THE DETERMINATION OF AVAILABLE FLANT FOOD IN SOILS. 117 XIII.-The Detepmination o f Ava.ilable Plant Food in Soils by the use of Weak Acid Solvents. By ALFRED DANIEL HALL, and PRANCIS JOSEPH PLYMEN. IN the analysis of soils, it has been customary of late years to employ a weak acid solvent in order to extract those mineral constituents, phosphoric acid and potash in particular, which are present in the soil in such a state of combination as to be readily taken up by the crop. The phosphoric acid and potash thus extracted have been termed the available,” as distinct from the total, amounts of the same substances which can be extracted by hot, strong hydrochloric acid, or other solvent, which completely breaks up the soil. It is claimed that better indications of the comparative richness or poverty of the soil and of the need or otherwise for special mineral manures can be obtained by determinations of the available rather than of the total constituents, the information supplied by the latter being often not in accord with the results of cropping.Although chemists are agreed generally about the value of weak solvents in the analysis of soils, considerable diversity of opinion exists as to the acid to use and the theoretical basis on which its action depends. Dilute acetic acid, originally suggested by H. von Liebig, was used by Dehhrain (Ann. Agron., 1891, 17, 445). An aqueous solution of carbon dioxide has been worked with in America, by Gerlach ( h n d w . Versuchs.-XtcLt., 1896, 46, 201) and by T. Schloesing (Compt. vend., 1900, 131, 149).Its adoption is obviously based upon the fact that the natural soil water, by which much of the nutrient matter of118 HALL AND PLYMEN : THE DETERMINATION OF the soil is conveyed to plants, largely owes it solvent power to carbonic acid. Petermann, in his examinations of Belgian soils (Rechedes de Chirnie et Physiologic, 1898, 3, 50), employs an ammoniacal solution of ammonium citrate for the determination of available phosphoric acid ; he regards it as 44 veritable reactif de groupe,” distinguishing between the mineral phosphate of lime and the precipitated phosphates of lime, iron, and alumina which will rapidly come into action in the soil, Hydrochloric acid of various strengths has been used ; the American Association of Official Agricultural Chemists has recommended a solution of fifth-normal strength; trials have also been made in America with hundredth-normal hydrochloric acid.Emmerling (Bied. Centy., 1900, 29, 7 5 ) has recommended a solution of oxalic acid of 1 per cent. strength for the purpose of distinguishing between phosphoric acid combined with the alkaline earths and that combined with the sesquioxides. HoEmeister (Landw. Ve~*suchs.-Xtat., 1898, 50, 363) suggests an amrnoniacal solution of humic acid for estimating the relative values of different forms of phosphoric acid, and Maxwell (J. Amer. Chem Xoc., 1899, 21, 415), in his examination of Hawaiian soils, used a 1 per cent. solution of aspartic acid, which was found to dissolve “phosphoric acid, lime, potash, and other bases out of the soil in almost the exact proportions that these elements have been found in the waters of discharge and in which they are removed by cropping.” T.Schloesing, jun, (Compt. rend., 1899, 128, 1004), working with dilute nitric acid of various strengths, found that as the strength of the acid was increased, the amonnt of phosphoric acid dissolved first increased, then remained stationary during a certain range, and then began to increase again; a t which point, and not before, iron began t o appear in the solution. He concludes that this stationary pro- portion of phosphoric acid indicates the amount of readily available calcium phosphates and that the beginning of the attack upon the ferric phosphate marks the point at which all the available phosphoric acid has passed into solution.But of all the dilute acids, none has been more widely applied to the determination of “available” plant food than a 1 per cent. solution of citric acid, as described by Dyer in a communication to this Society (Trans., 1894, 65, 115), the 1 per cent. citric acid solution being taken as approximating both to the nature and average strength of the natural solvent, the root sap. It is, however, doubtful if sufficient data exist upon which to base any a pri0s.i decision as to the best acid and strength to employ; theAVAILABLE PLANT FOOD IN SOILS. 119 state of combination of the phosphoric acid and potash in the soil, the nature of the root sap, and the part i t plays in obtaining mineral matter from the soil as compared with that which enters the plant by osmosis from the natural soil water, are all too imperfectly known to provide a theoretical basis for a method of analysis.I n the present state of our knowledge, these processes can only be tested by com- paring the conclusions to which they lead with the results obtained by cropping the soil ; indeed, the crop alone can measure the material avaiIable in the soil. It was in the hope of obtaining some critical results with regard to the various acids suggested for determining the available constituents in the soil that the authors have obtained a number of soils which have been the subject of field experiments, and submitted them to the action of certain of the acids indicated above. As a rule, abnormal soils have been chosen, that is, soil which are markedly deficient in available phosphoric acids or potash, as indicated by the large returns which could be obtained by the application of one or other of these substances in the shape of manure.By the kindness of Sir J. Henry Gilbert, the authors further were enabled to examine seven samples from the Broadbalk Field at Rothamsted, which had been under wheat and continually manured in the same way for forty-two years. Sir Henry Gilbert was good enough to furnish the authors with material drawn from seven sharply contrasted plots on this classic field, sufficient for duplicate determina- tions of both the phosphoric acid and potash dissolved by all the solvents to be described later. Determinations were made of both phosphoric acid and potash in the Eroadbalk soils and in four other cases ; the nine remaining soils were only analysed for one constituent.Arising out of the work, determinations were also made of the calcium carbonate and the organic matter in each soil, and a few other determinations were made t o ascertain what degree of variation might be introduced by the strength of the acid employed an3 thequantity of calcium car- bonate present. The Xoils Examined. The soil samples from the Broadbalk Field, Rothamsted, were taken in October, 1893; the plots had then grown wheat continuously for fifty years and the same manures had been applied to each plot year by year, with one exception, for forty-two years (J. Bo3. Agric. Xoc. Eng., 1884, 20, 391). The following table shows the numbers under which the plots are described in the Rothamsted Memoirs, the manures per acre per rtnnum, and the average yield of grain and straw :120 HALL AND PLYMEN: THE DETERMINATION OF Grain.Straw. No. of plot. Manure per acre per annum. Bushels. Cwt. 2b Farmyard manure, 14 tons .................. 34-8 321 5 Minerals only ................................ 14; l2i 7 Minerals + 400 pounds ammonium salts 32% 32% Ch Minerals + 275 pounds sodium nitrate.. 34f 384 3 Unmanured continuously ................. 122 108 6 Minerals + 200 pounds ammonium salts 24 21+ 16 Minerals + 800 pounds ammonium salts, 13 years - Unruanured, 19 years.., ..................... 