n f 1 THE COMPARATIVE AGRICULTURAL VALUE OF INSOL- UBLE MINERAL PHOSPHATES OF ALUMINUM, IRON AND CALCIUM BY JACOBUS STEPIIANUS MARAIS University of Illinois Reprinted prom Soil Science, Vol. XIII, No. 5, May, 1922 Digitized by the Internet Archive in 2010 with funding from The Library of Congress http://www.archive.org/details/connparativeagricOOnnara THE COMPARATIVE AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES OF ALUMINUM, IRON, AND CALCIUM BY JACOBUS STEPHANUS MARAIS Pass B.A. University of Cape of Good Hope, 1916 Honors B.A. University of Cape of Good Hope, 1917 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY In Agronomy IN THE GRADUATE SCHOOL of the UNIVERSITY OF ILLINOIS 1921 H' W g'!' '' ' ^ '•'• i ' !n t^tU » < '-t mmm of conq«ess DOCUIMjSNTS Repiinted from Soil Science Vol . XIII, No. 5, May, 1922 NJ- THE COMPARATIVE AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES OF ALUMINUM, IRON, AND CALCIUMi JACOBUS STEPHANUS MARAIS University of Illinois Received for publication June 6, 1921 INTRODUCTION "Phosphorus is the only element that must be purchased and returned to the most common soils of the United States. Phosphorus is the key to per- manent agriculture on these lands." This statement of C. G. Hopkins (29) emphasizes the extreme importance of the phosphorus problem in modern agriculture; especially at the present time when the seriousness of the world food situation is making an urgent appeal to agriculturists to increase and to maintain permanently the fertility of all tillable soils. The acute shortage of transportation facilities has placed farmers, not conveniently situated near phosphate-producing centres, at a disadvantage with regard to procuring phosphorus at other than exorbitant prices. This has resulted in a world-wide prospecting for phosphate deposits and has caused considerable speculation as to the feasibility of utilizing iron and aluminum phosphates for agricultural purposes. In spite of the fact that a considerable amount of work had been done that demonstrates the value of aluminum and iron phosphates, the general belief is that they have little significance from an agricultural point of view. The fact that they are practically useless for acid phosphate manufacture, com- bined with their low solubility in citric acid and ammonium citrate solutions is probably the main cause for the popular conception of their agricultural value. There are also numerous statements by eminent scientists scattered through- out the literature in which aluminum and iron phosphates are referred to as being particularly unavailable as plant-food. The fleeting action of super- 1 Thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Agronomy in the Graduate School of the University of Illinois. The author wishes to express his obligations to Prof. Robert Stewart for valuable suggestions and criti- cisms when the experiments were first planned; to Prof. A. L. Whiting and Dr. E. E. De Turk for help and advice during the progress of the later experiments and for reading the manuscript; and to Messrs. J. C. Anderson and W. Green for assistance rendered in the greenhouse in preparation and care of the pot cultures. 355 SOIL SCIENCE, VOL. XIII, NO. 5 356 JACOBUS STEPHANUS MARAIS phosphates on soils rich in aluminum and iron oxides, for example, is ascribed to the conversion of this phosphate into aluminum and iron phosphates. It was the object of the experiments reported in this paper to determine the comparative values of various phosphates of aluminum, iron and calcium which occur in nature, and simultaneously to determine how they are affected by diverse collateral treatments. REVIEW OF LITERATURE Some fundamental considerations A fundamental fact, which has a very important bearing on the phosphate problem in soils was brought to light by the work of Schloesing and Kossovitsch. In 1899, Schloesing (67) demonstrated the fact that plants can obtain their phosphorus from very dilute solutions, solutions containing only 1 to 2 mgm. phosphoric anhydride per liter. This emphasizes the importance of naturally dissolved phosphates in the soil solution for plant nutrition. Kossovitsch (37) repeated these experiments, verified Schloesing's results and showed simul- taneously that the relative feeding powers of plants do not rest solely on their ability to utilize the phosphorus occurring in dilute solutions. Flax, when compared with mustard and peas, has but feeble powers to utilize the phosphorus of tricalcium phosphate rock, but was shown to make good growth on a nutrient solution, which contained only 1 . 3 mgm. of phosphoric anhydride per liter. From the work of Schloesing one might at first conclude that the plant roots exert a solvent action on phosphates. Sachs (65) in 1860, demonstrated that plants roots were capable of corroding marble plates. In 1896, Czapek (13) conducted extensive investigations to deter- mine whether roots excrete or secrete acids, which might function in dissolving plant-food. Eventually he concluded that carbonic acid was the only acid given off in considerable quan- tity by live roots of plants. In 1902, Kossovitsch (37) demonstrated clearly that the plant roots themselves and not the nutrient solution were responsible for obtaining phosphorus from phosphorite. The following device was employed by him to determine this factor: Plants were grown in two sets of cylinders. In the one set, sand mixed with tricalcium phos- phate was used as a medium for the plants to grow in. Five liters of nutrient solution were passed daily through each cylinder. In the second set, pure sand was used as a medium for growth. As in the above case, five liters of nutrient solution were added daily with the excep- tion that the nutrient solution was first made to pass through another cylinder containing a mixture of quartz sand and tricalcium phosphate and in which no plants were growing. If the nutrient solution acted as a solvent of the phosphate, the plants in the second set of cylinders should have made a fair growth. The plants grew well in the first set and made hardly any growth in the second, proving that if the nutrient solution exerted any solvent action on the tricalcium phosphate, its action was very slight and that the action of the roots themselves was a very much more important factor. In 1911, Prianishnikov (61) made the claim that iron and aluminum phosphates were gradually decomposed by water and that root excretions do not play the important r6le in assimilation of these phosphates that has usually been ascribed to them. Varying ability of plants to assimilate phosphorus from insoluble phosphates In 1893, Balentine (3) working at the Maine Agricultural Experiment Station reported that Graminae were benefited more by acid phosphate than by redondite and rock phosphate, and that plants of the Cruciferae family were especially strong feeders on rock phosphate. Two years later, Merrill and Jordan (42) placed the four botanical families studied in the order given below as regards their foraging powers for insoluble phosphates. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 357 1. Leguminosae as represented by peas and clover. 2. Cruciferae as represented by turnips and ruta-bagas. 3. Graminae as represented by barley and corn. 4. Solanaceae as represented by tomatoes and potatoes. The insoluble phosphates employed in this investigation were Florida rock phosphate, iron phosphate, and aluminum phosphate. Kossovitsch at various times between 1898 and 1910 made mention in his writings concern- ing the feeding powers of different species of plants. In 1901 (36), he commented on the strong feeding powers of buckwheat and mustard when grown with phosphorite as a source of phosphorus. In a later publication (39) in which he summarized his work on the utiliza- tion of phosphorite by mustard, clover, oats, and flax, he placed these plants in the order in which they are here mentioned as regards their powers to utilize phosphorite. It should be observed that this order is somewhat similar to that put forth by Merrill and Jordan. Kosso- vitsch (38) also tried to correlate the feeding powers of plants with their ability to excrete carbonic acid, but the difference in the amounts excreted did not Justify the drawing of any definite conclusions. Schreiber (68) experimented with eleven species of the Graminae, nine of the Leguminosae, three of the Cruciferae, and eleven miscellaneous plants. The Leguminosae, the Cruciferae, and buckwheat utilized mineral phosphates to a considerable extent, whereas the Graminae, flax, tobacco, carrots, asparagus, beets, and potatoes showed little solvent powers. Wheeler and Adams of Rhode Island (82, 83), Prianishnikov (56, 57), Bonomi (5), Ged- roits (23), Chirikov (Tschirikov) (11), Semushkin (69), and Soderbaum (72), are among other workers who have drawn attention to the individuality of plants with respect to the topic under discussion. In nearly all these cases, their results agree in a general way with those of Merrill and Jordan. The work of the above investigators will be considered later in connection with another phase of our problem. Emil Truog (76, 77) has propounded a theory to explain the individuality of plants with regard to their feeding powers. Plants with a high calcium content he stated, have a rela- tively high feeding power for the phosphorus in phosphorites. For plants with relatively low calcium content, the reverse is true. Clover, alfalfa, peas, buckwheat, and several of the Cruciferae have high calcium content and are, therefore, according to this theory, powerful feeders on insoluble phosphates. Corn, rye, oats, wheat, and millet fall in the opposite class. A calcium oxide content of less than 1 per cent may be considered low. In another publica- tion (78), Truog claimed that high internal acidity of roots is accompanied by high feeding powers for calcium. Logically then, plants with roots of high internal acidity are capable of utilizing insoluble phosphates with greater success than plants with roots of relatively lower internal acidity. It is clear tliat the individuality of the plants is a large factor when the availability of phosphates is being considered. Effect of soil on availability of insoluble phosphates In studying this question three characteristics of soil have been considered by workers: 1. Mechanical composition. 2 . Amount of organic matter in soil. 3. Reaction of the soil. It is generally held (41) that it is preferable to use bone meal and basic slag on warm sandy soils. Soluble phosphates are put to better use on heavier clay soils. Wheeler and Adams (83) claimed that the addition of three-fourths to one ton of limestone per acre removes the drawback of using soluble phosphates on light sandy soils. On peat and muck soils, the first applications of soluble phosphates are ineffective, due to their entering into colloidal combinations, but after these demands have been met, their effects are noticeable. Con- cerning the reaction, predominant opinion asserts that soluble phosphates are employed with the greatest success on calcareous soils (14, 27, 49). Hilgard (27) in his celebrated work, 358 JACOBUS STEPHANUS MARAIS "Soils," made the following statement, ". . . .in the presence of high lime per- centages, relatively low percentages of phosphoric acid and potash may nevertheless prove adequate; while the same or even higher amounts, in the absence of satisfactory lime per- centages, prove insufficient for good production." Paturel (49), Deherain (14) and others claimed that unless sufficient lime be present, the phosphoric acid is fixed by aluminum and iron oxides into unavailable combinations. On the other hand, this view appears contra- dictory to the observations of Schloesing, fils, regarding the solubility of phosphoric acid in the presence of carbonate of lime (66), but natural conditions seem fully to justify Hilgard's conclusions. Numerous investigators found aluminum phosphates to be verj' beneficial to plant growth provided they were employed on soils well supplied with lime. Results in Mary- land (50), France (1), and Rhode Island (82, 83) all show that favorable results with alumi- num phosphate have always been obtained when the phosphate is used in connection with lime or on soils naturally calcareous. When tricalcium phosphate is employed, the best immediate results seem to be obtained on soils not saturated with bases (24) or on soils well supplied with organic matter (28, 82). Effect of nitrogen compounds on availability of insoluble phosphates Prianishnikov and a large number of other Russian workers have studied very carefully the effect of various nitrogen compounds on the availability of insoluble phosphates. All the results agree in general that ammonium sulfate enhances the availability of insoluble phosphates and that ammonium nitrate likewise increases the availability, but to a lesser extent. Sodium nitrate either has no effect or depresses the availability. Calcium nitrate is similar in its effect to sodium nitrate, but less marked. These results are due to inher- ent properties of the salts themselves and not to their conversion into other compounds, for example the formation of nitric acid as the result of nitrification of ammonium salts. Kosso- vitsch (36) was responsible for the classic work in regard to the effect of ammonium salts. In experiments in which the possibility of nitrification being a factor was carefully prevented, he confirmed in all instances the deductions of Prianishnikov. Wheeler and Adams (83) commenting upon Warington's work (81) seem to be of the opinion that with aluminum phosphates results would have been established which would be the reverse of those given above. The fact that nitrification materially affects the availability of insoluble phosphates has been definitely established by the investigations of Hopkins and Whiting (30). Soder- baum (72) checked up Prianishnikov's deductions. He believed that the physiological reac- tion of the accompanying nitrogenous fertilizer plays an important part, but claimed that other factors, such as kind of plant, soU and other collateral treatments used, may lessen or even reverse the influence of this factor. This point is well brought out by Chirikov (10) who found that when calcium nitrate replaced ammonium sulfate in his buckwheat cultures, the yields were not reduced, but increased. Nedokuchaev (46) working with different crops, oats and flax, reported that yields were lower where calcium nitrate was used in lieu of ammonium sulfate. On the whole Prianishnikov's deductions seem to be accurate, but we should bear in mind that no hard and fast rule can be laid down. In work on the availability of phos- fates, the accompanying nitrogenous fertilizer is a factor that must be remembered, espe- cially