The weight P of the full flask having been determined, as much as possible of the solution, is decanted by means of a very small siphon, of which the flow is moderated by fixing a rubber tube with a pinch-cock to its lower extremity. After the decantation is complete, the flask is again weighed, giving the weight of P′; the weight of liquid taken, therefore, is equal to P − P′. To determine the total weight of the liquid, throw upon a filter the earthy residue insoluble in the acid, and after washing and drying it determine its weight r. The weight of the empty dry flask p is also determined. The total weight of the soil will be, therefore, P − r − p. The part of the liquid which was extracted from the flask, and upon which the analytical operation is to be conducted is represented by the fraction P − P′
P − r − p. This method avoids washing and evaporation which would be of very long duration. It rests upon the supposition that the solid matter from which the liquor is separated has no affinity for the dissolved substances, and that the total of these substances has passed into the liquor, and that the solution is homogeneous.

In the liquor first decanted as described before, the potash is estimated. This liquor contains in addition to potash, soda, lime, magnesia, iron and aluminum oxids, as well as phosphoric, sulfuric, and hydrochloric acids. There is first added to it a little barium chlorid to precipitate the sulfuric acid. It is then heated to about 40° in a glass flask and some ammonium carbonate added in a solution containing an excess of ammonium hydroxid. By this process the lime and baryta are precipitated in the form of carbonates; the alumina and iron as oxids, and the phosphoric acid in combination with the last two bases. The magnesium carbonate is not precipitated because it is soluble in the ammonium carbonate with which it forms a double salt.

The employment of a gentle heat favors the formation of the precipitate of calcium carbonate in a granular form which lends itself easily to filtration. The contents of the flask are now thrown upon a filter and the insoluble residue washed. The filtrate contains the potash, soda, magnesia, ammonia, and nitric and hydrochloric acids. It is concentrated as rapidly as possible by heating in a flask, and afterwards the ammoniacal salts are destroyed by weak aqua regia and the whole is then transferred to a porcelain dish and evaporated to dryness. There is thus obtained a mixture of potassium, sodium, and magnesium nitrates, from which the potash is separated by means of perchloric acid in the manner already described.

Estimation of the Total Potash.—Beside the potash which can be dissolved by the boiling concentrated acids the soil contains potash combined with silicates, which becomes useful for plant life with extreme slowness. It is often of great interest to estimate the total potash contained in a soil, that is to say, the reserve for the future. In this case it is necessary to free entirely this base from its combinations by means of hydrofluoric acid. The operation is conducted upon two grams of earth previously ignited and reduced to an impalpable powder. The decomposition is conducted in a platinum capsule by sprinkling the sample with a few cubic centimeters of hydrofluoric acid, or solution of ammonium fluorid, and adding a few drops of sulfuric acid. It is then evaporated to dryness and dissolved in boiling hydrochloric acid. The part which remains insoluble is treated a second time by hydrofluoric and sulfuric and afterwards by hydrochloric acid. All of the potash is thus brought into solution. The estimation of the potash, after having obtained it in a soluble state, is conducted in the manner previously described.

Estimation of the Potash as Platinochlorid.—Instead of estimating the potash as perchlorate it can also be transformed into platinochlorid. This process gives as good results as the preceding one but it is necessary in all cases, to separate the magnesia. After having treated the soil as indicated in the case of the estimation of the potash as perchlorate, the separation of the sulfuric and phosphoric acids, of alumina and iron, of magnesia, and the destruction of the ammoniacal salts in the manner already described, there are finally left the alkalies potash and soda in the form of carbonates. These are transformed into chlorids by adding hydrochloric acid; afterwards they are evaporated to dryness and the mixture of the two chlorids weighed in order to determine what quantity of platinum chlorid it is necessary to add, in order that it be in excess. The quantity of chlorid to be added is calculated so as to be in sufficient quantity to saturate the whole of the chlorids weighed, whether they may be composed wholly of sodium or potassium. In this way there is a certainty of having an excess of platinum. The solution of platinum chlorid used should contain in 100 cubic centimeters seventeen grams of platinum. Each cubic centimeter of this solution will be sufficient for a decigram of the sodium and potassium double chlorids. After the addition of the platinum chlorid the mixture is evaporated in a capsule with a flat bottom, on a water-bath. It is important that the temperature should not exceed 100°. If the temperature should go above this there would be a tendency to form some platinum subchlorids insoluble in alcohol.

The evaporation is continued until the contents of the dish are in a pasty condition and form a rather solid mass on cooling. It is necessary to avoid a complete desiccation. After cooling, the residue is taken up by alcohol of 95° strength. It is allowed to digest with alcohol of this strength for some time, after having been thoroughly mixed and shaken therewith in order to obtain a complete precipitation of the platinochlorid. This digestion should take place under a small bell-jar resting upon a piece of ground glass. The evaporation of the alcohol is thus prevented. The mass is then washed by means of alcohol of the same strength and the liquors decanted upon a small filter placed within another filter of identical weight, which serves as a tare for it on the balance. The washing is prolonged until the filtrate becomes colorless. All of the particles in the dish should be brought upon the filter by means of a hair-brush. The filters are now dried at a temperature not exceeding 95° and the platinochlorid received upon the interior filter is weighed. The precipitate may also be washed from the small filter into the capsule in which it was formed by means of a jet of alcohol. The alcohol is evaporated and the precipitate weighed in the capsule. The weighing should be made rapidly on account of the hygroscopicity of the material. The weight obtained multiplied by 0.193 gives the corresponding quantity of potash in the soil.

Purification of the Oxalic Acid.—The commercial oxalic acid used in separating the magnesia, often contains lime, magnesia, and potash. When this reagent is used in a sufficiently large quantity in the estimation of the above substances, it is indispensable to free it entirely from them. This is secured by submitting the oxalic acid to successive recrystallizations which are obtained by dissolving it in warm water, filtering and leaving to cool. The mother waters are thrown away. After two or three successive crystallizations the traces of potash and magnesia have disappeared and the oxalic acid obtained after ignition leaves no trace of residue.

The purification may also be secured in the following manner: At a temperature of 60° a saturated solution of oxalic acid is made; the liquid is decanted, carried to the boiling point and filtered. Five per cent of nitric acid are added and it is allowed to cool. The crystals which are deposited are collected upon a funnel in which a plug of cotton has been placed, and are washed with a little cold water.

Purity of the Ammonium Carbonate.—The ammonium carbonate employed should not leave any residue whatever after volatilization. In general, it may be said of all the reagents employed in analyses and especially of those employed in large quantities, that it is indispensable to be sure that they contain no traces of the substances which are to be estimated. The acids, ammonia, etc., should always be examined with this point in view.

Estimation of the Soda.—It is often of interest to estimate the soda in the soil, not that it is an element of any great fertility but rather because it is hurtful when in excess. It is determined in the residue obtained in the estimation of potash and is estimated by difference. The weight of the mixture of sodium and potassium chlorids being known when the potash is determined, the weight of its chlorid is to be deducted from the weight of the two chlorids and thus the direct weight of the sodium chlorid is obtained.