274 28$ 10 years - - .................................... - Minerals + 550 pounds sodium nitrate, .................................... I n the above table, ‘‘ minerals ” stands for 200 pounds of potassium sulphate, 100 pounds of sodium sulphate, 200 pounds of magnesium sul- phate, and 3; cwt.of superphosphate (37 per cent. soluble phosphate) ; ammonium salts means equal parts of sulphate and chloride of ammon- ium containing about 43 pounds of nitrogen, which is also that con- tained in 275 pounds of sodium nitrate. If the quantities given above are translated into pounds of phosphoric acid and potash supplied and removed per acre per annum, the follow- ing approximate figures are obtained. They are partially taken from a recently published paper by Dyer on the phosphoric acid and potash in wheat soils of Broadbalk Field, Rothanisted (PM. Ty*ans., 1901, B. 194, 235-290), and are based on the manures supplied and the analyses of the grain and straw removed : Plot. 2b 3 5 6 7 16 9 C6 Phosphoric: acid.Supplied. Removed. 78 26 0 9.3 65 14 64 17 62 22 64 26 35 20 Potash. Siipplied. Removed. 235 50 0 15 104 23 108 33 107 51 108 50 50 43 Of the other soils, No. 1 is a clay soil from Essex furnished by Nr. T. S. Dymond. Some of the results obtained on this field in 1899 may be quoted as showing the response of the soil to dressings of phosphates : Sodium nitrate, 2 cwt. ............................. 3.3 8-2 9 ) ,, 4 cmt. superphosphate 17.8 25.4 Manure. Without lime. With lime. Other results with phosphatic manures, both in this year and 1900,AVAILABLE PLANT FOOD IN SOILS. 121 confirm the need for phosphates (see The Essex Field Experintents, 1901, I, 28).Soil No. 2 is a Welsh soil from Cardigan, selected by MI-. T. Parry as typical of the soils in that district which respond freely to dressings OF basic slag. The experimental plots in the same field showed ‘‘ astonishing results ” for a dressing OF 10 cwt. of basic slag, but, being in pasture, no weights can be given. Soils Nos. 3,6, and 10, were indicated by experiments carried out under the Bath and West of England Agricultural Society, in 1891, as likely to be deficient in available phosphoric acid, and were kindly procured for us by the occupiers, Mr. J. B. Till, of Park Farm, Thornbury, Gloucester- shire ; Mr. E. W. Drew, of Crichel, Wimborne, and Mr. W. H. Tremaine, of Trerice Manor, Grampound Road, Cornwall, from the fields which had been under experiment. The following extract from the report on the trials (J.Bath and West of Englccnd Agric. Soc., 1891-1892, [iv], 2, 264) shows the effect of phosphatic dressings on the mangold crop : 4 cwt. nitrate. Plot. Character of soil. With 4 cwt. nitrate. ’, superphosphate. 3 Gravelly loam ............... 6 32.3 6 Deep loam on chalk ...... 12.7 26 10 Stone rush .................. 8.7 19.7 Soil No. 4 is from strong land on the Weald Clay, near Marden, Kent ; the sample was taken from an arable field immediately adjoin- a hop garden which has been under experiment since 1895 by the South Eastern Agricultural College. The plots have always given large returns for the application of phosphates, as will be seen from the following table, giving the mean results 1895-1899 : Mean of 5 years’ crop, Plot.Manure per acre per annum. cwt. 1 Nitrogen, potash, 6 cwt. phosphates ......... 12.5 2 7 7 9 , 8 7, ......... 15.1 3 9 , 77 10 .......... 15.7 4 9 7 ,7 15 9 7 9, ......... 16.6 On the same soil, the omission of potash gave no consistent returns ; on three occasions, the plot receiving nitrogen, phosphates, and potash was superior by 9 per cent., 6 per cent., and 1 per cent. respectively; on two occasions, it was inferior by 15 per cent. and 11 per cent. ; hence we may fairly conclude that the soil can supply the potash re- quirements of an ordinary crop (see J. Soutit, Eastern, Agric. Coll., 1900, No. 5 is a sandy soil, resting on the Tunbridge Wells beds, near Frant, and is also taken from a field adjoining a hop garden which has been under experiment. In this case, phosphates above a certain No.10, 33).I22 HALL AND PLYMEN: THE DETERMINATION OF point give little return, but potash salts prodiice a great increase in the crop. Plot. Manure per acre per annuin. Mean crop. The table sets out three years’ results : 1 Rape dust 15 cwt. ( = 7 0 lbs. nitrogen)+O ................. 15.2 2 > 7 +basic slag 5 cwt. 15.1 3 ....................... 16.1 I ) 9 , 9 , 10 ,, 4 9 9 9 ) 9 9 15 ,, ........................ 15.4 5 7 7 9 , ,, 5 +potassium sulphate 5 cwt. 17.7 ........................ The only consistent increase in crop each year has been on the plot receiving potash, where the effect has also been noticeable in the character of the foliage (see J. South Eastern Agric. Coll., Zoc. cit.). by Mr.J. Alan Murray of the University College, Aberystwyth, and was taken from grass land on a light, alluvial loam a t Falcondale, which has been under experiment for 8 years, and has given marked returns for dressings of phosphatic manure. Taking the mean figures for 4 years, when phosphatic manures were applied, the excess of hay produced as compared with the plots receiving no phosphate was as follows (see Univ. Coll. A6erystwyth, Annual Report on Pield Experiments, 1900) : 336 lbs. per acre. J , 224 ,, ............ 518 ,, !, 336 7 > ............ 552 ,, ,, 85 ,, basic slag .................. 364 ,, .................. 713 7 7 9 9 170 77 .................. 777 ,, 7 7 255 ? Y Soil No. 7 was supplie For 112 lbs. superphosphate ............ > 7 77 77 ,, Soils 8 and 9 were from the garden at Hamel’s Park, Buntingford, Essex, belonging to Mr.H. Shepherd Cross, M.P., a soil notable for causing chlorosis in many species of plants grown there, especially in laurels, fruit trees, and chrysanthemums. Applications of superphos- phate had mitigated the onset- of the disease, but it is by no means certain that a deficiency in available phosphoric acid is the cause. Soils 11 and 12 were from the experimental plots of the South Eastern Agricultural College, at Wye ; the soil is a light loam resting on the chalk and as a rule shows no particular need for mineral manures. Soil 11 was from a plot which had for five consecutive years grown barley without manure. Soil 12 had also grown barley, but had received a general dressing of artificial manures, including 4 cwt.of superphosphate containing 26 per cent. of soluble phosphate and I+ cwt. of potassium sulphate. The following mean figures obtained with barley, oats, and grass in 1896 and 1897 serve to show the response the crop makes to mineral manures ; the various crops are reduced to a common standardAVAILABLE PLANT FOOD IN SOILS. 123 by calculating them on a basis of 100 for the plot with the complete manure. Plot. Maiiiires per acre. Relative crop. A. No manure ........................................................... ’73 B. Nitrogen + 2 cwt. superphosphate, no potash ................. 