^%ehaYior o^'nosphoriclici^^he Soil 
 
 A THESIS. 
 
 SUBMITTED TO THE UNIVERSITY FACULTY OF CORNELL 
 
 UNIVERSITY FOR THE DEGREE DOCTOR 
 
 OF PHILOSPHY 
 
 i: V 
 
 JAMES ADRIAN BIZZEEE 
 
 1903 
 
 ^RiSi OF 
 
 THE ITHACA, JOURNAL 
 
 ITMAGA. N. V. 
 
Behavior of Phosphoric Acid in the Soil 
 
 A THESIS. 
 
 SUBMITTED TO THE UNIVERSITY FACULTY OF CORNELL 
 
 UNIVERSITY FOR THE DEGREE DOCTOR 
 
 OF PHILOSPHY 
 
 BY 
 
 JAMES ADRIAN BIZZELL 
 
 w 
 
 1903 
 
 PRESS OF 
 
 THE ITHACA JOURNAL 
 
 ITHACA, N. Y. 
 
</^^'' 
 
 The author takes this occasion to acknowledge the kindness 
 of Professors G. C. Caldwell, h. M. Dennis and E. M. Chamot, 
 of the Department of Chemistry, under whose guidance the fol- 
 lowing work was carried out. 
 
 Ccrnell Uuiv. Lir-. Bxcban|f« 
 
CONTENTS 
 
 Historical 5 
 
 Summary 20 
 
 Experimental 
 
 I. Methods of Determining Phosphoric Acid 22 
 
 II. Relation between Phosphoric Acid and Compounds 
 
 of Iron and Calcium 23 
 
 (a) actionof lime on ferric phosphate 24 
 
 (b) action of lime on the phosphate formed 
 
 by fixation 27 
 
 (c) action of lime on soils 31 
 
 (d) fixation of phosphoric acid in the 
 
 presence of compounds of iron and 
 
 calcium 34 
 
 III. Behavior of Phosphotic Acid towards Humus 35 
 
 (a) absorption 
 
 (b) solution 40 
 
 IV. Absorption of Phosphoric Acid by Zeolites 41 
 
 V. Action of Ferrous Sulphate 44 
 
 Conclusions 46 
 
HISTORICAL 
 
 The fact that soils are capable of absorbing certain substances 
 from solution has been known for nearl}?^ a century. There has 
 been much discussion as to whom the credit for this discovery is. 
 due. The first authentic record appears to have been made by 
 Gazzeri in 1819. He says : "If extract of dung strongly colored 
 and containing nutritive matter is added to a clayey soil, the liquid 
 is rapidly decolorized. The soil takes hold of the substances in 
 solution, and forms with them compounds which are insoluble,, 
 but which are decomposed by the absorbing action of plants." 
 
 Bronner in his treatise on "Grape Culture in South Ger- 
 many" published in 1836, made similar statements regarding the 
 action of sand and garden soil, and HuxtableMn 1848, apparently 
 ignorant of previous experiments, repeated the work of Gazzeri. 
 A little later the work of these observers was extended somewhat 
 by Thomson'^ of England. 
 
 In 1850 Way^ made a systematic investigation of the absorb- 
 ing power of soils. His efforts were directed almost exclusively 
 toward ascertaining the cause and nature of the absorption of 
 bases. He published only one experiment on the absorption of 
 phosphoric acid. When a solution containing calcium acid phos- 
 phate was filtered through a soil, no trace of phosphoric acid was 
 found in the filtrate. 
 
 The next work bearing on the subject was that of Wicke*, 
 who found that pure marble kept in contact with a solution of 
 superphosphate evolved carbon dioxide, and the greater part of 
 the phosphoric acid was precipitated. 
 
 Thenard^ was the first to make any systematic study of the 
 absorption of phosphoric acid by soils. He succeeded in decol- 
 orizing the liquid from barnyard manure by the use of iron and 
 aluminum hydroxides, and calcium carbonate. He concluded 
 
 1 Jour. Roy. Agr. See, England. 
 
 2 Ibid., 1850, II, 68. 
 
 3 Ibid., 1850, II, 313. 
 
 4 Ann. Chem. Phar., 1856, 99, 97. 
 
 5 Compt. rend., 44, 819. 
 
'that the active constitueuts of manure are removed in the soil by 
 these compounds, and that the new compounds are decomposed 
 slowly for the use of the plant. In a subsequent investigation', 
 he always found phosphoric acid combined with aluminum and 
 iron, and never with calcium and magnesium. He added solu- 
 tions of calcium phosphate to the pure hydroxides of aluminum 
 aiid iron, and to soil, and found after a few days that the filtered 
 water contained calcium but no trace of phosphoric acid. 
 
 Deherain'^ examined a large number of soils. Some of these 
 contained phosphoric acid insoluble in water charged with carbon 
 dioxide, while others gave up phosphoric acid to this solvent. 
 Soils containing compounds insoluble in carbon dioxide water 
 gave up phosphoric acid to water containing calcium carbonate 
 ammonium carbonate. A solution of calcium carbonate in water 
 charged with carbon dioxide acted on iron phosphate producing 
 calcium phosphate. 
 
 The role of phosphoric acid in the soil was taken up by 
 L,iebig' in connection with his extended investigations on soil 
 absorption. He determined the solvent effect of sodium chloride, 
 sodium nitrate, and ammonium sulphate on phosphates of calcium 
 and magnesium. In all cases appreciable quantities of the phos- 
 phates were dissolved. When the solutions thus made were 
 added to soil the phosphoric acid was absorbed. Such changes 
 were thought to have a favorable influence in bringing about the 
 distribution of phosphoric acid in the soil. 
 
 Knop* confirmed the observations of Thenard regarding the 
 role of iron and aluminum hydroxides in soil absorption. He 
 showed that phosphoric acid was fixed in greater amount and 
 more rapidly when these compounds were added to soils, and 
 that soils containing large quantities had greater reverting prop- 
 erties than those containing small amounts. Data were not 
 accessible. 
 
 Peters' endeavored to ascertain the state in which phosphoric 
 .acid is held in soils. He manured a soil with bone dust and 
 
 1 Compt. rend., 46, 212. 
 
 2 Ibid., 47, 47, 988. 
 
 3 Am. Chem. Phar., 106, 185. 
 
 4 Jahresbericht, 1S65, 804. 
 
 5 Ann. Landw., 1867, 31. 
 
treated the mixture with different solvents. From his results he 
 concluded that phosphoric acid in soils is almost entirely com- 
 bined with iron and aluminum, and that from a solution of cal- 
 cium phosphate in carbon dioxide water, phosphoric acid is re- 
 moved b}' soils onh' when then the latter contain hj'droxides of 
 iron and aluminum. Soils deprived of these compoundsby treat- 
 ment with acids are apparently indifferent to the phosphate 
 solution. 
 
 In studying the fixing power of soils for phosphoric acid 
 Voelcker' placed weighed quanties of six different kinds of loams 
 in bottles, and added known quantities of superphosphate solu- 
 tion. He noted from time to time the amount of phosphoric acid 
 still remaining in solution. The hydroxides present in the soils 
 varied from zero to 17.38 per cent, and the carbonate of calcium 
 from 0.15 per cent to 67.5 per cent. The phosphoric acid was 
 fixed with especial ease by those soils which contained a good 
 store of calcareous matter. The power of the clay soils to render 
 the soluble phosphoric acid insoluble was far less than that of 
 chalky soils. 
 
 Warrington^ determined the effect of several substances on 
 the solubility of freshh'^ precipitated tri-calcium phosphate. He 
 found that one per cent solutions of ammonium chloride increased 
 the solubilit}' in pure water, but that its addition to water sat- 
 urated with carbon dioxide produced practically no effect. He 
 observed also that when an excess of calcium carbonate was pres- 
 ent, the amount of tri-calcium phosphate dissolved by carbon 
 dioxide water was excessively small. The water became satu- 
 rated with calcium carbonate, while onl^^ trace of the phosphate 
 entered into solution. Even a very small amount of calcium 
 carbonate was capable of producing this effect. 
 
 Two years later^ the same investigator carried out an elab- 
 orate experiment to show that the absorptive action of soil towards 
 phosphoric acid is due to the formation of insoluble phosphates, 
 by combination with hydroxides of iron and aluminum. Soils 
 freed from lime were treated with solutions of calcium phosphate 
 
 1 Jour. Roy. Agr. Soc, 1863, 24, 46. 
 
 2 Jour. Chem. Soc, 1866, 19, 296. 
 
 3 Ibid., 1868, 21, I. 
 
in carbon dioxide water. Almost all the phosphoric acid was re- 
 tained by the soil, while the greater part of the calcium went into 
 solution. The same results were obtained when hydroxides of 
 iron and aluminum were used instead of soil. 
 
 Beyer and Biedermann^ worked on the absorption of phos- 
 phoric acid by soils. Using a solution of sodium phosphate they 
 found no definite relation between the amount absorbed and the 
 quantities of iron, aluminum, and calcium contained in the soil, 
 though there was increased absorption with increased quantities 
 of these constituents. They determined the solvent action of 
 solutions of calcium sulphate, sodium chloride, sodium nitrate, 
 potassium chloride, and ammonium sulphate on the compound 
 formed by absorption. No appreciable effect was observed. 
 
 Trochot" in his observations on volcanic soils noticed that 
 fair amounts of lime and phosphoric acid were present, and that 
 the fertility of these soils was very high in spite of their shallow- 
 ness. The conclusion was drawn that the lime rendered the 
 insoluble phosphates available. 
 
 The fact that the phosphoric acid of superphosphate grad- 
 ually becomes insoluble in ammonium citrate solution was 
 observed by Millott^ and Joulie*. Millot found the change 
 to be due to the presence of large quantities of iron and alumium 
 oxides, while the results of Joulie seem to show that when an 
 insufficient quantity of sulphuric acid to completely decompose 
 the phosphate rock is used, (a condition which obtains in actual 
 practice) the change is due to the formation of tri-calcium phos- 
 phate. 
 
 These experiments were extended somewhat by Albert and 
 Vollbrecht\ They applied superphosphate to some soils con- 
 taining large amounts of lime, and to others containing small 
 amounts. With soils containing small quantities of lime, 6.5 
 per cent of the phosphoric acid of the superphosphate had become 
 insoluble in ammonium citrate after eight days, while with those 
 rich in lime, 15.4 per cent of the phosphoric acid had become 
 
 1 Chem. Centr., 1869, 945. 
 
 2 Compt. rend., 81, 1027. 
 
 3 Bied. Centr., 1875, 89—1879, 408. 
 
 4 Ann. Chim. Phys., Series 5, 1879, 244. 
 
 5 Bied. Centr., 18S0, 84. 
 
insoluble. These results were construed as proving the rever- 
 sion of phosphoric acid by soils to insoluble phosphate of lime. 
 In another experiment precipitated phosphates of iron and 
 aluminum were added to soils with the result that none of the 
 phosphoric acid became insoluble in ammonium citrate during 
 eight daj^s. 
 
 Petermann^ attempted to determine the crop producing 
 power of these various forms of phosphoric acid. The "citrate 
 soluble" was in many cases superior to the "water soluble," 
 though the results were not altogether satisfactory. The low 
 value of the water soluble was due apparently to leaching, while 
 the citrate soluble, being insoluble in water, was retained for the 
 use of the plant. 
 