A better way is to make a direct estimation. The soda is found entirely dissolved in the alcoholic solution obtained by washing the potash salt as before described, for the separation of the potassium platinochlorid. This alcoholic liquor is evaporated to dryness on a water-bath, in a bohemian flask of about 100 cubic centimeters capacity. The residue obtained consists of sodium platinochlorid and a little platinum chlorid. There is now fitted to the bohemian flask a cork stopper carrying two tubes. The apparatus is placed upon a water-bath and kept at about 100°. Through the tube which reaches to the bottom of the bohemian flask, a current of pure hydrogen is passed. The hydrogen passes off through the second tube. The hydrogen completely reduces the salts of platinum. In order that the decomposition may go on more rapidly a few drops of water are added. When the whole mass in the flask has become black owing to the separation of the platinum, it is shaken, evaporated to dryness and hydrogen passed through a second time. This operation is repeated three or four times, being stopped when the water no longer shows a yellow color. There is then in the flask only a mixture of reduced platinum and sodium chlorid. No trace of sodium chlorid has been lost because the temperature has never exceeded 100°. The sodium chlorid is dissolved by washing with water and filtered. The liquor, which must be absolutely colorless, is evaporated to dryness in a platinum capsule and weighed. There is thus obtained the weight of the sodium chlorid. For verification, the sum of the weight of potassium chlorid calculated from the platinochlorid and the weight of the sodium chlorid should be equal to the initial weight of the mixture of the two chlorids.

351. Potash Methods of the German Experiment Stations.^[225]—a. To one volume of air-dried fine earth which is obtained by sifting through a three millimeter sieve, two volumes of twenty-five per cent hydrochloric acid are added, or more if the soil contains much carbonate. The acid is allowed to act with frequent stirring for forty-eight hours at room temperature.

b. To one volume of the soil, as above prepared, are added two volumes of hydrochloric acid and allowed to stand for three hours with frequent shaking, at the temperature of boiling water.

c. (Halle method.) One hundred grams of the fine earth are treated with 500 cubic centimeters of forty per cent hydrochloric acid, made up to one liter with water and allowed to stand for forty-eight hours with frequent shaking. After filtering, a large aliquot part of the filtrate is evaporated for the estimation of the potash. The evaporated residue is washed into a half-liter flask in which the sulfuric acid is precipitated with barium hydroxid the flask filled to the mark and an aliquot part of the filtrate in a half-liter flask, treated with ammonium carbonate, filtered and the potash estimated as platinochlorid by the usual method.

352. Method of Raulin for the Estimation of Potash in Soils.^[226]—The process rests upon the very feeble solubility in aqueous solution of potassium phosphomolybdate, while sodium, magnesium, calcium, iron, and aluminum phosphomolybdates are more or less soluble. The process does not require complicated separation and permits of the treatment of a small quantity of soil, since the weight of the phosphomolybdate obtained is equivalent to nineteen times that of the potash.

The reagent is prepared by dissolving 100 grams of crystallized ammonium molybdate in as little water as possible and adding six and a half grams of neutral crystallized ammonium phosphate dissolved in a little water. Aqua regia is now added cold and some ammonium phosphomolybdate is precipitated. The mixture is heated, adding a little aqua regia from time to time, until the solution of the precipitate is accomplished. The whole is then evaporated to dryness, the final temperature of evaporation not being carried above 70°. Four hundred cubic centimeters of water are now added and five cubic centimeters of nitric acid, and the contents of the dish heated and filtered. The reagent is then ready for use.

The liquid to be used for washing the potassium phosphomolybdate is prepared by dissolving twenty grams of sodium nitrate in one liter of water, two cubic centimeters of pure nitric acid, and a mixture of about twenty cubic centimeters of the phosphomolybdic reagent and one and a half cubic centimeters of a solution of potassium nitrate containing eighty grams per liter, slightly heated in order to saturate the liquid with potassium phosphomolybdate. The solution is shaken, allowed to rest, and the liquid decanted.

For the preparation of the solution in which the potash is to be estimated, a sample of soil is carefully weighed of such magnitude as to contain about fifteen milligrams of anhydrous potash. The potash salts are dissolved by the usual processes and are separated from the largest part of the calcium, iron, and aluminum salts, and converted into nitrates. The solution is reduced to a volume of a few cubic centimeters and slightly acidulated with nitric acid. Four cubic centimeters of the phosphomolybdic reagent are added for every ten milligrams of anhydrous potash supposed to be present. The solution is evaporated to dryness at 50° and immediately brought upon very small weighed filters, of which each one is double, by using sixty cubic centimeters of the washing liquor mentioned above. The tared filter is likewise washed with the same liquid at 50° and weighed. The weight multiplied by 0.052 gives the anhydrous potash. This method for a direct precipitation of the potash salts does not have the merits of the perchlorate process and both are inferior in accuracy to the usual platinochlorid procedure.

353. Russian Method for Estimating Potash in Soils.^[227]—Ten grams of the soil are digested with 100 cubic centimeters of ten per cent hydrochloric acid on a steam-bath for twenty-four hours. After adding five cubic centimeters of nitric acid to the filtrate it is evaporated to dryness, taken up with dilute hydrochloric acid, filtered, the filtrate saturated with ammonia, the excess of ammonia driven off, again filtered, and the lime separated by ammonium oxalate.

The filtrate is treated with a little barium chlorid for the removal of sulfuric acid and afterwards with ammonium carbonate in excess, and digested for twenty-four hours. After filtering, the solution is evaporated in a platinum dish, the excess of ammonia driven off, the residue taken up with water, filtered, treated with hydrochloric acid, evaporated to dryness, and ignited at low heat. The residue is again dissolved in water, filtered, and the potash precipitated with platinum chlorid and estimated in the usual way.

354. Potash Method of the Italian Stations.^[228]—The potash in the soil should be determined in three forms; viz.,

1. Assimilable potash.

2. Potash soluble in concentrated acid.

3. Total potash.

For determination of the first, 100 grams of earth are put into a retort holding a liter and digested with dilute nitric acid.

For the analysis, an aliquot portion of the clear liquid is taken or weighed, and the determination of the potash is made by the common methods.

For an alternate method, from twenty to fifty grams of earth are put into a retort of 500 cubic centimeters, moistened with water, and nitric acid is gradually added. After one or two hours there are added from 200 to 300 cubic centimeters of water; the liquid is poured without filtering into a retort and the residue washed by decantation.

In the liquid, after the elimination of the other substances with barium chlorid, ammonium carbonate, etc., the potash is determined by the ordinary methods.

In the second case, by using warm concentrated acids, a portion of the insoluble silica is decomposed, but this decomposition is always partial and the quantity of the potash extracted depends upon the temperature, upon the concentration, upon the duration of the action, and upon the nature of the acid.