when we attempt to make generalizations from our results. Effect of lime on availability of insoluble phosphates When the effect on the availability of insoluble phosphates was considered, the influence of lime came up for discussion since the reaction of the soil and lime content of the soil are closely interrelated. Some further opinions on the effect of lime follows. Prianishnikov (61) divided the phosphates into two groups; the one, including tricalcium phosphate, bone meal, and phosphorite, consists of those of which the assimilation is markedly reduced by the lime; the other, including acid phosphate (mono- and di-calcium phosphates), Thomas slag, AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 359 mono-potassium phosphate, iron phosphate, and aluminum phosphate, consists of those unaffected by the addition of lime or even benefited by it. The studies were made in sand cultures. The crops employed were barley, peas, oats, wheat, and buckwheat. In all cases, however, where ammonium nitrogen was substituted for nitrate nitrogen, liming was bene- ficial. Shulov (71) in studies similar to those of Prianishnikov, determined that the assimil- ability of pure ferrous phosphate and vivianite was unaffected by lime; that of tricalcium phosphate, in the forms of bone meal and phosphorite, was adversely affected; and that of superphosphate, precipitated phosphate and Thomas slag was only slightly reduced. Gaither (21) explained the lack of harmful effects of lime upon the availability of soil phosphates as due to its action in replacing iron and aluminum in combination with phosphorus and so rendering the phosphates more soluble. Gaither used0.2iV nitric acid as a solvent for determining available phosphorus. Wheeler and Adams (84) pointed out that, in the phos- phate experiments at Rhode Island, iron and aluminum phosphates were more efficient than floats on limed land. This agrees with the findings of Prianishnikov. Effect of various solvents on the availability ofijisoluble phosphates It is beyond the scope of this work to enter into the controversy as to which solvents of phosphates can be used for determining their availability to plants. Some literature which has a bearing on this work is quoted. Risler (64) claimed that carbonic acid has much less solvent action on aluminum and iron phosphates than on calcium phosphates. Wagner (80) and later Storer (73) claimed that alkalies, such as sodium carbonate, ammonium carbo- nate, etc., can dissolve phosphates of iron and aluminum. Cameron and Bell (9) claimed to have proved that soil phosphates are decomposed or hydrolyzed by water with formation- of other phosphates containing relatively more of the base. Zecchini (85) reported that alu minum and iron phosphates are very insoluble except in alkaline solution. Gedroits (22) worked on solubility of phosphates in 2 per cent acetic and citric acids. The relative solu- bilities in acetic acid were tricalcium phosphate, aluminum phosphate, ferric phosphate, in the order named; in citric acid dicalcium phosphate and aluminum phosphate were equally soluble, ferric phosphate less soluble. In growing plants in sand culture with these phos- phates, the aluminum phosphate pots gave the highest yield, tricalcium phosphate was second, andiron phosphate pots a close third. Truog (75) questions the whole idea of employing chemical solvents as a means for determining the availability of different phosphates, basing his deductions on favorable results obtained with phosphates of aluminum and iron, which are, as a general rule, less soluble than calcium phosphate in such solvents. Elliot and Hill (16) had before this arrived at the same conclusions. Fraps (19), on the other hand, pro- posed 0.2 iV nitric acid as the solvent to indicate the available supply of phosphorus in the soil. He asserted that in pot experiments, the phosphoric acid removed by the crops is closely related to the quantity of "active" phosphoric acid. "Active" phosphoric acid is defined as that amount which dissolves in . 2 iV nitric acid. Several workers have indicated the value of dehydrating aluminum phosphate to render it more valuable as a fertilizer. The investigators at the Rhode Island Agricultural Experi- ment Station have always included roasted redondite in their comparative phosphate tests and have drawn attention to the value of dehydration. Morse (44) found that roasting increased the solubility of aluminum phosphate in neutral ammonium citrate, but pot and field tests failed to verify the laboratory indications of availability. Pilon et al (54) described a method for roasting double phosphates of iron and aluminum in order to render the com- bined phosphoric acid soluble in ammonium citrate. Fraps (20) pointed out that ignition increases the solubility of wavellite, dufrenite, and variscite inO.2 iV^ nitric acid about ten times and makes them almost completely soluble in 12 per cent hydrochloric acid. Peterson (51) conducted similar investigations and showed that heating wavellite for five hours at 200°C. increased the solubility of the phosphoric acid 4 to 50 per cent and heating to 240°C. increased the solubility to 100 per cent. Dufrenite, when heated at 200°C., was but slightly increased in solubility. 360 JACOBUS STEPHANTJS MARAIS Views concerning the comparative availability of phosphates of aluminum, iron and calcium Below we have simply an enumeration of claims and counter-claims as to the comparative values of aluminum, iron and calcium phosphates. Many of the statements decrying the value of aluminum and iron phosphates were based not on experimental work planned to test this particular point, but were the outcome of efforts to explain puzzling irregularities in the behavior of superphosphates and acid phosphates. Very many workers, too, reported on the topic under discussion as a side issue of a large problem and very often such work failed to effect a fair comparison because the individual phosphates probably display their opti- mum availability under unlike conditions. Merrill (43) reported that in most cases crude Florida rock phosphate outyielded Redonda phosphate. Paturel (49) advised that lime be applied to soils high in oxides of aluminum to prevent the fixation of phosphorus by them. Morse (44), as has already been pointed out, studied the solubility of aluminum phosphates and the effect of dehydration of them and showed that, while the solubility in neutral ammonium citrate was greatly increased, field tests failed to demonstrate a resulting increase in availability. Hilgard (27), as quoted in a former paragraph, stated that in the presence of high lime percentages, relatively low per- centages of phosphoric acid and potash may nevertheless prove adequate. This seems to indicate that Hilgard preferred calcium and magnesium as carriers of the phosphate in the soil to other bases. Deherain mentioned an experiment in France in which the action of superphosphates was very fleeting, due, supposedly, to the phosphoric acid passing into combination with iron and aluminum and so rendering the phosphate incapable of use as plant-food. Wheeler and Adams (83) predicted that soluble phosphates were not likely to have as good after effects on unlimed soil rich in iron and aluminum oxides as would bone meal and basic slag for the reason that the phosphoric acid would be fixed as aluminum and iron phosphates, in which forms plants cannot secure it readily. Gaither (21) study ng the effect of lime on the solu- bility of soil constituents declared that lime renders the insoluble phosphates in the soil soluble by replacing iron and aluminum, which are in combination with phosphorus. Pfeiffer and Blanck (53) analyzed the effect of alumina and silicic acid gels on the assimila- tion of phosphoric acid by plants and obtained results which showed that both gels reduced yields of plants as well as their phosphoric acid content. The experiment was conducted with sand fertilized with 3 gm. of basic potassium phosphate and soil extract. Bishop (4) worked with soybeans in pot cultures and concluded that soluble phosphates were not more desirable than Florida soft rock, iron and aluminum phosphates. Balentine (3) and later Merrill and Jordan (42), all of the Maine Agricultural Experiment Station working with sand cultures, found that acid phosphate gave the best returns in all cases and especially with the Graminae. Redondite, a phosphate or iron and aluminum, gave better results with Graminae than rock phosphate, but in all other cases the reverse was true. In the second report when these investigators worked with a larger variety of plants, they stated that acid phosphate was best, but the insoluble forms were utilized to a considerable extent and that Florida rock phosphate, on the whole, was better than iron and aluminum phos- phates, except for barley, corn, turnips, and potato tubers. The plants used in the investiga- tion were peas, clover, turnips, ruta-bagas, barley, corn, tomatoes, and potatoes. Andouard (1) worked with a calcareous soil and deduced that aluminum phosphate was readily available to plants. Burkett (7) obtained very favorable results with raw and roasted redondite. Gedroits (22), in pot culture with soil, declared that aluminum phosphate gave better yields than calcium phosphate and the latter better yields than iron phosphate. Director Pat- terson (50) of the Maryland Agricultural Experiment Station, made the following state- ment: "The iron and alumina phosphates proved in all cases to be valuable sources of phos- phoric acid, and it would seem that they deserve a higher rank as a fertilizer than that usually accorded them." AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 361 Nagaoka (45) employed phosphates on rice fields exhausted by continuous cropping. All the phosphates gave large increases in yield. Table 1 gives the relative jdelds, double super- phosphates being taken as 100. Bonomi (5), in comparing aluminum phosphate with mineral phosphate, superphosphate, and Thomas slag, reported that aluminum phosphate gave large increases in yield with both clover and wheat, but that superphosphate was always superior to it; spring wheat yields with aluminum phosphate was smaller than those with Thomas slag, but with clover the reverse was true. Elliot and Hill (16) showed that from weights of crops produced in pot experi- ments, iron and aluminum do not fix phosphoric acid in forms unavailable to plants; as a matter of fact, they claim that iron and aluminum phosphates produce more plant growth than the calcium compounds do. For this reason, they denounced the solvents used by chemists for determining the reversion of phosphates as useless for the purpose. TABLE 1 Relative yields of rice as influenced by various phosphates (from Nagaoka) 1. Double super phosphate 2. Ferric phosphate 3. Ferrous phosphate 4. Aluminum phosphate . . 5. Calcium phosphate .... FIRST SECOND THIRD FOURTH YEAR YEAR YEAR YEAR 100 100 100 100 140 141 399 58 87 88 194 44 92 145 514 103 117 110 161 118 AVER- AGE 100 185 103 216 127 Shulov (71) worked with vivianite — a ferrous phosphate— a pure ferrous phosphate, alumi- num phosphates, tricalcium phosphate, and superphosphates in sand cultures. In all cases, the iron and aluminum phosphates proved highly efficient as fertilizer and increasing amounts of lime up to 1 per cent produced very little depressing effect on their action. Eaguley (2) compared normal orthophosphates of calcium, iron and aluminum on oats, peas, and Swedish turnips grown on artificial soil of sand and chalk. As a general rule, iron and aluminum phosphates proved more efficient than calcium phosphates. Peterson and Truog (52), in pot cultural work, demonstrated that freshly precipitated and dried ferric phosphate served as a better source of phosphorus for oats than did rock phosphate, while for rape, the results were exactly the reverse. Truog (75) later made the following statement: "Contrary to the general belief that aluminum and iron phosphates are relatively unavailable to plants, nine out of ten plants tested made better growth on aluminum phosphate than on calcium phos- phate, and six better growth on ferric phosphate." EXPERIMENTAL These experiments were planned to determine whether or not it is desir- able to employ mineral phosphates of aluminum and iron as sources of phos- phorus. Studies were made comparing their value as sources of phosphorus with that of calcium phosphate in various forms both natural and artifi- cial. Simultaneously efforts were made to determine what conditions would cause these phosphates to be of the greatest value for crop growth. Description of materials used The aluminum phosphates employed were lazulite from near Death Valley, Inyo county, California, wavellite from Cumberland county, Pennsylvania and Saldanha phosphate from the Cape Province in South Africa; the iron 362 JACOBUS STEPHANUS MARAIS phosphates were dufrenite from near Vesuvius, Rockbridge county, Virginia; and vivianite from Leadville, Colorado; the calcium phosphates were Florida rock and Laingsburg phosphate from the Cape Province, South Africa. Some of the wavellite was obtained from Montgomery county, Arkansas. Besides these phosphates there were also used bonemeal and acid phosphate. In the sand cultures disodium hydrogen phosphate in solution replaced the acid phosphate. Table 2 gives the analyses of the various phosphates. TABLE 2 Composition of phosphates employed in experiments KIND OF PHOSPHATE Lazulite Wavellite Saldanha Dufrenite Vivianite Florida hard rock Laingsburg Bonemeal Acid phosphate . . PHOSPHORUS ALUMINUM IRON per cent per cent per cent 13.72 16.20 5.04 10.04 17.40 2.48 9.14 16.30 1.61 12.07 1.60 40.20 9.11 1.20 22.40 14.70 4.17 1.53 14.01 2.92 2.59 12.52 0.00 Trace 7.01 0.81 0.40 per cent 0.97 0.23 1.06 0.11 0.10 26.40 31.90 27.10 14.70 All the aluminum phosphates are basic phosphates, i.e., they have aluminum hydrate associated with the phosphate and all of the phosphates are more or less hydrated. Lazulite has the additional property of being completely insoluble in acids. Hot aqua regia acting on lazulite for an hour fails to dis- solve more than a trace of phosphoric acid. Wavellite and Saldanha phos- TABLE 3 Essential plant-food elements per acre of 2,000,000 pounds of water-free soil or approximately the surface layer of 6% inches over one acre PLANT rOOD ELEMENTS Phosphorus Potassium Nitrogen Limestone requirement by Hopkins' method in pounds of CaCOs per acre BROWN SILT LOAM YELLOW SILT LOAM Ihs. lbs. 1,096 706 32,240 29,180 4,287 1,942 400 2,949 phates dissolve readily in acids. Infrenite is a basic ferric phosphate con- taining a trace of magnesium. The formula usually ascribed to it by geologists is, FeP04-Fe(OH)3. Vivianite crystallizes in the monoclinic form and is a hydrated ferrous phosphate with the formula Fe3(P04)2 -81120. The Florida hard rock is rather high in aluminum as compared to the usual run of phosphate from this source. The Lainsburg phosphate contains quite an appreciable quantity of calcium carbonate. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 363 The pot cultures were conducted in 1-gallon glazed earthen-ware pots drained by a hole in the bottom of the pot and capable of holding 10 pounds of soil. In most of the soil cultures a light phase of brown silt loam from the University Farm at Urbana, lUinois, was used. As far as is known the soil had never been cultivated and had never received soil treatment of any kind. The soil is known to respond readily to applications of phosphorus. In later experiments a yellow silt loam soil was introduced. This soil came from near Vienna, in Johnson county, Illinois. In all soil cultures 10 pounds of soil were used per pot and in the sand cultures 12 pounds of sand. Experiment 1 This experiment was planned to test the comparative effects of the phos- phates on crops and the effect of lime and gyspum on their availability. The experiment was begun in the spring of 1920. Brown silt loam was used and treated as described in table 4. The pots were planted to buckwheat and annual white sweet clover. All the buckwheat pots were numbered as in table 4; the sweet clover pots were given the same numbers as the buckwheat pots but had "rv;" prefixed to the number. Each treatment was carried out in duplicate. The planting oc- curred on February 6, 1920. The sweet clover seed was inoculated. Twenty buckwheat seeds and thirty sweet clover seeds were planted in each pot. After the seeds were up the plants were gradually thinned so that at the end of 4 weeks only the seven strongest buckwheat plants were left in each pot and the ten strongest sweet clover plants in each of the sweet clover pots. Much cloudy weather was experienced and this combined with the short days made growing conditions in the greenhouse unsatisfactory. It was noticed that the buckwheat especially was looking decidedly poor. In order to insure the elimination of all factors tending toward depression of growth it was thought advisable to start a new series of cultures in which the buckwheat would receive an application of 1.84 gm. of calcium nitrate, the equivalent of 100 pounds of nitrogen to the acre. In all other respects the same plan of treatment was followed, also: Series 600 corresponded exactly with series 100 Series 700 corresponded exactly with series 200 Series 800 corresponded exactly with series 300 Series 900 corresponded exactly with series 400 Series 1000 corresponded exactly with series 500 In addition, eight control pots were planted to determine the effect of limestone gypsum and calcium nitrate. The treatment applied to these pots and the yields obtained are shown in table 5. The planting of this series began on February 22, and was completed on February 24. 364 JACOBUS STEPHANUS MARAIS w 1 ^ '-I On ON 00 CO Os ON ^ CO "* CO ^ d 00 WW" « > M O O p. CN lO rfi ■<* MD CN -* 00 10 CN CM 00 1-H 00 CN w 1171 ■rt CS '-H ■^ CO CO 1 1 ■?i. -" bo 10 10 10 »o 10 10 »0 Tt< o a -* CM 00 On 0\ 10 00 »-l Os '^ to 10 NO '^ ^ < 5i ON On 00 00 00 T-l -< On «M ^ CO OS 00 Ov Os o w ,-1 .rH T-H l-H . o\ 10 00 -w 10 T-H ON <>- On Os t^ CS CN CN 00 ON is £, On On 00 9 On T-l ^-H CN OS CN ■■-1 CO OS 00 CO On O o w h-] T-l ^^ >-l •^ 1-1 T-H »-f to ^ . On 00 Cn ■■* 00 CS Os CS ON ON 00 00 ■n-i tJ< « 1 od '-I Os CN CNI OS On 00 .^ .^ tH -.-1 T-H 1-1 T-H '"* ^ < < < < m eq PQ M u CJ CJ < < < < o ■.— 1 (N ro ^ ^ CN CO ■^ '-t CM CO Tt< CN CO -^ to '-I tH 00 ^ ^ CO 00 On t^ 0^ d d 0^ rH CN '-I CO 10 SO -^^ h. ■.-H CN •^ T-1 CO 10 (N NO T-4 10 <>I -^ 1 1 5g o ■?>, J; o lO I/O "^ 10 10 ID 10 to "0 10 10 p.^ c3 to ■* 00 Os CO CN 00 00 NO Os 10 1-H a 0^ <=> ^ d> di T-i ■* CN rt< CO CO OS 00 On 00 B < ■■-1 ■^ ^- »-i —* -^ 'r-f ■^ ■rt — ( T-l l-H \4 CJ sa . OS NO 00 00 CO ^ d -h' 00 !>• OS ■rt rjl -rt >0 NO ^ ^ ^ T-H CO 00 00 On CN 00 00 Os ON c? o s CO 1 ■< , ( '. \ T 4 '"''"' '^ '"' rt ^ ^ ^ 1 4^ . ^ IT) "0 (T) SO t^ CO 00 t^ UO so 00 CN 1 s EP. S ^ --H CN -^ tN tJ< "-1 CO CO OS 00 Os t^ M < < < < W pq CP M U U CJ CJ <: < < < ^ -l ■>-( 1-1 •.— 1 CS csl (^q CN &« K (U W H 1 .i .a to K Td x) TS td 'CJ C3 c CI <=^ « ^o >o ^ c 3 ■ ^ ' ^ •^ ■' H U rt O W o s o C/3 a; a t3 SI 1-^ rt ^ an P 1 AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 365 On 1^ O ^0 t-~ J:^ VO t^ ■•-I lO TjH vo 00 T-l f^ Tjl TiH o ■^ j>; lO NO \0 lO 0\ to O T-i '-H ^ CN tH ^ ^ ^ '^ ^ "^ T^ t:Jh Tfi irj in rh •rii -^ IT) ir> ■<-* tr^ (^ !>• •r-^ (6 -^ d> On l>- CS O •^' CO "^ to ro O to NO CO to NO to ■t^ -^ On O fO to to NO 0\ 00 NO J:^ CN fO NO to O 00 T-H .,-1 C^J -.-< # t|h tH NO CN Tt* (N t— NO On to NO r^ oo 00 o to ■* f^ to CO CO CO CO 00 CO On to ■* T^i O O CO -sti On t- CS ■* CN ^ CN O O t^ t^ OO NO On Tt O CO 1^ 00 On CO to rjH to ir^ to to rj< CO O On O CO to •^ to cs On CS On r^ to CO CO (N csi •* to On OO NO O to 00 On to I~~ to NO t^ t^ pq pq pq pq '-I CS CO "* oooo CN CN CS CN < < < < ■<-i CN CO rfi OOOO CO CO CO CO pq pq pq pq ■^ CM CO •* OOOO CO CO CO CO ■^ CN CO ■<* OOOO ^ T^ ^ ^ OOOO to to to to xl X) Ti a a a O a a m en a o 1) S a a 3 3 thing imestone ypsum ypsum a 'A ^ o o !z; h-I o o ^ ^ O O ^o g ^5 a a :z; H^i O o o .3 a, bo pq ci lo o ^ eu a, S. a >> a "^ o u .s ni CI, >H in 1) X) Pl^ 3 3 S ^ O nS ra ,fl u T3 nJ T-l 2Fe(OH)3 + 3H2CO3 or (7) FePOi + 3H6 N02-^Fe(N02)3 +H 3PO4 Chemically, one would expect that the availabihty of tricalcium phosphate would be suppressed by the action of limestone. Carbonic acid and nitrous acid produced in soil will react in part at least with the limestone. (8) CaCOs + H2CO3-* CaH2(C03)2 (9) CaCOa + 2HN02-^ Ca(N02)2 + H2CO3 Apart from this factor the introduction of the common calcium ion will tend to force the equilibrium of the following equation to the left rather than to the right. (10)Ca3(PO4)2+ 2H2CO3 ?± CaH2(P04)2+CaH2(C08)2 Truog (77) was able to demonstrate by pot cultures that the introduction of soluble calcium ions into the soil solution tended to lower the rate of assimila- tion of phosphorus from tricalcium phosphate. This applies especially to plants which do not feed heavily on calcium. 370 JACOBUS STEPHANUS MARAIS In the literature survey, Hilgard (27) was quoted as stating that in calcareous soils relatively smaller percentages of phosphoric acid will suffice for good plant growth than in acid soils. Truog (77) referring to observations of a decrease in growth of cereals due to the addition of lime carbonate in pot cul- tures, makes the following statement "This decrease in availability is undoubtedly due to a condition which is temporary. In becoming acid a soil goes into a condition which takes years to develop, and the addition of lime carbonate causes many profound changes, some of which may affect the availability of the phosphorus. The veiy favorable results obtained by investigators in long continued field experiments involving the use of ground limestone is strong evidence that any unfavor- able result at the start is due to temporary conditions." If we consider the fact that in a soil, and even a calcareous soil, there is con- siderably more aluminum, as a rule, than calcium, we cannot but believe that during the ages of weathering to which soils have been subjected, a considerable quantity of phosphorus has gone into combination with aluminum. This phosphorus will be readily available to plants in a calcareous medium as had already been explained. Is it not the aluminum phosphate in the soil rather than the calcium phosphate that has caused Hilgard to express the opinion quoted above? On the other hand, the favorable results obtained in field experiments as the result of long continued use of limestone may be explained in the following manner. Legumes in general grow better in limed soils. Good farm practice would, therefore, result in the incorporation of more organic matter in the soil and especially of more highly nitrogenous organic matter. The limestone creates conditions favorable for biological activity in the soil. The organic matter is more rapidly decomposed and hence there is rapid pro- duction of carbon dioxide and nitrous acid. These acids may readily produce acid zones in the soil. In such a heterogeneous mass as the soil, it is not difii- cult to conceive of acid and alkaline or neutral zones in close proximity. These zones will naturally not be stationary. Acid zones will continuously be formed and again destroyed. In the acid zones, tricalcium phosphate will be dis- solved and rendered available to plants; in the alkaline zones, aluminum phos- phates will be hydrolyzed and rendered available to crops so that, even there, soluble phosphorus will not be lacking entirely. Lime, as such, un- doubtedly reduces the availability of tricalcium phosphate but due to its effect on the organic matter and on the biological activities of the soil, it acts indirectly as a liberator of phosphorus. Gypsum was added to certain pots in an endeavour to stimulate root growth in the plants and so improve the feeding capacity of the plants. If an in- crease was to be expected one would have looked for it in connection with the use of aluminum and iron phosphates. With the calcium phosphates, the introduction of the common ion calcium would result in a reduction of yield according to Truog (77). No such reduction can be said to have been observed. The gypsum seems to have been without effect of any kind. It may be pointed out, too, at this time that the choice of calcium nitrate as a nitrogen fertilizer AGklCULTURAL VALUE OP INSOLUBLE MINERAL PHOSPHATES 371 was probably unfortunate in that it may have caused a reduction in yields on the calcium phosphate pots, due to the introduction of the common ion cal- cium. On the other hand the Russian work (10) has shown that calcium nitrate is the best form of nitrogen to apply for buckwheat. It is decidedly superior to sodium nitrate. The use of ammonium salts was avoided since it would have introduced the factor of extremely rapid nitrification and the copious production of acids. Buckwheat and sweet clover were chosen as crops because of their reputation as strong feeders on insoluble phosphates. Experiment 2 In order to test the various phosphates under conditions where no soil phosphorus was present it was thought advisable to compare their action in sand culture. Buckwheat and sweet clover were again chosen as the crops to be grown. The phosphates, lime and gypsum were applied by thoroughly mixing them in the sand. The rest of the required plant-food nutrients were added in a culture solution composed of 10 cc. of each of the following solutions and the mixture diluted to a liter. 164 gm. of calcium nitrate in 2500 cc. 50 gm. of potassium sulfate in 2500 cc. 20 gm. of magnesium sulfate in 2500 cc. 0.01 gm. of ferric chloride in 2500 cc. One liter of nutrient solution was added at the time of planting, another liter after 3 weeks, a third liter 2 weeks later, and thenceforth a liter was applied every week. The same pots were employed as in the former experiment and the same quantities of the phosphates, gypsum, and limestone were employed. Each pot contained 12 pounds of sand. The rate of application of fertilizers were therefore: Phosphates, 1000 pounds of 65 per cent rock phosphate per acre Gypsum, 200 pounds per acre Limestone, 1 ton per acre The fertilizers applied in the solid form were thoroughly incorporated into the sand. All the pots were planted in duplicate. The sweet clover pots had an "x" prefixed before each number. The buckwheat pots were planted on March 13, 1920, and harvested on May 13. The sweet clover pots were planted on March 13, 1920, and harvested on May 20. Through an error, one of the pots in the buckwheat series did not receive any phosphorus and had to be discarded. As in the former experiment 30 seeds were planted in each pot, and as time went on the weaker plants were pulled out until finally the buckwheat pots each contained 7 plants and the sweet clover pots each 10 plants. On April 2, April 23, and May 4, all the pots were sprayed with nicotine sulfate solution to kill thrips with which the plants had become infested. 372 JACOBUS STEPHANUS MARAIS TABLE 7 Treatments applied to various pots NUMBER OF POT TREATMENT IN ADDITION TO NUTRIENTS NUMBER or POT TREATMENT IN ADDITION TO NUTRIENTS I201A Laz.* 1041A FL, R. 1202A Laz., L 1042A Fl. R., L 1203A Laz., G 1403A Fl. R., G 1204A Laz., G, L 1404A Fl. R., G, L 1201B Wav. 1401B Lgg- 1202B Wav., L 1402B Lgg., L 1203B Wav., G 1403B Lgg, G 1204B Wav., G,L 1404B Lgg., G, L 1201C Sal. 1501 Ac. P. 1202C Sal., L 1502 Ac. P., L 1203C Sal., G 1503 Ac. P., G 1204C Sal., G, L 1504 Ac. P., G, L 1301 A Duf. 1601 Bone 1302A Duf, L 1602 Bone, L 1303A Duf., G 1603 Bone, G 1304A Duf., G, L 1604 Bone, G, L 1301B Viv. 1302B Viv., L 1303B Viv., G 1304B Viv., G, L * For explanation of abbreviations see table 4. t Included in each liter of nutrient solution at rate of 10 cc. of solution containing 26 gm. Na2HP04 in 2500 cc. DISCUSSION AND RESULTS OF EXPERIMENT 2 The weights of the crops produced are recorded in tables 8 and 9. This experiment bears out even more markedly than the former one the relative abihty of buckwheat and sweet clover to assimilate phosphorus from the various sources employed. On the unlimed pots bonemeal was on a par in value to disodium hydrogen phosphate in solution. Vivianite proved to be an excellent source of phosphorus. As in the first experiment, the addition of lime and gypsum had no effect on its availability. The average yield from the vivianite pots was equal to the average yield of all the Florida phosphate and Laingsburg pots. On the unlimed pots the tricalcium phosphate minerals proved superior to vivianite; on the limed pots, inferior. In the buckwheat series, wavellite and Saldanha phosphates in the limed pots were on a par with Florida and Laingsburg phosphates in the unlimed pots. In the sweet clover series this does not hold. The sweet clover made considerably better growth with the aluminum phosphates on the limed pots than on the unlimed pots, AGRICULTURAL VALUE OP INSOLUBLE MINERAL PHOSPHATES 373 TABLE 8 Weights of crops of buckwheat produced in sand culture WEIGHT OF CROP 1 POT NUMBER TREATMENT IN ADDITION TO NOT^IENTS YIELD First pot Second pot Average gm. gni. gm. per cenl pot 1501 = 100 per cent 1201A Laz.