UMASS/AMHERST 31EDbbQD56SSDfi7 wmm mill: /ij fe:! ii IMP*: . "iilfiiliiii iilliilifiii:;.:';.'...;: Digitized by the Internet Archive in 2010 with funding from Boston Library Consortium IVIember Libraries http://www.archive.org/details/studyofsoilpotasOOcurr T^reai^iiir of tiw* DECS 1912 REPRINTED FROM -A^jriciiltTaral ORIGINAL COMMUNICATIONS, EIGHTH INTERNATI@^IJeo:# CONGRESS OF APPLIED CHEMISTRY. Vol. XV— Page 51. '"' A STUDY OF SOIL POTASSIUM Bt B. E. Cubry and T. O. Smith Durham, N. H. Until recent years the theory of soil fertilization has rested almost entirely upon the practice of adding plant food or amend- ments to the soil in the form of fertilizers and manures for the pur- pose of increasing production. In practice the ultimate aim is to obtain an increased production at a profit. With our present knowledge of soil management, tillage, and fertilization it is a simple proposition to increase yields but it is not a simple prop- osition to secure profitable increased yields. It is doubtful from a practical point of view whether the maximum yield is the most profitable. The practice of applying fertilizers is followed because the amount of plant food in the soil is considered insufficient or unavailable for the needs of the crop. For special cases as in market gardening and greenhouse culture where high-priced crops are grown the relatively small cost of fertilizers becomes almost negligible. In other crops such as the average farmer grows the difference of a few dollars makes the difference between profit and loss in an operation on an acre. The aggregate of values of farm crops is largely derived from the ordinary crops grown under ordinary conditions. In New Hampshire one of the large cash crops is hay. The farms have generally been allowed to run down and become unpro- ductive. The soil is apparently as good as it ever was but through neglect production has decreased. A larger acreage is devoted to grass than to any other crop. This condition has become estab- lished through the natural adaptibility of the soil to grass culture and through certain economic factors due to location and labor problems. The heavy horse and ox power once so common on New Hampshire farms is now a condition of the past. Partly because of this the land does not get the cultivation it once had. 61 52 Original Communications: Eighth International [vol. Commercial fertilizers have in part been substituted for tillage in order to maintain production under conditions of lessened tillage. A good many general observations lead one to believe that the present method of using commercial fertilizers in the production of hay is often unprofitable. There seems to be very little doubt that in the common use of fertilizers a very large proportion of the plant food added is never converted into profitable plant production. It is doubtful if the use of commercial fertilizers can long be substituted for tillage. There are other important soil and plant growth factors which enter into profitable produc- tion and which cannot be eliminated by the use of fertilizers alone. The soils of New Hampshire are extremely varied in charac- ter and are for the most part of granitic origin. Almost any type may be found between the limits of sand and pure boulder clay. However, the sands in many instances seem to be rich in f eld- spathic and other minerals. Many of the types have been con- siderably washed while others seem to have been formed in place. Areas of limestone soil are very limited and are found in only a few sections. In connection with the use of commerical fertilizers in the production of grass some very interesting facts have been found. Many of the grass fertilizers carry a large amount of potassium. Also the tillable soils are on the average very rich in potassium. Those soils which are relatively poor in potassium contain large amounts in the aggregate. Some of the boulder clay soils con- tain as high as 3f % K2O. Assuming that an acre foot of soil weighs three million pounds, such a soil would contain fifty-two and one-half tons of K2O per acre. Some of the medium clay soils carry about 2% K2O. while the light soils carry still less. Some very sandy soils have been found which contain as much as 1% K2O. Such soils carry a considerable amount of mineral in connection with the sand. With these large quantities of potassium present in the soil there came the question of the use of potassium in commercial fertilizers. Is it not reasonable to think that the soil under proper tillage conditions could supply enough potassium