The method of moistening twenty to fifty grams of earth with water and adding, thereto, concentrated nitric acid of 1.20 density, in such a manner that the earth shall be completely saturated, may also be employed. Then the temperature is kept at 100° during two hours. In the solution, the potash is determined as usual.

In the third case the soil is to be decomposed by hydrofluoric and sulfuric acids, or by fusion with alkaline carbonates, and the total potash determined by one of the standard methods.

If it is desired to adopt a general method for the determination of the potash the following points must be carefully considered:

1. The quantity of the earth to be examined.

2. The state of humidity or dryness of the same.

3. The quantity, nature, and concentration of the acid.

4. The quantity of the water.

5. The duration of the treatment.

355. Method of J. Lawrence Smith for Potash.—This method, designed especially for mineral analysis, has been fully approved by the general experience of analysts.

The principle of the method^[229] depends upon the decomposition of silicates on ignition with calcium carbonate and ammonium chlorid. The object of this mixture is to bring into contact with the mineral, caustic lime in a nascent state at a red heat, the caustic lime being soluble to some extent in calcium chlorid at a high temperature. Pure calcium carbonate, made by precipitation of marble, should be used.

The ammonium chlorid should be prepared by taking crystals of pure, sublimed sal ammoniac, dissolving in water, and filtering, and evaporating the solution until small crystals are deposited, the solution being well-stirred until one-half or two-thirds of the whole has crystallized. The mother-liquor is poured off while still hot, and the crystals dried on an asbestos filter at ordinary room temperature.

A special platinum crucible should be used in the Smith method, but the common crucible, especially if very deep, can be employed. The special crucible is of about double the usual length. Smith recommends a crucible ninety-five millimeters in length, diameter at top twenty-two millimeters, at bottom sixteen millimeters, and weighing thirty-five to forty grams. The object of the long crucible is to have the part of the bottom containing the silicate subjected to a high heat, while the top of the crucible is at a much lower temperature, thus preventing the loss of alkalies by volatilization.

Method of Analysis.—The samples of soil or silicate containing the alkalies are well pulverized in an agate mortar, and from one-half to one gram of the finely pulverized material taken for analysis. This is carefully mixed with the same weight of finely powdered sal ammoniac and the mineral and sal ammoniac rubbed well together in a mortar. Eight parts by weight of calcium carbonate are next added in three or four portions, and the whole intimately mixed after each addition. The contents of the mortar are emptied on a piece of glazed paper and then introduced into the crucible, which is tapped gently upon the table until the contents are well settled. It is then fixed in the furnace which is used for heating, and a small bunsen burner is placed beneath the crucible, and the heat applied just about at the top of the mixture and gradually carried toward the lower part until the sal ammoniac is completely decomposed, which requires from four to five minutes. The heat is then applied by means of a blast-lamp and the crucible kept at a bright red heat for from forty to sixty minutes. The crucible is allowed to cool, the contents detached and placed in a platinum or porcelain dish of about 150 cubic centimeters capacity, and sixty to eighty cubic centimeters of distilled water added. The solution of the flux may be hastened by heating the water to the boiling point. The crucible and its cover are also well washed with hot water until all matter adhering to them is dissolved. After the slaking of the mass it is best to continue the digestion with hot water for six or eight hours, although this is not absolutely necessary. The contents of the crucible are filtered and washed well with about 200 cubic centimeters of water. The filtrate contains in solution all the alkalies of the mineral, or soil, together with calcium chlorid and caustic lime. A solution of pure ammonium carbonate containing about one and one-half grams of the pure salt is added to the filtrate. This precipitates the lime as carbonate. The dish containing the material is placed on a water-bath and its contents evaporated to about forty cubic centimeters. Two additional drops of ammonium carbonate are added, and a few drops of caustic ammonia, to precipitate any lime which may be redissolved by the action of the ammonium chlorid solution on the calcium carbonate. Filter on a small filter and wash with as little water as possible and collect the filtrate in a small beaker. The filtrate contains all the alkalies as chlorids, together with a little ammonium chlorid. Add a drop of ammonium carbonate solution to be sure all the lime is precipitated, evaporate on a water-bath in a deep platinum dish, in which the alkalies are to be weighed. The dish should have from thirty to sixty cubic centimeters capacity, and during the evaporation should never be more than two-thirds filled. After the evaporation has been completed the dish is slowly heated and then gently ignited over a gas-flame to drive off any ammonium chlorid which may be present. During this process the platinum dish may be covered with a thin piece of platinum to prevent any possible loss by the spitting of the salt after the ammonium chlorid has been driven off. The heat should be gradually increased until it is brought to a point a little below redness, leaving the cover off. The platinum dish is again covered, and when sufficiently cooled placed on a balance and weighed.

If lithium chlorid be present it is necessary to weigh it quickly as the salt being very deliquescent takes up moisture rapidly. The alkalies may now be separated in the usual way.

If the sample under examination contains magnesia the residue in the capsule should be dissolved in a little water and sufficient pure lime-water added to render the solution alkaline. It should then be boiled and filtered. The magnesia will, in this way, be completely separated from the alkalies. The solution which has passed through the filter is treated with ammonium carbonate in the manner first described, and the process continued and completed as above mentioned.

If it be suspected that the whole of the alkalies have not been obtained by the first fusion, the residue upon the filter can be rubbed up in a mortar with an amount of ammonium chlorid equal to one-half the weight of the mineral, mixed with fresh portions of calcium carbonate and treated exactly as in the first instance. Any trace of alkali remaining from the first fusion is thus recovered in the second one.

Method of Heating the Crucible.—The apparatus used by Smith for igniting the crucible is shown in Fig. 67. It consists of an iron filter-stand HG, a clamp, ED, carrying the muffle NC, attached by the supports AB, and heated by the lamp F. The muffle NC is a chimney of sheet iron, eight to nine centimeters long, ten centimeters high, the width at the bottom being about four centimeters on one side and three centimeters on the other. It is made with the sides straight for about four centimeters and then inclining toward the top so as to leave the opening at the top about one centimeter in width. A piece is cut out of the front of the chimney of the width of the diameter of the hole in the iron support and about four centimeters in length, being semi-circular at the top, fitting over the platinum crucible. Just above this part of the chimney, is riveted a piece of sheet iron in the form of a flattened hook, N, which holds the chimney in place by being slipped over the top of the crucible support; it serves as a protection to the crucible against the cooling effects of the currents of air.

356. International Method for Assimilable and Total Potash.—In the International Congress of Chemists held in Paris in 1889,^[230] the discrimination between the assimilable and total potash was declared to be of prime importance. Unfortunately no method is known by which the potash which is present in the soil in a state suited to the wants of plants can be determined with approximate accuracy. In general, that portion which is given up to weak acids may be assumed to be available. In the treatment of soils with weak acid, as pointed out in the Congress, it is demonstrable that with a 0.05 to 0.1 per cent nitric acid solution, the quantity of potash which goes into solution increases by continued stirring of the mixture with the time of action of the acid up to a certain maximum which is reached in from three to four hours, and after that, it is not changed even when the strength of the acid mixture is increased to two per cent. From this time on, concentrated acids withdraw from the soil which has already been exhausted by the weak acid, a new quantity of potash. The soils which have been exhausted by concentrated acids yield also an additional quantity of potash when they are treated with hydrofluoric acid, or melted with barium or sodium carbonate. Potash, therefore, appears to exist in the soil in various forms.