 Solubility determinations of the compounds supposed to 
 exist in superphosphate were made by Albert and Wagner'. Di- 
 calcium phosphate was found to be soluble in water charged with 
 carbon dioxide. Sulphates and chlorides appeared to have very 
 little solvent action, while nitrates and carbonates were more 
 eifective. The solubility of precipitated phosphates of iron and 
 aluminum in carbon dioxide water was a great deal less than 
 that of precipitated calcium phosphate. Drying these salts les- 
 sened their solubility. Regarding the absorption of phosphoric 
 acid by soils they found a solution of the phosphates in carbon 
 dioxide water similar in its action to a solution of "soluble phos- 
 phoric acid." Absorption was most complete in clays poor in 
 lime, less complete in calcareous clays and organic soils, and 
 very small in sands poor in lime. An increase of humus in the 
 soils rich in lime and clay was always accompanied by an increase 
 in absorptive power. 
 
 In contradiction to the work just quoted, Fiedler^ states that 
 solutions of nitrates dissolve less phosphoric acid from soils than 
 pure water. His results showed absorption to be favored slightly 
 by sodium nitrate. A solution of sodium nitrate did not affect 
 the solubility of tri-calcium phosphate but it acted on di-calcium 
 phosphate and phosphates of iron and aluminum. 
 
 1 Bied. Centr., iSSo, 87. 
 
 2 Ibid., 1880, 640. 
 
 3 Landw. Versuchs-Stat., 26, 135. 
 
lO 
 
 Kostitcheff' working with pure compounds, obtained a de- 
 composition of calcium carbonate in the presence of water and 
 phosphates of iron and aluminum. Carbon dioxide was evolved 
 and calcium phosphate was formed. The change took place even 
 when there was present a large excess of iron with respect to the 
 phosphoric acid. 
 
 Tuxen^ investigated two soils — a sand and a clay — in order 
 to ascertain the effect of chlorides of sodium and potassium, and 
 sodium nitrate on the absorption of plant food. Phosphoric acid 
 was readily taken up in the presence of the alkalies. Potassium 
 chloride was the most effective, causing an increase in absorption 
 of from 25 per cent to 40 per cent. Solutions of these salts were 
 found to have little effect on the solubility of soil phosphoric acid. 
 
 Krocker and Grahl^ carried on manuring experiments with 
 phosphates of various kinds. Oats and beets seemed to be little 
 benefited b}^ the soluble phosphates when used alone. When used 
 in conjunction with ammonium sulphate there was decided in- 
 crease in the crop, the greatest yield being obtained by the use of 
 bone meal. 
 
 Similar experiments were made by Hoffmeister*. The vari- 
 ous phosphates were applied as a top dressing. The results are 
 interesting. In a sandj^ soil poor in lime the soluble phosphoric 
 acid did not descend more than ten inches. Di-calcium phos- 
 phate remained unaltered during the experiment, and the mono- 
 calcium phosphate was not converted into the tri-basic form. 
 
 Fleischer and Kissling^ studied the influence exerted by cer- 
 tain salts on the solubility of the phosphates present in peaty 
 soils. Among the salts used were the sulphate, chloride, and 
 carbonate of potassium,, the sulphate and chloride of calcium, 
 sodium nitrate, ammonium sulphate, and kainite. Potassium 
 Sulphate was very effective, while sodium nitrate and kainite 
 were far less so. 
 
 Gladding® claims to have been the first to prove by direct 
 
 1 Bull. Soc. Chim., (2) 1880, 34, 341. 
 
 2 Landw. Versuchs Stat., 27, 107. 
 
 3 Bied. Centr., 1S82, 1154. 
 
 4 Ibid., 1881, 813. 
 
 5 Ibid., 1883, 153. 
 
 6 Chem. News, 50, 16, 27. 
 
II 
 
 laboratorj^ experiment, that the greater part of the phosphoric 
 acid reverted in soils is in the form of iron and aluminum phos- 
 phates. On adding a solution of superphosphate to three repre- 
 sentative soils he found after several days that all the phosphoric 
 acid was soluble in ammonium citrate at a temperature of 65 
 degrees C. "This" the author states "shows that most of the 
 phosphoric acid was present in the form of iron and aluminum 
 phosphates, since tri-calcium phosphate is much less readily at- 
 tacked by this reagent than these phosphates." The author 
 criticizes results obtained by Albert and Vollbrecht\ from which 
 these investigators concluded that the phosphoric acid not dis- 
 solved by ammonium citrate at 40 degrees C had become insoluble 
 tri-calcium phosphate. Unfortunately Gladding did not give the 
 composition of the soils used in his experiments. 
 
 The composition and action of superphosphate have been 
 studied from many points of view. Weilandt^ found it to be very 
 completely absorbed by marl. The action was rapid. Marge- 
 stein^ showed its value to be increased by mixing with wood 
 ashes. He grew potatoes, mustard, barley, and maize on a 
 diluvial sandy soil. Quantities of ashes up to 25 per cent were 
 beneficial. Analyses of soils u.sed were not given. 
 
 From his results Joffre* referred reversion to the formation 
 of phosphate of iron. He determined its coefficient of solubility 
 in water, in water charged with carbon dioxide, and in solutions 
 of different salts. In all cases it was less sensitive to these 
 substances than tri-calcium phosphate. Vegetation tests showed 
 it to be less valuable. 
 
 Thompson" determined the absorptive power "of sand, as well 
 as that of the same sand containing known quantities of ortho- 
 clase, calcium carbonate, iron and aluminum hydroxides, of 
 mixtures of calcium carbonate and orthoclase, and of calcium 
 carbonate and hydroxides of iron and aluminum. He also 
 noticed the effect of sodium chloride and potassium nitrate on 
 the process of absorption. The original publication was not 
 
 1 Bied. Centr. iS8o, S7. 
 
 2 Landw. Versuchs-Stat., 34, 207. 
 
 3 Bied. Centr., 1888, 225. 
 
 4 Bull. Soc. Chim., 47, 312. 
 
 5 Inaug. Dissertation, Dorpat, 1S90. 
 
accessible but abstracts give the following conclusions. "Sand 
 offers no resistance to the extraction of the phosphoric of super- 
 phosphate by water. Addition of ortlioclase produces no effect. 
 Calcium carbonate combines quickly with soluble phosphoric 
 acid. Iron and aluminum hydroxides are also active in retain- 
 ing phosphoric acid. The compounds formed with iron and 
 aluminum are more stable towards salt solutions than tri-calcium 
 phosphate. Sodium chloride solutions ( i per cent and 2 per 
 cent) dissolve less phosphoric acid from superphosphate than 
 pure water. In the presence of calcium carbonate and hydroxi- 
 des of iron and alumium the salt solutions dissolve more than 
 pure water." 
 
 The action of carbon dioxide on tri-calcium phosphate alone 
 and in the presence of ferric hydroxide was studied by Geogievice\ 
 Tri-calcium phosphate suspended in water was decomposed by 
 carbon dioxide, but the reaction was far from being complete. 
 When ferric hydroxide was also present decomposition took place, 
 but the phosphoric acid immediately combined with the iron. 
 Under some conditions the whole of the phosphoric acid was with- 
 drawn from the calcium salt. These results confirm the observa- 
 tions of many previous workers, namely, that all of the phosphoric 
 acid in calcium phosphate when applied to the soil finally becomes 
 converted to phosphate of iron. 
 
 Some interesting facts regarding the assimilation of phos- 
 phoric acid by crops were obtained by Deherain' on the experi- 
 mental plots at Grignon. He noticed that the sum of the phos- 
 phoric acid in drainage and crops in ten years, was not equal 
 to the phosphoric acid not extracted by acetic acid. He con- 
 cluded that the calcium phosphate was changed to iron and 
 aluminum phosphates after a time. A soil was mixed with tri- 
 calcium phosphate and placed in a seltzogene. After a few days 
 no phosphoric acid was dissolved by water charged with carbon 
 dioxide, although tri-calcium phosphate is soluble in this mix- 
 ture. He does not state whether the soil contained lime. Ferric 
 and aluminum phosphates were assimilated b}^ oats but not by 
 wheat. 
 
 1 Monatsh., 1891, 12,566. 
 
 2 Ann. Agron., 17, 445. 
 
13 
 
 Kellner^ made some observations on the action of lime as a 
 manure on paddy fields. Complete infertility had resulted in 
 several places. He attributed part of this to loss by leaching. 
 Among other things the action of phosphates in two soils was in- 
 vestigated. He mixed in from 0.25 percent to 5 per cent of 
 quicklime, and after two weeks added a solution of potassium tri- 
 hydrogen phosphate. Lime apparently caused an increase in the 
 soluble phosphoric acid. The maximum effect was obtained with 
 I per cent to 2.5 lime. After two months still more phosphoric 
 acid had become soluble. 
 
 Gerlach'* repeated Thompson's work'^ on the absorption of 
 phosphoric acid by soils. The results here recorded show that 
 clay, peat, and sand which have been extracted with hydrochloric 
 acid have no power of absorbing free phosphoric acid, sodium 
 phosphate, and superphosphate. Calcium and magnesium car- 
 bonates absorb phosphoric acid, while iron and aluminum oxides 
 do so very completely. In the first case the compounds formed 
 are comparatively soluble and are completeh' extracted by water 
 charged with carbon dioxide. The compounds formed with iron 
 and aluminum are insoluble in water even in the presence of car- 
 bon dioxide, but are more or less dissolved by the prolonged 
 action of organic acids. He endeavored to determine the form 
 in which phosphoric acid is absorbed. A solution of mono-cal- 
 cium phosphate was placed in contact with calcium carbonate. 
 Di- and tri-calcium phosphates were formed depending on the 
 relative amounts of the materials used. When iron and aluminum 
 hydroxides were the ab.sorbents, not only was the phosphoric acid 
 absorbed, but the calcium was also retained. Di-calcium phos- 
 phate and an iron phosphate were formed. The author concluded 
 that normal ferric phosphate is formed when sufficient ferric 
 hydroxide is present. A basic phosphate of iron was also con- 
 sidered a possibility. 
 
 Stocklasa* estimated the yearly loss of calcium carbonate 
 from various soils. The results are of interest as bearing on the 
 
 1 Bull. Coll. Agr. Tokyo Imp. Univ., 1891, 9, i. 
 
 2 I,andw. Versuchs-Stat., 1895, 46, 201. 
 
 3 Inaug. Dissertation, Dorpat., 1890. 
 
 4 BiecL Centr., 1S95, 82. 
 
14 
 
 loss of phosphoric acid. Four soils — loam, marl, clay and humus 
 — were analysed and the total phosphoric acid compared with 
 that found in the drainage water. From the data thus obtained 
 the annual loss was estimated. The results showed a loss of 
 9146 grams per hectare in the case of the clay, while the humus 
 lost 21995 grams per hectare. The humus contained the smallest 
 per cent of phosphoric acid, but lost the largest amount. 
 