93 87 E. Nitrogen + 2 cwt. superphosphate, 2 cwt. pol,assium sulphate 100 Soil No. 13 was supplied to the authors by Mr. J. L. Duncan, B.Sc., from his farm at Birgidale Knock, Rothesay, N.B. It is a deep, alluvial loam, in good heart, but gave extraordinary returns for potash in some experiments with turnips carried out by Professor J.Patrick Wright in 1895. Nitrogen, potash, + 1 cwt. D. Nitrogen + 2 cwt. potassium sulphate, no phosphoric acid ... Nitroge n and Manure, nil. Phosphate only. phosphate. sulphate of potash. Crop, nil. 5.9 8.9 19.8 tons. (See Repovts on Mcmuring, &c., Glusgow and West of Scotland I’echnical College, 1 89 6 .) The Dilute Acids Used. Since the 1 per cent. solution of citric acid is so widely used, es- pecially among chemists in this country, for the determination of available phosphoric acid and potash, it was taken as the basis of comparison, and the other acids, as far as possible, were reduced to the same strength.This seemed preferable to using the other arbitrary strengths which have been suggested, such as 1 per cent. acetic acid, 1 per cent. and one-fifth normal hydrochloric acid, especially as pre- liminary experiments showed that the strength of the acid is a factor in the amounts of phosphoric acid and potash dissolved. Citric acid solution containing 10 grams of the pure crystallised acid per litre is approximately one-seventh normal and is equivalent to a solution of acetic acid containing 8-57 grams per litre and one of hydrochloric acid containing 5.2 grams per litre. The ammonium citrate solution cannot be compared in strength with the other solvents; it is made up according to Petermann’s formula, and used for the estimation oE phosphoric acid only : 1 litre contains 87.1 grams of ammonium citrate, rendered alkaline by 9.2 C.C.of strong ammonia (sp. gr. O*SSO); 500 C.C. are digested with 50 grams of the soil for 1 hour a t a temperature of 35-40’, with constant shaking. As a source of water charged with carbonic acid, recourse was had t o the ‘6 sparklet ’’ bottles of commerce ; one of the larger sized bottles holds conveniently 50 grams of soil and 500 C.C. of water. Into this a sparklet charged with liquid carbon dioxide was broken in the usual124 rIALL AND PLYMEN: THE DETERMINATION OF way, the c w % . " the bottle were allowed to stand for a week and shaken from time to time as with the other weak acids. The larger spsrklets mere found by trial to contain about 4.5 grams of carbon dioxide, so that the solution within the bottle would contain a little less than 9 grams per litre, and be approximately 0.4 normal.After opening the bottle, as soon as the first effervescence has subsided, the solution must be rapidly filtered and the filtering completed before all the free carbon dioxide has diffused out of the liquid. When chalk is present in the soil, a strong solution of calcium bi- carbonate is produced in the sparklet bottle, and precipitation of cal- cium carbonate begins when the solution is brought into contact with the atmosphere. Preliminary tests showed that solutions of acid cal- cium phosphate and calcium bicarbonate can exist together until the excess of carbon dioxide is expelled, when calcium phosphate is pre- cipitated.However, the first portions of calcium carbonate precipi- tated during filtering, although mixed with a little fine clay, showed no appreciable amount of phosphoric acid. The three acids, citric, acetic, and hydrochloric, of the same titre, together with carbonic acid water, were used on the soils for the estimation of both the phosphoric acid and potash. itlethods of Analysis. The soil samples were all air-dried, gently broken in a mortar with a wooden pestle, and passed through a sieve having round holes 3 mm. in diameter. The stones retained by the sieve were rejected, the fine earth that passed the sieve was used for analysis without any further preparation. I n the case of the soils from the Broadbalk Field, the samples had already been put through a wire sieve with meshes -;P inch apart.The 3 mm. round sieve took out a few more stones, amounting to about 24 grams from each sample of 3 pounds, or, approximately, 1.8 per cent. Except in the case of the ammonium citrate and the carbonic acid solutions, 200 grams of the air-dried soil were put into a Winchester quart bottle with two litres of the dilute acid, the bottle was kept stoppered and shaken whenever convenient during 7 days at the ordi- nary temperature of the room. At the end of this period, the solution was filtered and an aliquot part of the extract (generally 500 c.c.) was evaporated to dryness and ignited. For the determination of phosphoric acid, the residue was attacked with hydrochloric acid, evaporated to dryness, and ignited very gently to render the silica insoluble.It was then taken up with dilute nitric acid, a few grams of ammonium nitrate were added,AVAILABLE PLANT FOOD IN SOILS. 125 with 50 C.C. of a solution of ammonium mcl?bdate, containing 60 grams of molybdic acid per litre. The volume of the nitric acid solution was always brought to 50 C.C. before adding the ammonium molybdate, in order that the work should always be carried out under uniform conditions. The mixture was well stirred and allowed to stand in a warm place, not exceeding 40°, for 24 hours. The phosphomoly bdic acid, after washing with ammonium nitrate solution, was dissolved by ammonia into a tared basin, evaporated to dryness, ignited gently over an Argand burner, and weighed. The resulting material was assumed to contain 3.794 per cent.of phosphoric acid. In potash determinations, the ignited residue from the evaporated solution was taken up with weak hydrochloric acid and the potash determined by Tatlock’s method as described by Dyer (loc. cit., p. lal), the potassium platinichloride being sometimes weighed as such, and sometimes converted into metallic platinum. The so-called ‘‘ total ” potash and phosphoric acid were determined on portions of the same soils that were ground until they would pass through a woven sieve of 1 mm. mesh. Twenty grams of such soil were extracted with 70 C.C. of strong hydrochloric acid containing 20.2 per cent. of pure acid (that is, the acid which results on boiling the con- centrated acid under ordinary atmospheric pressure) for 48 hours on a water-bath in a loosely stoppered flask.The amount of calcium carbonate is calculated from the amount of carbon dioxide evolved on treating the soil with dilute acid by a method described in another communication (this vol., p. 81). Some of the carbon dioxide may be derived from magnesium car- bonate, but as the factor that is wanted is the amount of “base” available in the soil, it is not necessary to attempt to differentiate between calcium and magnesium carbonates. All the figurea given are calculated as percentages on the soil in an air-dry condition; the amount of water each soil loses a t 100’ is also given. I. PHOSPHORIC ACID RESULTS. Soils from the Byoadbalk Piela?. I n the table on p. 126, the results obtained by the action of the various acids employed on the soils from tbe seven plots of the Broadbalk wheat field are set out.(1). A first inspection of the figures shows that in general citric acid dissolves the most, ammonium citrate a little less, hydrochloric acid comes next in order, then acetic acid, the carbonic acid charged water dissolving least of all. This order of solvent power is preserved in each plot. Taking the means of the quantities dissolved from the six manured plots, 2b, 5, 6, 7, 9a, and 16, it will be seen that the citricHALL AND PLYMEN: THE DETERMINATION OF Dung ........................... Unmanurccl .................. Minerals + 200 lb. ammon- ium salts. ................... Minerals + 400 lb. ammon- ium salts.. ................... Minerals + 275 lb. sodium nitrate.. ......................Minerals+ 800 lb. am-> Minerals only ............... 126 - Plot. ___ 2b 3 5 6 7 9a 0.0477 0 *0080 0.0510 0.0446 0'0402 0.0295 TABLE I. BI ail nri 11 g. Citric. I I--- ... ...... monium salts 13 years Unmanured Minerals+ 550 1b.sodinm nitrate ........ .10 years ITCI. Acetic. 0'0224 0.0021 0.0360 C-0264 0.0243 0.0070 0.0051 0'0166 0'0011 0.0098 0'0086 0.0067 0.0032 0'0016 0'0095 0'0005 0.0058 O-OU31 0*0030 0'0021 0.0011 0.0433 0.0069 0.0388 0.0283 0.0266 0.0197 0*0141 0.209 0.114 0'228 0.195 0'191 0.164 0.157 acid dissolves about ten times as much as the carbonic acid, about five times as much as the acetic acid, and twice as much as the hydrochloric acid (Table 11). I n the case of the unmanured plot, the ratios are of the same order. TABLE Ir.I P206 dissolved from I I Solvent. 1 Six manured plots. 