First. In the form of indecomposable silicates which have, agriculturally perhaps, very little interest.

Second. In the form of silicates which are more basic than those just mentioned. These silicates are attacked by strong acids and give up probably every year a portion of their potash to vegetation.

Third. In a form which is easily soluble in weak acids and consequently directly assimilable by plants.

In view of the fact that it would be of interest to chemists and agronomists to establish certain methods of investigation so as to be able to obtain comparative results, it was decided to adopt the original method recommended by Gasparin for the estimation of the potash decomposable by concentrated acids. This method consists in the treatment of the soil with boiling aqua regia until the sand which is not decomposed, is white.

Determination of the Fineness of the Earth which is Used for Analysis.—For the estimation of potash, the soil should be divided as finely as possible, and passed through a sieve of thirty meshes to the centimeter. The decomposition is then completed in two hours, while if a sieve of only ten perforations per centimeter is used, the acid must be allowed to work for twelve hours.

The determination of the potash after solution, is accomplished by any of the standard methods.

357. Method of Tatlock as Used by Dyer.—Attention was called in paragraph 328 to the estimation of the total plant food in the soil by extraction of the sample with citric acid. Dyer first determines the total potash by Tatlock’s method which is as follows:

To determine potash ten grams of fine dry soil are treated with ten cubic centimeters of hydrochloric acid and evaporated to dryness on the water-bath, the residue taken up with another ten cubic centimeters of acid, warmed, diluted with water, boiled, filtered, and washed. The filtrate and washings are concentrated and gently incinerated to get rid of organic matter, and the residue redissolved in hydrochloric acid, and evaporated slowly with a considerable quantity of platinum chlorid. If the evaporation be conducted slowly, the potassium platinochlorid settles out well, despite the iron, aluminum, and calcium salts, and is easily washed with some more platinum chlorid solution, followed by alcohol. The application of this modification of the platinum chlorid process to solutions containing comparatively minute quantities of potash amid an overwhelming excess of iron, aluminum, and calcium salts is probably new to many chemists. It works admirably, and obviates the necessity for removing iron, aluminum, calcium, magnesium, etc., with the necessary use of ammonia, and the tedious processes of concentration and final volatilization of the ammonium salts; but, of course, the process cannot be employed if soda also is to be determined.

The potash, soluble in hydrochloric acid, having been thus determined, the undissolved siliceous matter is incinerated, weighed, and finely ground in an agate mortar. A weighed portion of it is then, as in the Smith method, mixed with a large bulk of pure calcium carbonate and a little ammonium chlorid and heated, beginning with a low temperature, rising slowly to bright redness. The mass is then boiled with water, washed, incinerated, reground, mixed with some more ammonium chlorid, and again heated, boiled, and washed. The process is repeated and the filtrates from all the treatments concentrated, the calcium being removed as carbonate, and the potash determined in the filtrate, after evaporation and incineration at a low temperature, by means of platinum chlorid.

Five hundred cubic centimeters of the citric acid solution of the soil, made as described in paragraph 328, corresponding to fifty grams of soil, are evaporated to dryness in a platinum dish and ignited at a low temperature. The residue is dissolved in hydrochloric acid filtered and washed, and the filtrate again evaporated to dryness and treated again as just described. The potash is then determined as above.

358. Estimation of Total Alkalies and Alkaline Earths.—To properly determine the exact amount of these substances in a sample of soil it is necessary first to remove the silica. This is accomplished in the process of Berthelot and André^[231] by intimately incorporating with the sample, in a state of very fine powder, four or five times its weight of ammonium fluorid. The mixture, in a platinum dish, is moistened with strong sulfuric acid and allowed to stand for a few hours. It is then gently heated until all fumes of hydrofluosilicic acid have disappeared, but the mass is not raised to a red heat. If there is any doubt about the complete decomposition of the silica the treatment is repeated.

At the end of the operation there remain only sulfates without excess of sulfuric acid. The sulfates likely to be present are of potash, soda, lime, magnesia, alumina, and iron. The separation of these bodies is conducted in the ordinary manner.

Fusing the soil with potash does not give reliable results but it can be used in certain cases for the rapid estimation of alumina and iron. In this case after the separation of the silica in the ordinary way the iron can be determined as ferric oxid.

The iron can also be directly determined by reducing to the ferrous state and titrating with potassium permanganate.

Comparison of Fluorin Method with Common Methods.—To establish the difference in the data obtained by the old and new processes samples of the same earth were treated by Berthelot and André by different methods with the following results:

The impossibility of getting all the alkalies and oxids into solution by even the prolonged action of a boiling acid is clearly set forth in the above table. Boiling sulfuric acid might do a little better but would not give correct results. Lime alone of the elements in the soil can be correctly determined by solution in boiling hydrochloric acid, a circumstance due to the fact that lime is found chiefly as carbonate, sulfate, and phosphate in the soil, and these compounds are easily soluble in hot hydrochloric acid with the exception of the sulfate. Even lime could not be thus determined in soils containing silicates rich in lime. The other mineral elements cannot be determined by the wet method. This is due to the forms in which they occur, being mostly silicates of different composition, with excess of silica.

As to the silicates they may be divided into two groups. The first of these are the hydrated silicates, resembling the zeolites, capable of being completely decomposed by boiling acids. The first group of silicates is doubtless of greater importance to vegetable life than the second since it would, doubtless, give up its alkalies with greater ease. This distinction is, however, arbitrary. It is, in fact, impossible to place on one side the soluble and on the other the insoluble silicates. This distinction represents only the unequal degrees in the speed of decomposition of the different silicates contained in the primitive rocks under the influence of atmospheric agents, the soil being nothing more than the products of the decomposition of these rocks with vegetable mold. The second group is insoluble in acids.

That part of the silicates least decomposed at any given moment will be attacked more easily by acids, while that portion whose decomposition has been pushed furthest will be more slowly attacked. The action of the acid will grow more feeble as the time of contact is prolonged, and after a time a point is apparently reached where the results are nearly constant. But it is evident that this distinction is purely conventional and bears no necessary or even probable connection with the quantity of alkali really assimilable by plants.

Vegetables, moreover, exert on a soil, for the extraction of its alkalies and other matters, chemical reactions peculiar to themselves, altogether distinct from the tardy action of atmospheric agents and still more distinct from the rapid action of mineral acids.

It is well known with what energy, it ought to be said with what admirable instinct, plants take from the soil the least traces of phosphorus, of sulfur, of potash, of iron, and other substances necessary to their sustenance.

These specific actions of vegetables on the soil merit, in the highest degree, the attention of analysts and agronomists. Their intervention plays a most important part in the restitution to the soil, by means of complementary fertilizers, the mineral elements removed by vegetable growth.