 Schreiber^ grew oats followed by turnips on sandy, humus, 
 and loamy soils using two fertilizers, viz : ( i ) a mixture of di- 
 calcium phosphate, calcium sulphate, and magnesium carbonate, 
 (2) a mixture of sodium phosphate, and carbonates of calcium 
 and magnesium. The first mixture gave the best results in 
 every case. The author thought that the diminished action in 
 the second case was due to the precipitation of phosphoric acid 
 by the carbonates of calcium and magnesium. His results with 
 peaty soils showed that the humus phosphoric acid, soluble in 
 alkaline ammonium citrate, was almost useless for vegetation. 
 In some cases the humus acted on the assimilable phosphoric 
 acid in a manner analagous to calcium carbonate. 
 
 Joffre^ made vegetation experiments to determine the relative 
 values of superphosphate, tri-calcium phosphate, and the ferric 
 phosphate obtained from a sample of manure containing very 
 little tri-calcium phosphate. Crops of mustard were grown on 
 plots which differed only with regard to the nature of the phos- 
 phate present. The ferric phosphate was little better than no 
 fertilizer, while the other phosphates gave a large increase. 
 
 On mixing basic slag and superphosphate with soil, Smora- 
 wski and Jacobson^ noticed that the phosphoric acid originally 
 soluble in water was rapidly converted to the citrate soluble 
 form, and this compound underwent no further change. 
 
 Stoklasa* called attention to the fact that ferrous salts in the 
 presence of phosphoric acid soluble in water, give rise to the 
 production of di-tri-ferric phosphate, unless there is an excess of 
 free phosphoric acid present. With soluble phosphoric acid 
 
 1 Exp. Sta. Rec, 1895, 804. 
 
 2 Bull. Soc. Chim., 1896, (3) 15, 
 
 3 Bied. Centr., 1896, 580. 
 
 4 Ann. Agron., 1S97, 23, 588. 
 
15 
 
 aluminum salts do not form compounds analagous to those of 
 ferrous and ferric salts, but behave like salts of calcium and 
 magnesium. No data were given. 
 
 Prianischnikoff' working on the relative values of mineral 
 phosphates, made sand culture experiments with cereals manured 
 with phosphorites. The phosphoric acid was only slightly avail- 
 able. Under the same conditions certain other plants such as 
 peas, lupines, buckwheat, and mustard seemed to be able to util- 
 ize the phosphoric acid. The "podzols" (soils containing a 
 large amount of fine siliceous and organic matter) were appar- 
 ently able to render phosphorites available for cereals and other 
 plants. The same was true of forest and peaty soils. Some of 
 the black soils, however, were without action on the phos- 
 phorites. 
 
 The results of numerous experiments conducted by Ullmann 
 and Grimm'^ showed that for months after the application of 
 superphosphate, phosphoric acid soluble in water passed through 
 a depth of soil equal to ten inches. The authors found only a 
 portion of the magnesia, lime, and oxides of iron and aluminum, 
 available for retaining phosphoric acid. The mechanical fixa- 
 tion appeared to depend on the amount of fine sand present, 
 especially sand of the fineness of dust. 
 
 A series of experiments, on the use of lime, extending over 
 several years, have recently been carried out at the Rhode Island 
 Experiment Station^ The practical results obtained in field 
 tests by the use of lime in conjunction with phosphates are of 
 particular interest, and apparently contradictory to a majority of 
 the writers on agricultural chemistry. Because of its peculiar 
 interest, the work is quoted here in some detail. Crops were 
 grown for four successive years, and the average yields on the 
 limed and the unlimed plots recorded. The results on the yields 
 of ha}^ are given herewith as being typical of those obtained 
 with the crops grown. The figures represent pounds per acre. 
 
 1 Ann. Agron., iSgg, 25, 177. 
 
 2 Cheni. Ind., 1900, 23, 61. 
 
 3 R. I. Agr. Expt. Sta. Bull., No. 58. 
 
i6 
 
 Forms of Phosphoric Acid Applied. lyimed. Unlimed. Gain from 
 
 I<iming. 
 
 Dissolved Boneblack 19,837 9.820 10,016 
 
 Dissolved Bone 19,281 8,564 10,716 
 
 Dissolved Phosphate Rock 20,205 8,951 11,253 
 
 Fine Ground Bone 22,012 11,855 io,i57 
 
 Basic Slag Meal 20,400 13, ^93 7,206 
 
 Floats 20,525 10,560 9,965 
 
 Alumina Phosphate (raw) 14,387 5,042 9,345 
 
 Alumina Phosphate (ignited) 19,481 4,930 i4,55i 
 
 No Phosphoric Acid 15, 737 2,547 13,190 
 
 Double Superphosphate 17, 937 4,752 13,184 
 
 It will be seen that there was a wonderful gain in the crop, 
 in all cases, resulting from the use of lime. The acidulated phos- 
 phates appeared to be helped more by the lime than the unacidu- 
 lated forms. The authors suggested that lime might have been 
 beneficial by correcting the acidity of the first. No experiments, 
 however, were made to ascertain whether the lime would liberate 
 phosphoric acid from the insoluble phosphates of the soil. 
 
 Some rather peculiar results were obtained by Schreiber^ on 
 the action of calcium carbonate on mineral phosphates. Addi- 
 tion of calcium carbonate to the phosphates decreased the action 
 of the latter on the crops grown, the greater the amount, the 
 greater the injury. In some cases the effect extended to the 
 next year. The action of basic slag was not affected. 
 
 Kellner and Bottcher^ obtained similar results, when using 
 superphosphate, basic slag, and bone meal for oats. Addition of 
 calcium carbonate caused reduced yields in every case, but this 
 decrease was less with the superphosphate and slag than with 
 the bone meal. Comparing the amounts of increase due to phos- 
 phoric acid applied, the addition of calcium carbonate was not 
 unfavorable when used in conjunction with slag and super- 
 phosphate. With bone meal the addition of calcium carbonate 
 reduced the yields on the average by 67 per cent. No explana- 
 tion was ofi"ered for the action of the latter. 
 
 The relation between humus and the phosphoric acid of the 
 soil, has not received very much attention. Incidentally, obser- 
 
 1 Bied. Centr., 1900, 162. 
 
 2 Deut. I,andw. Presse, 1900, 27, 665. 
 
17 
 
 vations have been made on the retentive and solvent powers of 
 some organic soils, but in most cases the work was designed for 
 the investigation of other problems. 
 
 Fleischer and Kissling\ in a study of the action of moorland 
 soils on insoluble phosphates, found that the effect was to render 
 a portion of the phosphate soluble in water, amounting in one 
 case to 5.5 per cent of the total. At the same time a portion was 
 reduced to the di-calcium salt, and in one compost as much as 17 
 per cent of the total was brought into this form. When the ratio 
 of soil to phosphate was widened, there was an increase in soluble 
 salt. This increase varied directly with the time of contact. 
 There appeared to be a limit beyond which the soluble salt became 
 reduced. 
 
 In 1892 Berthelot and Andre^ made some experiments on the 
 absorption of phosphoric acid by artificial humic anhydride pre- 
 pared from sugar. Solutions of phosphates of sodium and am- 
 monium were mixed with the acid anhydride and allowed to 
 stand in the cold for twenty-four hours. The absorption from 
 these salts were very small. In most cases practically no phos- 
 phoric acid was retained. 
 
 The action of humic acid in the soils on the solubility of 
 various natural phosphates was taken up by Minssen and Tacke^ 
 Many of the insoluble phosphates were acted on in the presence 
 of the free humic acid of the soil. When lime was added the 
 acid was neutralized and the solvent power thereby largely 
 destroyed. The quantity of phosphoric acid dissolved increased 
 with an increase in the quantity of free humic acid present. The 
 numerical agreement, however, was affected by the reabsorption 
 of the phosphoric acid made soluble. 
 
 Snyder^ prepared humus by mixing soil with such sub- 
 stances as clover, flour, straw, saw dust, and sugar, and allowing 
 to ferment for a 3'ear. At the end of that time the humus pro- 
 duced contained more phosphoric acid than was originally present 
 in the humus- forming material, indicating that some of the phos- 
 phoric acid of the soil had united with the humus. By "humus" 
 
 1 Bied. Centr., 1883, 155. 
 
 2 Ann. Chim. Phys., 27, 196. 
 
 3 L,andw. Jahrb., 1S9S, IV, 392. 
 
 4 Minn. Agr. Expt. Sta. Bull., No. 53. 
 
i8 
 
 is here meant the sokible organic matter extracted with a 2 per 
 cent solution of ammonia, after treating the soil with a dilute 
 solution of hydrochloric acid. The "humic phosphoric acid" is 
 the phosphoric acid present in the extractive material. Humic 
 phosphoric acid proved to be one of the most valuble forms. 
 
 Dumont' investigated the absorbing power of several organic 
 soils containing humus and lime. Mono-calcium phosphate was 
 applied in aqueous solution before and after calcination of the 
 soil. Calcination seemed to decrease the immediate, but increased 
 the final absorption of phosphoric acid. These results were con- 
 strued as proving the combination of humus with phosphoric 
 acid. The quantities of lime present appeared to have same 
 effect on the absorption, but the relationship of lime, humus, and 
 phosphoric acid was not well defined. Mono-calcium phosphate 
 was also applied to precipitated humus. Absorption took place 
 in the ratio of humus 10, to P2O5I. Precipitated humus was not 
 defined. 
 
 The use of ferrous sulphate in agriculture, has not been the 
 object of much experimental enquir3\ Griffiths^ however, ob- 
 tained some very interesting practical results by its use, and his 
 work is taken up here because of its bearing on the behavior of 
 phosphoric acid in the soil. 
 
 The material was applied at the rate of 50 lbs. per acre. 
 Among the crops grown were ha}- , mangels, beans, potatoes, and 
 wheat. Applied with farmyard manure, its action was injurious, 
 the poisonous effect being due to the formation of ferrous sul- 
 phide. The crops in nearly all other cases were increased largely 
 and the percentage content in phosphoric acid was increased in all 
 cases. The ferrous sulphate was not oxidized rapidly in the soil, 
 its presence being proven six weeks after application. Ferrous 
 sulphate was used in conjunction with superphosphate in all tests. 
 
 The subject was taken up a little later by Delacharlony and 
 Destremx'\ . They found that the sulphate was productive of in- 
 creased yields only when the soil contained small quantities of 
 iron, the increase ceasing when the ferric oxide reached 3 per 
 
 1 Compt. rend., 1901, 132, 435. 
 
 2 Jour. Chem. Soc, Trans. 1884, 71.— 1886, 114, — 1887, 215. 
 
 3 Bied. Centr., 1889, 9. 
 
19 
 
 cent, and was detrimental when 4 per cent was reached. The 
 increase varied with different crops, from 5 per cent to 139 per 
 cent. 
 
 Cazeneuve and Nicolle' studied the solvent action of ferrous 
 sulphate on some of the phosphatic fertilizers in common use, 
 namely, slag, mineral phosphate, bones, superphosphate, and di- 
 calcium phosphate. After eight days the first three had suffered 
 no change in the amount of phosphoric acid soluble in acetic acid 
 and ammonium citrate. With superphosphate half of the water 
 soluble had become reverted. With the di-calcium phosphate, 
 the phosphoric acid soluble in acetic acid and ammonium citrate 
 had increased, but the change was not to the water soluble form. 
 