for the needs of the crop without the addition of potassium from artificial xv] Congress of Applied Chemistry 53 sources? It is a remarkable fact that where potassium is added in comparatively large quantities as a fertilizer constituent that that amount is very insignificant when compared with the amount of potassium present in the soil under natural conditions. To illustrate : sixty pounds of K2O is higher than the average applica- tion per acre. In a soil which carries 2% K2O the amount per acre foot aggregates thirty tons. In comparison with sixty thousand pounds, siTjty becomes a very small quantity: under such conditions one pound is added to one thousand pounds already present. In this connection it is interesting to note that the soil very rapidly renders the applied potassium insoluble. For instance, when dilute solutions containing potassium salts are percolated through columns of soils the potassium is removed from solution and changed into a comparatively insoluble form. If the salt is the nitrate, chloride, or sulphate the acid radical remains in solution as the acid radical of some new salt. When potassium is percolated through as at phosphate both the base and acid radicals are removed from solution and no new salt appears in the percol- ate as in the first inst-ginces. The same conditions? hold when potassium salts come in con- tact with the soil in otlier ways. In order to show these reactions five soils fairly reprfjs-entative of the different types found in New Hampshire were selected. An analysis of these soils follows : 54 Original Communications: Eighth International [vol. Table I Sofl No. 1 2 3 4 5 Loss on Ignition 2.660 3 5.33 5.830 4.905 7.260 Moisture .770 1.419 0.516 .908 1.681 Si02 81.050 74.090 72.020 71.850 62.340 P,Oa .070 .084 .073 .085 .089 FezOs 1.448 2.856 2.896 2.688 3.428 AI2O3 9.017 12.195 13.446 14.702 18.638 CaO .756 .722 .764 .806 .956 K2O 1.630 1.730 1.910 2.470 2.990 N .081 .186 .290 .148 .257 Total 97.482 98.615 98.745 98.462 97.609 Na, Mg, etc., not deter- mined. 2.518 1.385 1.255 1.538 2.391 Sample No. 1 is a light sandy soil, No. 2, a light clay loam, No. 3 a sandy loam, No. 4, a heavy clay loam, and No. 5 a heavy boulder clay. Known amounts of potassium chloride in solution were added to these soils and the moisture content brought up to twenty per cent, by the addition of water. The samples were thoroughly mixed, placed in sealed jars, and allowed to stand for several weeks. They were then shaken with water and the amount of soluble potassium determined by the chloroplatinate method. XV Congress of Applied Chemistry 55 The results follow: Table II No. of Soil Grams K2O as K CI per Kg. soil Total grams K2O recovered Total grams K2O retained in the soil 1 .9354 .7012 .2342 .6236 .4471 .1765 .3118 .2024 .1094 .1559 .0072 .0587 2 .9354 .6173 .3181 .6236 .3872 .2364 .3118 .1871 .1237 .1559 .0841 .0718 3 .9354 .5517 .3837 .6236 .3460 .2776 .3118 .1724 .1394 .1559 .0709 .0850 4 .9354 .4400 .4954 .6236 .3101 .3135 .3118 .1240 .1878 .1559 .0561 .0998 5 .9354 .3243 .6111 .6236 .2712 .3524 .3118 .0779 .2339 .1559 .0437 .1122 The table shows that large amounts of potassium are taken up and held by the soils. Also, that the amount of potassium so held increases with increase in the clay content of the soil. 56 Original Communications: Eighth International [vol. If di-potassium phosphate is substituted for potassium chloride under the above conditions the same results are observed as is shown in the following table : Table III No. of Son Grams K2O as K2HPO4 per Kg. of soil Total grams K2O recovered Total grams K2O retained in the soil 1 2 3 4 5 3.1840 3.1840 3.1840 3.1840 3.1840 1.6238 .7378 .4458 .7472 .5520 1.5602 2.4462 2.7382 2.4368 2.6320 A further examination of the water extract of the soil treated with potassium chloride shows that the solubility of the acid radical is not affected by the soil. All the chloride appears in the solution as potassium chloride and as the acid radical of a new salt. The new salt or salts are chiefly chlorides of calcium and magnesium. If the acid radical is the sulphate or nitrate the same conditions of solubility occur. A possible exception is the case where more calcium sulphate is formed than can dissolve in the given volume of water. Salts of iron and aluminum must in some cases be formed in the process of these reactions. Their rare appearance in the solution is probably due to alkalinity of the water extract which in many instances is sufficient to cause these salts to hydrolyze and the bases to reprecipitate. XV ' Congress of Applied Chemistry 57 Different conditions of solubility are observed in the case of the phosphates. The extent to which they are taken from solution and retained by the soil is shown in the following table : Table IV No. of Soil Grams P2O6 as K2HPO4 per Kg. of soil Total grams P2O5 recovered Total grams P2O5 retained in the soil 1 2.4000 1.1985 1.2015 2 2.4000 .4806 1.9194 3 2.4000 .0846 2.3154 4 2.4000 .8355 1.5645 5 2.4000 .5553 1.8447 When soils react with potassium chloride, nitrate, etc., the reaction must be chemical because the amounts of new salts which appear in the water extract are equivalent to the amount of potassium removed. When soils react with potassium phos- phate both the base and acid radical are removed from solution. No new salts appear in solution and the water extract gives no evidence of the nature of the reaction. By analogy, however, the reaction must be chemical. When potassium exchanges places with calcium and magnesium these in turn form phosphates which have a very low solubility and do not appear in the solution. This is further substantiated by the fact that where large amounts of calcium sulphate are formed not all of the acid radical appears in solution. At present there are no data at hand which show the relation between bacterial activities in the soil in New England and in other sections. The probabilities are that such action is com- paratively small and that the rate of activity of bacterial action increases with increase of temperature. Plant growth is greatly affected by change in temperature. Bacterial activities undoubt- edly are affected by temperature changes in much the same way. This may account in part for the slow nitrification of the organic matter in the soil and afford an explanation why active nitrogen- 58 Original Communications : Eighth International [vol. ous fertilizers are so uniformly effective. The organic matter in the soils of old grass fields nitrifies slowly also because of lack of tillage and consequent poor aeration. The field of soil bacteriol- ogy is still almost untouched and affords some very interesting sources for speculation in the problems of soil fertility under these conditions. In order to secure as much information as possible about the potassium in these soils both laboratory and field observations, have been made. It has been shown in a preceeding table how the soil behaves toward potassium salts when applied as soluble fertilizer constituents. The amount of water-soluble potassium in many different samples of soil has been determined to show if possible whether there is any definite relation between the total potassium content of the soil and the amount of water-soluble potassium in the same soil. Such a relation cannot be established with any satisfactory assurance because it has not been possible to secure soils which differ only as regards the total amount of potassium. The total amount of potassium is usually in proportion to the amount of clay in the soil. The amount of clay has considerable effect on the nature of the organic matter. The amount and nature of organic matter apparently affects the solubility of potassium and other mineral constituents; also, some soils carry a large portion of their potassium in the form of minerals. This is particularly true of the sandy soils. The solubility of potassium in mineral form must be different from that in the form of clay. This must be true because of the influence of the clay itself. Also the past treatment of the soil may influence the solubility of potassium depending upon cultivation and whether or not com- mercial fertilizers have been used. XV Congress of Applied Chemistry 59 The solubility of potassium in pure ground feldspar for instance is very different depending on whether the solvent consists of pure water, water carrying calcium hydroxide, or such salts as sodium nitrate, calcium sulphate, etc., and whether the mineral is pure or mixed with clay. These facts are shown by data in the follow- ing tables: Table V Reagent added Amt. of reagent added in grams cc water Amount of K2O liberated in grms. Aver- age Amount of K2O liberat- ed by action of reagent 180 . 0064 1 .0069 \ .0067 .0070 CaO 1 .OI22I CaO 2 .0156 \ .0149 .0082 CaO 2 .0168^ CaS04 1 . 0076 1 CaS04 2 .0080 \ .0091 .0024 CaS04 3 .0118 J NaNOs NaNOs 1 2 .0120 1 .0121/ .0120 .0053 (NH4)2S04 (NH4).S04 1 2 .OI20I .0132/ .0126 .0059 Na^COs Na,C03 1 2 . 0096 1 .0109/ .0103 .0036 Na2HP04 Na2HP04 1 2 .0109 1 . 