359. Estimation of Lime by the French Method.—The quantity of lime contained in the soil varies within wide limits. Sometimes this base is entirely absent to such a degree that it is even impossible to discover feeble traces of it. Sometimes it composes almost the whole of the earthy mass. Lime is found in the soil principally in the state of carbonate. It is also found combined with organic matter under the form of humates, with sulfuric acid, etc. It is customary to estimate the lime as a whole, without distinguishing between the different states in which it exists. The quantity of material which is used in the French method^[232] varies in proportion to the amount of calcareous matter contained in it. For a soil which contains a large amount of lime, one or two grams would be sufficient for the analysis. For a soil which is poor in calcareous matter ten or even twenty grams must be taken. The quantity of lime dissolved differs according to the strength of the acids employed and length of contact of the acid with the soil. The calcium carbonate, the sulfate, the nitrate, and the humate rapidly pass into solution when treated with acid as above, but this is not the case with calcium silicates which are attacked much more slowly. Sometimes the silicates give only an insignificant increase in the amount of lime, and in this case it is immaterial what process of solution is employed. For simplicity it is best to adopt the method of solution in boiling concentrated nitric acid, prolonging the boiling for a period of five hours. This method of operation is sufficient to bring into solution at one treatment, not only the lime, but also the potash and magnesia. After having heated with acid for the necessary time there are added in the capsule in which the solution took place ten cubic centimeters of nitric acid and fifty cubic centimeters of water. The mixture is heated, collected upon a filter and the residue washed. To the filtrate, the volume of which should be from 400 to 500 cubic centimeters, a sufficient quantity of ammonia is added to render it slightly alkaline. There is formed a precipitate of alumina and of iron oxid containing phosphoric acid and also sometimes a trace of the lime combined with the same acid. In order to keep the whole of the lime in solution it is necessary to add a little acetic acid, about ten cubic centimeters more than is necessary to neutralize the ammonia which has been added in excess. If the liquid is turbid on account of the presence of the iron and aluminum phosphates it is necessary to filter it. There is afterwards added a slight excess of ammonium oxalate in solution, and the whole is left for twenty-four hours in order that the calcium oxalate may deposit. Indeed, the complete precipitation is not always immediate, and especially in the presence of magnesia it takes place with slowness. The calcium oxalate is collected upon a filter and washed with hot water. To determine the quantity of the lime the best procedure consists in transforming the oxalate into carbonate by a careful ignition, and afterwards heating in a Schloesing or Leclerc furnace for four or five minutes. The oxalate for this purpose should be contained in a covered platinum crucible. By this method the calcium carbonate is transformed into calcium oxid, in which form it is weighed rapidly to avoid absorption of moisture.

In laboratories which have no means of securing so high a temperature as is mentioned before, the lime may be weighed as sulfate. For this purpose the calcium oxalate is transformed into carbonate by ignition in a platinum crucible. Afterwards it is treated with nitric acid until the carbon dioxid is completely driven off. The platinum crucible is now covered with a funnel which is afterwards washed in order to bring back into the dish the small drops which have been projected in the process of boiling. An excess of sulfuric acid is added and evaporated to dryness on a sand-bath. Afterwards, in a muffle, the temperature is carried to a feeble redness until the vapors of sulfuric acid are all driven off. The lime is weighed in the form of sulfate, and the weight multiplied by 0.412 gives the lime contained in the quantity of earth analyzed.

In special researches in which it is desired to avoid attacking the siliceous pebbles of the soil, the concentrated nitric acid is replaced by dilute nitric acid in slight excess, and heated for a few moments only. The calcium carbonate is then dissolved with the other calcareous salts not combined with silica in the rock products. The analysis is continued in other respects as just described.

360. Estimation of the Actual Calcium Carbonate.—The lime which is found in the state of carbonate plays one of the most important rôles in the chemical phenomena which take place in the soil. It is often of great importance to determine it. The most certain process is to estimate the carbon dioxid which is disengaged from the carbonate under the influence of an acid and to receive this gas in a jar graduated to measure it by volume. The flask recommended by the French Commission for this purpose contains about 300 cubic centimeters. The neck of the flask is connected with a condensing tube of about one centimeter interior diameter, which is cooled by a current of water.

According to the presumed richness in calcium carbonate varying quantities of earth are taken for analysis, from as little as half a gram for soils which are rich in carbonate, up to five or even ten grams for soils which are poor in carbonate. The apparatus is connected with a mercury pump for the purpose of exhausting the air as completely as possible therefrom. For this purpose the flask in which the carbonate is disengaged is made in the shape of a tubulated retort. Through the opening into the retort, a narrow tube is introduced and connected with a small funnel by means of a rubber tube supplied with a pinch-cock. When the retort has been connected with the mercury pump a slight vacuum is produced and the pinch-cock is opened and forty cubic centimeters of distilled water allowed to enter. The pinch-cock is closed soon enough to retain a portion of the water in the funnel. The retort is then heated and a vacuum partially produced by means of the pump. When the flask is boiling, the steam drives out the air. A refrigerating jacket is connected with the tube leading from the retort to the pump by means of which the steam is condensed and falls back into the flask. After some minutes of boiling, a vacuum is produced; the lamp is then taken away and a cylinder, graduated at 100 cubic centimeters and filled with mercury, is placed over the lower orifice of the pump, and there is introduced into the apparatus, by the funnel above described, some hydrochloric acid in small quantities, but sufficient only to saturate the whole of the carbonate in the sample of soil taken. Usually three or four cubic centimeters will be sufficient. The acid should be added in such quantities as to prevent the production of any large amount of foam. If frothing should be excessive a little oil can be added to the flask. The whole of the carbon dioxid produced in the reaction is withdrawn by means of the mercury pump and collected in the graduated jar. Towards the end of the operation the flask is heated anew in order to produce an ebullition which is continued for some time. The volume of gas collected is measured after making the proper corrections for pressure and temperature. Afterwards the carbon dioxid which has been produced is absorbed by two or three cubic centimeters of a solution of potash of 42° baumé. This potash is introduced into the graduated jar by means of a pipette bent into the form of a ᥩ in the lower portion. If the whole of the gas is not absorbed the volume which remains is read, and this is subtracted from the original volume after having made the proper corrections for pressure and temperature. The difference gives the quantity of carbon dioxid contained in the amount of earth employed. From this the actual weight of the calcium carbonate is computed. This official French method does not appear to possess any advantage in accuracy to the usual absorption method and is far more complicated.

361. Estimation of the Active Calcareous Matter in Soils.—Like other soil elements, the calcium carbonate exists in different degrees of fineness and availability in the soil. It must be admitted that the fine particles play the most important rôle. The calcium carbonate, which exists in large fragments, presents only a circumscribed surface and remains almost inactive, although it is easily corroded by the rootlets of plants. It is possible to estimate in a rapid way, the quantity of fine carbonate in the soil, considering that in a time relatively short, feeble acids act upon calcareous matter proportionally to the surface which it presents, and that it attacks, therefore, especially the finest particles. By measuring the amount of carbon dioxid set free under the action of dilute acids it is possible to estimate the content of available calcareous matter in the soil.