 The action of ferrous sulphate on crops was investigated by 
 Boiret and PatureP. Seedlings of peas and oats were used in 
 water culture experiments. Even with very dilute solutions, the 
 plants were poisoned, owing to the formation of free sulphuric 
 acid. Iron citrate and citric acid had the same effect. Tests 
 were then made with soils containing known quantities of lime. 
 The sulphate was injurious except when there was present a suf- 
 ficient amount of lime to neutralize the free acid formed. Analysis 
 of the crops showed that ferrous sulphate had no effect on the 
 amount of phosphoric acid in the product. In these experiments 
 the sulphate was not used in conjunction with other fertilizers, as 
 in the work of Griffiths already quoted. 
 
 1 Chem. Centr., IV, (2) 1892, 121. 
 
 2 Ann. Agron., 18, 418. 
 
20 
 
 SUMMARY 
 
 So far as the writer has been able to ascertain, the literature 
 cited above includes practically all the work relating directly to 
 the subject under investigation. Briefly stated, the following 
 conclusions have been drawn from the work referred to : 
 
 I. When soluble phosphates are added to soils the phos- 
 phoric acid is usually changed to insoluble forms. The action in 
 most cases is due to chemical change, by virtue of the presence 
 of compounds of iron, aluminum, and calcium, and possibly 
 magnesium. These compounds are easily decomposable salts 
 which can be extracted with hydrochloric acid. 
 
 II. Hydrated oxides of iron and aluminum, and carbonate 
 of calcium unite with phosphoric acid very rapidly. The phos- 
 phates of iron and aluminum formed are from two to five times as 
 insoluble in pure water as calcium phosphate. They are also less 
 soluble in water containing carbon dioxide and various alkali 
 salts. 
 
 III. It is the tendency of phosphoric acid to form the more 
 dijBiculty soluble compounds in the soil, but not all the phosphoric 
 acid appears to be so combined, nor is such combination always 
 effected with with great rapidity. The complete fixation in 
 some cases does not take place for several months. 
 
 IV. Mixed with calcium carbonate, in the presence of soil 
 water, phosphates of iron and aluminum undergo decomposition 
 with the formation of calcium phosphate. With certain soils, 
 calcium hydroxide also apparently has a solvent action on the 
 phosphates of iron. Conditions affecting the equilibrium between 
 these compounds have not been determined. 
 
 V. Absorption may be influenced to some extent by the 
 presence of some of the alkali salts. The effect is not very well 
 defined. In some cases the solubility of the soil phosphoric acid 
 seems to be increased in the presence of these salts. 
 
 VI. In some cases the humus in soils has the effect of ren- 
 dering insoluble phosphates more available to plants. The action 
 
21 
 
 seems to be due to free humic acid, rather than to salts of this 
 acid. This point, however, has not been determined. 
 
 VII. Humus is believed to possess great absorptive prop- 
 erties though experimental evidence supporting this view is 
 exceedingl}^ limited and not very conclusive. The relation be- 
 tween humus and phosphoric acid is not well defined. 
 
 VIII. Ferrous sulphate in conjunction with phosphates,, 
 applied to some soils is beneficial to crops. It seems to increase 
 the availability of the phosphoric acid. In large quantities on 
 acid soils, it acts as a poison to plants. The injurious effect is 
 due in part to the formation of sulphuric acid. The same effect 
 may be produced in neutral soils, and soils not sufiiciently basic 
 to neutralize all the acid formed. 
 
 IX. Sand and orthoclase offer no resistance to the extrac- 
 tion of soluble phosphoric acid by pure water. 
 
EXPERIMENTAL 
 
 I. Methods of Determining Phosphoric Acid. 
 
 The work about to be described necessitated a number of 
 phosphoric acid determinations. Hence it was thought advisible 
 in most cases to employ the Optional Volumetric Method of the 
 Association of Official Agricultural Chemists^ In order to test its 
 accuracy, this method was compared with the Official Gravimetric 
 Method^ on a variety of materials typical of those to be con- 
 sidered in the experiments following. The methods were carried 
 out as follows : 
 
 Volumetric Method. The sample was dissolved in from 15 
 to 30 c.c. of strong hydrochloric acid and from 3 to 10 c.c. of nitric 
 acid. The solution was cooled, diluted to 250 c.c, and an aliquot 
 part taken for the determination. This portion was treated with 
 10 c c. of nitric acid, nearly neutralized with ammonia, diluted 
 to about 100 c.c, and heated in the water bath to 65 degrees C. 
 Molybdic solution was then added (the amount depending on the 
 amount of P.^Oj present) the mixture stirred, allowed to stand 
 thirty minutes, and filtered on an asbestos filter. After thor- 
 ough washing with cold water, the yellow precipitate was re- 
 turned to the precipitating beaker, dissolved in standard potas- 
 sium hydroxide, a few drops of phenolphthalein added, and the 
 excess of alkali titrated with standard nitric acid. Each cubic 
 centimeter of the standard potassium hydroxide was equivalent to 
 I mo- P O 
 
 Cravimetric Method. The sample was dissolved in from 
 15 to 30 c.c of strong hydrochloric acid and from 3 to 10 c.c. of 
 nitric acid. The solution was cooled diluted to 250 c.c, and an 
 aliquot part corresponding to 0.3 gram taken for analysis. The 
 solution was treated with 10 c.c. of nitric acid, nearly neutralized 
 with ammonia, diluted to about 100 c.c, and heated to 65 degrees 
 
 1 U. S. Dept. Agr., Div. Chem., Bui. No. 46, 13 (revised). 
 
 2 Ibid., 10. 
 
23 
 
 C in a water bath. Molybdic solution was then added, the whole 
 digested for one hour at 65 degrees C, filtered and washed thor- 
 oughly with cold water. The precipitate was dissolved on the 
 filter with ammonia and hot water, and washed into a beaker to 
 a bulk of not more than 100 c.c. The solution was nearly neu- 
 tralized with hydrochloric acid, cooled, and magnesia mixture 
 added, drop by drop, from a burette, stirring vigorously. After 
 fifteen minutes, 30 c.c. of ammonia (0.96 sp. gr. ) were added, 
 and the whole allowed to stand over night. The precipitate was 
 then collected on a filter, washed with 2.5 per cent ammonia 
 until free from chlorides, dried, ignited to whiteness and 
 weighed. The quantity of P^Oj was calculated from the weight 
 of the Mg,P,0,, using the factor 0.6396. The following results 
 were obtained. 
 
 Material Grams P2O5 
 
 Gravimetric Volumetric 
 
 "Insoluble" from Superphosphate 0.0145 0.0147 
 
 "Insoluble" from Superphosphate 0.0146 0.0147 
 
 Ferric Phosphate 0.1146 0.1137 
 
 Ferric Phosphate 0.1149 0.1140 
 
 Aluminum Phosphate o. 1137 o. 1143 
 
 Aluminum Phosphate 0.1138 0.1146 
 
 Calcium Phosphate 0.1361 0.1369 
 
 Calcium Phosphate 0.1367 0.1374 
 
 The results by the two methods agree closely, the difference 
 being within the experimental error usually allowed for such 
 determinations. With the larger quantities of phosphoric acid, 
 the gravimetric method was used as a check on the volumetric. 
 The results in all cases, therefore, are strictly comparable. 
 
 II. Relation Behveen Phosphoric Acid, and Compounds of Iron, 
 
 and Calcium. 
 
 Recent results obtained by the use of lime in conjunction 
 with phosphates of various kinds, have suggested new possibili- 
 ties with regard to the relation between phosphates of iron and 
 calcium in the soil. Lime has hitherto been regarded, partly as 
 a direct plant food, partly as an amendment in liberating plant 
 food, especially potash, but principally as an agent for improving 
 
I 
 
 24 
 
 the physical condition of soils. With regard to the action of phos- 
 phoric acid in the presence of calcium (lime and calcium carbonate) 
 and iron, the literature is somewhat contradictory. The reaction 
 to some extent appears to be a reversible one. The mass law, 
 which of late years has served to clear up many unexplained 
 phenomena, undoubtedly plays a ver)^ important part. 
 
 It occurred to the author that the beneficial effect of lime 
 might be due, to some extent, to the chemical action of this sub- 
 stance in rendering phosphates more available to crops, and that 
 the behavior of phosphoric acid in the presence of compounds of 
 calcium and iron, might depend on the relative quantities of the 
 last two present. The experiments designed to study these 
 points are described under the following heads : 
 
 (a) Action of lime on ferric phosphate. 
 
 (b) " " " " the phosphate formed by fixation. 
 
 (c) " " " " soils, 
 
 (d) Fixation of phosphoric acid in the presence of com- 
 
 pounds of iron and calcium. 
 
 (a) ACTION OF LIME ON FERRIC PHOSPHATE. 
 
 The object of the experiment was to determine the effect of 
 lime on the availability of precipitated and natural ferric phos- 
 phates. 
 
 The solvent selected for this purpose was the i percent citric 
 acid solution proposed by Dyer\ for the determination of avail- 
 able plant food in soils. This reagent is the onlj^ one of the 
 many solvents proposed for determining available plant food, that 
 has received any favor among agricultural chemists. While it is 
 still of doubtful efficiency in determining absolute availability, in 
 determining relative values, for which purpose it was used in the 
 present work, it has been of unquestioned service. Dyer pro- 
 posed this reagent as representing the average root juice acidity 
 of a large number of agricultural plants examined by him. 
 
 Precipitated ferric phosphate was prepared by treating fer- 
 ric chloride with di-sodium phosphate. The resulting precipitate 
 was washed thoroughly with cold water and the product air 
 
 I Jour. Chem. Soc. Trans., 1894, 115. 
 
25 
 
 dried. Phosphoric acid was determined by the gravimetric 
 method, and ferric oxide by the vohimetric bi-chromate method. 
 The following results were obtained : 
 
 Fe^Os 41.53% 
 
 F,0, 38.20% 
 
 H2O 20.08% 
 
 Correcting for hygroscopic moisture, the figures agree with 
 the formula FeP0,.2H,0. 
 
 The only natural ferric phosphate that could be obtained 
 was a sample of Dufrenite containing 16.01 per cent of P.O5. 
 This is a basic ferric phosphate. 
 
 The solubilit}^ of these materials was first determined by 
 placing I gram in a small flask, adding 100 c.c. of i per cent 
 citric acid solution, and allowing to stand twenty-four hours at 
 room temperature, shaking once each hour. The insoluble resi- 
 due was then filtered off, washed with cold water, and the phos- 
 phoric acid determined. The following results were obtained : 
 
 Precipitated Ferric Phos 
 
 Precipitated Ferric Phos 
 
 Dufrenite 
 
 Dufrenite 
 
 The effect of lime on the solubility of these phosphates was 
 then determined. For this purpose i gram of each of the ma- 
 terials was placed in a flask, pure slaked lime (free from phos- 
 phoric acid) added, and enough water to make a thin paste. 
 After twenty-four hours the, contents of the flask were treated 
 with 100 c.c. of I per cent citric acid solution and an additional 
 amount of the crystallized acid to neutralize the excess of lime. 
 The solubility in the citric acid was then determined as in the 
 first instance. The following table contains the results. 
 