* 1.20 1.20 1.20 7.8 1201B Wav. 4.76 5.70 5.23 34.2 1201C Sal. 5.79 5.09 7.44 48.6 1301A Duf. 2.26 2.14 2.20 14.4 1301B Viv. 9.77 9.81 9.79 64.0 1401A Fl. R. 9.23 9.40 9.32 60.9 1401B Lgg- 12.66 10.44 11.55 75.6 1501 Ac. P. 14.87 15.72 15.30 100.0 1601 Bone 15.02 15.47 15.25 99.7 per cent pot 1502 = 100 per cent 1202A Laz., limestone 1.46 1.38 1.42 9.5 1202B Wav., limestone 9.45 9.98 9.72 65.1 1202C Sal., limestone 9.49 8.92 9.21 61.7 1302A Duf., limestone 2.17 2.26 2.23 14.9 1302B Viv., limestone 9.42 10.04 9.73 65.2 1402B Fl. R., limestone 6.39 6.97 6.68 44.7 1402B Lgg., limestone 6.00 6.73 6.37 42.7 1502 Ac. P., limestone 14.93 14.93 100.0 1602 Bone, limestone 12.19 11.49 11.84 79.9 per cent pot 1503 = 100 per cent 1203 A Laz., gypsum 1.17 1.23 1.20 8.0 1203B Wav., gypsum 5.69 6.31 6.00 40.3 1203C Sal,, gypsum 5.63 5.40 5.52 36.9 1303 A Duf., gypsum 2.12 2.28 2.20 14.7 1303B Viv., gjrpsum 10.01 9.39 9.70 64.9 1403A Fl. R., gypsum 10.59 11.03 10.81 72.3 1403B Lgg-, gypsum 10.10 11.50 10.80 72.2 1503 Ac. P., gypsum 14.71 15.18 14.95 100.0 1603 Bone, gypsum 14.37 13.65 14.01 93.6 per cent pot 1504 = 100 per cent 1204A Laz., limestone, gypsum 1.39 1.55 1.47 9.6 1204B Wav., limestone, gypsum 9.47 10.16 9.82 64.4 1204C Sal., limestone, gypsum 10.37 9.09 9.73 63.8 1304A Duf., limestone, gjnpsum 2.01 2.41 2.21 14.5 1304B Viv., limestone, gypsum 10.40 11.29 10.85 71.2 1404A Fl. R., limestone, gypsum 6.24 6.73 6.49 42.6 1404B Lgg., limestone, gypsum 6.16 5.80 5.98 39.2 1504 Ac. P., limestone, gypsum 15.81 14.66 15.24 100.0 1604 Bone., limestone, gypsum 11.60 9.18 10.39 68.2 ■ For explanation of abbreviations see table 4. 374 JACOBUS STEPHANUS MARAIS TABLE 9 Weights of crops of sweet clover produced in sand cultures WEIGHT or CROPS POT NUMBER TREATMENT IN ADDITION TO NUTRIENTS YIELD First pot Second pot Average gm. gm. gm. per cent pot xlSOl = 100 per cent X1201A Laz.* 2.18 2.29 2.24 19.7 X1201B Wav. 2.51 2.61 2.56 22.5 X1201C Sal. 2.6 2.86 2.73 24.0 X1301A Duf. 2.40 2.11 2.31 20.3 X1301B Viv. 8.53 9.82 9.18 80.8 X1401A Fl.R. 10.37 10.54 10.46 92.1 X1401B Lgg- 10.13 10.06 10.15 89.3 xl501 Ac. P. 11.74 10.98 11.36 100.0 xl601 Bone 12.53 13.02 12.78 112.5 per cent pot xl 502 = 100 per cent X1202A Laz. limestone 3.01 3.10 3.56 26.6 X1202B Wav., limestone 6.72 6.50 6.61 49.4 X1202C Sal., limestone 7.35 7.38 7.37 55.0 X1302A Duf., limestone 2.26 2.09 2.18 16.3 X1302B Viv., limestone 9.20 8.41 8.81 65.8 X1402A FI. R., limestone 8.63 8.40 8.52 63.6 X1402B Lgg., limestone 7.31 7.74 7.53 56.2 xl502 Ac. P., limestone 13.42 13.36 13.39 100.0 xl602 Bone, limestone 10.19 10.42 10.31 76.9 per cent potxl503 = 100 per cent X1203A Laz., gypsum 2.00 1.92 1.96 15.6 X1203B Wav., gypsum 3.14 2.89 3.02 24.1 X1203C Sal., gypsum 2.67 2.81 2.74 21.9 X1303A Duf., gypsum 2.37 2.17 2.27 18.1 X1303B Viv., gypsum 10.54 9.18 9.86 78.6 X1403A Fl. R., gypsum 12.04 11.09 11.57 92.3 X1403B Lgg., gypsum 9.68 9.71 9.70 77.4 xl502 Ac. P., gypsum 12.61 12.47 12.54 100.0 xl602 Bone, gypsum 13.35 12.72 13.04 103.9 per cent potxl504 => 100 percent X1204A Laz., limestone, gypsum 2.98 2.89 2.93 21.8 X1204B Wav., limestone, gypsum 6.71 7.22 6.97 51.9 X1204C Sal., limestone, gypsum 8.13 8.17 8.15 60.7 X1304A Duf., limestone, gypsum 2.42 2.27 2.35 17.5 X1304B Viv., limestone, gypsum 9.79 9.81 9.80 73.0 X1404A Fl.R., limestone, gypsum 8.86 8.28 8.57 63.9 X1404B Lgg., limestone, gypsum 6.80 7.45 7.13 53.1 xl504 Ac. P. limestone, gypsum 13.29 13.54 13.42 100.0 X1604 Bone, limestone, gypsum 10.01 10.76 10.39 77.4 For explanation of abbreviations see Table 4. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 375 but the yields were not as large as those obtained on the unlimed pots fer- tilized with the tricalcium phosphate minerals. The Florida and Laingsburg phosphates, as with the buckwheat, proved inferior in the limed pots to the same phosphates in the unlimed pots, but in this case equally as good as wavel- lite and Saldanha on the limed pots. Lazulite and dufrenite behaved as they did in the first experiment, proving themselves poor sources of phosphorus. A fact to be recorded and probably of some significance is that in the earlier stages of growth of the buckwheat the big differences in total growth on the aluminum phosphate were not so much in evidence. It was during the last 4 or 5 weeks of growth that the plants on the limed pots displayed a greater rate of growth than those on the unlimed pots. Plate 1, figure 1, shows the buckwheat at the age of 6 weeks. The effects of liming is plainly evident where the calcium minerals were applied but not nearly so well marked where aluminum minerals were used. Figure 2 shows that where sweet clover was grown liming showed very marked influence from the very beginning. The pot marked xl705 was one of a series that was discarded because of the series becoming infected with red spider. This pot received a complete nutrient solution in which the phosphorus was supplied in the form of monocalcium phosphate. Besides this the pot was limed and treated with 14 gm. of Florida rock phosphate, i.e., rock phosphate at the rate of 7 tons per acre. These observations tally with the explanation as to assimilability of the phosphates of aluminum and calcium, i.e., in an unlimed medium the availa- bility of aluminum phosphate will decrease as time goes on, whereas the effect of lime on the calcium phosphates will be in evidence immediately. This is considered strong evidence in favor of the explanation as to the effect of lime on the availability of aluminum phosphate. It is remarkable that similar results have been obtained with buckwheat and sweet clover. Both crops, of coarse, are known to be heavy feeders on phosphates; but, on the one hand, buckwheat has a rather limited rooting sys- tem while sweet clover, on the other hand, has a very extensive rooting system. It seems, therefore, that the two plants should vary considerably in feeding power or else in the manner in which they feed. It is possible to conceive of the idea that the sweet clover may have been injured by aluminum on the alu- minum phosphate series. Soluble aluminum in any form would be injurious to sweet clover. This perhaps explains why sweet clover did not respond as well as buckwheat to treatment with aluminum phosphate. Sweet clover roots probably excrete more carbonic acid than do buckwheat roots. Alu- minum phosphate is not as readily dissolved by carbonic acid as is tricalcium phosphate. Buckwheat may feed more heavily on phosphorus rendered soluble by hydrolysis by reason of a more rapid removal of phosphorus from the root-hairs to the growing parts of the plant. It must be borne in mind that in these sand cultures, microorganisms do not play the part that they do in soils. None of the pots were inoculated with soil infusion and no nitrifiable material was added. In all probability all pots were infected with 376 JACOBUS STEPHANUS MARAIS some kind or kinds of organisms but it is not likely that any organisms that could affect the availability of phosphorus appreciably could have been present, or even if they had been present could have exerted any influence, so that the plants had to obtain phosphorus by one of the following methods: 1 . The solution of phosphates in the nutrient medium. 2. By hydrolysis and consequent solution of phosphates. 3. By solvent effect of acid root excretions, which would be, according to Czapek (13), chiefly through the agency of carbonic acid. Phosphorus brought into solution by the first two methods should be equally available to both crops. The phosphorus obtained by the third method would depend on the individuality of the crop in regard to the quantity of the car- bonic acid excreted. The differences in feeding power between the two crops under the conditions of the experiment would in all probability have to be ascribed to the rate of carbonic acid excretion, unless there is a difference between the plants in the rate at which phosphorus is translocated from the root hairs to the growing parts of the plants. It was thought that in sand culture we would be able to duplicate Truog's (77) results with regard to the effect of soluble calcium salts on the availa- bility of the tricalcium phosphate minerals (i.e., reduce it); but gypsum, as in soil cultures, appeared to have no effect on the growth of the plants. It is, of course, possible that due to the large excess of calcium already present in the form of calcium nitrate the additional effect of the calcium ions from the gypsum was too smaU to register an appreciable difference in crop growth. Experiment 3 The purpose of this experiment was to determine the effect of nitrification of urea on the availability of the various phosphates both in soil and sand cultures. The experiment was commenced in the fall of 1920. The media for growth employed were the brown silt loam and the yellow silt loam described in the first section of this paper and pure quartz sand. Throughout the experiment phosphorus and limestone were applied in the same quantities as in former experiments. In the case of the sand cultures the nutrient was applied at the same rate as in previous sand cultures. Where the pots received urea, 0.75 gm. was applied to each pot, i.e., such a quantity was added that, if all the nitrogen in it were converted to nitric acid, enough acid would be formed to displace exactly all the phosphoric acid from the tricalcium phosphate ap- plied to each of the pots treated with it. The pots, which did not receive urea, received an application of 2.05 gm. of calcium nitrate, i.e., nitrogen in equal quantity to that added to the pots receiving urea. The treatments were applied to all the pots and water added to the optimum amount in each pot. The sand pots each received 50 cc. of a soil infusion. The pots were then left unplanted for 14 days, the object being to allow the urea to be nitrified before the germination of the seed. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 377 TABLE 10 Treatments applied to individual pots TREATMENT BROWN SILT LOAM SERIES POT NtJMBERS YELLOW SILT LOAM SERIES POT NUMBERS SAND CULTURE SERIES POT NUMBERS Laz., Ca(N03)2* Laz., Ca(N03)2, L Laz., urea Laz., urea, L llA 12A 13A 14A 61A 62A 63A 64A lllA 112A 113A 114A Wav., Ca(N03)2 Wav., Ca(N03)2, L Wav., urea Wav., urea, L IIB 12B 13B 14B 61B 62B 63B 64B lllB 112B 113B 114B Sal., Ca(N03)2 Sal., Ca(N03)2, L Sal., urea Sal., urea, L lie 12C 13C 14C 61C 62C 63C 64C lllC 112C 113C 114C Duf., Ca(N03)2 Duf., Ca(N03)2, L Duf., urea Duf., urea, L 21 22 23 24 71 72 73 74 121 122 123 124 Fl. R., Ca(N03)2 Fl. R., Ca(N03)2, L Fl. R., urea Fl. R., urea, L 31A 32A 33A 34A 81A 82A 83A 84A 131A 132A 133A 134A Lgg., Ca(N03)2 Lgg., Ca(N03)2, L Lgg., urea Lgg., urea, L 31B 32B 33B 34B 81B 82B 83B 84B 131B 132B 133B 134B Bone, Ca(N03)2 41 91 141 Bone, Ca(N03)2, L 42 92 142 143 144 Bone, urea Bone, urea, L 43 44 93 94 Ac. P., Ca(N03)2 Ac. P., Ca(N03)2, L Ac. P., urea Ac. P., urea, L 51 52 53 54 101 102 103 104 151 152 153 154 Ca(N03)2 Ca(N03)2, L Urea Urea, L Check 1 " 2 " 3 " 4 Check 1 " 2 " 3 « 4 No check group ■ For explanation of abbreviations see table 4. 378 JACOBUS STEPHANUS MARAIS The materials added to the pots were ground together in a mortar and thoroughly mixed before they were incorporated in the growing media. Thor- ough mixing of the media and the material was effected. Five series of pots were used, two containing brown silt loam, one contain- ing yellow silt loam, and two containing sand. For each soil, one series was planted to wheat. The two extra series were planted to annual white sweet clover and each of these pots was numbered the same as its corresponding wheat pot except for an "x" prefixed to the number. The pots in the sand series each received in addition a nutrient solution containing magnesium sul- fate, potassium sulfate, and ferric chloride. The concentrations of the salts in this solution, the manner and time of application are exactly as described under experiment 1. The treatments applied to the individual pots in all the series are indicated in table 10. Sweet clover pots in the first series were planted on October 25 and 26, 1920; the wheat pots, on October 27 and 28, 1920; and a second series of pots containing sand were planted to annual white sweet clover on November 7, 1920. DISCUSSION AND RESULTS OF EXPERIMENT 3 Due to poor light conditions during the winter months all the plants made very slow growth so that harvesting occurred only toward the middle of March. During the growing season the greenhouses were fumigated on two occasions with "nicofume" to rid the plants of aphis. On three separate occasions spraying with nicotine sulfate was resorted to in order to kill the thrips with which the plants had become infested. The last 6 weeks of the growing periods when the days were becoming longer the plants grew most rapidly. The sweet clover, especially, remained stunted to a considerable extent in the earlier grow- ing period. On March 4, the wheat on the sand cultures was harvested. On March 13 and 14, the wheat on the soil cultures was harvested and on March 18, all the sweet clover pots were harvested. The crops were kept in paper bags dried in an oven at 105°C. and weighed. The weights of crops obtained are recorded in tables 11, 12, 13, 14, and 15. From the above tables the percentage increase in growth as a result of the treatments can be determined. In general, the yields from the pots not treated with urea substantiate the findings of the first experiment with respect to the comparative availability of the various phosphates and the effect of lime on the assimilabihty of the phosphorus. With wheat, acid phosphate gave easily the best results on both limed and unlimed soil, while wavellite and Saldanha phosphates on limed soil and bonemeal on unlimed soil were second in their effect. Florida rock and Laingsburg phosphates were effective on the un- limed soil but failed to produce much increased growth on the limed pots. The sweet clover yields on the brown silt loam series showed that, even where pots had been limed, the tricalcium phosphates were a better source of phos- AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 379 TABLE 11 Weight of wheat from brown silt loam series YIELD OF CROP INCREASE POT NUMBER TREATMENT OVER First pot Second pot Average CHECK 1 Sw. gm. gm. per cent llA 5.63 5.73 5.68 -1.6 IIB Wav., Ca(N03)2 6.68 6.20 6.53 13.2 lie Sal., Ca(N03)2 7.57 8.42 7.99 38.3 21 Duf., Ca(N03)2 5.67 5.58 5.63 -2.4 31A Fl. R., Ca(N03)2 8.37 8.92 8.65 49.9 31B Lgg., Ca(N03)2 8.69 8.31 8.50 47.3 41 Bone, Ca(N03)2 9.57 9.07 9.32 61.5 51 Ac. P., Ca(N03)2 10.72 10.31 10.52 82.3 Check 1 Ca(N03)2 alone 5.73 5.81 5.72 INCREASE OVER CHECK 2 per cent 12A Laz., Ca(N03)2, L 6.32 6.41 6.37 10.4 13.6 12B Wav. Ca(N03)2,L 8.64 9.42 9.03 56.5 61.5 12C Sal , Ca(N03)2, L 8.57 9.42 9.00 56.3 61.0 22 Duf., Ca(N0s)2, L 5.71 5.46 5.59 -3.1 32A Fl. R., Ca(N03)2, L 7.09 7.17 7.13 23.6 27.5 32B Lgg., Ca(N03)2,L 6.23 6.26 6.25 8.3 11.8 42 Yone., Ca(N03)2, L 8.18 7.84 8.01 38.8 43:3 52 Ac. P., Ca(N03)2, L 10.03 10.67 10.35 79.4 85.2 Check 2 Ca(N03)2, L 5.71 5.46 5.59 -3.1 INCREASE OVER CHECK 3 per cent 13A Laz., urea 12.80 13.27 13.04 125.9 25.0 13B Wav., urea 14.95 14.15 14.55 152.2 39.5 13C Sal., urea 14.71 15.88 15.30 165.1 46.7 23 Duf., urea 13.69 15.31 14.50 153.2 39.0 33A Fl. R., urea 20.82 19.70 20.26 251.1 94.2 33B Lgg., urea 19.68 19.97 19.83 243.7 90.1 43 Bone, urea 18.50 19.82 19.16 232.1 83.7 53 Ac. P, urea 16.41 16.84 16.63 188.2 59.5 Check 3 Urea alone 10.09 10.77 10.43 80.7 INCREASE OVER CHECK 4 per cent 14A Laz., urea, L 10.73 10.75 10.74 86.3 23.6 14B Wav., urea, L 10.96 12.00 11.48 98.9 31.9 14C Sal., urea, L 11.74 12.54 12.14 110.4 39.7 24 Duf., urea, L 12.07 11.42 11.80 104.5 35.7 24A Fl. R., urea, L 17.76 16.55 17.16 197.4 97.5 34B Lgg., urea,L 14.80 14.03 14.42 149.9 65.9 44 Bone, urea, L 17.25 16.80 17.03 195.1 95.9 54 Ac. P., urea, L 15.35 15.24 15.30 165.2 76.1 Check 4 Urea,L 8.48 8.89 8.69 50.6 380 JACOBUS STEPHANUS MARAIS TABLE 12 Yields of sweet clover from brown silt loam soil series YIELD OF CROA INCREASE POT NUMBER TREATMENT OVER X CHECK 1 First pot Second pot Average gm. gm. gm. per cent xllA Laz., Ca(N03)2 4.V8 4.41 4.60 -3.4 xllB Wav., CaCNOs)! 