The apparatus of Mondesir is used for this purpose by the French chemists. It is composed of a tubulated flask of about 600 cubic centimeters capacity. The interior tubulature carries a manometer fixed by means of a stopper. This is formed of a rubber tube, terminated by a glass tube, whose extremity is united to a little rubber bag, very flexible, placed in the interior of the flask.

Graduation of the Apparatus.—If the apparatus is new it is necessary to begin by graduating it. The rubber bag is filled with water, the air being carefully excluded, in such a way that the level of the water comes just a little above the bend in the tube. There are placed in the flask 125 cubic centimeters of water and it is shaken for a few seconds. The flask and the manometer are then unstoppered and the level of the water in the manometer is made to equal the level of the water in the flask. With a rubber ring the level of the water in the manometer tube is marked. The manometer is then stoppered. There are then added to the flask two-tenths gram of pure calcium carbonate. The flask is closed and shaken for a minute. There are then added, enclosed in a little piece of filter paper, six-tenths of a gram of pulverized tartaric acid and the flask immediately closed and shaken several times. The manometer tube is then uncorked and moved until the level of the water reaches the point marked before. The difference in level after the height of the water remains constant is then read. The depression in the level observed, corresponds to two-tenths gram of pure calcium carbonate.

362. Estimation of the Available Calcareous Matter in the Soil.—There is introduced into the flask of the apparatus a quantity of soil varying in amount in accordance with the content of carbonate which it is supposed to contain. There are added 125 cubic centimeters of water and the flask is shaken for a minute. As in the test given before, the level of the water in the manometer is then made to correspond to that of the water in the flask. The level in the manometer is marked as before with a rubber band, and the manometer is then closed. There are then added, contained in a piece of filter paper, two grams of pulverized tartaric acid and the operation is finished as described before. The amount of tartaric acid added, in general, should be three times as much as the amount of calcium carbonate supposed to be contained in the earth. The pressure in the manometer being proportional to the quantity of carbon dioxid disengaged, it is easy to calculate the quantity of calcium carbonate in a state of fine division contained in the soil taken for the test.

In order to fill the rubber bag it is necessary to put it in its proper place in the apparatus. The flask is filled with water in order to flatten the rubber bag and expel the air from it. It is then closed with a cork. Afterwards, with the aid of a small funnel and with a copper wire placed in the tube, the lower extremity of which descends just to the elbow, the air in the tube is replaced by water. The operation is finished by uncorking the flask and inclining it or shaking it after a partial vacuum has been established. It is useless to attempt to drive off the last particles of the air. The rubber bag should have a content of about double the volume of the whole of the interior of the manometric tube. In the washing which is necessary between two successive operations, it is well to fill the flask entirely with water in order to expel all the carbon dioxid which it may contain.

The same remark may be made of this method of determination as was made of the last one. In the present case, however, the operation is not quite so complicated. When the apparatus is once arranged, it will admit of rapid determinations.

363. Lime Method at the Riga Station.—Ten grams of the non-ignited sample of the fine earth are digested with 100 cubic centimeters of ten per cent hydrochloric acid, in a 250 cubic centimeter erlenmeyer for twenty-four hours on the steam-bath, with frequent shaking. The filtrate, with washings after the addition of five cubic centimeters strong hydrochloric acid, is evaporated to dryness in a porcelain dish and the residue taken up with dilute hydrochloric acid. After filtering, ammonia is added in excess, the excess removed by evaporation, and the mass is again filtered. In the filtrate, the lime is thrown down with ammonium oxalate, filtered, ignited, and weighed as calcium oxid.

The above method cannot give exact results chiefly because more or less lime may be carried down with the phosphoric acid. Also if manganese be present it will be thrown down with the lime. These errors are compensatory, but only by chance could the compensation lead to exactness. It would be better in all cases to remove the iron and alumina in such a way as would avoid loss of time.

364. Estimation of Assimilable Lime.—In the determination of the total lime in soils or even of that part present as carbonate, it is not to be assumed that the quantity assimilable by plants is known; particles of lime minerals in soils are corroded only superficially by the rootlets of plants and any process which would attack only the superficies of the lime particles would thus more nearly resemble the activity of the solvent forces of plant growth. Oxalic acid is a reagent of this kind, attacking only the surfaces of lime particles. Reverdin and de la Harpe guided by this fact have based a method for determining the amount of lime present in the soil in an available state on the solvent action of oxalic acid.^[233] After the total lime content has been determined, twenty grams of the soil sample are covered with 200 cubic centimeters of a solution containing in molecular proportion a known quantity of sodium oxalate and carbonate. The mixture is digested on the water-bath for one hour. By this treatment all lime minerals are converted superficially into oxalate while particles containing magnesia are not affected. After filtering and washing well, the filtrate and wash-waters are acidulated with hydrochloric acid. If any precipitate of organic matter be produced separate it by filtration. Treat the filtrate with a slight excess of sodium acetate by which process the excess of hydrochloric acid is replaced with acetic after which the oxalic acid may be separated by treatment with calcium chlorid and subsequently titrated with potassium permanganate in presence of excess of sulfuric acid. The oxalic acid obtained, deducted from the quantity originally present will give the amount consumed on the surfaces of the lime particles and consequently the amount of lime corresponding thereto which may be considered as available for plant growth.

365. Method of the Halle Station for Lime.^[234]—a. In Phosphates, Limestones, etc.—Four grams of the prepared substance are heated with fifty cubic centimeters of hydrochloric acid and five cubic centimeters of nitric acid, in a porcelain dish on the water-bath to dryness, and left for a few hours at 105° for the purpose of separating the silicic acid. The dry residue is moistened with hot water and a few drops of hydrochloric acid, and allowed to stand for some time with frequent stirring. The contents of the dish are then washed into a half-liter flask, filled up to the mark and the separated silicic acid removed by filtration. If the silicic acid is not taken into account, the solution can be made directly in a half-liter flask.

After filtration, an aliquot part of the filtrate is neutralized in a 500 or 250 cubic centimeter flask with ammonia, again acidified with a few drops of hydrochloric acid and allowed to stand six hours at least, in the cold, with ammonium acetate. For each four grams of the substance fifty cubic centimeters of an ammonium acetate solution are used, made by dissolving in one liter of water 100 grams of ammonium acetate. If phosphoric acid is present in excess, iron and aluminum oxids are precipitated completely as phosphates. If iron and aluminum oxids are in excess, the excess must be precipitated by ammonia. If it is feared that in the subsequent precipitation of the lime by ammonium oxalate there may be still some phosphoric acid in solution, before precipitation with ammonium acetate the proper amount of ferric chlorid is added and the iron is afterwards precipitated with ammonia. It is certain that in the presence of oxalic acid and phosphoric acid the lime is precipitated as oxalate, but should it be feared that traces of calcium phosphate are precipitated with the iron and aluminum phosphates the precipitate of iron and aluminum phosphates may be dissolved in hydrochloric acid, neutralized with ammonia, again acidified and a second time precipitated with ammonium acetate and the filtrate added to that first obtained.