 P2O5 sol. in 
 
 Per 
 
 cent of 
 
 r cent citric acid. 
 
 
 total. 
 
 12.20% 
 
 
 31-9 
 
 12.05% 
 
 
 31-5 
 
 0.07% 
 
 
 0.4 
 
 0.09% 
 
 
 0.5 
 
26 
 
 Material. I,ime. P2O5 sol. in i per Per cent of 
 
 Prec. Ferric Phos i gram 
 
 Prec. Ferric Phos i gram 
 
 Prec. Ferric Phos 0,5 gram 
 
 Prec. Ferric Phos 0.5 gram 
 
 Prec. Ferric Phos 0.25 gram 
 
 Prec. Ferric Phos 0.25 gram 
 
 Dufrenite i gram 
 
 Dufrenite i gram 
 
 The following table shows the comparative solubility of the 
 
 phosphoric acid (PjOJ in i per cent citric acid, with and without 
 the addition of lime : 
 
 Material. 
 
 Prec. Ferric Phos.. 
 Prec. Ferric Phos.. 
 Dufrenite 
 
 cent citric acid. 
 
 total. 
 
 24-95% 
 
 65.3 
 
 24.90% 
 
 65.2 
 
 35.80% 
 
 93-7 
 
 35.80% 
 
 93-8 
 
 37- 00% 
 
 96.9 
 
 36.90% 
 
 96.6 
 
 0.45% 
 
 2.8 
 
 0.47% 
 
 2.9 
 
 I'ithout lime. 
 
 
 With lime. 
 
 
 
 I gr- 
 
 0-5 gr. 
 
 0.25 gr. 
 
 12.20% 
 
 24-95% 
 
 35-80% 
 
 37.00 
 
 12.05% 
 
 24.90% 
 
 35-90% 
 
 36.90 
 
 0.07% 
 
 0.45% 
 
 
 
 
 0.09% 
 
 0.47% 
 
 
 
 
 It is seen from the above tables that the action of lime on 
 the natural ferric phosphate was not verj- marked. In the case 
 of the precipitated ferric phosphate the lime caused a considerable 
 increase in the vSolubility. The larger quantities did not cause 
 as great an increase as the smaller quantities. This fact indi- 
 cated the formation of the more soluble calcium phosphates with 
 the smaller quantities. With an excess of lime, as in the experi- 
 ment in which i gram was used, tri-calcium phosphate was 
 probably formed, and this compound is much less soluble in i 
 per cent citric acid than the di- and mono-calcium compounds. 
 
 It is to be remarked that the reaction was fairly rapid. 
 Within one hour after the addition of lime, a red compound began 
 to separate. This compound was probably ferric hydroxide. An 
 attempt was made to separate the products formed in the reaction. 
 Various solvents were tried, but it was found impossible to effect 
 a separation of calcium and ferric phosphates in the presence of 
 ferric hydroxide. In the course of these trials a 5 per cent solu- 
 tion of neutral potassium oxalate was observed to act on precipi- 
 tated ferric phosphate, dis.solving both the iron and the phos- 
 
27 
 
 phoric acid in the proportion in which they existed in the 
 original compound. This action of potassium oxalate was made 
 the basis of a qualitative test for the presence of ferric phosphate, 
 in the presence of ferric hydroxide. The solution was found to 
 have no solvent action on ferric hydroxide. The compound 
 formed with ferric phosphate appeared to be a double oxalate, 
 very soluble in water. 
 
 Ferric phosphate was treated with varying quantities of 
 lime. The products of the reaction were treated with the solu- 
 tion of potassium oxalate, filtered, and the filtrate tested for 
 iron. The absence of iron in the oxalate solution was taken as 
 showing the absence of ferric phosphate in the mixture treated 
 with this solvent. 0.2 gram lime was found to be the smallest 
 amount capable of completely converting i gram of ferric phos- 
 phate. 
 
 The behavior of potassium oxalate suggested the use of this 
 reagent for determining the available phosphoric acid in phos- 
 phatic materials. A few preliminary trials on its solvent action 
 on calcium and iron phosphates, under different conditions, were 
 made. The results were not sufficiently satisfactory to indicate 
 the usefulness of this reagent. 
 
 (b) ACTION OF LIME ON THE PHOSPHATE FORMED BY FIXATION. 
 
 It is fairly well established that soluble phosphates when 
 applied to soils tend to pass into comparatively insoluble forms. 
 This reversion is caused largely by hydrated oxide of iron. It 
 was considered important therefore, to know the effect of lime on 
 the compound thus formed. 
 
 To this end two grams of pure ferric hydroxide were treated 
 in a small flask with 50 c.c. of an aqueous solution of superphos- 
 phate, shaken, and allowed to stand ten days with occasional 
 agitation. The insoluble residue was then filtered off and washed 
 with cold water until phosphoric acid ceased to be given up. 
 The filtrate and washings were then analyzed and the phos- 
 phoric acid absorbed calculated. Several residues were obtained 
 in this way. Two of these were treated with 100 c.c. of i per 
 cent citric acid solution, and the solubility determined as de- 
 
28 
 
 scribed under ferric phosphate. Four others were treated with 
 0.5 gram of lime, enough water added to make a thin paste, and 
 the mixture allowed to stand one and nine days. The solubility 
 in I per cent citric acid was determined, sufficient crystallized 
 citric acid being added to correct the basicity due to lime. The 
 superphosphate solution originally added contained 0.1340 gram 
 of P2O5. Under "residues" in the following table are included 
 the amounts of PjO^ absorbed : 
 
 P2O5 in Ivitne added, 
 
 residues. 
 
 P2O.5 sol. in 
 
 Per Cent 
 
 cent citric acid. 
 
 soluble. 
 
 0.0214 grs. 
 
 17.I 
 
 0.0224 grs. 
 
 179 
 
 0.0544 grs. 
 
 43-7 
 
 0.0535 grs. 
 
 43 -o 
 
 0.0499 grs. 
 
 40.1 
 
 0.0495 grs. 
 
 39-7 
 
 0.1249 grs. 
 
 0.1248 grs. 
 
 0.1244 grs. 0.5 grs. one day 
 
 0.1240 grs. 0.5 grs. one day 
 
 0.1244 grs. 0.5 grs. nine days 
 
 0.1244 grs. 0.5 grs. nine days 
 
 Incidentally a similar experiment was made using neatral 
 ammonium citrate solution (1.09 sp.gr.) as the solvent. For 
 various reasons this solvent was not used in the soil investiga- 
 tions. The preceding plan was followed except that the action 
 of lime was allowed to continue for two days. The results were 
 as follows : 
 
 P2O.3 in L,ime added, 
 
 residues 
 
 o. 1217 grs. 
 
 o. 1211 grs. 
 
 0.1223 grs. 0.5 grs. two days 
 
 0.1224 grs. 0.5 grs. two days 
 
 The results show that lime materially increased the solubility 
 of reverted phosphate of iron in i per cent citric acid. The action 
 appeared to be complete in one da^^ The .solubility of the re- 
 verted phosphate in ammonium citrate solution was not increased 
 by the addition of lime. 
 
 The composition of this phosphate of iron has not been deter- 
 mined. It is much more insoluble in i per cent citric acid, and am- 
 monium citrate than the normal ferric phosphate. That none of 
 the latter compound was formed in the fixation experiment, was 
 indicated by the following tests : 
 
 P2O.-, sol. in 
 
 Per Cent 
 
 am. citrate. 
 
 .soluble. 
 
 0.0535 grs. 
 
 43-9 
 
 0.0534 grs. 
 
 44.0 
 
 0.0536 grs. 
 
 43-« 
 
 0.0527 grs. 
 
 43-0 
 
29 
 
 (i ). Ferric hA'droxide and a solution containing phosphoric 
 acid in large excess of the amount necessarj- to form the normal 
 salt, were allowed to react for weeks. The amount of phosphoric 
 acid absorbed was then determined. 
 
 Fe(OH); used 0.2500 grs. 
 
 P2O-, absorbed 0.0275 " 
 
 P.^.O, required to form FePOj 0.1664 " 
 
 Even in the presence of a large excess of phosphoric acid, 
 the ferric hydroxide absorbed little more than one tenth of the 
 amount necessary to form the normal salt. 
 
 (2). The compound formed was treated with the 5 per 
 cent solution of potassium oxalate already described. After 
 filtering the filtrate was tested for iron, the absence of which 
 showed the absence of normal ferric orthophosphate. 
 
 The compound formed by fixation is evidently a highly basic 
 phosphate of iron. The compound appears to be formed even 
 when only a limited amount of the base is present. In the soil, 
 the bases would usually be present in large excess with respect 
 to the phosphoric acid. 
 
 In the experiments just described, the lime was brought into 
 intimate contact with the phosphate of iron. Its action there- 
 fore was probably at a maximum. It was thought advisable to 
 supplement this with artificial soil experiments under conditions 
 which usually obtain in soils, conditions presumably less favora- 
 ble to the action of lime. 
 
 For this purpose 200 gram portions of clean sand were ex- 
 tracted with hydrochloric acid and washed free from acid with 
 cold water. A blank experiment showed that the product pos- 
 sessed no absorptive properties for phosphoric acid. Each por- 
 tion was mixed with 4 grams of ferric hydroxide and placed in a 
 six inch cylinder. Superphosphate solution was then added and 
 absorption allowed to continue for ten days. During the mean- 
 time the mixture was kept moderately moist. After ten days 
 the phosphoric acid unabsorbed was washed out with cold water 
 and determined, the amount absorbed being calculated from the 
 data thus obtained. The mixture then represented a soil con- 
 taining 1.5 per cent Fe.Og, and 0.12 per cent of P.p^. 
 
30 
 
 For this work and the soil investigation to be described later 
 two solvents were selected, namely, i per cent citric acid solu- 
 tion, and hydrochloric acid. The first has already been re- 
 200 -^ 
 
 ferred to , but for soil work the conditions attending its use were 
 modified somewhat. 
 
 Hydrochloric acid was recently proposed by Moore^ for 
 
 the determination of available mineral plant food in soils. In the 
 work referred to pot experiments were made with a large variety 
 of soils, and the amounts of phosphoric acid extracted by the 
 solvent in question agreed closely in nearly all cases with the 
 actual amount removed by the crops. 
 
 The solubility determinations with the reagents selected 
 were carried out in the following manner. The ratio of solvent to 
 substance was 500 c.c. to 100 grams. These were placed in a 
 litre glass stoppered bottle, and kept at a temperature of 40 de- 
 grees C in a water bath for exactly five hours. The bottles were 
 shaken every fifteen minutes. After the digestion the whole was 
 shaken and emptied on to a folded filter sufificienlly large to hold 
 the entire contents of the bottle. After draining, 400 c.c. of the 
 filtrate representing 80 grams of the soil, were evaporated to 
 dryness. In the case of the citric acid the residue was treated 
 with magnesium nitrate solution, evaporated to dryness, and 
 ignited until all organic matter was destroyed. These residues 
 were taken up with water and nitric acid and the phosphoric 
 acid determined. 
 
 In all the experiments a preliminary digestion was made 
 with 20 grams of soil and 100 c.c. of the solvent, in order to 
 determine the basicity of the soil. A correction was then made 
 in the strength of the acids so as to reduce the solvent action to- 
 a uniform basis. 
 