1 Unmanured plot. Citric acid ..................................... Ammonium citrate ....... - .................... Hydrochloric acid ............................. Acetic acid ..................................... Carbonic acid ................................... 0.0390 0.0080 0'0202 0.0021 0 0077 0'0011 0*0005 0.0042 0.0285 i 0.0069 i (2). The ratios in which the various acids dissolve phosphoric acid are not the same for each plot, as will be seen from a consideration of the following table (111), where the results are recalculated as per- centages of the '' total " phosphoric acid, that is, the amount dissolved by strong hydrochloric acid from each soil. It is now seen that as the total phosphoric acid in the soil diminishes, so does the fraction which is soluble in any of the acids, Citric acid disaolvesmore than 20 percent, of the total phosphoricacid in the soil from the dunged plot and from the plots receiving minerals alone or minerals and ammonium salts; the percentage drops to 13.3 in the soil from plot 16, which had been for some time unmanured and at other timesAVAILABLE PLANT FOOD IN SOILS.Ammonium citrate ..,I 90.8 Hydrochloric acid ...... 46'9 Acetic acid ............... 34.8 Carbonic acid ............ 19.9 I27 86.2 76.1 26'2 70.6 13.7 ~ 19'2 6'2 I 11.4 TABLE 111. Plot. I 2b. I 3. 7. 9a. 16. 5. i 6. I 0.228 1 0.195 I I ___ I Oe209 Total 1)hosphoric acid. 1- -~ Perceiitages of total dis-; solved by :- I Citric acid .............. ,j 22.8 Ammonium citrate ...30'7 Acetic acid ............... 7-92 Carbonic acid ...... .....,I 4.53 Hydrochloric acid ...... , 10.f 0.114 0.191 0'164 _. .- 0.157 I 7-02 6.05 1.84 0.965 0'439 , 22.4 ~ 22.9 17.0 ' 14.5 15'1 13.5 4.30 ~ 4-41 2-54 ~ 1.64 21 '1 13'9 12 *7 3.51 1-57 18.0 12'0 4'28 1 -95 1 -28 13 '3 8.98 3'25 1 -02 0.701 drained OF minerals by the use of heavy dressings of nitrogenous manures, and still further drops to 7 per cent. in the soil from the unmanured plot. With the other acids, the same progression is observed. The crops first remove the more soluble portion of the phosphoric acid within the soil, and on those plots where the phos- phoric acid has been reduced by cropping, the residue is in a com- paratively insoluble form, attacked with increasing difficulty by the dilute acids employed.(3). 1 n order to compare the relative powers of attack possessed by the acids on the different plots, it is convenient to take as a standard for each plot the amount dissolved by the citric acid and reduce the results given by the other acids to this basis. The following table is thus obtained : TABLE IV. Amount dissolved by 1 Plot 2b .I 3. 1 5. 6. 7. 9%. I-l-l----- 100 Citric acid .............. .I 100 1 100 1 100 100 I100 --I- /- I-/-- _~ 63 -5 59'2 19.3 7.0 66.2 60 '4 16.7 7.5 66.8 23 *7 10.8 7 *1 of the It is clear that some difference exists between the actions various acids; if a given acid has twice the solvent power of another in dealing with one soil, i t does not follow that the same ratio mill be preserved on passing to a soil of a different type.The solution of hydrochloric acid has about two-thirds the solvent power of the citric acid in dealing with soil from the group of plots 5, 6, and 7, which receive minerals alone or with ammonium salts ; one-128 HALL AND PLYMEN: THE DETERMINATION OF half with the soil from 2b, which contains much organic matter ; and less than one-fourth with the soils from plots 9a and 16, where the minerals have been accompanied by nitrate. When compared with citric acid, acetic acid also dissolves a smaller proportion of the phosphates in the soils from the nitrated plots 9a and 16, but a higher proportion than usual when dealing with the dunged plot 5. Carbonic acid dissolves a fairly constant proportion of the phosphates dissolved by the citric acid except in dealing with the dunged plot, when its solvent powers are comparatively high.The attack of ammonium citrate is relatively speaking at its best in dealing with the dunged plot and with the continuously un- manured plot. (4). Turning now to the practical question, which acid yields results most in accord with the past history of the plots, it will be convenient to arrange the results in a fresh form. I n the following table (V), the amount of phosphoric acid dissolved from plot 5 (minerals only) will be taken as the standard of comparison, thus showing the varia- tion caused by the plots in the case of each acid. Plot 5 is chosen for the standard, as it has been continually manured with minerals, and but scantily cropped owing to the absence of nitrogen; it should therefore contain the greatest amount of '' available " phosphoric acid.TABLE V. Plot 5 . . . . a , ,, 2b ... , , 6 ...... , , 7 .. I . . . ,, 9a ... ,, 16 ...... , , 3 ...... Total. 100 91 *9 85.5 83.8 71.9 68.9 50.0 Citric. 100 93.5 73 '0 68-8 57'8 40'8 15'7 Ammonium citrate. 100 112 a7 '4 78.8 50 '8 36.4 17 -8 HCI. 100 62 *2 73.3 67.5 19'4 14'2 5 - 8 Acetic. 100 169 87.8 68'4 32 *7 16'3 11.2 c** 100 164 53.5 51.7 36.2 19'0 8'6 It is seen that all the weak solvents give more trustworthy information about the soil than the strong hydrochloric acid does. With the strong hydrochloric acid, the variation in passing from the richest plot, 5, continuously manured with superphosphate and very scantily cropped, to the poorest plot, 3, which has been cropped without manure for 50 years, is only 100 : 60, whereas with other acids the ratio varies from 00 : 1 7 4 to 100 : 5.8.With a few exceptions, each of the cids would set the plots in the same order of fertility ; the ratios ofAVAlLABLE PLANT FOOD IN SOILS. 129 -~ 1275 1193 1115 1005 738 520 165 attack shown by citric acid and ammonium citrate are fairly similar, those of acetic and carbonic acids are still more alike. Acetic and carbonic acids and ammonium citrate rate 2b, the dunged plot, as richer than 5, the plot which receives minerals only. Hydrochloric acid rates the dunged plot very low, below 6 and 7, receiving mineral manures with ammonium salts ; hydrochloric acid also rates 9a, the nitrated plot, very lorn, extracting less than one- third as much from this plot as from plots 6 and 7, whereas citric acid would make this plot almost as rich as 6 and 7.With the variable factors introduced by the long-continued use of dung, ammonium salts, and nitrate respectively, it would be difficult to say which of these plots would be shown by crop- ping as relatively the richest in phosphoric acid; the surplus of the phosphoric acid supplied as manure over that removed in the crop during the last 42 years gives some figures wherewith to form an opinion, but one that does not take into account the different states of combination into which the phosphoric acid has entered in the soil. The following table compares the surplus of phosphoric acid added to the soil during the last 50 years with the amounts removed from each plot by the various acids, assuming for the fine earth down to the depth of 9 inches, an average weight of 2,500,000 lbs.per acre. The figures are in pounds per acre. - _ ~ - ~ 970 1082 707 665 492 TABLE VI. 900 560 660 607 175 127 52 Plot 5.. . . . I ,, 2 b . . , , , 6 ...... , , 7 ..... I ,, 9a ,. ,, 16 ....., , , 3 ..... - _ _ _ 245 415 215 167 80 40 27 *5 Surplus P,O, retained by soil. 2582 2619 2355 1985 1885 765 - 467 P,05 dissolved by I 1 HC1. I Acctic. CO,. Citric. ' Amnionium citrate. __.~_ 145 237 77 -5 75 52.5 27 *5 12.5 The following table shows the calcium carbonate and the loss on The loss on ignition includes ignition of the soils under consideration. both organic matter and water of hydration, but as the latter is likely t o be constant in dealing with soils from the same field, the variations in the loss on ignition represent pretty nearly the varia- tions in the amount of organic matter present.VOL. LXXXI K130 HALL AND PLYMEN: THE DETERMINATION OF 6 . 