5.44 5.70 5.57 17.0 xllC Sal., Ca(N03)2 4.65 5.48 5.06 6.3 x21 Duf., Ca(N03)2 4.70 5.47 5.09 6.9 x31A Fl. R., Ca(N03)2 7.93 7.15 7.54 58.4 x31B Lgg., Ca(N03)2 7.46 7.26 7.36 54.6 x41 Bone, Ca(N03)y 10.57 9.66 10.12 112.6 x51 Ac. P., Ca(N03)2 6.72 7.90 7.31 53.6 xCheck 1 Ca(N03)2, alone 4.80 4.22 4.76 INCREASE OVER X CHECK 2 per cent xl2A Laz., Ca(N03)2, L 5.45 4.95 5.20 9.5 -0.4 xl2B Wav., Ca(N03)2,L 6.00 6.46 6.23 30.9 19.3 xl2C Sal., Ca(N03)2, L 6.84 6.13 6.49 36.3 24.3 x22 Duf., Ca(N03)2,L 5.44 5.31 5.38 13.0 3.1 x32A Fl. R., Ca(N03)2, L 6.68 7.05 6.87 44.3 31.6 x32B Lgg., Ca(N03)2,L 6.70 6.23 6.47 35.9 23.9 x42 Bone, Ca(N03)2, L 8.70 7.80 8.25 73.3 58.0 x52 Ac. P., Ca(N03)2, L 10.69 9.56 10.13 112.8 94.1 xCheck 2 Ca(N03)2, L 5.12 5.32 5.22 9.7 INCREASE OVER X CHECK 3 per cent xl3A Laz., urea 4.87 4.53 4.70 -1.3 13.3 xl3B "Wav., urea 5.62 6.21 5.92 24.4 42.7 xl3C Sal., urea 5.11 5.24 5.18 8.8 27.2 x23 Duf., urea 7.21 6.80 7.01 47.3 68.9 x33A Fl. R,, urea 8.12 8.37 8.25 73.3 98.8 x33B Lgg., urea 6.24 7.67 6.96 46.2 67.7 x43 Bone, urea 7.33 8.03 7.68 61.3 85.1 x53 Ac. P., urea 7.55 7.23 7.39 55.3 33.3 xCheck 3 Urea alone 4.38 3.92 4.15 -12.8 INCREASE OVER X CHECK 4 per cent xl4A Laz., urea, L 6.72 5.91 6.41 34.7 19.1 xl4B Wav., urea, L 7.34 6.30 6.82 43.3 26.8 xl4C Sal., urea, L 6.73 6.92 6.83 43.3 26.8 x24 Duf., urea, L 6.24 6.42 6.33 32.9 17.7 x34A Fl. R., urea, L 9.64 8.99 9.33 96.0 73.4 x34B Lgg., urea, L 9.67 8.94 9.31 95.6 73.1 x44 Bone, urea, L 9.90 10.86 9.38 97.1 74.4 x54 Ac. P., urea, L 10.89 11.74 11.32 137.8 110.4 xCheck 4 »JlCa, Li. 5.29 5.47 5.38 13.0 AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 381 phorus for this plant than aluminum and iron phosphates. The effect of liming was concordant with the former findings; it was in the relative powers of assim- ilating the various phosphates that the plants differed. Wheat uses alumi- num phosphates more readily than does sweet clover. Sweet clover is a heavy TABLE 13 Yields of wheat from pots in yellow silt loam series. {A large number of crops were lost due to the oven, in which they were being dried, becoming overheated and crops being charred) YIELD OF CROPS INCREASE POT NUMBER TREATMENT OVER CHECK 1 First pot Second pot Average gm. gm. gm. per cent 61A Laz., Ca(N03)2 5.36 4.87 5.12 -1.2 61B Wav., Ca(N03)2 7.48 7.11 7.30 40.8 61C Sal., CaCNOs)? 5.87 5.63 5.75 11.0 81B Lgg., Ca(N03)2 7.23 7.11 7.17 38.4 Check 1 Ca(N03)2, alone 5.18 5.18 INCREASE OVER CHECK 2 per cent 62A Laz., Ca(N03)2,L 6.12 5.85 5.98 15.4 2.9 62B Wav., Ca(N03)2, L 7.37 9.05 8.21 58.5 41.3 62C Sal., Ca(N03)2, L 8.75 8.44 8.60 66.0 48.0 82B Lgg., Ca(N03)2,L 6.40 5.90 6.15 18.7 5.9 Check 2 Ca(N03)2, L 5.81 5.81 12.2 INCREASE OVER CHECK 3 per cent 63A Laz., urea 6.14 6.29 6.22 20.1 0.2 63B Wav., urea 6.35 6.53 6.44 24.3 3.7 63C Sal., urea 7.89 10.22 9.06 74.9 45.9 83B Lgg., urea 9.10 10.07 9.59 85.1 54.4 Check 3 Urea alone 6.21 6.21 20.0 INCREASE OVER CHECK 4 per cent 64A Laz., urea, L 8.26 7.71 7.99 54.2 30.1 64B Wav., urea, L 9.69 8.44 9.07 75.1 47.5 64C Sal., urea, L 10.07 10.35 10.21 97.1 49.8 84B Lgg., urea, L 8.42 7.97 8.20 58.4 33.5 Check 4 Urea, L 5.92 6.31 6.12 18.1 feeder on calcium, wheat a light feeder. It is to be expected, therefore, that in the presence of lime or even in the absence of lime sweet clover would be capable of assimilating the phosphorus of tricalcium phosphate more readily than would wheat. Truog (77) is of the opinion that oats feed heavily on the natural phosphates of the soil because of their large fibrous root system. 382 JACOBUS STEPHANUS MARAIS Wheat, which is very similar to oats in its general structure and manner of feeding, should therefore also feed heavily on soil phosphorus. The results of TABLE 14 Yields of wheat from pots in the sand series YIELD OF CROP POT NUMBER TREATMENT IN ADDITION TO NUTRIENT SOLUTION First pot Second pot Average gm. gm. gm. lllA Laz., Ca(N03)2 1.90 1.28 1.59 lllB Wav., Ca(N03J2 3.66 4.05 3.86 lllC Sal., Ca(N03)2 3.16 2.22 2.69 121 Duf., Ca(N03)2 1.72 2.04 1.88 131A Fl. R., Ca(N03)2 2.00 1.95 1.98 131B Lgg., Ca(N03)2 2.18 2.07 2.13 141 Bone, Ca(N03)2 3.11 2.26 2.69 151 Ac. P., Ca(N03)2 10.04 20.21 19.63 112A Laz., Ca(N03)2, L 1.53 1.31 1.42 112B Wav., Ca(N03)2,L 5.40 5.01 5.21 112C Sal., Ca(N03)2, L 4.06 4.16 4.11 122 Duf., Ca(N03)2,L 2.23 2.04 2.14 132A Fl. R., Ca(N03)2, L 1.65 1.54 1.60 132B Lgg., Ca(N03)2, L 2.30 2.12 2.21 142 Bone, Ca(N05)2, L 2.38 2.16 2.27 152 Ac. P., Ca(N03)2, L 17.62 19.40 18.51 113A Laz., urea 1.60 2.10 1.85 113B Wav., urea 4.89 5.21 5.05 113C Sal., urea 6.01 6.00 6.01 123 Duf., urea 2.46 2.67 2.57 133A Fl. R., urea 4.30 5.44 4.87 133B Lgg., urea 4.19 3.90 4.05 143 Bone, urea 7.58 7.58* 153 Ac. P., urea 19.34 20.02 19.68 114A Laz., urea, L 1.90 1.25 1.58 114B Wav., urea, L 4.77 5.02 4.90 114C Sal., urea, L 6.20 6.25 6.23 124 Duf., urea, L 1.64 2.00 1.82 134A Fl. R., urea, L 2.22 2.22* 134B Lgg., urea, L 2.40 2.30 2.35 144 Bone, urea, L 2.68 2.68* 154 Ac. P., urea, L 19.59 19.97 19.78 * Weights from only one pot available. The duplicate plants had died soon after ger- mination, presumably from the toxic effect of either the urea or ammonia formed from it. the above experiments justify his conclusions, for the wheat grown in sand culture made but poor growth. On the other hand, the wheat has responded very markedly to phosphate treatment on the soils. It seems logical to be- AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 383 lieve that the more rapid development of a root system in the soil due to the presence of some readily available phosphorus accounts for the greater ability TABLE 15 Yields of sweet clover from the sand series YIELD OF CROP POT NUMBER TREATMENT IN ADDITION TO NUTRIENT SOLUTION First pot Second pot Average gm. gm. gm. xlllA Laz., Ca(N03)2 0.84 0.53 0.69 xlllB Wav., Ca(N03)2 4.00 5.20 4.60 xlllC Sal., Ca(N03)2 5.26 5.53 5.40 xl21 Duf., Ca(N03)2 1.33 1.03 1.18 xl31A Fl. R., Ca(N03)2 7.86 10.23 9.05 xl31B Lgg., Ca(N03)2 9.90 9.73 9.82 xl41 Bone, Ca(N03)2 10.27 9.97 10.12 xl51 Ac. P., Ca(N03)2 11.46 10.17 10.82 xll2A Laz., Ca(N03)2, L 1.19 1.14 1.17 xll2B Wav., Ca(N03)2, L 6.22 6.10 6.16 X112C Sal, Ca(N03)2, L 6.20 7.93 7.07 xl22 Duf., Ca(N03)2, L 1.00 1.03 1.02 xl32A F1.R., Ca(N03)2,L 8.23 9.35 8.79 xl32B Lgg., Ca(N03)2, L 7.73 7.05 7.39 xl42 Bone, Ca(N03)2, L 7.20 7.04 7.12 xl52 Ac. P., Ca(N03)2,L 11.57 10.84 11.21 xll3A Laz., urea 0.48* 0.12* xll3B Wav., urea 0.56* 0.77* xll3C Sal., urea * * xl23 Duf., urea * * xl33A Fl. R., urea 7.62 * xl33B Lgg., urea 8.45 9.39 8.92 xl43 Bone, urea 10.49 >*> 10.49t xl53 Ac. P., urea 7.42 4.32* 7.42t xll4A Laz., urea, L 1.12 1.43 1.28 xll4B Wav., urea, L 3.24 3.54 3.39 X114C Sal., urea, L 4.84 5.46 5.15 xl24 Duf., urea, L * 1.79 1.79t xl34A Fl. R., urea, L * * xl34B Lgg., urea, L 9.27 8.04 8.66 xl44 Bone, urea, L 8.38 * 8.38t xl54 Ac. P., urea, L 11.70 10.92 11.31 * Part or all of plants died within first 3 weeks, t From one pot only. to use phosphorus applied to soil. The poor growth in sand cultures, where only insoluble mineral phosphates are present, is due to the inability to de- velop a root system in the early growth stages. Sweet clover, with its high 384 JACOBUS STEPHANUS MARAIS feeding capacity for calcium, finds enough phosphorus in early stages of growth to develop an extensive root system; and, therefore, in later stages is better equipped to forage for its phosphorus. Sweet clover, of course, always has a very much more extensive root system than wheat. The comparison would be more plain if we could compare buckwheat and wheat. It must be borne in mind, however, that the application of calcium nitrate may have reduced the availability of phosphorus for wheat more than for clover, which is capable of utilizing more calcium. It would be interesting to grow wheat in sand cul- ture with insoluble phosphates and supplying it with a very small quantity of soluble phosphorus just after the germination of the plants. THE ACTION OF TIREA Some remarkable results have been obtained as a result of tlie action of urea. The object of the addition of urea was to determine the effect of ni- trification on the availability of the phosphates. It was an ideal source of organic matter to use because of its being free from phosphorus and conver- tible only into nitrous (or nitric) and carbonic acids, thus leaving no mineral residue in the soil. Furthermore, urea is nitrified very rapidly. The urea is con- verted into ammonium carbonate by the urea organisms present in most soils. The ammonium carbonate is rapidly transformed into nitrous acid and carbonic acid. On the brown silt loam where urea was used without phosphate treat- ment very curious results were obtained. With wheat, very large increases in growth were evident, more in the unlimed pots than in the limed ones. With sweet clover, urea had practically no effect, while on the yellow silt loam series the effect of urea was small. It is evident that in the brown silt loam the acid production as a result of nitrification resulted in the liberation of a considerable quantity of plant food. The reduction in yield of wheat where lime too was applied lends strength to this statement. The failure of response by sweet clover is the result of toxic effect on the plants by the acids produced. The lesser effect of urea on wheat in yellow silt loam series is due to the poor quahty of the soil and the inability of the soil to nitrify the urea as rapidly as the brown silt loam did. There is even a possibility that the urea remained unchanged in the soil long enough to injure the young wheat seed- lings. No such injury was visible at any time, however. THE EPEECT OF THE UREA ON THE AVAILABILITY OF THE PHOSPHATES The effect of the urea on the availabihty of the phosphates themselves was very remarkable. In many instances yields were almost trebled. Wheat benefited considerably more than sweet clover. The increases in yield due to various treatments are recorded in tables 11, 12, and 13. It is evident that urea exerted its greatest influence on the tricalcium phosphates. Where sweet clover was grown, no, or only small, increases were observed with alu- minum phosphates. On the yellow silt loam series, lazulite was the only AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 385 aluminum phosphate which benefited by the presence of the urea. An out- standing feature is the remarkable benefit derived by sweet clover from urea and dufrenite in the unlimed series, especially as dufrenite has proved to be a poor source of phosphorus thus far. The last column on the right of tables 11, 12, and 13 give the increases in yield that may be ascribed as due to the phosphorus alone. It is very doubtful whether the figures can be said to repre- sent the influence of the phosphates because the factors regulating growth are much too intricate to be separated in such an arbitrary fashion; but at the same time they show the percentage increase of growth as due to combined efforts of the urea and phosphates as against the influence of the urea or the calcium nitrate. Plate 2 shows photographs of the various wheat pots which give a clear picture of the effect of the various treatments. The tremendous increase in growth can only be ascribed to the ability of the urea to render plant food available as a result of its transformation into acids. The small quantity of soil in each pot and hence the relatively small quantity of available plant- food in each pot combined with the large number of plants growing in each pot made it difficult for the plants to grow to any considerable size without the addition of fertilizing materials. The roots made a tangled mass in the soil penetrating into every nook and corner of the pots. Hopkins and Whiting (30) showed that nitrite bacteria could dissolve seven times as much phos- phorus from rock phosphate as would be required by a growing plant in a medium where this phosphate was the only supply of a base. The urea was in intimate contact with the phosphorus and, therefore, admirably situated for the acids produced to act on the phosphates, rather than on other soil materials. All the pots were well stocked with nitrogen. It is evident then that in the pots not treated with urea the plants suffered from phosphate starvation, while in the pots treated with urea, phosphorus was dissolved in plenty and the extensive network of roots in the pots ensured the utilization of a large proportion of the phosphorus thus placed at the disposal of the plants. An observation which lends support to this is the fact that the wheat, growing on the urea-treated pots, developed heads and seed, while only in isolated cases did any of the heads develop at all on any of the other pots. Seed pro- duction and early maturity are coupled with good phosphorus supply. The depression in yields where lime was used together with urea and phosphate and especially tricalcium phosphates, must be explained as due to the neutrali- zation by the lime of part of the acids formed. The smaller response to urea by aluminum phosphate is probably due to the fact that aluminum is a questionable base for the nitrifying organisms which are known to respond to calcium and magnesium as bases. In the sweet clover series the formation of soluble aluminum salts was in all probability detrimental to the plants to which they are known to be very toxic. The difference in yield between check 3 and check 4 in table 12 is large enough to conclude that urea alone had injured the plants but that in the pres- 386 JACOBUS STEPHANUS MARAIS ence of lime it was beneficial. The results from the sand series bear out the above statements. Urea without lime benefited the wheat by supplying at least some soluble phosphorus; on the limed pots the solvent action of the acids was removed so