 Portions of the artificial soil described were air-dried and the 
 solubility of the phosphoric acid determined. Lime (one and 
 two grams) was mixed with other portions and the action allowed 
 to go on for seven days, the mixture being kept moist during the 
 meantime. These mixtures were then air-dried and solubility 
 determinations made. The following results were obtained. 
 
 I Jour. Am. Chem. Soc. 1902, 79. 
 
31 
 
 SOLVENT — I PER CENT CITRIC ACID SOLUTION 
 
 P2O5 in I.ime added. P2O5 sol. in Percent 
 
 residues. i per cent citric acid. soluble. 
 
 0.2530 grs. 0.0622 grs. 24.5 
 
 0.2540 grs. 0.0611 grs. 24.0 
 
 0.2488 grs. I gram 0.1060 grs. 42.6 
 
 0.2483 grs. I gram 0.1070 grs. 43.0 
 
 0.2527 grs. 2 grams 0.1232 grs. 48.7 
 
 0.2517 grs. 2 grams 0.1220 grs. 48.4 
 
 n 
 
 SOLVENT — HYDROCHLORIC ACID 
 
 200 
 
 P2O5 in I,inie added. P2O5, sol. in Per Cent 
 
 residue. i per cent citric acid. soluble. 
 
 0.2530 grs. 0.0033 grs. 1-3 
 
 0.2540 grs. 0.0038 grs. 1.4 
 
 0.2488 grs, I gram.. 0.0044 g^s. 1.7 
 
 0.2483 grs. I gram 0.0052 grs. 2.0 
 
 0.2527 grs. 2 grams 0.0048 grs. 1.8 
 
 0.2517 grs. 2 grams 0.0052 grs. 2.0 
 
 From the results it is seen that lime increased the solubility 
 in I per cent citric acid considerably as in the preceding experi- 
 ments. The increase is somewhat less, however, with the arti- 
 ficial soil. 
 
 With resrard to the solubility in hydrochloric acid, the 
 
 ° 200 
 
 addition of lime produced little change. The differences are 
 within the range of experimental error. It is to be remarked 
 that the solvent action of the hydrochloric acid was very small. 
 
 (c) ACTION OF LIME ON SOILS. 
 
 For the purpose of studying the effect of lime on the phos- 
 phoric acid originally present in soils, two samples of the latter 
 were taken. The soil designated "A" was a sandy loam under 
 cultivation. Soil "B" was a clay loam not under cultivation. 
 
 Samples were prepared for analysis according to the methods 
 of the Association of Olficial Agricultural Chemists'. 
 
 I U. S. Dept. Agr., Div. Cheni., Bui. No. 46, 71. 
 
32 
 
 The determination of total phosphoric acid was made,, the 
 method being to weigh out 2 grams of soil into a platinum dish 
 and ignite to drive off organic matter. Put in i or 2 c.c. of 
 hydrofluoric acid and allow the soil to come in contact with the 
 acid slowly to avoid loss by sputtering. After the violent action 
 has ceased, place on a water bath and evaporate to dryness. After 
 repeating this operation once or twice, take up with a little nitric 
 acid and water and determine phosphoric acid. This method 
 was rapyd and gave closely agreeing results. In this way the 
 following results were obtained : 
 
 Soil "A." Soil"B." 
 
 Total P,0, 0.145% 0.155% 
 
 Complete analyses of the acid soluble materials in these were 
 then made according to the methods of the Association of Official 
 Agricultural Chemists\ with the following results : 
 
 DIGESTION IN HYDROCHLORIC ACID I. II 5 SP. GR. 
 
 Constituent. Soil "A." Soil "B." 
 
 Insoluble matter 87.85% 83.53% 
 
 Potash (K2O) 0.74% 0.80% 
 
 Soda (NaoO) 0.20% 0.41% 
 
 Lime (CaO) 0.11% 0.14% 
 
 Magnesia (MgO) : 0.55% 0.78% 
 
 Oxide of Manganese (MnO) 0-03% 0.02% 
 
 Ferric Oxide (Fe.Os) 1-74% i-9i% 
 
 Alumina (AI2O3) 4-00% 5.72% 
 
 Phosphoric Acid (P2O.O 0.12% 0.10% 
 
 Sulphuric Acid (SO3) 0.04% 0.06% 
 
 Carbonic Acid (CO2) 0.26% 0.18% 
 
 Volatile matter 4- 17% 6.51% 
 
 Portions of 200 grams of the air dried soils were then mixed 
 thoroughly with lime (one and two grams), the whole placed in 
 C5dinders, moistened and allowed to stand exposed to the atmos- 
 phere for seven days. The portions were then again air dried, 
 and weighed. The solubility of the original and limed portions 
 
 in I per cent citric, and hydrochloric acids was then deter- 
 
 200 
 
 I U. S. Dept. Agr. Div. Chem., Bui. No. 46, 72. 
 
33 
 
 mined according to the method described under artificial soils, 
 the quantities taken being calculated on a basis of the original 
 air dry soil. The quantities of phosphoric acid in the following 
 tables are calculated to this basis, and are therefore comparable 
 in all cases. 
 
 SOIL "a." solvent — I PER CENT CITRIC ACID 
 
 Lime added. 
 
 None 
 
 None 
 
 One gram 
 
 One gram 
 
 Two grams 
 
 Two grams 
 
 '2O.5 soluble. 
 
 Per cent ol 
 
 
 total. 
 
 0.028% 
 
 19-3 
 
 0.030% 
 
 20.7 
 
 0.031% 
 
 21.3 
 
 0.032% 
 
 22.0 
 
 0.037% 
 
 255 
 
 0.038% 
 
 26.2 
 
 SOIL "b." solvent — I PER CENT CITRIC ACID 
 
 Lime added. P2O 5 soluble. Per cent of 
 
 total. 
 
 None 0.0068% 4.3 
 
 None 0.0067% 4-3 
 
 One gram 0.0132% 8.5 
 
 One gram 0.0130% 8.3 
 
 Two grams 0.0100% 6.4 
 
 Two grams 0.0102% 6.5 
 
 n 
 
 SOIL A." SOLVENT HYDROCHLORIC ACID 
 
 200 
 
 Lime added. P2O5 soluble. Percentof 
 
 total. 
 
 None 0.0030% 2.0 
 
 None ■ 0.0030% 2.0 
 
 One gram 0.0043% 2.9 
 
 One gram 0.0041% 2.8 
 
 . Two grams 0.0040% 2.8 
 
 Two grams 0.0038% 2.7 
 
 n 
 
 SOIL "B." SOLVENT HYDROCHLORIC ACID 
 
 200 
 
 The amount of phosphoric acid was too small to be deter- 
 mined. The addition of lime (one and two grams), seemed to 
 produce no effect on the solubility of the phosphoric in this soil. 
 
34 
 
 The two soils used in the experiments described contained 
 practically equal amounts of total phosphoric acid, and phosphoric 
 acid soluble in hydrochloric acid 1.115 sp. gr. , while the avail- 
 bility of that in the cultivated soil, as shown by the weak acid 
 solvents, was about five times as great as that in the uncultivated 
 soil. The chemical analyses of the two present no differences 
 that could serve as an explanation of the difierence in availability. 
 This difference, therefore, is probably due, partly to the state of 
 cultivation, and partly to the physical characteristics of the soils. 
 
 From the results obtained it is doubtful whether hydro- 
 
 200 
 
 chloric acid would prove a suitable solvent for the determination 
 
 of available plant food in all soils. Comparing the two soils, 
 
 however, the results with it indicate in a general way the same 
 
 relative availabilitj' of the phosphoric acid as is shown by the 
 
 solubility in i per cent citric acid. 
 
 In regard to the action of lime, the solubility in citric acid 
 was increased slighth^ in both soils. The solubility in hydro- 
 chloric acid was not increased in the case of the uncultivated 
 soil. This soil evidently held the phosphoric acid in a very 
 insoluble form. 
 
 It is interesting to note the difference in action of lime on 
 natural and artificial soils. A comparison is made in the follow- 
 ing table. 
 
 SOLUBILITY OF P^Oj IN I PER CENT CITRIC ACID 
 
 Without lime. With lime. 
 
 Soil "A" 19.3% 21.3% 
 
 Soil "A" 20.7% 22.0% 
 
 Artificial soil 17.1% 43-7% 
 
 Artificial soil 17-9/1) 43-o% 
 
 The figures represent per cent of the total phosphoric acid. 
 It is evident that the natural soil was less susceptible to the 
 action of lime than the artificial product. 
 
 (d) FIXATION OF PHOSPHORIC ACID IN THE PRESENCE OF 
 COMPOUNDS OF IRON AND CALCIUM 
 
 While the tendency of soluble phosphates, applied to soils, 
 is to pass into the insoluble phosphate of iron, several investi- 
 
35 
 
 gators, notably Deherain, KostitchefF, and others have observed 
 that calcium carbonate reacts with ferric phosphate forming cal- 
 cium phosphate, the action taking place even in the presence of 
 an excess of ferric hydroxide. Whether this was the reverted or 
 the normal ferric phosphate was not stated. 
 
 It occurred to the author that the presence of a relatively 
 large quantity of calcium compounds might interfere with the 
 formation ' of reverted phosphate of iron. A short experiment 
 was made in order to ascertain whether the presence of calcium 
 carbonate would influence the fixation of phosphoric acid by 
 ferric hydroxide. 
 
 Tri-calcium phosphate was dissolved in water charged with 
 carbon dioxide, and 200 c.c. of the solution containing 0.0176 
 gram of P2O5 were added to i gram of ferric hydroxide, with and 
 without varying quantities of calcium carbonate. The action was 
 allowed to continue for ten days, after which the mixtures were 
 filtered. The solubility of the residues in ammonium citrate solu- 
 tion was determined by adding 50 c.c. and allowing to stand 
 twenty-four hours in the cold. This residue was filtered and 
 washed with cold water. The results are contained in the fol- 
 lowing table : 
 
 Material mixed with 0.0176 grs. P2O5. P2O5 insoluble in 
 
 ammonium citrate. 
 
 I gram CaCOg 0.0018 grs. 
 
 I gram Fe( OH) 3 0.0145 grs. 
 
 I gram Fe(OH)3 plus i gramCaCO.j 0.0152 grs. 
 
 I gram Fe(OH)3 plus 2 grams CaCOg 0.0147 grs. 
 
 In the blank determination above practically all the phos- 
 phoric acid was extracted by ammonium citrate. The presence 
 of calcium carbonate (one and two grams) apparently had no 
 effect on the solubility of the compound formed by the reversion 
 of phosphoric acid. 
 
 III. Behavior of Phosphoric Acid towards Humus 
 
 That humus possesses the power of absorbing phosphoric 
 acid from solution has not been proven. Organic soils have been 
 known to effect solution of insoluble phosphates in some cases, 
 
36 
 
 while in other cases soils containing a large amount of humus ex- 
 hibited a strong attraction for phosphoric acid. 
 
 In studjang the retentive power of humus for phosphoric 
 acid it was thought best to work with the artificial preparation, 
 thereby eliminating the possibility of the interference of any other 
 absorbing material. 
 