2-50 3.75 2-03 TABLE VII. 7. 9a. ~ - _ _ _ _ 2'62 4'17 4'44 4'49 1.92 2.06 1 2b. 1 3. 1 5. --I---- I---/- Calcium carbonate . . , . . . 3 -86 3-55 3'67 Loss on ignition .........I 6'21 I 3'32 1 3'65 Hygroscopic water lost at 100" ..................... 2'26 1'92 1'85 16. 3.03 4 -34 2'33 The amounts of either calcium carbonate or organic matter present in the soils do not shed any consistent light on the different rates of attack shown by the solvents employed. The amount of calcium carbonate present is in no case sufficient to neutralise the acids, for which purpose about 15 grams of the carbonate would be required.Much of the calcium carbonate in the soil of plots 6 and 7 has been removed by the continual use of ammonium salts, and this may ex- plain why the hydrochloric acid dissolves far more from these plots than from the nitrated plot 9a, which is richest in calcium carbonate. Phosphoric mid-Broadbnlk Field. I_ 00.5 r C m c p HCl 1 Acelic I Carbonrc 1 Ammon Cirrate T G l c> I0AVAILABLE PLANT FOOD IN SOILS. 131 Carbonic. But the acetic acid, the solvent action of which is little affected by variation in the calcium carbonate present, also dissolves less from 9a than from plots 6 and 7. On the other hand, the dunged plot is rich in calcium carbonate and is comparatively resisted to hydrochloric acid, yet it is the plot which yields the most to acetic acid.It is noticeable that the citric acid and ammonium citrate solutions contain considerable quantities of organic matter, silica and salts of iron and aluminium. The same mineral materials are attacked by the hydrochloric acid, but are not present to any appreciable extent in the solutions in acetic and carbonic acids. The comparative action of the various acids may be most clearly seen in the diagram on p. 130, where the heights of the vertical columns are proportionate to the amounts of phosphoric acid dissolved in each case. For purposes of comparison, the total phosphoric acid soluble in strong hydrochloric acid is added, Gut plotted to the smaller scale of one-tenth. Phosphok Acid Results on other Soils.The following table shows the percentages of phosphoric acid dis- solved by each of the acids from the soils 1 to 12 previously described, arranged according to the total amount of phosphoric acid they contain : Amlnoniulr citrate. TABLE VIII. Soil. 1 2 3 4 5 6 7 8 9 10 11 12 Citric. _~.___^_ 0-0055 0.0085 0'0100 0'0029 0'0082 0.0033 0.0233 0-0210 0'0085 0.0071 0'0240 0.6420 HC1. 0 -0024 0'0013 0'0035 0'0021 0'0031 0'0003 0-00435 Acetic. 0'0007 0.0007 0'0016 0.0007 0'0011 0'0003 0 *0006 0.0067 0 0016 0.0033 0'0013 0.0030 0.0017 0.0023 0-0008 0*0011 0'0019 0'0080 0.0295 0.0128 i 0'0104 I 0.0099 ' 0'0122 1 0,0182 i 0.0210 0'0040 0'0016 0-0022 ' 0-0081 0-0022 0.0019 i 0.0022 I 0.0089 0'0360 0.0120 0.0089 ~ 0.0540 0'0167 1 0'0056 0*0014 ' 0.0166 Strong HC1.Total. 0.073 0.089 0.089 0.104 0'110 0.112 0.118 0'121 0.142 0.145 0.152 0.163 (5). It is at once seen that the order in which the soils are arranged according to the total phosphoric acid is not the order of their relative richness in 'I available " phosphoric acid as judged by any one of the dilute solvents. This is only to be expected, considering the very differ- ent types of soil here brought together. The results generally afford K 2132 HALL AND PLYMEN: THE DETERMINATION OF strong confirmation of the practical value of dilute solvents in judging of the need of a given soil for a phosphaticmanure. With three excep- tions, all the soils contain more than 0.1 per cent. of total phosphoric acid, which has Seen regarded as sufficient for fertility ; yet the cropping tests of these soils show that; onIy two, 11 and 12, are at all properly furnished with phosphoric acid.If, on the contrary, Dyer’s limit of 0.01 per cent. of phosphoric acid soluble in 1 per cent. citric acid be taken as a criterion, the two latter soils are sharply dis- tinguished from the rest, as containing 0.024 and 0.042 per cent. re- spectively, and the others with two exceptions would be rated as in need of phosphoric acid. With acetic acid as a solvent and a limit of 0.0025 per cent. of phos- phoric acid soluble, all the soils except the two, 11 and 12, known to be provided with phosphoric acid, would be rated as in need of a phos- phatic manuring. (6). The action of the different acids can be best reviewed by plot- ting them as before, and also by recalculating the results in terms of the amounts dissolved by citric acid from each soil, compare Table IX (p.133) with Table IV (p. 127). Table X. (p. 133) shows the calcium carbonate, the hygroscopic moisture, and the loss on ignition for each soil. I n Table XI (p. 133) the soils 1 to 12 are arranged as the Broadbalk soils in Table V (p. 128) ; that is, one soil is taken as a standard of comparison (in this table, No 5, which is known to be very slightly if a t all in need of phosphatic manuring); the phos- phoric acid dissolved by each acid from this plot is called 100, and the amounts dissolved by the same acid from the other plots are reduced to this standard. An inspection of the diagram (p. 134) shows that citric, acetic, hydro- chloric and carbonic acids agree, with one or two exceptions, as to the comparative richness in available phosphoric acid of any plot.The vertical columns representing the acids rise and fall together in pass- ing from plot to plot, as was the case with the Broadbalk soils. The ammonium citrate, however, gives results essentially different ; i t rates soil 2 as better than 3, the other acids make 3 distinctly richer than 2 ; again, i t rates 4 below 5, contrary to the relative position assigned to these two soils by the other acids and by cropping experiments. From all the soils 1-8, 10, and 12, ammonium citrate extracts more than citric acid, a result never obtained with any of the Broad- balk soils. The high and irregular results given by ammonium citrate as compared with the other acids may probably be attributed to the comparative richness of these soils in organic matter and their poverty in calcium carbonate.The soils, 2, 4, 6, 7, and 8, which are ratedSoil. 1. I 2. 1 3. i 4. ~ i 5. i 6. I 7. 1 1 9. i l O . l l 1 . TABLE IX.-Percentages of P,O, dissolved by other solvents calculated on that taken up by citric acid. 12. Citric acid ....................... Amnionium citrate ............ Acetic acid ........................ Carbonic acid,. .................... Hydrochloric acid.. ............. 100 144 58-9 100 1 100 100 347 138 362 1;:; 1 35.0 73.2 15'9 24'4 15-3 29'6 60.6 \ 100 100 121 ~ 370 37'8 ' 9.1 13'9 ' 9'1 27'9 24.0 Calcium carbonate ............... 0.03 trace 0'04 1 0.01 0'08 Loss on ignition ...............5.93 10.43 3.68 4'74 3.09 Hygroscopic water.. ............. 4'26 6'34 3*13 2'34 2'73 I ! 100 137 32.7 4.5 8 -3 0.21 0.03 6.01 9'10 4.87 2.75 7. ____- 162 184 140 52% 48-0 107 . - Citric acid ........................ Ammonium citrate ............ Hydrochloric acid.. ............. Acetic acid.. ..................... Carbonic acid.. ................... Total ............................. 8. 256 212 216 138 110 83-0 100 100 31 -9 7-5 9'1 the loss 0.01 9 '19 4'74 100 94 *8 47 *1 18.9 25.6 TABLE XI.-Percentages of dissolved P,O, calculated on plot 5. on ignition. 3.00 0.03 1 4'59 5.09 ~ 7.11 4'08 3'98 j 2.76 1 2-06 5. 1 1. --I-- 100 67 *3 100 61.4 100 66.1 2. 104 298 42.3 57'0 56 '8 80'6 3. 122 129 113 139 129 81.3 4. 35.0 67 -7 61 '4 76.0 ' 94'2 105 6. 40.2 9.7 26.3 34'9 123 102 I 100 100 126 69.2 31.1 1 69'6 9.104 129 141 129 81'4 95.2 100 129 85.7 28% $ 21.2 t? M cd t? P 3'32 3 4.01 q 1.87 8 tJ L m I I 0 86.0 90 *o 70 *6 96 '1 170 132 293 168 539 491 138 62 '0 512 545 1161 1053 390 148 +134 HALTA AND PLYMEN: THE DETERMINATION OF u3 0 0 t- I I 1 m l-l 0 0 0 0 0 Phosphoric oxide, P,Os, per cent.AVAILABLE PLANT FOOD IN SOILS. 