that the calcium phosphates showed hardly any benefit from the addition of urea, but with the aluminum phosphates the lime rendered the aluminum phosphates available so that but small differences were noticeable. Wheat is not so susceptible to the toxic action of aluminum salts of acidity. In the sweet clover series urea and phosphate without lime caused havoc amongst the plants. All the pots treated with aluminum phosphate had their plants severely injured or killed. Even some pots in limed series suffered. This points strongly in the direction that sweet clover was injured by soluble aluminum salts or by acids, or by ammonium or ammonium nitrite. NITROGEN AND PHOSPHORUS CONTENT WITH AND WITHOUT UREA The analyses of some of the crops grown in this experiment (table 16) in- dicate that the differences in growth are due to the supply of available phos- TABLE 16 Analyses of wheat from sand series POT NUMBER TREATMENT NITROGEN PHOSPHORUS per cent per cent lllC Sal., Ca(N03)2 2.37 0.048 112C Sal., Ca(N03)2,L 2.03 0.054 113C Sal., urea 2.42 0.076 114C Sal., urea, L 2.23 0.056 131B Lgg., Ca(N03), 1.96 0.046 132B Lgg., Ca(N03)2,L 2.04 0.047 133B Lgg., urea 2.12 0.057 134B Lgg., urea, L 1.95 0.047 141 Bone, Ca(N03)2 2.01 0.048 142 Bone, Ca(N03)2, L 2.02 0.050 143 Bone, urea 2.03 0.059 144 Bone, urea, L 2.03 0.046 phorus rather than to any other causes. The analyses were of wheat grown in the sand series — both nitrogen and phosphorus content were determined. Experiment 4 The purpose of this experiment was to determine the availability of chemi- cally pure phosphates of aluminum, iron, and calcium and the effect of igni- tion on the availability of mineral phosphates of calcium, iron, and aluminum. In the first part of the experiment, the chemically pure phosphates were compared, with and without the effect of lime, and in order to insure that the calcium phosphate was not placed at a disadvantage by the use of calcium nitrate, comparative pots were planted in which ammonium sulfate was used AGRICULTt'EAL VALUE OF INSOLUBLE MLN'ERAL PHOSPHATES 387 as a source of nitrogen. With the aluminum and iron phosphates a small quantity of calcium silicate was added as a source of calcium where ammonium sulfate was used. TABLE 17 Yields of buckwheat on pots with chemically pure phosphates and ignited mineral phosphates WEIGHT OF CROP POT >njiIBER TREATMENT IN ADDITION TO NTTRIEXT SOLUTION First pot Second pot Average ivi. gm. gm 1 Al. P., (NH4)2S04,CaSi03 3.00 3.30 3.15 2 Al. P., (NH4)2S04, L,CaSi03. 4.67 4.64 4.66 3 Al. P., Ca(N03)2 6.20 6.63 6.42 4 Al. P., Ca(X03)2, L 6.92 7.00 6.96 11 Fe. P., (NH4)2S04,CaSi03 4.47 4.55 4.51 12 Fe. P., (XH4)2S04, L,CaSi03 4.71 4.35 4.53 13 Fe. P., Ca(N03)2 5.39 6.20 5.80 14 Fe. P., Ca(N03)2, L 6.27 5.67 5.97 21 Ca. P., (XH4)2S04 4.31 4.25 4.28 22 Ca. P., (NH4)2S04, L 3.16 3.18 3.17 23 Ca. P., Ca(N03)2 6.65 6.92 6.79 24 Ca. P., Ca(X03)2, L 4.51 4.83 4.67 31 Laz. I., Ca(N03)2 3.33 3.65 3.49 32 Laz. I., Ca(N03)2, L 4.76 4.70 4.73 2>2> Laz., Ca(N03)2 1.56 1.40 1.48 34 Laz., Ca(N03)2, L 2.14 2.57 2.36 41 Wav.L, Ca(N03), 5.96 5.87 5.92 42 Wav. I., Ca(N03)2, L 7.49 8.05 7.77 43 Wav., Ca(N03)2 2.80 2.47 2.64 44 Wav., Ca(N03)2, L 3.59 3.18 3.39 51 Sal. I., Ca(N03)2 5.93 6.06 6.00 52 Sal. I., Ca(X03)2, L 7.89 7.87 7.88 53 Sal., Ca(X03)2 3.03 3.50 3.27 54 Sal., Ca(X03)2, L 4.95 4.04 4.50 61 Duf. I., Ca(X03)i 2.24 2.29 2.27 62 Duf.I., Ca(N03)2 L 2.00 2.29 2.15 63 Duf., Ca(X03)2 1.40 1.86 1.63 64 Duf., Ca(N03)2, L 1.60 1.76 1.68 71 Fl. R. I. , Ca(X03)2 6.31 5.96 6.14 72 Fl. R. I. , Ca(X03)2, L 4.80 4.56 4.68 73 Fl. R., Ca(X03)j 6.11 6.21 6.16 74 Fl. R., Ca(X03)2, L 4.48 4.77 4.63 * Key to Abbreviations: Al. P., Aluminum phosphate; Fe. P., pure ferric phosphate; 1., ignited. The medium used was pure quartz sand which received the same nutrient solution at the same intervals as described in experiment 3. The limestone and other salts were applied in the same manner and quantities as in the former experiments. SOIL SCIENCE, VOL. XIII, NO. 5 388 JACOBUS STEPHANUS MARAIS These pots were planted on January 17, 1920, and harvested March 29. On three separate occasions the plants were dusted with tobacco dust to con- trol the thrips with which they had become infested. The buckwheat plants were thinned until 6 remained in each pot. The crops, when harvested, were placed in paper bags, dried in an oven at 105°C., and weighed. Table 17 gives the weights of crops obtained. DISCUSSION AND RESULTS OF EXPERIMENT 4 The weights of crops obtained are well in accord with what one might expect from the results of experiment 1. Calcium nitrate is evidently a far better form of nitrogen for buckwheat than is ammonium sulfate. Even with calcium phosphate, calcium nitrate proved to be the better form of nitrogen in spite of the fact that ammonium sulfate is supposed to enhance the assimilability of tricalcium phosphates. It is fortunate that calcium nitrate was chosen for the latter half of the experiment as a source of nitrogen. Of the pure phos- phates, we may safely state aluminum phosphate is as available as calcium phosphate and that these two are only slightly superior to ferric phosphate as a source of phosphorus for buckwheat. The effect of lime on the availa- bility of the pure aluminum phosphate was not so much in evidence as it was with the mineral phosphates of aluminum. This is to be expected. The aluminum phosphate is free to hydrolyze readily at the beginning, having no aluminum hydrate associated with it. Only as time goes on and the aluminum hydrate begins to accumulate does the lime exert its influence on the availa- bility of the phosphates. With iron and calcium phosphates the effect of lime resulted as expected. The slight effect of lime on pure aluminum phos- phate, the failure to affect ferric phosphate at all and the great depression in availability of pure tricalcium phosphate is cited again as strong evidence in support of the theory described under the first experiment as to the effect of lime on the availability of aluminum, iron, and calcium phosphates. In connection with the relative availabilities of the three types of phosphates to crops, it must be borne in mind that a plant with great ability to utilize tricalcium phosphate was employed, as has been demonstrated time and again. It is very probable that had wheat, oats, millet, flax, or some such low-feeding- power plant been used instead of buckwheat, aluminum and iron phosphates would have shown up to better advantage since their phosphorus is rendered available more readily by hydrolysis than by the action of roots through the agency of carbonic acid. Acids have less solvent action on aluminum and iron phosphates than on calcium phosphates as has been demonstrated in experiment 3. Ignition of the phosphates has had a remarkable effect on the assimil- ability of the phosphorus. It may be stated here that the phosphates were ignited at a bright red heat for 5 hours. Saldanha phosphate and wavellite lost about 15 per cent of their weight; the other three phosphates about 5 AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 389 per cent. This loss of weight was, of course, taken into account when apply- ing the phosphates so that all phosphates were applied on an equal phosphorus basis. Ignition of the three aluminum phosphate minerals resulted in doub- ling the crops obtained with them as compared with the unignited minerals. Smaller but still significant increases in yields were realized by igniting dufren- ite. Ignition has had no influence on the availability of the calcium phosphate used. The increased availability of the aluminiun and iron phosphates may be due to two causes: 1. Dehydration of the mineral and dehydration of the aluminum and ferric hydrates associated with the phosphate. 2. Destruction of the physical structure of the minerals. The first cause seems to be the more logical one. As has been explained in experiment 1, the basic phosphates are less available than the non-basic ones. With the lazulite, wavellite, and Saldanha phosphates we have associated a quantity of aluminum hydrate. Ignition converts the hydrates into oxides: 2Al(OH)3^Al203 + 3H2O With dufrenite we have associated ferric hydrate which is dehydrated: 2Fe(OH)3-»Fe202 + 3H2O Now aluminum and ferric oxides are far less soluble than aluminum and ferric hydroxides, so that they would, therefore, exert a less depressing effect on the availability of the phosphates. No doubt these oxides will slowly again be converted into hydrates in the soil, but this hydration is slow and furthermore in the case of aluminum hydrate, as hydration proceeds, the lime will precipi- tate the aluminum as calcium aluminate. 2A1(0H)3 + 3Ca(OH)2->Ca3Al206 + 6H2O This precipitation will also be more effective than when the lime has to act on the aluminum hydrate en masse. The inability of lime to remove the ferric hydrate explains the smaller effect of ignition on the availability of the dufrenite. In the unlimed pots treated with ignited aluminum phosphate and in all the pots treated with ferric phosphate, the process of hydration of the oxides is gradual so that most of the alumina and ferric oxide will at first occur as partially hydrated oxides: AIO(OH), Al20(OH)4, FeO(OH), Fe20(OH)4, and numerous others. These partially hydrated oxides are not as soluble as the fully hydrated ones and would therefore exert less influence on the solubility of the phosphates, their hydrolysis, and final assimilation by the plants. The destruction of physical structure of the minerals may, of course, be an important consideration. The solubility of the minerals may readily be greatly altered by destruction of the crystalline structure. The natural 390 JACOBUS STEPHANUS MARAIS solubility of the minerals will greatly affect the rate of hydrolysis of the various minerals. There are large possibilities for accounting for many riddles with regard to phosphates on the basis of crystalline structure of the minerals. Quartaroli (63) claims the existence of two dicalcic phosphates which he Ca represents schematically by Ca/Ca/(HP04)2, and „ is amorphous and transformable into Ca(H2P04)2 and the second is crystalline and not transformable into Ca(H2P04)2. He suggests the possibility of four forms of tricalcium phosphates which he represents schematically as: U) / Ca / Ca / Ca / (P04)2 (HP04)2. The first (B) (C) (D) Ca Ca Ca Ca Ca Ca (P04)5 (P04)5 Ca Ca Ca (P04)5 (A) would be gradually transfonnable into dicalcium phosphate, then mono- calcium and finally into phosphoric acid; (B) would be transformable into the crystalline type of dicalcium phosphate and would not be able to produce any monocalcium phosphate; (C) would pass from tricalcium phosphate to the monocalcium phosphate without yielding the dicalcium phosphate and finally form phosphoric acid; (D) cannot be converted into di- or monocalcium phosphate but passes directly to phosphoric acid. Quartaroli has proven the presence of two lithium phosphates. He claims that phosphorites are mixtures of the four forms of calcium phosphate. Aluminum and iron phosphates would lend themselves to the production of similar isomers. Perhaps the variability inter se in availability of calcium phosphates, aluminum phosphates, and iron phosphates is due to the varying proportions of the different isomers in the several minerals. Ignition may or may not alter the proportions of the various isomers and so exert its effect on the availability of the various phosphates. The differences in the availability of lazulite, wavellite, and Saldanha phos- phates, that of Florida rock and Laingsburg phosphate, etc., may easily be due to the proportions of the various possible isomers as suggested by Quartaroli. It is very remarkable that ignited wavellite and Saldanha phosphates on the limed pots should have produced larger crops than even any of the pure phosphates. The availability of lazulite was much increased as a result of the ignition but in no form did the lazulite prove nearly as good as wavellite or Saldanha phosphates. This proves that there is a fundamental difference between the aluminum phosphates in these three minerals. Neither the crystalline structure nor the presence of aluminum hydrate can be designated AGRICULTURAL VALUE OF LNSOLUBLE MLNERAL PHOSPHATES 391 as the reason. It seems that Quartaroli has made a very notable contribution to our understanding of the phosphates we deal with in agriculture. Plate 3, figure 2, shows the relative growth made by the buckwheat with the ignited and imignited aluminum phosphate minerals. Experiment 5 In this experiment an attempt was made to illustrate some factors, which affect the availability of phosphates. The first factor studied was the solubility of alumimmi phosphate in an alkaline solution. The plan followed was similar to that of Kossovitsch (37) described on page 356 in the survey of literature. Three pots were planted. Inside each pot there was placed a porous pot made of bauxite. The porous pot had the same depth as the gallon pots used throughout this experiment and had a diameter of about 4 inches. This pot allowed the penetration of crystalloids in solution but effectively withstood root penetration through its walls. In pot 1, both the inner and the outer pots were filled with sand. The outer pot received an application of lime and ferric chloride. The rest of the plant- food materials were applied in a nutrient solution containing monocalcium phosphate, potassium sulfate, and magnesium sulfate. In pot 2, the inner pot received an application of 10 gm. of wavellite well mixed with the sand and the outer pot, an application of lime and ferric chloride. The rest of the plant nutrients were added in the form of a nutrient solution containing potassium sulfate, magnesium sulfate, and potassium carbonate. The nutrients were all applied in the manner and amounts already described, except that one-third of the potassium sulfate was replaced by an equivalent in potassium of potassiima carbonate. In pot 3, was a dupHcate of pot 2, except that the wavellite was applied in the outer pot instead of the inner pot. Inoculated annual white sweet clover seed was sowed in the outer pots. All the pots were planted in duplicate. The nutrient solutions were applied only through the inner pot. All the water added was applied to the inner pot, so that the soil solutions reaching the plant roots all passed through the walls of the porous pot. Planting occurred on March 23, 1920, harvesting on May 27, 1920. During the first month of the growing period, it was not thought that the plants in pot 2 would survive. On pot 1, the sweet clover grew luxuriantly. On pot 3, fairly good growth was obtained. At the end of the month the plants in pot 2 suddenly began growing, those near the porous pot first, those farther away in succession