For the further estimation the filtrates are united and a quantity corresponding to a given part of the original sample, and being in volume from fifty to one hundred cubic centimeters is made slightly acid with acetic acid and while hot precipitated with dilute ammonium oxalate. The filtrate must contain acetic acid since calcium oxalate is best precipitated from a slightly acetic acid solution. The filtering of the calcium oxalate should not take place until from six to twelve hours after precipitation, and during this time it should stand in a warm place. Filter paper of the best quality should be used for the purpose.

The dried precipitate is brought into a platinum crucible together with the filter; the filter is first incinerated over an ordinary bunsen and the calcium oxalate converted into calcium oxid by ignition for fifteen minutes over the blast. It is then cooled in a well-closed desiccator and weighed as oxid. If in the precipitation of the iron and aluminum phosphates sodium acetate be employed instead of ammonium acetate, the precipitation must take place hot and filtration also be accomplished on a hot filter.

b. Estimation of Lime in Soils.—For the estimation of lime in soils there may be used either the acid soil-extract, prepared as under the direction for the estimation of potash, or twenty grams of the soil may be treated with hydrochloric acid and a few drops of nitric acid, and evaporated to dryness in a porcelain dish and the silicic acid separated as described for the estimation of lime in phosphates and limestones. In the case of soils, iron and aluminum oxids can be precipitated directly with ammonia since the small quantity of phosphoric acid usually contained in soils is not sufficient to influence in any way the estimation of the lime. For example suppose there is 0.10 per cent of phosphoric acid contained in a soil. In case the whole of this phosphoric acid is taken down with the lime it would only amount to about 0.10 per cent of calcium oxid precipitated as phosphate. This case, however, is very improbable since it is much more likely that the iron and aluminum phosphates will be precipitated and the whole of the phosphoric acid be carried down with them instead of being precipitated with the lime.

The precipitation of the lime and its subsequent treatment are to be conducted as just described.

366. Estimation of the Magnesia.—Magnesia is a much more rare element in the soil than lime. It is usually necessary to operate upon considerable quantities of earth in order to determine the magnesia with any degree of precision. From ten to twenty grams of the soil are taken. The decomposition is accomplished as in the case of lime. A few drops of barium nitrate are added for the purpose of precipitating any sulfuric acid present. Some ammonia and ammonium carbonate are added to precipitate the iron and aluminum oxids, the lime and the excess of barium introduced, as well as the phosphoric acid. The operation is best conducted on a dilute solution having a volume of from 400 to 500 cubic centimeters. The solution from which the lime has been precipitated, contains with the magnesia, large quantities of ammoniacal salts which it is necessary to destroy. For this purpose the solution is concentrated in a flask until its volume is about ten cubic centimeters. About ten cubic centimeters of nitric acid are added and the whole brought to the boiling point. Afterwards a few drops of hydrochloric acid are added. Continuing the heating, hydrochloric acid is added in small portions and, from time to time, some nitric acid until the bubbles indicating the setting free of gaseous nitrogen, resulting from the action of the nascent chlorin upon the ammonia, have completely ceased to appear. The whole is then evaporated on a sand-bath in a porcelain dish in order to separate the silica. The residue is taken up by water containing a few drops of nitric acid. It is filtered and evaporated to dryness in a covered porcelain dish. Upon the residue four or five grams of oxalic acid, in a state of powder, are placed. A little water is added in such a way that the moist mass covers entirely the matter in the dish. In order to avoid all losses there is placed upon the dish a funnel which serves as a cover. The dish is heated on a sand-bath, but when the film which is formed begins to break there are added from time to time, a little more oxalic acid and water until there is no longer any disengagement of the vapor of nitric acid. Afterwards it is evaporated to dryness and the heat raised to a low redness. The magnesia is found in a free state or mixed with alkalies. It is washed with a small quantity of water and collected upon a very small filter paper. The filter paper is dried, burned, the ignition carried to redness and afterwards cooled and weighed. In order to test the purity of the magnesia it is transformed into sulfate by the addition of a few drops of sulfuric acid. The excess of sulfuric acid is driven off by heating moderately by means of a gas-burner moving it in a circular manner round the bottom of the capsule and lifting the cover from time to time in order to allow the vapors of sulfuric acid to escape. The weight of the magnesium sulfate should correspond to that of the magnesia from which it was formed.

Magnesia exists most often in the soil in the state of carbonate or silicate. In this last state it is especially abundant in some soils, such as those which are derived from mica schists, serpentines, etc. In treating earth of this last quality with concentrated, nitric acid there is dissolved also a notable part of the magnesia of the silicates. If, however, it is treated for some minutes only with dilute hydrochloric acid the amount of magnesia present as carbonate alone can be estimated separately.

367. Estimation of Magnesia in Soils.—Method of the Halle Station.—For the estimation of magnesia the sample of soil or fertilizer is brought into solution in the same way as is given for the estimation of lime. After the separation of the silicic acid, the iron and alumina are precipitated with sodium acetate. In the case of phosphoric fertilizers, ferric chlorid should first be added in order that the excess of phosphoric acid shall be in all cases certainly combined with the iron. After this the lime is separated as usual with ammonium oxalate. After the precipitation of the lime, the magnesia is precipitated in an ammoniacal solution with sodium phosphate and the ammonium magnesium phosphate estimated exactly as in the case with the estimation of phosphoric acid, as magnesium pyrophosphate.

A simpler method for the estimation of magnesia consists in precipitating it as ammonium magnesium phosphate in the presence of a solution of ammonium citrate, the other bases remaining in solution. In this case the operation is carried on in an inverse way as described under the estimation of phosphoric acid, the proper quantity of the acid solution being neutralized with ammonia and after the addition of sodium phosphate, the required quantity of citrate solution added and a further excess of ammonia supplied.

368. Estimation of Manganese.—The estimation of manganese in the presence of Fe₂O₃, Al₂O₃, CaO, etc., presents peculiar difficulties. In ordinary alluvial clays the quantity of manganese is proportionately small and its estimation may be neglected. In volcanic clays the quantity of manganese, in proportion to the lime and magnesia, is much larger. The method used for estimating manganese is that of Carnot.^[235] The hydrochloric acid extract of the soil is evaporated to dryness and heated with potassium bisulfate in order to destroy the organic substance, the neutralized solution of the residue precipitated with twenty cubic centimeters of hydrogen peroxid solution and thirty cubic centimeters of ammonia. The colorless filtrate gives, with nitric acid and bismuth peroxid, no trace of reaction for manganese. The precipitate, washed by decantation, is carried into a carbon dioxid apparatus and treated with oxalic acid and dilute sulfuric acid. From the amount of carbon dioxid obtained, the quantity of manganese is calculated on the supposition that the precipitate corresponds to the formula Mn₆O₁₁.