135 comparati+ely high by ammonium citrate, are rich in organic matter, 2 and 7 being the only pasture soils in the group, and 8 an artificially made soil. No. 5, which is rated low by ammonium citrate, is excep- tionally deficient in organic matter. The quantities dissolved by acetic and carbonic acids are very similar; it is to be noticed that acetic acid dissolved slightly less than carbonic acid from the soils 1-10, which are, with the exception of 9, short in calcium carbonate, but that it obtained the larger amount of phosphoric mid from soils 11 and 12 and from the Broadbalk soils which contain more than I per cent.of calcium carbonate. On close inspection of the figures many differences are evident in the mode of attack of the various acids, which when followed up on a number of soils will provide information as to the forms i n which the phosphoric acid of the soils is combined. The authors, however, wish in this communication to confine themselves to the question of which dilute acid yields results most in accord with the known history of the soils, and is therefore most likely to be useful in judging an unknown soil. (7). A few figures may be here inserted showing the effect of varia- tion in the strength of the acid used, and of additions of calcium carbonate to the soil.Dyer (Zoc. cit.) has already given figures showing that an increase in the strength of the acid results in more phosphoric acid going into solution ; the authors' results are in the same sense : TABLE XII. Solvent. Citric acid 0.2 normal ....................... ,, 1 per cent. ........................ ,, 0.1 normal ........................ Percentage of P,05 dissolved. Soil 7. Soil A. 0.0198 0'0133 0.0084 0.0424 0'0349 0.0206 Soil A does not appear elsewhere in this paper, but was chosen as one rich in phosphoric acid and calcium carbonate, but poor in organic matter, and thus a complete contrast to soil 7. Soil 7 was further mixed with varying amounts of calcium carbonate, obtained by grinding Iceland spar to a fine powder, and subjected to the action of citric, acetic, and carbonic acids, with the following results :136 HALL AND PLYMEN: THE DETERMINATION OF Soil only ._......... , ... .. ... . ........ ... + 2 per cent. calciiini carbonate , , 3 7 + 5 9 J ,, ,, 9 9 + I 0 9 , 9 , ,, TABLE XIII. 0.0133 0’0006 0 ‘0090 0 -0009 0.0056 0.0006 0*0007 0.0007 i Phosphoric acid. I 1 - - I Soil 7. I Citric. I Acetic. I- I- co,. __ 0*0011 0.0007 0 *0009 0.0009 Potash. _____ Citric. - 0.0148 0.0092 0.0092 - Acetic. - 0’00714 0.00706 0.00710 These trials were not pushed further; the citric acid as it was neutralised by the calcium carbonate dissolved less and less phosphoric acid, until with 10 per cent.of calcium carbonate (more than is requisite for complete neutrality), the amount of phosphoric acid dissolved approximated to that dissolved by carbonic acid only. The solution effected by carbonic acid is independent of the calcium carbonate present, and that effected by acetic acid approximately so, because the liberated carbonic acid is an equally efficient solvent. Review of Results. (8). On reviewing the whole of the results, it seems very improbable that any distinction of kind can be drawn between (‘ available ” and ‘6 non-available ” compounds of phosphoric acid in the soil ; that is, there is not a compound or group of compounds available,” which can be wholly removed by the plant or dissolved by Rnn acid before the remaining compounds are attacked.Were this the case, those soils which contain only n limited amount of ‘‘ available ” phosphoric acid would yield all of it or none to a given solvent, and the strength of the solvent would be without influence on the result when the time limit is large. On the contrary, the amount of phosphoric acid dissolved varies with both the nature and strength of the acid. There is no reason for regarding the phosphoric acid dissolved by the citric acid solvent as the available ” phosphoric acid in the soil rather than that which is dissolved by the acetic acid. A soil which contains much or little ‘ I available ” phosphoric acid according to one acid would be rated in the same way by another acid, even when the absoluteamounts dissolvedare ten times as great in one case as in another.The individual acids possess a certain selective power for different combinations of phosphoric acid and attack the different types of soils with more or less vigour, but in the main the relative action of all the acids on all the soils is alike,AVAILABLE PLANT FOOD IN SOILS. 137 The phosphoric acid of a soil must not be looked on as existing in certain compounds A, B, C, D, &c., of which A and B are insoluble and unavailable, C and D as “available ” ; rather A, B, C, D, &c., repre- sent compounds possessing in each soil a coefficient of solubility, vary- ing with the acid and with their own physical condition. The latter factor affects all the acids alike, and combined with the absolute quantity of the phosphoric acid in the soil determines the “ available ” phosphoric acid.The available phosphoric acid measured by a given acid depends on the coefficient of solubility possessed by the acid and the relative proportions of A, B, C, D, &c., in the soil. As soils of the same type contain A, B, C, D, &c., in roughly the same proportion, the latter factor is eliminated and the amounts of available phosphoric acid from different soils as measured by any one of the acids will be proportional to the phosphoric acid which is really (‘ available,” so that all the acids will show roughly the same relations between the soils. Again, a soil may contain di- and tri-calcium phosphates, ferric and aluminium phosphates, and organic compounds of phosphorus like nuclein and lecithin; it would be no gain to discover a reagent which would dissolve the di- and tri-calcium phosphates only and leave the rest, for the physical conditions of these phosphates may render them less ‘‘ available ” to the plant than the other com- pounds of phosphorus present which happen to be in a favourable physical or mechanical condition for solution.On this view the hope must be abandoned of finding any particular acid which will dissolve out the ‘‘ available ” phosphoric acid and leave the rest ; in the results obtained by any acid, the factors are too numerous and variable to admit OF exact discussion; because of its complexity, the method becomes empirical and the best acid is that which most accords with experience. (9). I n forming a conclusion as to the most suitable solvent, three things should be taken into account : (a) The amount of phosphoric acid dissolved should show a wide variation in passing from soil to soil, so as t o discriminate sharply between rich and poor soils.The largest quantity of phosphoric acid dissolved by strong hydrochloric acid from any one of the soils examined is 0.228 and the smallest 0.0727 per cent. ; other things being equal, variations of this order would not discriminate so well between the soils as the variations exhibited by citric acid, which lie between 0.051 and 0.0029, or of acetic acid, which lie between 0.012 and 0*0003 per cent. ( b ) The amount of phosphoric acid dissolved from normal soils should be sufficient for exact estimation, so that the variations ex- hibited may be of a different order of magnitude from the experimental error, which is inevitably large.138 HALL AND PLYMEN: THE DETERMINATION OF (c) The variations in the amount of phosphoric acid dissolved should so follow the known history of the soils that the reaction of an un- known soil to phosphatic manures can be predicted from its