 To this end humic acid was prepared from sugar according 
 to the method described by Berthelot\ 600 grams of sugar fur- 
 nished 100 grams of the air dry material. In one portion the 
 moisture was determined by drj-ing to constant weight at 100 de- 
 grees C. Other portions were subjected to combustion analysis. 
 The following analysis is calculated on the basis of dry matter. 
 
 Carbon 61.48% 
 
 Hydrogen 4-74% 
 
 Oxygen 33-78% 
 
 The phosphoric acid to be absorbed was applied in the form 
 of an aqueous extract of superphosphate, thus approximating 
 conditions which obtain in actual practice. Portions of i gram 
 of the air dry humic acid were placed in a small flask, 50 c.c. of 
 a solution containing 0.1550 gram P2O5 and 0.0185 gram calcium 
 added to each, and the mixtures allowed to stand one, four and 
 eight days with occasional shaking. At the end of these periods 
 the insoluble matter was filtered off and washed with cold water 
 until the wash water gave no test for phosphoric acid. The 
 quantities of phosphoric acid and calcium in the filtrate and 
 washings were then determined. Calcium was determined by 
 the volumetric permanganate method. The results are found in 
 the following table. 
 
 No. Duration of contact. P2O5 soluble. Ca soluble. 
 
 1 one day o.i540grs. 0.0173 grs. 
 
 2 one day 0.1545 grs. 0.0180 grs. 
 
 3 four days 0.1535 grs. 0.0185 grs. 
 
 4 four days o.i55ogrs. 0.0188 grs. 
 
 5 eight days o.i550grs. 0.0180 grs. 
 
 6 eight days 0.1550 grs. 0.0177 grs. 
 
 As is shown by the above results, neither the phosphoric 
 acid nor the calcium were retained by the insoluble humic acid. 
 
 I Chim. Veg. Agr., 4, 123. 
 
37 
 
 It was not expected that chemical absorption would take place. 
 The experiment was suggested by the statements of many 
 writers to the effect that humus is active in retaining plant food 
 by virtue of its physical structure. The data given above indicate 
 that insoluble free humic acid exerts no influence in rendering 
 phosphoric acid insoluble in water. 
 
 The organic acids present in the soil do not, however, exist 
 to any great extent in the free state except in bogs and morasses. 
 Soils suitable for agricultural purposes contain little or no free 
 acid except carbonic acid. In most cases they give an alkaline 
 reaction. Humic acid exists as humates of the bases predomi- 
 nating in the soil. These compounds may be divided into two 
 classes, namely, (i) those insoluble or but slightly affected by 
 water, including humates of calcium, magnesium, iron, and 
 aluminum, and (2) those soluble in water including humates of 
 the alkalies. Representative of the first class, calcium humate 
 was selected for study. 
 
 This compound was prepared by mixing 10 grams of the 
 humic acid already described with one litre of lime water con- 
 taining 2.42 grams of CaO, and allowing the mixture to stand 
 for five days with occasional shaking. The insoluble residue 
 was then filtered off and washed with cold water until the wash- 
 ings were neutral. On being subjected to combustion analysis 
 the product showed the following composition, calculated on the 
 basis of dry matter. 
 
 Carbon 55-48% 
 
 Hydrogen 4-ii% 
 
 Oxygen 34-8i% 
 
 Calcium 5-6o% 
 
 This preparation was treated with a solution of superphos- 
 phate in the manner already described for humic acid. The solu- 
 tion contained, 
 
 P2O5 o. 1550 grs. 
 
 Ca 0.0185 grs. 
 
 The quantities of phosphoric acid remaining unabsorbed by i 
 gram of the calcium humate are given below. 
 
38 
 
 Duration of contact. P2O5 soluble. Ca soluble. P^O.-j absorbed. 
 
 'One day o. isSsgrs. 0.0309 grs. 0.0165 grs. 
 
 One day 0.1385 grs. 0.03^2 grs. 0.0165 grs. 
 
 Four days 0.1360 grs. 0.0278 grs. 0.0190 grs. 
 
 Four days o. 1365 grs. 0.0283 grs. 0.0185 grs. 
 
 Eight days 0.1340 grs. 0.0259 grs. 0.0210 grs. 
 
 Bight days 0.1345 grs. 0.0260 grs. 0.0205 grs. 
 
 The above figures show that absorption of phosphoric acid 
 took place to a limited extent. It occurred to the author that a 
 small amount of calcium carbonate might have been formed in the 
 preparation of the calcium humate, in which case the absorption 
 could not be ascribed to the latter compound. Carbon dioxide 
 was determined in a portion of the preparation therefore, the 
 amount obtained being 0.64 per cent. Calculated to calcium 
 carbonate the one gram of the material contained 0.0142 gram of 
 the material contained 0.0142 gram of the carbonate, which in 
 turn would be sufficient to react with 0.0207 gram of P^O^ form- 
 ing the mono-calcium salt. This salt is decomposed by water with 
 the formation of the di-calcium phosphate which is not soluble in 
 water. The absorption therefore is to be referred to the presence 
 of calcium carbonate. 
 
 The increase of the calcium in solution was due to the for- 
 mation of some of the mono-calcium phosphate from the free 
 phosphoric acid present in the superphosphate solution. 
 
 It is to be inferred from these facts that calcium humate does 
 not take part in the absorption of phosphoric acid from super- 
 phosphate. 
 
 As the majority of the humus compounds found in the soil 
 contain nitrogen, it was thought advisable to include some of 
 these in the present work. The compounds used were soluble 
 and insoluble ammonium humate. 
 
 For the preparation of these 25 grams of humic acid were 
 treated with 500 c.c. or ammonia solution containing 14 3 grams 
 of ammonia, the flask closed, shaken and allowed to stand five 
 days. The insoluble residue was then filtered off and washed 
 until free from ammonia. When dried the product w^eighed 1 1 
 grams showing that a little more than half of the humic acid had 
 been dissolved by the ammonia. The soluble portion was put 
 
39 
 
 into an evaporator and allowed to stand in the cold until all free 
 ammonia had evaporated. The solution was reserved for further 
 investigation. 
 
 The insoluble portion was subjected to a combustion analy- 
 sis. The "nitrogen as ammonia" represents the amount of 
 nitrogen expelled by boiling with magnesia. 
 
 Carbon 62.29% 
 
 Hydrogen 5.62% 
 
 Oxygen 28.81% 
 
 Nitrogen (total) 3.28% 
 
 Nitrogen (as ammonia) 108% 
 
 Although humus varies in its content of nitrogen (depend- 
 ing on the material from which it is formed) the above agrees 
 fairly well with the average analysis of humus extracted from 
 soil by various solvents. Regarding the form in which nitrogen 
 is held in this compound, Berthelot has shown that besides 
 ammonia, it is present most probably as amide. This fact is in 
 accordance with the data obtained when working with the nat- 
 ural product. 
 
 The action of this substance on superphosphate was investi- 
 gated. The material was treated with a solution of superphos- 
 phote in the manner already described for humic acid. The 
 solution contained, 
 
 P2O5 o. 1550 grs. 
 
 Ca 0.0185 grs. 
 
 The quantities of phosphoric acid remaining unabsorbed by 
 I gram of the insoluble ammonium humate are given below. 
 
 No. Duration of contact. P2O5 soluble. Ca soluble. 
 
 1 one day 0.1540 grs. 0.0185 grs. 
 
 2 one day 0.1545 grs. 0.0178 grs. 
 
 3 four days o.i54ogrs. 0.0191 grs. 
 
 4 four days 0.1540 grs. 0.0183 grs. 
 
 5 eight days 0.1540 grs. 0.0178 grs. 
 
 6 eight days 0.1540 grs. 0.0187 grs. 
 
 The results show that the insoluble ammonium humate ex- 
 erted no influence in retaining the phosphoric acid of superphos- 
 phate. The content in calcium was also not affected. 
 
40 
 
 The soluble ammonium humate already referred to contained, 
 
 Nitrogen (total ) 20. 25% 
 
 Nitrogen (as ammonia) 9-65.% 
 
 calculated on a basis of dry matter. With a solution of super- 
 phosphate, this substance produced a brown flocculent precipitate. 
 Thinking that this precipitate consisted of calcium humate, 
 another portion of the ammonium humate solution was treated 
 with lime water. No precipitate was formed. Neither calcium 
 chloride or ferric chloride produced any change. On adding 
 pure orthophosphoric acid a brown precipitate was formed appar- 
 ently identical with that formed with the superphosphate solu- 
 tion. These facts indicated the formation of an organic com- 
 pound containing phosphoric acid. An attempt was made to 
 separate this compound, but it was exceedingly difficult to wash, 
 and apparenth^ underwent partial decomposition. 
 
 From the foregoing data it appears that insoluble humus offers 
 no resistance to the extraction of the phosphoric acid of super- 
 phosphate by water. The soluble humus appears on the other 
 hand to form an insoluble compound with phosphoric acid, and 
 thus acts as an absorbent. To what extent such absorption may 
 take place was not determined. 
 
 (b) SOLUTION 
 
 The solvent action of two of these humus compounds was 
 considered briefly. For this purpose precipitated tri-calcium 
 phosphate, humic acid, and soluble ammonium humate were 
 selected. The experiments were carried out as follows. 
 
 Portions of 0.5 gram of pure precipitated tri-calcium phos- 
 phate were placed in small flasks, and to some of these was added 
 X gram of insoluble humic acid, and 10 c.c. of water. To other 
 portions were added 10 c.c. of the solution of ammonium humate. 
 The contents of the flasks were shaken occasionally for five days, 
 after which the residues were filtered off, washed with cold 
 water, and the phosphoric acid determined in the filtrate and 
 washings. 
 
 A blank experiment was made with 0.5 gram of precipitated 
 tri-calcium phosphate and 10 c.c. of water treated as described 
 
41 
 
 for the other experiments. The results of these tests are given 
 in the following table. 
 
 Material. P2O5 soluble in water. 
 
 Tri-calcium phos. 0.5 gr none 
 
 Tri-calcium phos. 0.5 gr none 
 
 Tri-calcium phos. 0.5 gr. am. humate ioc.c._ none 
 
 Tri-calcium phos. 0.5 gr. am. humate ioc.c._ none 
 
 Tri-calcium phos. 0.5 gr. humic acid i gr 0.0030 gr. 
 
 Tri-calcium phos. 0.5 gr. humic acid i gr 0.0028 gr. 
 
 The ammonium humate was without action on the tri- 
 calcium phosphate. The free humic acid rendered a portion of 
 the phosphoric acid soluble in water, though the action was not 
 very great. These facts are in accordance with the observations 
 of other workers, namely, that humus is active in rendering 
 insoluble phosphoric acid available to plants. The results pre- 
 sented herewith indicate that this action is due to the humic 
 acid and not to the compounds of humic acid, namely, the 
 humates. 
 
 IV. Absorption of Phosphoric Acid by Zeolites 
 
 There is abundant evidence to show the existence in soils of 
 easily decomposable silicates analagous to the zeolites. The 
 power of these substances to fix the bases potassium, sodium, 
 calcium, ammonia, and magnesium has been thoroughly investi- 
 gated by a number of workers. 
 