369. Estimation of the Manganese by the French Method.—Manganese exists in all plants and its presence in small quantities seems necessary to vegetation. The method of estimation adopted by the French Commission is the one proposed by Leclerc and is applicable even when the base exists in small quantities.

Twenty grams of the soil are taken and the organic matter destroyed by incineration. In a flask of 200 cubic centimeters capacity, are placed thirty cubic centimeters of water and, little by little, some hydrochloric acid for the purpose of decomposing the calcium carbonate. When effervescence has ceased ten cubic centimeters of the same acid are added and boiled for half an hour, filtered, washed, and the wash-water and filtrate evaporated to dryness in a porcelain dish. Afterwards there are added twenty cubic centimeters of nitric acid of one and two-tenths density, and ten cubic centimeters of water. The liquor is boiled with constant shaking. Afterwards there are thrown in, in two or three portions, ten grams of lead dioxid. The boiling is stopped just at the moment when all the lead oxid is introduced into the liquor and the mixture is then shaken vigorously. The manganese is transformed by this treatment into a highly oxygenized compound having a deep rose coloration. It is transferred immediately afterwards to a graduated cylinder of 100 cubic centimeters capacity, with the wash-waters the volume is completed to 100 cubic centimeters and it is vigorously stirred with a rod, flattened at its extremity, in order to obtain a complete homogeneity of the liquid. The stirring rod is withdrawn and the liquid left to settle. At the end of some minutes the principal part of the liquid is clear, and it is decanted by means of a pipette graduated at fifty cubic centimeters, and this quantity of the clear liquid is poured into a small glass precipitating jar to which is added immediately, with constant stirring, a solution of mercurous nitrate from a graduated burette. The addition of the nitrate is arrested at the moment when the rose color of the liquor disappears, and the volume of the mercurous nitrate employed is read from the burette.

It is now necessary to determine the strength of the mercurous nitrate, that is the quantity necessary to decolorize one milligram of manganese. For this purpose dissolve by means of five cubic centimeters of hydrochloric acid 150 milligrams of manganese dioxid, which is prepared perfectly pure by means of precipitation. When the solution is complete evaporate to dryness, add one cubic centimeter of sulfuric acid and heat on a sand-bath until white fumes of sulfuric acid appear. Redissolve in water and make the volume up to 100 cubic centimeters. Each cubic centimeter of this solution should contain one milligram of manganese. Take five cubic centimeters of this solution, equivalent to five milligrams of manganese, treat in a capsule with twenty cubic centimeters of nitric acid and ten cubic centimeters of water, afterwards add ten grams of lead dioxid, carrying on the operation exactly as described above. Fifty cubic centimeters, taken as before described, are then decolorized by the solution of mercurous nitrate, and thus it is easy to calculate the quantity of manganese which corresponds to one cubic centimeter of the mercurous nitrate employed. By a simple proportion the quantity of manganese contained in the twenty grams of earth to be analyzed is calculated.

The mercurous nitrate is prepared by dissolving five grams of crystallized mercurous nitrate in one liter of water; it is allowed to repose for some time and is preserved in a well-stoppered flask.

370. Estimation of Iron.—Iron, in general, is quite abundant in the soil where it is met with, principally in the state of anhydrous sesquioxid or the hydrated sesquioxid of silicates. Some soils, however, only contain iron in small proportions and it can happen that the introduction of iron as a fertilizing element may be useful. Plants assimilate iron only in small quantities, but it appears to be indispensable to their development and to the proper functional activity of their assimilating faculties. The method of estimation which is recommended is based upon the decoloration of potassium permanganate by iron in the ferrous state. The following description, based on the method proposed by the French Commission, will illustrate the process to be followed.

Ten grams of the soil are ignited in a porcelain capsule until all organic matter is destroyed. The ignited mass is then introduced into a flask of 100 cubic centimeters capacity with thirty cubic centimeters of hydrochloric acid and fifteen cubic centimeters of water. It is boiled for about half an hour. The iron oxid is dissolved and is found in solution in the form of ferric chlorid. After filtering and washing, the volume of the filtrate is reduced by evaporation to about twenty-five cubic centimeters. The liquor is afterwards placed in a flask of from 100 to 150 cubic centimeters capacity, which is closed by a stopper carrying a tube furnished with a valve destined to prevent the re-entrance of the air. Ten cubic centimeters of dilute sulfuric acid are added from a mixture containing twenty cubic centimeters of strong acid and eighty cubic centimeters of water. Afterwards the iron is reduced to the ferrous state by introducing into the flask, in quantities of about five decigrams, metallic zinc and waiting after each addition until the portion last added is dissolved before adding another. This addition of zinc is continued until the iron is all reduced. When this point is reached and the last portion of zinc added is dissolved, the contents of the flask are transferred rapidly to a precipitating glass of about one liter capacity, in which there has been placed a little lately boiled but cold water. The flask is washed several times with cold water, previously boiled, to remove from it all traces of oxygen. The volume is made up to 500 cubic centimeters, and afterwards, without any loss of time, by means of a graduated burette and with constant stirring, a solution of potassium permanganate is added which is stopped exactly at the moment when the liquor begins taking on a light rose tint. The quantity of permanganate employed is read from the burette and is proportional to the amount of iron contained in the soil. A blank operation is made for the purpose of detecting traces of iron which the zinc may contain. If, as often happens, the soil contains a large amount of iron it is advisable to use only one gram of it for this operation. The aspect of the earth will indicate in general if it be very ferruginous.

Preparation and Standardization of the Permanganate Liquor.—In one liter of water are dissolved ten grams of crystallized potassium permanganate and the quantity of iron which corresponds to one cubic centimeter of this liquor is determined. It may be well enough to remark that this liquor does not remain constant and it is necessary to titrate it from time to time. For this purpose pure iron is taken. Piano wire may be used, being almost pure iron. One-tenth of a gram of this wire is dissolved in a flask in the manner recommended for treating the soil and with the same quantities of acid and water. When the solution is complete it is transferred to the flask to be estimated. It is made up to one liter and permanganate added, just as in the case before mentioned, until the rose color persists. There is thus determined the quantity of iron which corresponds to each cubic centimeter of the permanganate, and by a simple proportion, the quantity of iron contained in the soil analyzed is determined.

The Italian agricultural chemists proceed essentially in the same manner in determining the iron in soils, first igniting the sample and afterwards extracting the iron in the ferric state with boiling hydrochloric acid, reducing with hydrogen, and titrating with potassium permanganate.

The following are the reactions which take place:

Fe₂O₃ + 6HCl = Fe₂Cl₆ + 3H₂O.

Fe₂Cl₆ + 2H = 2FeCl₂ + 2HCl.

10FeCl₂ + K₂Mn₂O₈ + 16HCl = 5Fe₂Cl₆ + 2KCl + 2MnCl₂ + 8H₂O.

Sulfuric may take the place of hydrochloric acid in the above reactions.