analysis.For this reason, the action of the acid should not be markedly affected by other variable constituents in the soil, such as calcium carbonate and organic matter. Ammonium citrate fails to meet the last requirement ; although when dealing with soils of one type, like the Broadbalk soils, its results fall into line with those given by the other solvents, yet with the other soils the indications provided by tlie analysis do not agree with experience.Soils 2, 7, and 8 yield comparatively large quantities of phosphoric acid to ammonium citrate solution and would be rated as sufficiently supplied with phosphoric acid, but 2 and 7 respond freely to phosphatic manures. Soils 4 and 6 yield more phosphoric acid than 5, which is quite contrary to the crop results. These discrepancies are due to the solubility of the humus containing phosphorus compounds in the alkaline ammonium citrate solution, thus introducing material of a different order of solubility, and as the ammonium citrate solution offers no compensating advantages it may be dismissed as uneuitable. Hydrochloric acid presents many anomalies of attack; it has very little solvent power for phosphoric acid when dealing with soils 1-10 which are poor in calcium carbonate; for example, it can only dissolve 0.0031 per cent.from soil 5, which is fairly provided with phosphoric acid as judged by the crop, whereas it can get 0.0021 per cent. from the unmanured plot at Rothamsted, and as much as 0.0167 per cent. from soil 11, the poorish chalky Wye soil which had been unmanured for 5 years. The Broadbalk plot 9a, which receives minerals and sodium nitrate, is rated very low ; it yields only three times as much phosphoric acid as the continuously unmanured plot, and less than one-third as much as the corresponding plot 6, which receives ammonium sulphate instead of sodium nitrate. The duuged plot is also rated as inferior to the plots receiving minerals and ammonium salts. On the whole, the results obtained with hydrochloric acid are difficult to reconcile with experience, and present no features which would justify its recommendation in place of citric acid.Wccter chcwged with curbonic acid is so similar in its action to acetic acid, both in the relative and absolute amounts dissolved from the various soils, that the greater convenience of using the latter acid would cause it to be preferred. The choice thus becomes narrowed down to acetic and citric acids. Of these two, acetic acid better satisfies the first condition laid down With the above, the variations in the amounts dissolved are larger.AVAILABLE PLANT FOOD IN SOILS, 139 Broadbalk soils they range from 169 to 11.2, against 100 to 15.7 for citric acid (Table V). On the other soils they range from 1053 t o 26.3, against 512 to 35 for citric acid (Table XI).AS regards the second criterion, the quantities of phosphoric acid dissolved by the acetic acid are very small, one-tenth t o one-fifth of the amount dissolved by citric acid. The limit to be taken as indicating the need for phosphatic manuring would be about 0.002 per cent., which means the determination of only 0.001 gram of phosphoric acid in the 500 C.C. of solution commonly employed. On the other hand, the acetic acid solution is the easier t o manipulate, owing to the absence of iron, alumina, silica, and dissolved organic matter ; SO that the experimental error is not likely t o be greater than with citric acid, less indeed in unskilled hands. AS regards the interpretation of the results, it is clear that all soils deficient in calcium carbonate,as 1-8, are rated very low by acetic acid.I n such soils, much of the phosphoric acid is present as precipitated ferric and aluminium phosphates, which are left practically untouched by the acetic acid, yet there is no evidence that such phosphates are quite “ non-available ” for the crop. Soil 5 is a case in point ; acetic acid dissolves only 0.001 per cent. of phosphoric acid, yet the crops on this soil find no great need of phosphates. The Broadbalk soils are very clearly differentiated by acetic acid, the doubtful point being the comparatively low position attached to 9a and 16, the nitrate plots. The position assigned to these two plots and to 5 in the other group makes it difficult to accept acetic acid as the most ‘( critical ” solvent.Considering the results yielded by citric acid, some difficulty of interpretation attaches to soils 2, 3, 7, and 8. Taking the limit of 0.01 per cent. of phosphoric acid suggested by Dyer, soils 7 and 8 are above the limit with 0,0133 and 0.021 per cent. respectively ; soil 3 is on the limit, and soil 2 is a little below with 0-0087 per cent. ; yet the field trials indicate a need of phosphates on soils 2, 3, and 7, probably on 8 also, although as an exceptional soil it is hardly comparable with the rest. Of all the soils examined, soils 2, 7, and 8 show the greatest loss on ignition ; 2 and 7 are old pastures, 8 is a made soil containing leaf mould, and as citric acid dissolves some of the organic matter of soils, i t is to this source that the high proportion of phosphoric acid yielded by these soils may be attributed.Probably the superior limit of 0.01 per cent. of phosphoric acid, as indicative of the need of phosphatic manuring, requires revision when dealing with pastures and other soils rich in organic matter. The results yielded by soil 5 also require a little explanation ; the citric acid solution only dissolves 0.0082 per cent., yet the crops show no exceptional response to phosphatic manuring. The soil is a very140 HALL AND PLYMEN: THE DETERMINATION OF light sandy loam, typical of many of the soils derived from coarse, ferruginous sandstones of secondary age. It contains very little calcium carbonate (0.08 per cent.) and little organic matter (loss on ignition 3-08 per cent,).The phosphoric acid must be largely present in this soil as ferric phosphate, and although citric acid is a better solvent than acetic acid in such cases, even the citric acid does not indicate all the phosphoric acid that seems to be ‘L available ” for crops. Gerlach (loc. cit.) has already indicated that typically sandy soils from which citric acid dissolves less than 0.01 per cent, of phosphoric acid may give little response to phosphatic manures. As regards the Broadbalk soils, the results yielded by citric acid are more in accord with our knowledge of the plots than those furnished by acetic and the other acids; in particular the plots receiving nitrate 9a and 16, though below all the others except the unmanured plot, are shown as still high above the limit which may be taken to indicate the need of phosphatic manuring.Reviewing the whole body of results, the authors consider the 1 per cent. solution of citric acid gives results which are most in accord with the known history of the soils. On soils well provided with cal- cium carbonate all the acids tried give very similar relative results, but this type of soil is rarely in need of phosphatic manuring, and the practical question for which the analysis is performed, whether the soil is in need of phosphatic manuring or not, usually arises in the case of soils poor in calcium carbonate. From these soils, acetic acid can extract so little that it reduces them all to practically the same level, whilst citric acid is able to dissolve the natural phosphates of iron and alumina in a manuer more in accord with the natural attack of crops.11. POTASH RESULTS. Methods of analysis based upon the solvent action of weak acids must be even more empirical, when dealing with the potash in soils than with the phosphoric acid, Certain definite compounds of phos- phorus, such as the organic residues, the phosphates of the sesqni- oxides, the