rate of one millimeter per second it follows that this particle will remain at rest if the liquid be set in motion upward with a like velocity. If the velocity be greater the particle will be carried upward and eventually out of the containing vessel. Such a particle is said to have a hydraulic value of one millimeter per second. If there be a perfect separation of a soil into its constituent particles and no subsequent flocculation, all the particles of one millimeter hydraulic value and less will be separated by a current of the velocity mentioned.

The general principles on which the separation rests, therefore, are the securing of the proper granulation of the sample and the maintenance of a fixed velocity of the current until the separation is finished. The separation must be commenced with a period of subsidence so as to remove first of all the suspended clay or impalpable particles. The velocity can then be increased in a certain fixed ratio to secure a separation into particles of any required hydraulic value.

214. Nöbel’s Apparatus.—One of the earliest methods of separating the soil particles by a moving liquid is that of Nöbel.^[147] The apparatus is shown in Fig. 28. The four separating vessels 1, 2, 3, 4 are of glass, pear shaped, and have a relative capacity of 1³, 2³, 3³, 4³, or 1 : 8 : 27 : 64. No. 4 has an outlet tube leading to the beaker B, of such a capacity as to allow the passage of just nine liters of water in forty minutes, constant pressure being maintained by means of a Mariotte’s bottle or of the constant level apparatus A, a, b, which is connected with the main water supply through the tube a by means of a rubber hose. The reservoir C should hold about ten liters. The sample of soil to be separated should be previously boiled and passed through a sieve having circular openings one millimeter in diameter. The flask in which the sample is boiled is allowed to stand for some time when the muddy supernatant liquid is poured into elutriator No. 2 and the remaining sediment washed into No. 1. No. 1 is filled with water by connecting it with the water supply and opening the pinch-cock p. The water is carefully admitted until the air is all driven out and Nos. 1 and 2 connected. The cock p is then opened and the vessels all filled, and the water allowed to run into B for forty minutes, the level being maintained uniformly at A.