 The work of Doelter (i), Friedel (2), Rinne (3), Clarke (4), 
 and others on the constitution and properties of the zeolites, sug- 
 gested to the writer the possibility of the power of these sub- 
 stances to absorb phosphoric acid. 
 
 For this investigation both the artificial and natural silicates 
 were used. The first was prepared by treating a solution of soda 
 alum with a solution of sodium silicate. The resulting precipi- 
 tate was washed thorougly with cold water, filtered and dried. 
 The product was difficult to wash, and seemed to undergo partial 
 decomposition even in cold water, with the elimination of sodium 
 silicate. The dry material was found to contain the following : 
 
42 
 
 Alumina 18.12% 
 
 Silica 57-21% 
 
 Soda 5.85% 
 
 Water 18.82% 
 
 Two grams of this substance were treated with superphos- 
 phate solution, allowed to stand seven days, filtered, washed, and 
 the phosphoric acid in solution then determined. The residues 
 were treated with 100 c.c. of i per cent citric acid solution and 
 allowed to stand for twenty-four hours. The phosphoric acid 
 dissolved by this reagent was then determined, with following 
 results : 
 
 No. P2O.5 absorbed. Per cent .sol. in i 
 
 per cent citric acid, 
 
 1 O.I340grs. 56.8 
 
 2 0.1340 grs. 57.4 
 
 The natural silicates used were selected with a view of ob- 
 taining as much variety in composition as possible. A few insol- 
 uble silicates were also used. 
 
 It is not within the province of this paper to enter into a 
 discussion of the constitution of these compounds. The empir- 
 ical formulas are given herewith. 
 
 Analcite NaAlSi.Ofi H2O 
 
 Chabazite (CaNaj )Al2Si40i , 6H,0 
 
 Halloysite H4Al2Si20., aq. 
 
 Heulandite . H4CaAl,(Si03)6 3H2O 
 
 Prehnite H2Ca2 Al.iSiO J,, 
 
 Pyrophyllite H2Al2(SiO.,)4 
 
 In order to determine the absorptive power of these sub- 
 stances for phosphoric acid, 2 gram portions of each w^ere placed 
 in small flasks, and to each were added 50 c.c. of a solution of 
 superphosphate containing o. 1550 gram Pfiy The experiment 
 was carried out in the manner described for the precipitated sil- 
 icate. Solubility determinations of the residues, in i per cent 
 citric acid were made in a similar way. A check determination 
 was made with washed sand. The following table contains the 
 data obtained. 
 
43 
 
 Material. P20t absorbed. Per cent sol. in i 
 
 per cent citric acid. 
 
 Sand none 
 
 Analcite 0.0305 grs. 46.5 
 
 Analcite 0.0295 grs. 44.7 
 
 Chabazite 0.0170 grs. loo.o 
 
 Chabazite 0.0165 grs. 100. o 
 
 Halloysite 0.0125 grs. 80.8 
 
 Hallo}'site 0.0125 grs. 80.S 
 
 Heulandite 0.0130 grs. loo.o 
 
 Heulandite 0.0125 g^s. loo.o 
 
 Prehnite 0.0125 grs. 77.6 
 
 Prehnite 0.0120 grs. 76.6 
 
 Pyrophyllite none 
 
 The results show that some of the natural silicates (zeolites) 
 possess an absorbtive power for phosphoric acid, while the arti- 
 ficial preparation did so to a marked degree. The compounds 
 formed appeared to be more soluble in i per cent citric acid than 
 the compound formed when ferric hydroxide was the absorbent. 
 
 The nature of this absorption is not understood. In the light 
 of our present knowledge of the zeolites present in the soil it was 
 considered impracticable to undertake to ascertain its nature. 
 
 It occurred to the writer that this absorption might be due 
 to the presence of free hydrated oxide of aluminum, in which case 
 the action could not properl}^ be referred to the silicates in ques- 
 tion. In order to determine this point, some representative sam- 
 ples of the materials used were strongly dehydrated over the 
 blast lamp and the absorptive power again determined with the 
 following results : 
 
 Material. Before dehydration After dehydration. 
 
 P2O5 absorbed. 
 
 Analcite 0.0305 grs. 0.0348 grs. 
 
 Analcite 0.0295 grs. 0.0348 grs. 
 
 Chabazite 0.0170 grs. 0.0170 grs. 
 
 Chabazite 0.0165 g^'S- 0.0175 grs. 
 
 Precipitated silicate o. 1340 grs. o. 1 170 grs. 
 
 Precipitated silicate o.i340grs. o.ii70grs. 
 
 With analcite dehydration seemed to increase the absorptive 
 power though only to a slight extent. Dehydration of the arti- 
 
44 
 
 ficial preparation decreased slightly the absorptive power, indi- 
 cating the presence of a small amount of aluminum hydroxide. 
 
 It is doubtful whether there exists in soils silicates identical 
 either with the natural or the artificial materials used in the 
 above work. But it is believed that the data here presented 
 serves to show that the zeolitic silicates in soils may take part 
 in the fixation of phosphoric acid. 
 
 V. Action of Ferrous Sulphate 
 
 The value of ferrous sulphate in agriculture appears to be 
 questionable. Regarding its action towards phosphoric acid, 
 two possibilities suggested themselves, namely, (i) the hydrol- 
 ysis and consequent solvent action of the sulphuric acid formed, 
 on the insoluble phosphates, (^2) the action in retaining the 
 phosphoric acid applied in a comparatively soluble form. 
 
 The solvent effect on several phosphatic materials was deter- 
 mined by placing i gram of each in a flask and treating with 
 I gram of ferrous sulphate, 100 c.c. of water, and shaking 
 occasionally for four days. The residues were filtered, washed, 
 and the filtrate tested for phosphoric acid. None was present in 
 any case. The residues were then treated with 100 c.c. of i per 
 cent citric acid, and the solubility of the phosphoric acid in this 
 reagent determined. The following table contains the data 
 showing the effect of ferrous sulphate on the solubility in i per 
 cent citric acid. 
 
 SOLUBIIvlTY IN I PER CENT CITRIC ACID 
 
 Material. 
 
 Tri-calcium phosphate . 
 Tri-calcium phosphate. 
 
 Ferric phosphate 
 
 Ferric phosphate 
 
 Apatite 
 
 Apatite 
 
 Wavellite 
 
 WavelHte 
 
 No FeSOi. 
 
 FeSOi 
 
 2O5 dissolved. 
 
 P2O-, dissolvt 
 
 26.40% 
 
 20.40% 
 
 26.20% 
 
 20.50% 
 
 12.05% 
 
 11.50% 
 
 12.20% 
 
 11.25% 
 
 1-47% 
 
 1.36% 
 
 1-47% 
 
 1.44% 
 
 trace 
 
 trace 
 
 trace 
 
 trace 
 
 LOFC. 
 
• 45 
 
 Ferrous sulphate did not increase the citric acid solubility of 
 the phosphates under investigation. In fact with the tri-calcium 
 the effect was to lessen the solubility. 
 
 The remarkable effect of ferrous sulphate when used with 
 superphosphate as was the case in the field tests made by Grif- 
 fiths, indicated that the substance converted the phosphoric acid 
 into forms more available than the reverted phosphate formed 
 from ferric hydroxide. 
 
 The precipitate formed on the addition of superphosphate to 
 ferrous sulphate was found to be easily soluble in i per cent citric 
 acid. In order to determine whether this compound is formed in 
 the presence of ferric hydroxide, 2 grams of the latter were 
 mixed with 0.5 gram of ferrous sulphate and 50 c.c. of super- 
 phosphate solution. Of the reverted phosphate of iron formed 
 from ferric hydroxide without the sulphate, 17.3 per cent of the 
 phosphoric acid was soluble in i per cent citric acid, while the 
 compound formed when ferrous sulphate was present showed 
 19.0 per cent citric acid solubility. 
 
 It is improbable therefore that ferrous sulphate can be of 
 any benefit as a retentive agent for phosphoric acid, especially 
 when there is sufficient ferric hydroxide present. 
 
46 
 
 CONCLUSIONS 
 
 Summarizing briefly, the writer thinks that the following 
 conclusions are supported by the foregoing work. 
 
 I . Lime increases considerably the solubility of precipitated 
 ferric phosphate in i per cent citric acid solution. The nature 
 of the calcium phosphate formed depends on the relative amount 
 of lime added. The reaction takes place with moderate rapidity. 
 The solubility of dufrenite is increased by the addition of lime, 
 though not to a very marked degree. 
 
 II. Reverted phosphate of iron, formed when phosphoric 
 acid is absorbed by ferric hydroxide, is not identical with normal 
 ferric phosphate. It is much less soluble in i per cent citric 
 acid. The addition of lime to reverted phosphate of lime in- 
 creases its solubility in i per cent citric acid. The solubility in 
 
 hydrochloric acid is increased only to a slight extent. The 
 
 200 ^^ 
 
 solubility in neutral ammonium citrate solution is not increased. 
 
 Reverted phosphate of iron appears to be a highh' basic compound. 
 
 III. Lime applied in moderate amounts to the soils used, 
 increased the solubility of the phosphoric acid, in i per cent citric 
 
 acid, though the increase was small. The solubilitv in 
 
 200 
 
 hydrochloric acid was affected to a less degree, and in one of the 
 
 solis used no action occurred. The letter reagent, as a solvent 
 
 for the available phosphoric acid in soils, is of doubtful utility'. 
 
 IV. The phosphoric acid originally present in soils is 
 much less susceptible to the action of chemical agents than re- 
 verted phosphate of iron. The latter probably becomes more 
 resistant with age. 
 
47 
 
 V. The presence of calcium carbonate in relatively large 
 amounts, appears to have no effect on the solubility, in am- 
 monium citrate, of the compound formed by fixation. 
 
 VI. Insoluble humic acid, calcium humate, and insoluble 
 ammonium humate offer no resistance to the extraction of the 
 phosphoric acid of superphosphate by water. Soluble ammonium 
 humate forms with superphosphate solution, an insoluble com- 
 pound whose composition and properties have not been deter- 
 mined. The absorptive property of humus is due probably to 
 the humus compounds soluble in water. 
 
 VII. Soluble ammonium humate does not effect solution of 
 precipitated tri-calcium phosphate. Insoluble humic acid renders 
 a portion of the phosphoric acid soluble in water. 
 
 VIII. Some of the zeolites and similar hydrous double sil- 
 icates are capable of abstracting phosphoric acid from solution. 
 The property is not destroyed by strong ignition. The com- 
 pounds formed are comparatively soluble in i per cent citric acid. 
 
 IX. The solubility of tri-calcium phosphate, ferric phos- 
 phate, apatite, and wavellite, in water and in i per cent citric 
 acid, was not increased by the addition of ferrous sulphate. 
 With tri-calcium phosphate, the solubility in i per cent citric acid 
 was decreased, indicating that the application of ferrous sul- 
 phate under some circumstances may be detrimental. 
 
 X. Ferrous sulphate forms with superphosphate a compound 
 soluble in i per cent citric acid. In the presence of ferrous sul- 
 phate, and ferric hydroxide, the compound formed with super- 
 phosphate, has practically the same solubility in i per cent citric 
 acid as the compound formed when ferrous sulphate is absent. 
 The latter possesses therefore no especial value as a retentive 
 agent for phosphorics acid. 
 
i 
 
%