until all the plants were growing. Ten plants were finally left in each pot. The yields are given in table 18. Enough growth was obtained on pot 2 to give confidence that the plants obtained phosphorus. Thus, phosphorus must have been dissolved by the 392 JACOBUS STEPHANUS MARAIS nutrient solution and diffused through the walls of the porous pots. In alkaline soils, then, aluminum phosphate will be dissolved by the soil solution. This fact is in accord with Storer's statement (73). Plate 4, figure 1, shows a photograph of the roots of the plants clustered around the porous pots. Chemotaxis is probably the cause for the location of the roots of the plants. The second factor studied was the solvent effect of plant roots on various phosphates. Thin, flat, smooth-surfaced plates of plaster of Paris were made. While the plaster was setting, phosphate was dusted onto one surface through a 200- mesh sieve and smoothed over the surface, so that the entire surface was cov- ered with a thin uniform layer of phosphate which was also firm and smooth. All portions of the plates not covered with phosphate were carefully painted with "asphaltum" paint. With a little practice very satisfactory smooth phosphate surfaces may be produced. Such plates were made using Saldanha TABLE 18 Yields of sweet clover WEIGHT OF CROPS POT NUMBER First pot Second pot Average gm. gm. gm. 1 14.25 13.62 13.94 2 3.27 2.67 2.97 3 7.87 8.7 8.31 phosphate, wavellite and Laingsburg phosphate. Each plate was placed vertically in a pot of sand moistened with a nutrient solution containing po- tassium sulfate, magnesium sulfate, and ferric chloride in the proportions already described in former experiments. The sand also received an appli- cation of limestone and gypsum, the latter at the rate of 200 pounds per acre. There were two plates of each phosphate. In the one case, a sweet clover plant about 6 inches tall was placed with its roots against the phos- phate surface, in the other the plant was set on the side opposite to which the phosphate was to be found. Eight weeks after planting, the plants were removed from the pots and the roots and the plates examined. Plates 4 and 5, accurately depict the effect of the plant roots on the phosphate sur- faces. The roots of the plants were matted all over the surface of the phos- phate, a large portion of which had been removed. The pitted appearance of the plates marked clearly the corrosive effect of the roots. The sweet clover plants using calcium phosphate developed slightly better than the other four plants but the use of only one plant precludes the drawing of any conclusions as to the best phosphate in this form. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 393 The plant roots could find phosphate only in very limited area. Nearly all the roots were confined to that area. The demonstration, while giving no actual proof, indicates that root contact with the insoluble phosphates is im- portant as a factor in the assimilation of such phosphorxis and that roots of sweet clover ha,ve considerable ability in rendering phosphates soluble, prob- ably as a result of acid excretion. It was noticeable that the roots growing against the phosphate plates were considerably flattened. GENERAL DISCUSSION The phosphorus of the soil occurs in the form of organic compounds and minerals; the latter chiefly as aluminum, calcium, iron, and magnesium phos- phates. These mineral phosphates may of course be in the form of complexes with organic matter as suggested by Peterson (51), especially if the phosphates are basic ones. Acid humic bodies and acid silicates may readily form com- pounds with basic phosphates. Truog (77) points out that these compounds may be "very resistant and insoluble compounds." It is the problem of agriculturists to furnish growing crops a plentiful supply of phosphorus to be drawn from the stock in the soil. Each of the various phosphorus-containing compounds has a different degree of availability and, what is perhaps more important, demands certain special conditions for its maximum availability which vary with each type of phosphate. The ideal practice for the farmer is to obtain such conditions as will yield him the largest quantity of available phosphorus per acre. In order to prevent the deteriora- tion of the land, this involves maintaining and generally, too, increasing the stock of phosphorus in his soil. The necessity of growing and plowing under legumes to add nitrogen to the soil involves almost invariably the use of lime- stone. Few legumes thrive in acid soils. Apart from the question of growing legumes it is a known fact that the organisms involved in transforming or- ganic nitrogen into the nitrate form thrive best on calcareous media. Cen- turies of profitable employment of lime is proof enough of the value and the necessity of its use. Drainage, cultivation, liming, and incorporation of or- ganic matter with the soil are essential farm practices on most of the arable soils of the world. These practices must be followed. The kind and amount of phosphorus compounds to be applied, the time and manner of application, the kind of crops grown, the use of catch crops, the employment of other fertilizers not involving phosphorus are the factors which the farmer may use in order to make the best use of the phosphorus of the soil and of insuring a good supply of phosphorus to the crops he grows. The investigations reported in this paper have shown that plants can utilize aluminum, iron, and calcium phosphates to some extent. Certain forms of each of these phosphates are better than others; certain conditions improve, other conditions impair,' the availabiHty of these phosphates. Under all conditions, however, plants are able to obtain some phosphorus from any of the minerals used. The greater the stock of phosphorus in the soil, then the 394 JACOBUS STEPHANUS MARAIS greater the amount the plants can obtain. The greater amount of surface exposed to agencies tending to dissolve the phosphorus, and the greater amount of contact of phosphate with the plant roots are two favorable factors which would multiply the effectiveness of the extra phosphate. Insuring the presence of a large quantity of phosphorus in the soil is the solution of the fundamental soil problem. Hopkins (29) claims that good farming practice renders 1 per cent of the phosphorus in the surface layer available every year. Twenty-three hundred pounds of phosphorus per acre would insure the availability of sufficient phosphorus to produce maximum crops of such plants as corn, oats, alfalfa, wheat, etc. If Hopkins' claim is true, the first step in the solution of the phosphorus problem of the soil would be to raise the phosphate stock of the soil to the above amount. The choice of the type of phosphorus to add is the second problem. The experiments reported above would indicate that as an average, calcium phos- phates are to be preferred. In a soil well stocked with limestone, however, aluminum phosphates may be more desirable. There is no doubt that very satisfactory results may be obtained by the use of aluminum phosphate. The price of the material would be the big factor in determining the choice of phosphates. Aluminum phosphates should never be used on acid soils unless, of course, lime is applied at the same time. In fact it would be prefer- able to apply lime together with the aluminum phosphate so that the two may be in intimate contact in the soil. Tricalcium phosphates are used with greater effect on acid soils. Iron phosphates have a doubtful value. If the phosphate is not basic it may be applied to advantage but the lasting effects will be much lower than that for calcium phosphates or aluminum phosphates on limed soils. In choosing phosphates to apply to soil, discretion should be used. No phosphate material should be used without a preliminary test. The low assimilability of the phosphorus in dufrenite and lazulite is a warning against indiscriminate buying of these phosphates. Aluminum, iron, and calcium phosphates vary as to the manner in which they are rendered available in the soil. It would perhaps be a good policy to apply both aluminum and calcium phosphates to the soil so as to make full use of all the reactions which tend to place phosphorus at the disposal of plants. Considering the time of applying phosphorus, it would be wise to apply phos- phorus for the green-manuring crop especially if clover, sweet clover, rape, mustard, or some such heavy feeder on phosphorus is used. These crops will then place the phosphorus they have used at the disposal of the money crops following. This practice would be especially desirable where aluminum phos- phates are used With calcium phosphates, it would perhaps be more de- sirable to plow the phosphate into the soil with the green-manuring crop in order to utilize to the fullest extent the acids produced during nitrification of the nitrogenous material and at the same time placing the phosphorus in intimate contact with the big source of carbonic acid production. The urea experiment reported above is further support of the results of Hopkins and AGRICULTUIIAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 395 Whiting (30) regarding the efifect of nitrification on the availability of phosphates. Truog (77) makes a suggestion for a rotation of crops in Wisconsin in which he introduces white mustard and rape as catch crops, the former being planted after wheat harvest, the latter, the following year in corn at the last cultivation. The third year clover is seeded in the oats, the fourth year the field remains in clover. This suggestion is an admirable one in the direction of keeping the soil well supplied with organic matter and in using other crops to help the weak feeding Graminae to obtain readily available phosphorus. Acid phosphate and soluble phosphates, in general, are usually too expensive to have a place in building up the phosphorus stock of a soil. They can be used with effect in another direction. If small top dressings of this phosphate be used, they will serve to give the young seedlings a rapid start so that they wUl develop a strong root sj^stem which will then function in feeding the plant in later growth stages. This practice should be used only in connection with a system in which adequate provision is made for stocking the soil with phos- phorus. If not, the practice will prove to be one of the best ways of rendering a poor soil poorer. CONCLUSIONS 1. Mineral phosphates of aluminum and iron are valuable sources of phos- phorus for plants; under certain conditions they are superior to calcium phos- phate, under others inferior. 2. Nitrification of urea with the consequent production of acids acts very favorably in assisting plants to obtain phosphates of almninum, iron, and cal- cium for food. 3. Chemically pure phosphates of aluminum and iron are as readily avail- able to the plants tested as is pure calcium phosphate. 4. Mineral phosphates of aluminum and iron are not as readily available as the pure phosphates of the same metals due to the fact that most of them are hydrated basic phosphates. 5. Igniting the minerals, thereby dehydrating the bases associated with the phosphates and destroying the crystalline structures of the minerals, removes the drawback against the use of mineral phosphates of aluminum and iron. 6. Aluminum phosphates, whether chemically pure or in mineral form, ignited or unignited, always display their maximum effect in a calcareous medium. 7. The effect of iron phosphates is neither enhanced nor depressed by the addition of limestone under the conditions of the experiment. 8. Under the conditions of the experiments, where chiefly neutral growing media were used, tricalcium phosphates were affected adversely by the addi- tion of limestone. 9. An alkaline soil solution dissolves aluminum phosphate and aids the plant in obtaining its phosphorus for food. 396 JACOBUS STEPHANUS MARAIS 10. Contact of the roots of plants with mineral phosphates is a very im- portant factor in the assimilation of the phosphorus by plants for food. (1 (2 (3 (4: (5 (6: (7 (8 (9: (lo: (11 (12 (13 (14: (15 (16: (17 (18 (19: (2o: (21 REFERENCES Andouard, a. 1895 Etude sur la valeur agricole du phosphate d'alumine du Grand Connetable. /w Ann. Agron., v. 21, p. 171-181. Bagtjley, a. 1912 The phosphate nutrition of plants. In Jour. Agr. Sci., v. 4, no. 3, p. 318-322. Balentine, W. 1893 Investigations on the foraging powers of some agricultural plants for phosphoric acid. In Me. Agr. Exp. Sta. Ann. Rpt., 1893, p. 13. Bishop, W. H. 1893 Report on pot experiments with phosphates. In Del. Agr. Exp. Sta. 6th Ann. Rpt., p. 193-202. BoNOMi, Z. 1906 Aluminum phosphate as compared with mineral super-phosphate and Thomas slag. In Ann. R. Staz. Sper. Agr. Udine, v. 8, p. 26-32; ahs. in Exp. Sta. Rec, v. 18, p. 219. Burgess, J. L. 1920 The relative availability of acid phosphate and the native soil phosphates in the presence of pulverized limestone. Part I. In Bui. N. C. Dept. Agr., v. 41, no. 3, p. 24. Btjrkett, C. W. 1898 Thomas slag versus redondite as a fertilizer. 7n N. H. Agr. Exp. Sta. 10th Ann. Rpt., Bui. 59, p. 189-190. BuRLisoN, W. L. 1916 Availability of mineral phosphates for plant nutrition. In Jour. Agr. Res., no. 6, p. 485-514. Cameron, F. K., and Bell, J. M. 1907 The action of water and aqueous solutions upon soil phosphates. U. S. Dept. Agr. Bur. Soils Bui. 41. Chirikov, F. V. 1913 The solvent action of ammonium salts on phosphorites in sand. In Iz. Resul't Veget. Opytov. Lab. Rabot, v. 9, p. 436-440; ahs. in Exp. Sta. Rec, V. 26, p. 321. Chirikov, F. V. 1914 The assimilation of the phosphoric acid phosphorite by plants under different conditions of nutrition. In Iz. Resul't Veget. Opytov. Lab. Rabot., Moscov. Khov. Inst., v. 10, p. 149-171. Chirikov, F. V., and Khardin, N. V. 1914 On the decomposition of calcium phos- phate by acetic acid. In Iz. Resul't Veget. Opytov. Lab. Rabot., Moscov Selsk. Khoz. Inst., V. 22, no. 1, p. 104-114. Czapek, F. 1896 Fiir Lehre von dem Wurzel-ausscheidungen. In Jahrb. Wiss. Bot., V. 29, p. 321. Deherain, p. p. 1902 Traite de chimie agricole, p. 506, Masson and Cie, Paris. Dyer, B. A chemical study of the phosphoric acid and potash contents of the wheat soils of Broadbalk Field, Tothamsted. In Phil. Trans. Roy. Soc. London, v. 68, p. 11. Elliot, W. B., and Hill, H. H. 1910 Contributions to the study of phosphoric acid in soils and fertilizers. In Va. Agr. Exp. Sta. Ann. Rpt. 1909-10. Fraps, G. S. 1909 Active phosphoric acid and its relation to the needs of the soil for phosphoric acid in pot experiments. In Tex. Agr. Exp. Sta. Bui. 126. Fraps, G. S. 1909 Interpretation of soil analysis with respect to phosphoric acid. - In Proc. 26th Ann. Conv. Assoc. Off. Agr. Chem., 1909. Fraps, G. S. 1910 Relation of pot experiments to the active phosphoric acid of the soil. In Jour. Indus. Engin. Chem., v. 2, no. 8, p. 350-352. Fraps, G. S. 1911 Effect of ignition on solubility of soil phosphorus. In Jour. Indus. Engin. Chem., v. 3, no. 5, p. 335. Gaither, E. W. 1910 Effect of lime upon the solubility of soil constituents. In Jour. Indus. Engin. Chem., v. 2, no. 7, p. 315-316. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 397 Gedroits, K. K. 1903 Chemical methods for the determination of the fertility of soils with respect to phosphoric acid. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), no. 4, p. 403-432; ahs. in Exp. Sta. Rec, v. 15, p. 519. Gedroits, K. K. 1907 Availability of phosphoric acid of different phosphates. In Tradui Selsk. Khoz. Khim. Lab. St. Petersburg, v. 6, p. 343-408; abs. in Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.) v. 11, no. 2, p. 228-230, also in Exp. Sta. Rec, v. 24, p. 324. Gedroits, K. K. 1911 The action of phosphorites on soils saturated and unsaturated with bases. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), v. 12, no. 4, p. 529-546; ahs. in Exp. Sta. Rec, v. 26, p. 30. Hebert, U. a. 1897 Les phosphates de I'Herault. In Ann. Agron., v. 23, p. 114. Hendrick, J. 1919 The phosphates in industry. In Jour. Soc. Chem. Indus., v. 38 no. 9, p. 155-157. HiLGARD, E. W. 1914 Soils, p. 365. MacMillan, New York. Hoffmeister, W. 1898 Ein neues Losungsmittel zur Unterscheidung der Phos- phorsauren in verschiedenen Phosphaten. In Landw. Vers. Sta., v. 50, p. 363. Hopkins, C. G. 1910 Soil Fertility and Permanent Agriculture, p. 183. Ginn, New York. Hopkins, C. G., and Whiting, A. L. 1916 Soil bacteria and phosphates. 111. Agr. Exp. Sta. Bui. 190. Iakushkin, I. v., AND KivoKOKOV, P. I. 1913 Some properties of phosphorites from Sengliee. In Iz. Resul't Veget. Opytov. Labs. Rabot, v. 9, p. 160-163; ahs. in Exp. Sta. Rec, v. 34, p. 329. Iakushkin, I. V. 1914 The assimilation of phosphoric acid by cereals. In Izv. Mos- cov. Selsk. Khov. Inst. (Ann. Inst. Agron. Moscou), v. 22, no. 1, p. 66-84; ahs. in Exp. Sta. Rec, v. 37, p. 323. Jordan, W. H. 1913 Studies in plant nutrition: 1, N. Y. Agr. Exp. Sta. Bui. 358. KiLLNER, O., Sakano, H., Sato, D., and Shinjo, S., 1890. Researches on the action of lime as a manure, with special regard to paddy fields. In Bui. Col. Agr., Tokyo Imp. Univ., no. 9, p. 1-25; ahs. in Exp. Sta. Rec, v. 2, p. 762-765. KossoviTSCH, P. S. 1900 Report of the agricultural chemical laboratory of the ministry of agriculture and imperial domains in St. Petersburg for 1898. St. Petersburg: Dept. Agr., 1900, p. 242; ahs. in Selsk. Khoz. i. Lyesov., 1900, 199, p. 237-238; also in Exp. Sta. Rec, vol. 13, p. 235. KossoviTSCH, P. S. 1901 Ammonium salts as a source of nitrogen for plants. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), v. 2, no. 5, p. 625-638 KossoviTSCH, P. S. 1902 The role of the plant in dissolving the plant food of the soil. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), v. 3, no. 2, p. 145-180. KossoviTSCH, P. S. 1904 Quantitative determination of the carbon dioxide secreted by roots during growth. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), V. 5, p. 482-493; abs. in Exp. Sta. Rec, v. 16, p. 1048. KossoviTSCH, P. S. 1909 Plant, phosphorite, and soil according to experiments in the agricultural laboratory of St. Petersburg. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), v. 10, no. 6, p. 782-842. Krober, E. 1909 tJber das Losichwerden der Phosphorsaure aus wasserunloslichen Verbindungen unter der Einwirkung von Bacterien and Hefen. In Jour. Landw., V. 57, p. 5-80. KtJHN, J. 1897 Experiments on the fertilizing action of the phosphoric acid of bone meal. In Fiihling's Landw. Ztg., v. 46, no. 16, p. 471-482; also in Ber. Landw. Inst. HaUe, v. 13, p. 198-208. Merrill, L. H., and Jordan, W. H. 1895 Investigations on the foraging powers of some agricultural plants for phosphoric acid. In Me. Agr, Exp. Sta. Rpt., 1895, p. 10-18. 398 JACOBUS STEPHANUS MARAIS (43) Merrill, L. H. 1898 Box experiments with phosphoric acid from diEferent courses. In Me. Agr. Exp. Sta. Ann. Rpt., 1898, p. 64-74. (44) Morse, F. W. 1903 The effect of moisture on the availability of dehydrated phos- phates of alumina. In Jour. Amer. Chem. Soc, v. 25, p. 280-288. (45) Nagaoka, M. 1904 On the action of various phosphates upon rice plants. In Bui. Col. Agr., Tokyo. Imp. Univ., v. 6, no. 3, p. 215-261 ;a65.?«Exp. Sta. Rec, V. 16, p. 555. (46) Nedokuchaev, N. 1913 On the influence of calcium nitrate and ammonium sul- phate on the assimilation of phosphoric acid of raw phosphates. In Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), v. 14, no. 1, p. 1-21; abs. in Exp. Sta. Rec, V. 29, p. 318. (47) Otryzanev, a. 1906 Influence of lime in connection with different phosphatic fertilizers. In Vyestnik. Selsk. Khoz., no. 37; ahs. in Zhur. Opuitn. Agron. (Russ. Jour. Exp. Landw.), v. 8, no. 2, p. 204^206. (48) Parker, F. W., and Truog, E. 1920 The relation between the calcium and the nitrogen content of plants and the function of calcium. In Soil Sci., v. 10, no. 1, p. 49-56. (49) Paturel, G. 1902 Contribution a I'etude de la dissolution de I'acide phosphorique dans les liquides du sol. In Ann. Agron., v. 28, no. 8, p. 385-398; abs. in Exp. Sta. Rec, v. 14, p. 127. (50) Patterson, H. J. 1907 Fertilizer experiments with different sources of phosphoric acid. Md. Agr. Exp. Sta. Bui. 114. (51) Peterson, P. P. 1911 Effect of heat and oxidation on the phosphorus of the soil Wis. Agr. Exp. Sta. Res. Bui. 19. (52) Peterson, W. H., and Truog, E. 1913 Condition of phosphorus in soil and avail- ability of various phosphates. Wis. Agr. Exp. Sta. Bui. 228. (53) Pfehter, Th., and Blanck, E. 1911 tjber die Wirkung eines Zusatizis von Toner- degel zum Boden auf die Ausnutzung der Phosphorsaure durch die Pflanzen. In Mitt. Landw. Inst. Breslau, v. 6, no. 2, p. 315-324. (54) PiLON, ET AL 1906 Treatment of double phosphates of iron and aluminum in order to render the combined phosphoric acid soluble in ammonium citrate. French patent, 368521, July 30, 1906; abs. in Jour. Soc. Chem. Indus., v. 25, no. 24, p. 1227. (55). Prianishnikov, D. N. 1905 On the effect of ammonium salts on the assimilation of phosphoric acid by higher plants. In Ber. Deut. Bot. Gesell., v. 23, no. 1, p. 8017. (56) Prianishnikov, D. N. 1906 On the relative value of different phosphates. In Landw. Vers. Sta., v. 65, p. 23-54. (57) Prianishnikov, D. N. 1906 On the factors which influence the fertilizing action of difi&cultly soluble phosphates. Abs. in Chem. Ztg., v. 30, no. 37, p. 438. (58) Prianishnikov, D. N. 1909 Influence of calcium carbonate and ammonium sul- phate on the assimilation of phosphoric acid (by plants) from different sources. Paper presented at the Internat. Cong. Appl. Chem., London, abs. in Jour. Soc. Chem. Indus., v. 28, no. 13, p. 722. (59) Prianishnikov, D. N. 1909 On the behaviour of aluminum and iron phosphate. In Izv. Moscov Selsk. Khoz. Inst. (Ann. Inst. Agron., Moscou, v. 15, no. 1, p. 16-23; abs. in Exp. Sta. Rec, v. 22, p. 129. (60) Prianishnikov, D. N. 1909 The influence of calcium carbonate on the action of different calcium phosphates. In Izv. Moscov Selsk. Khoz. Inst. (Ann. Inst. Agron. Moscou), 15, no. 1, p. 1-15; abs. in Exp. Sta. Rec, v. 22, p. 130. (61) Prianishnikov, D. N. 1911 The influence of calcium carbonate on the action of different phosphates. /« Landw. Vers. Stat., v. 75, no. 5-6, p. 357-376; abs. in Chem. Zentbl., v. 2, no. 22, p. 1706. AGRICULTURAL VALUE OF INSOLUBLE MINERAL PHOSPHATES 399 (62) Prianishnikov, D. N., et al 1912 Experiments with different phosphates. In Izv. Moscov Selsk. Khov. Inst. (Ann. Inst. Agron. Moscou), v. 18, no. 1, p. xiii, 1-179. (63) QuARTAROLi, A. 1919 Phosphorites, superphosphates, tetraphosphates. In Staz. Sper. Agr. Ital., v. 52, p. 416-^35; abs. in Chem. Abs., v. 15, no. 6, p. 916-917. (64) RiSLER, E. 1898 Geologie Agricole, p. 29. Berger, Levrault et Cie, Paris. (65) Sachs, J. 1860 Auflosung des Marmors durch Mais-Wurzeln. In Bot. Ztg., v. 18, p. 117-119. (66) SCHLOESESTG, Th., eils 1899 Etude sur I'acide phosphorique dissous par les eaux du sol. In Compt. Rend. Acad. Sci. (Paris), v. 127, no. 4, p. 236-239; no. 6, p. 327-329. (67) ScHLOESiNG, Th., fils 1899 Etude sur I'acide phosphorique dissous par les eaux du sol. In Ann. Sci. Agron., v. 16, p. 316-359. (68) ScHREiBER, C. 1897 Attention called to experiments reported by Schreiber. In Rev. G6n. Agron., 1902, v. 11, no. 7-8, p. 334^341. (69) Semxtshkin, V. V. 1914 The influence of calcium carbonate on the solution of iron phosphate in acetic acid and its availability to plants. In Izv. Rezul't Veget. Opytov. Lab. Rabot (Rec. Trave. Lab. Agron.) Moscov. Selsk. Khoz. Inst. (Ann. Inst. Agron. Moscou), v. 22, no. 1, p. 85-103; abs. in Exp. Sta. Rec, v. 37, p. 324. (70) Shxtlov, I. S. 1909 Sand cultures with raw phosphates. In Izv. Moscov. Selsk. Inst. (Ann. Inst. Agron. Moscou), no. 1, p. 32-73; abs. in Exp. Sta. Rec, v. 22, p. 128. (71) Shulov, I. S. 1911 Sand cultures with different phosphates. 7ra Izv. Moscov. Selsk. Khoz. Inst. (Ann. Inst. Agron. Moscou), v. 17, no. 2, p. 177-198. Abs. in Exp. Sta. Rec, v. 26, p. 427-428. (72) Soderbaum, H. G. 1915 Certain factors which influence the fertilizing action of slightly soluble phosphates. In K. Landtbr. Akad. Handl. och Tidskr., v. 54, no. 5, p. 448-477. In Meddel. Centralanst. Forsokav. Jordbraksomradet, no, 112, p. 36. Abs. in Chem. Abs., v. 9, no. 23, p. 3319. (73) Storer, F. H. 1887 Agriculture, v. 2, p. 169. Schribners, New York. (74) Truog, E. 1912 Factors influencing the availability of rock phosphate. Wis. Agr. Exp. Sta. Res. Bui. 20. (75) Truog, E. 1914 Availability of phosphate to various crops. Wis. Agr. Exp. Sta. Bui. 240, p. 23. (76) Truog, E. 1915 A new theory regarding the feeding power of plants. In Science, n. s., v. 41, no. 1060, p. 616-618. (77) Truog, E. 1916 The utilization of phosphates by agricultural crops, including a new theory regarding the feeding power of plants. Wis. Agr. Exp. Sta. Res, Bui. 41. (78) Truog, E. 1918 Soil acidity: Its relation to plant growth. In Soil Sci., v. 5, p. 169-195. (79) Ulbricht, W, 1892 Ueber die dugende Wirkung des Redonda, Alto Veto und Los Roques Phosphats. In Landbote, 1892, no. 11, p. 79. (80) Wagner, P. 1871 Das Verhalten der Phosphorsaure in Erdboden. 7« Jour. Landw,, v. 19, p. 102-103. (81) Warington, R. 1900 The comparative value of nitrate of sodium and sulphate of ammonium as manures. In Jour. Roy. Agr. Soc, v. 61; p. 300-346. (82) Wheeler, H. J., Adams, G. E. 1906 A comparison of nme different phosphates upon Umed and unlimed land with several varieties of plants. R. I. Agr. Exp. Sta. Bui. 114. (83) Wheeler, H. J., Adams, G. E. 1907 Continued test of nine phosphates with different plants. R, LAgr. Exp. Sta. Bui. 118. 400 JACOBUS STEPHANUS MARAIS (84) Wheeler, H. J. 1910 After effects of certain phosphates on limed and unlimed land. In Jour. Indus. Engin. Chem., v. 2, no. 4, p. 133-135. (85) ZeCCHINI, M. 1900 A new phosphatic manure; prepared phosphate of alumina. In Chem. Industriale, Jan. 1900; ahs. in Chem. News, v. 81, p. 210-212. PLATE 1 Fig. 1. Buckwheat six weeks old in sand culture. Pot number Treatment 1201 A Lazulite 1202A Lazulite and lime 1203 A Lazulite and gypsum 1204A Lazulite, lime and gypsum 1301 A Dufrenite 1302 A Dufrenite and lime 1303A Dufrenite and gypsum 1304A Dufrenite, lime and gypsum Pot number Treatment 1201C Saldanha 1202C Saldanha and lime 1203C Saldanha and gypsum 1204C Saldanha, gypsum and lime 1201B WaveUite 1202B Wavellite and lime 1203B Wavellite and gypsum 1204B Wavellite, gypsum and lime Fig. 2. Sweet clover six weeks old Pot number Treatment xl705 Mono calcium phosphate xl202C Saldanah phosphate and lime xl201C. Saldanah phosphate alone xl202B Wavellite and lime xl201B Wavellite alone xl302B Vivianite and lime in sand cultures showing effect of lime Pot number Treatment xl402B Laingsburg phosphate and lime xl401B .. .Laingsburg phosphate alone xl402 A . . . Florida Rock phosphate and lime xl401A . . .Florida Rock phosphate alone xl601 Bonemeal and lime xl602B . . . Bonemeal alone AGRICULTURAL VALUE OF INSOLUBLE PHOSPHATES JACOBUS STEPHANL'S HARMS PLATE 1 ^i m if IIS '^o^A fSOlS '502A 'SOM ^^04^ Fig. 1. i^sf ^ J ji"' '\ Fig. 2. 401 SOIL SCIENCE, VOL. XIII, NO. 5 PLATE 2 Fig. 1. Wheat on brown silt loam series showing the effect of urea on the availability of aluminum and iron phosphates. Fig. 2. Wheat on yellow silt loam series showing the effect of urea on the availability of the calcium phosphates. 402 AGRICULTURAL VALUE OF INSOLUBLE PHOSPHATES JACOBUS SIEPHANUS MARAIS PLATE 2 Fig. 1. ■L J| PH(BPHATE»C Fig. 2. 403 TLATE 3 Fig. 1. Wheat and clover in sand series showing the best pot with each of various phosphates. Fig. 2. Buckwheat on sand culture showing effect of ignition on availability of aluminum phosphates. 404 AGRICULTURAL VALUE OF INSOLUBLE PHOSPHATES JACOBUS STEPHANUS M.4RAIS PLATE 3 I m m P js ^1'' SALDANNAhIdUFRENITE J FLORIDA RDCK I ROCK PHOSPHATE i30.NlE MEAL ' '. &C1DF I I'IG. 1. Fig. 2. 405 PLATE 4 Fig. 1. Roots of sweet clover clinging to the porous pots. Pot on right received soluble phosphorus. Fig. 2. Effect of sweet clover roots on smooth surface of wavellite. 406 AGRICULTURAL VALUE OF INSOLUBLE PHOSPHATES JACOB as STEPHANUS MARAIS PLATE 4 Fig. 1. WAVELLITE ALUMINUM PHOS. Fig. 2. 407 PLATE 5 Fig. 1. Effect of sweet clover on smooth surface of Laingsburg phosphate (rock phosphate). Fig. 2. Effect of sweet clover on smooth surface of Saldanah phosphate. 408 AGRICULTURAL VALUE OF INSOLUBLE PHOSPHATES JACOBTfS STEPHANUS MARAIS PLATE 5 i^l?A ' \ ^ Fig. 1. Fig. 2. 409 VITA The author of this paper was born June 6, 1896, at Pretoria, Transvaal, South Africa. From 1903 to 1909, he attended the pubHc schools at Malmesbury, in the Cape Province; from 1909 to 1912, the boys' school at Pretoria in the Transvaal;, in 1913, he attended the boys' school at Stellenbosch. In December, 1916, he obtained the Pass B.A. Degree from the University of Cape of Good Hope, and a year later, the Honours B.A. Degree from the same university. In 1918, he entered the Graduate School of the University of Illinois where he was a student up to June, 1921. LIBRARY OF CONGRESS 002 756 358 2 SOIL SCIENCE VOLrME 13, NUMBER 5, MAY, 1922 CONTENTS D. J. Healt and p. E. Karraker. The Clark Hydrogen-Electrode Vessel and Soil Measurements 323 Selman a. Waksman. Microorganisms Concerned in the Oxidation of Sulfur in the Soil: III. Media used for the Isolation of Sulfur Bacteria from the Soil. 329 William Mather. The Effect of Limes Containing Magnesium and Calcium upon the Composition of the Soil and upon Plant Behavior 337 J. S. Marais. The Comparative Agricultural Value of Insoluble Mineral Phosphates of Aluminum, Iron and Calcium 355