0087 / .0107 .0040 60 Original Communications: Eighth International [vol. Table V shows the amount of potassium going into solution in a given volume of water upon stirring the feldspar with some of the more common fertilizer constituents. Figures showing the solubility of potassium when feldspar is stirred with clay and calcium oxide follow: Table VI Bottle Grams Feldspar Grams Clay Gramn CaO. cc water Soluble K,0 in grams 1 25 0.0 180 .0014 2 25 2. 180 .0011 3 30 25 1. 180 .0072 4 30 25 2. 180 .0062 5 30 0. 180 .0067 6 30 0. 180 .0072 7 30 1. 180 .0151 8 30 2. 180 .0158 The data in Table V and Table VI were obtained by placing the different combinations in a thermostat at room temperature and stirring until there was no further action. These data show conditions which may be met with in soil studies and why it is difficult to eliminate them when making comparisons. The mineral feldspar has a certain solubility as regards the potassium when the solvent is pure water. The addition of calcium oxide, sodium nitrate and other salts increases the solubility very greatly. The presence of clay decreases this effect. All soils except the sands contain some clay, therefore the solubility of the minerals in the soil is affected by the clay and is different from what would be expected from the pure mineral. At present, it has not been determined whether the clay reduces the effect of the solvent or whether the effect is on the potassium after it is acted upon by the solvent. The prehminary part of this work was begun by looking into the field conditions to determine if possible whether anything xv] Congress of Applied Chemistry 61 could be found from this point of attack. A series of plots were fertilized with fairly heavy amounts of nitrogen in nitrate of soda, phosphoric acid in acid phosphate, and potassium in the form of sulphate and chloride. These plots have been under observation for the past five years. To date, but little information has been obtained from these plots excepting that the fertilizers disappear very rapidly after they are applied. It is an easy matter to find the nitrate and ammonium salts for sometime after they are applied but for all practical purposes it is safe to say that the potassium and phosphoric acid disappear from soluble forms after the first good rainfall and in an ordinary application they cannot be found by chemical methods now in use. This is in harmony with the results obtained in the experiments which have already been discussed and which show the chemical reaction which takes place between soils and potassium salts. Like the sulfate and chloride the nitrate radical remains soluble until it finally dis- appears through decomposition or some other destructive process. Because of the chemical change which takes place in the case of potassium fertilizers this plan of attack could not give any very definite observations. Also, the relative amounts of soil and fertilizing elements were so different that no very satisfactory results could be had in this way to determine the effect of the fertilizers on the soil constituents. On this account the effects of different salts were studied under conditions which were subject to closer control. In order to meet these conditions, solutions of various salts of known strengths, were percolated through columns of soils, the rate of percolation being controlled by means of capillary tubes and the height of the water level in the containing vessel. The flow was adjusted at the rate of about 90 c.c. per 24 hours and maintained at that rate throughout the period of observation. The strength of the solutions was made uniform on the basis of the potassium equivalent. The changes effected in the passage through the soil were determined by studying the percolate. In this way it was possible to determine what had been taken from and what added to the original solution. Preliminary observations showed that all the chloride, nitrate and sulfate radicals were left in solution; for that reason these 62 Original Communications : Eighth International [vol. radicals were not considered further in this discussion. Except- ing phosphates all the solutions were destructive in respect to the soil itself; but, in the process of destruction new bases were sub- stituted for the ones removed. The results when potassium chloride solution, 0.35 grams per liter, is percolated through 500 gms. of soil are shown in the fol- lowing table : Table VII No. of Soil i i i> o i Q) O M 9 s H 9a§ 9io 9a§ oa§ 9a§ 2-02 wis M2" 3-5 M|§ III a st3 ill a ¥-« oa5 oa« oa« OBS, oaa oa-3 1 .0152 .0078 .0050 .0030 .0020 .0018 .0609 2 .0193 .0096 .0068 .0064 .0040 .0024 .0861 3 .0205 .0089 .0074 .0061 .0058 .0032 .0928 4 .0250 .0104 .0080 .0078 .0058 .0028 .1057 5 .0280 .0139 .0090 .0080 .0067 .0043 .1305 While lime and small amounts of iron and aluminum became soluble a certain amount of potassium was removed from solu- tion and retained in the soil. The data show that some of the soils remove a large part of the potassium from the first portions of the percolate. As more solution is percolated through the soils smaller quantities of potassium are retained. While these soils contain naturally large amounts of potassium they remove addi- tional amounts from solution. The soils richer in clay and also in potassium retain larger amounts than the lighter soils which are relatively poorer in clay and potassium. It has been shown in Table VII that when potassium chloride is percolated through columns of soil potassium is removed from solution and retained in the soil. The solubility of the soil potas- sium is therefore not increased by such a solution. A number of solutions of different salts were percolated through the soils to determine what effect they might have on the solubility of the soil potassium. For these experiments solutions were made of XV Congress of Applied Chemistry 63 sodium nitrate, sodium chloride, sodium carbonate and acid phosphate. The strength of these solutions was the same as that of the potassium chloride used in Table VII on the basis of the potassium equivalent. The amount of soil was also five hundred grams. In order to determine the effect of calcium carbonate and calcium sulphate these salts were mixed dry with the soil at the rate of 1 gram per 100 grams of soil. Distilled water was perco- lated through these and the percolate examined for potassium. The results follow : Table VIII No. of Soil Parts per million of potassium in first 300 cc. of percolate of dis- tilled water through 500 grams soil Parts per million of potassium in first 300 cc. of percolate of dis- tilled water through 500 gms. soil plus 1% calcium carbonate Parts per million of potassium in first 300 cc. of percolate of dis- tilled water through 500 grams, soil plus 1% calcium sulfate 1 75 92 224 2 287 323 442 3 179 68 383 4 58 72 140 5 136 50 172 The data in Table VII as well as other observations made in this laboratory show that calcium oxide and calcium carbonate do not liberate potassium from these soils. Some observations have indicated a decreased solubility. A limited number of observations with calcium sulfate indicate that small amounts of potassium are made soluble. These determinations have been made colorimetrically and the evidence is not as positive as that obtained in the following tables by the gravimetric method. 64 Original Communications : Eighth International [vol. Table IX. Sodium Nitrate Solution No. of Soil .9-s •So .s-s .9 § .go .S§ O 6 OS O 6 9S| o S Og- Mo^ M82 MoS M§^ M^-3 mO c3 ,n«i ^ a"o S-a ° a':^o a fl « a ja"© aSs 0«a a O^ ft 2|S 2 > o- O d ft 03 1> a Ceo Table XI. Sodium Carbonate Solution .s ° og iCO oj O I cor? fr. r^ tn t^ *rr 1 .0029 .0036 .0015 .0007 .0005 . 0003 .0190 2 .0027 .0036 .0013 .0006 .0007 .0002 .0182 3 .0032 .0023 .0020 .0011 .0006 .0004 .0192 4 .0039 .0037 .0025 .0016 .0014 .0007 .0276 5 .0051 .0046 .0024 .0018 .0012 .0006 .0314 Table X. Sodium Chloride Solution ' !a<» •« o a o .B-z .ss s ° .S§ .9*° O 6 o§ 9§ 9§- o§ o§ OS No. of SoU a"o Ota ft MO a) Mo2 a"o Ota ft a fl s § ^ ft OS'S O a ft ago a > 2 O 0) a Grams K' first 3600 percolate (Approx.) 1 .0021 .0017 .0010 .0004 .0005 .0003 .0120 2 .0017 .0022 .0010 .0005 .0003 .0004 .0122 3 .0024 .0024 .0009 .0008 .0002 .0004 .0142 4 .0025 .0026 .0018 .0006 .0008 .0006 .0178 5 .0042 .0037 .0015 .0017 .0008 .0010 .0258 No. of Soil a- a"o a>« a 6 a o .gs •« o 06 OS Oci 9S| %2 O § 082 «oS M85 MoS M§S M.a mO C3 mO 03 a" o ItjO a""o a s S3 a.ja'o aSs Ota ft S.t5 ti O't^ ft M 2^^ OS'S Oa ft OH .So O o (n CD ffj o a'" o ft s-g g ft Ota ao- 1 .0019 .0015 .0011 .0005 .0004 .0003 .0114 2 .0021 .0013 .0008 .0009 .0003 .0004 .0116 3 .0025 .0020 .0010 .0010 .0004 .0004 .0146 4 .0022 .0019 .0014 .0007 .0005 .0007 .0148 5 .0025 .0024 .0022 . 0018V .0008 .0006 .0206 XV] Congress of Applied Chemistry 65 Table XII. Acid Phosphate Solution No. of Soil a°o .So 98 oil .s-s 9§l a d !3 OS'S d"o III .8 8 gSs if! Co Grams KiO in first 3600 cc. of percolate (Approx.) 