neutral and acid phosphates of calcium and magnesium, exist in the soil, and are, to some extent, differentially attacked by the various solvents, but the potash compounds are far more com- plex and indefinite. I n addition to more or less weathered silicates, like felspar and glauconite, there are indefinite compounds formed when humus and clay withdraw potash from the solution produced by the weathering of potash minerals or the application of manures, Even the amount of potash dissolved by strong hydrochloric acid from a soil is a purely conventional figure, dependent ou the strengthAVAILABLE PLANT FOOD IN SOILS.141 of the acid and the length of attack; the Broadbalk soils, for example, yield about 0.5 per cent. of potash to strong hydrochloric acid, but the total potash contained in the soil from plot 5 , as determined after breaking up the soil completely with ammonium fluoride, amounted to 2-26 per cent. The tables below show the results yielded by the soils from the same seven plots of the Broadbalk Field at Rothamsted, and by five other soils previously described ; the results are also set out graphically on p. 143 in the same manner (compare p.131) as were the phosphoric acid results. TABLE XIV. Potash-soils from Broadbalk Field. - 13iot. 2b 3 5 6 7 Da, Manuring. Dung .......................... Unmanured. .................... Minerals only.. ................ Minerals 200 Ib. ammouiuni salts ........................... Minerals 400 lb. ammoniuin salts ........................... Minerals 275 lb. sodium nitrate ........................ Minerals 800 lb. ammon- ium salts .13 years Unmanured .19 years Minerals 550 lb. sodium nitrate.. ......... .10 years I ........ ..... Citric. 0'0400 OmO043 0.0458 0.0322 0-0233 0 *0272 0.0203 HCI. 0'0684 0'0147 0.0522 0.0487 0.0464 0'0414 0'0421 Acetic. 0'0451 0'0082 0.0307 0.0271 0-0240 0-0237 0'0184 arbonic 0.0380 0'0111 0.0215 0.0151 0.0091 0.0238 0.0145 - Strong HCI.0-453 0.380 0 '463 0.530 0.500 0'440 0.504 On examining the results yielded by the Broadbalk soils, it is noticeable that the amounts of potash dissolved by the different acids are very similar, much more so than with phosphoric acid. Citric acid dissolves ten times as much phosphoric acid as the water charged with carbonic acid, whereas hydrochloric acid, the most energetic solvent for potash, dissolves only about three times as muchas the weak- est, which is again carbonic acid. On the whole, each acid leads to the same conclusions with regard to the relative richness of the plots in '' available " potash, but citric acid shows the widest variation in passing from plot to plot; the ratio of 2b, the dunged plot, t o 3, the un- manured plot, is 9.3 : 1 for citric acid against 4.65 : 1 , 5 5 : 1, and 3.4 : 1 for hydrochloric, acetic, and carbonic acids respectively. The results with the Broadbalk soils would indicate that the citric acid is the most "critical'' solvent for '' available " potash in the soil.HALL AND PLYMEN : THE DETERMINATION OF - (0 L 0 C 0 5 ]e r( 0 0 e 0 0 0 0 Potash, K,O, per cent.AVAILABLE PLANT FOOD IN sons.Loss 011 ignition. TABLE XV. Potash dissohed by wet& acids from other soils. Water lost a t 100". Soil. -1- Citric. 0.0104 0.0156 0.0053 HCI. -- 0.0480 0'0580 0.0113 0.0154 0'0178 0*01'76 1 0.390 4.59 0'0241 0.378 3'32 0.0057 1 0.313 0'01 j Acetic. Carbonic/ 'EZg 1 CaCO,. 0-0059 j 0'0062 j 0.439 0'0053 , 0.0079 ' 0'592 1 I 0'08 0'02 143 4.08 4'01 4'74 3.09 12.53 2-06 1.87 3-13 2 3 4 13.04 11 12 4 s 13 Of the other soils examined, 11 and 12 should be compared together as soils freely supplied with calcium carbonate, whereas soils 4, 5, and 13 are notably deficient in this constituent.Soils 11 and 12 are from the plots, side by side, on the same field, shown by experi- ment not to be particularly in need of potash manuring. No. 11 had been cropped without manure for 5 years, during which time 12 had received each year a general manure containing 1; cwt. per acre of potassium sulphate. All the weak solvents show 12 as richer than 11 in ' I available " potash, whereas the strong hydrochloric acid would make them practically alike. The difference between them is most sharply drawn by citric acid; it is also noticeable that citric acid shows both plots as comparatively poor in '' available " potash, the other three acids would rate them as comparatively rich. Of the other three soils, field experiments have shown that 4, a strong clay, is in no need of potash manuring, but 5 and 13 gave very marked returns for potash dressings.Strong hydrochloric acid would make both 5 and 13 much richer in potash than 4 ; it dissolves 0.593 and 0.439 per cent. respectively from 13 and 5, against 0.313 per cent. from soil 4. Dilute hydrochloric acid would also set soil 4 below 5 and 13 in (' available " potash, acetic and carbonic acids would rate them alike, the differences between the various results being of the same order as the experimental error. Citric acid alone draws a sharp distinction between the soils ; it dissolves 0.025 per cent. from 4, and only 0.011 and 0.0085 per cent. respectively from the other two soils. The results with these five soils afford most striking evidence of the practical value of weak solvents as against extraction with a strong acid in judging of the requirements of n soil for a potash manure ; a t the same time, they indicate if may be necessary in the light of extended experience to adopt different limits for soils of different types, for example, soils rich or poor in calcium carbonate. 0*0050 0'0093 0'0250 0'0110 0'0085144 THE DETERMINATION OF AVAILABLE PLANT FOOD IN SOILS. Of the four weak acids employed, the authors regard citric acid as furnishing results most in accord with the history of the soils exam- ined. Xummar y. The authors have compared the amounts of phosphoric acid that could be extracted from nineteen different soils by a 1 per cent. solution of citric acid, by equivalent solutions of hydrochloric acid and acetic acid, by a saturated solution of carbonic acid, and by an ammoniacal solution of ammonium citrate respectively. Seven of these soils were from plots on the Broadbalk Field, Rothamsted, which had been continuously manured in the same manner for forty-two years previously; the re- maining twelve were soils of very varied origin, which had been the subject of crop experiments and whose reaction to phosphatic manuring was well marked. I n the same seven soils from the Broadbalk Field, the authors deter- mined the potash extracted by the same dilute solvents, with the exception of ammonium citrate ; five other soils of different origin, whose response or otherwise to potash manuring had been tested by experiment, were also examined in the same way. Determinations were also made of the phosphoric acid and potash dissolved after long digestion with strong hydrochloric acid, of the loss on ignition, and of the earthy carbonates present in each soil. The authors conclude :-(1). That no sharp line of distinction can be drawn between ‘‘ available ” and non-available phosphoric acid and potash in the soil, and that any process of determining the ‘ I available ” constituents is an empirical one, dependent on the strength and nature of the acid used. (2). That the weak solvents give information as to the requirements of a given soil for mineral manures of a far more trustworthy nature than that which is afforded by such a solvent as strong hydrochloric acid. (3). That of the acids examined, the 1 per cent. solution of citric acid gives results most in agreement with the recorded history of the soil, although there is evidence that the same interpretation cannot be put on results obtained from all types of soil. Soum EASTERN BGRICULTUKAL COLLEGE, WYE.