1 .0020 .0026 .0017 .0002 .0004 .0003 .0144 2 .0014 .0026 .0019 .0003 .0008 .0005 .0150 3 .0025 .0027 .0021 ; .0005 .0006 .0004 .0176 4 .0024 .0041 .0015 .0003 ,0010 .0007 .0200 5 .0043 .0042 .0018 .0005 .0010 .0009 .0254 Tables IX, X, XI and XII show that dilute solutions of sodium nitrate, sodium chloride, sodium carbonate and acid phosphate in contact with these soils liberate potassium in considerable amounts. The data in Table XII was secured to determine the effect of commercial acid phosphate in liberating potassium. A sample was found which contained no potassium and the soluble phosphate extracted with water. The free acid was neutralized with lime, the solution filtered and standardized. As already stated these observations could not be obtained from field work because of the relatively small amounts of fer- tilizers generally used. There is, however, no reason to believe that these same reactions do not take place under field conditions. The solutions used in these experiments while dilute are much more concentrated than soil solutions excepting under special conditions. Nitrate of soda and acid phosphate solutions are very active in their effect on potassium. This has a very practical bearing in relation to crops grown in connection with nitrate and acid phosphate fertilizers. In the field observations have been made to determine the relation between potassium fertilizers and the yield of the crop. Potassium salts have been used alone and also in connection with nitrogenous and phosphate constituents. For this work soils of uniform character have been selected and plots laid out and treated in various ways to show the specific effect of the potassium salts. The plots have been placed on different types of soil to 66 Original Communications : Eighth International [vol. afford an opportunity to show how different soils respond to the same treatment. The fertihzers have consisted of nitrate of soda, acid phosphate, potassium sulphate and potassium chloride and have been applied singly and in combinations as a top dressing on grass fields. One or more check plots have received no fer- tilizers. This provided a means for observing the value of potas- sium under all ordinary conditions. In general the effect of a com- bination of fertihzers has not been additive. To show the results on different types of soils the yields on a number of series follow. The results show average yields for a period of two or more years on the same field: The following five tables show the influence of the various fertilizing constituents when applied as a top dress- ing under ordinary field conditions to soils representing the five types already discussed. Whenever used the fertilizers were apphed at the rate of 120 lbs. K2O, 40 lbs. N, and 50 lbs. PaOs per acre. The plots to which these fertilizers were applied had received no fertilizers for more than a year and were covered with a good timothy sod from which, in most cases, at least one crop of grass had been removed. Table XIII. Yield on a Sandy Soil Nitrate of Soda, Nitrate of Soda Nitrate of Soda Muriate of Acid Phosphate Muriate of and Acid and Muriate Potash and and Muriate Check Potash Phosphate of Potash Acid Phosphate of Potash 1.155 1.155 1.575 1.330 1.120 2.170 Table XIV. Yield on a Light Sandy Loam .945 .875 1.715 1.890 1.085 2.625 Table XV. Yield on a Light Clay Loam .931 .805 2.660 2.415 .945 2.975 Table XVI. Yield on a Heavy Clay Loam 1.505 1.330 2.537 3.110 1.910 2.905 Table XVII. Yield on a Heavy Boulder Clay 1.702 1.616 2.079 2.089 1.819 .2140 xv] Congress of Applied Chemistry 67 In general the yields from the different plots on the various types of soils were uniform in a number of respects. There is considerable variation in the yields which come from the plots on the different types of soils but these may be due more to seasonal differences than to productiveness of the soil itself or influence of the fertilizers. The effect of muriate of potash when used alone or in connec- tion with acid phosphate produced no increased yield of hay. A considerable increase in yield was produced when muriate of potash was used with nitrate of soda. This increase, however, is due to the effect of the nitrate of soda. Some increase in produc- tion is obtained when muriate of potash is used in connection with nitrate of soda and acid phosphate. This increase is greatest on the lighter soils. On heavier soils the potassium has very little effect under any conditions. The following data show the average yield obtained by the use of chloride and sulphate of potassium on a series of plots over a period of five years. Table XVIII. Compaeative Yields for the Chloride and Sulphate of Potassium Check Sulphate of Potassium Chloride of Potassium 1.565 1.595 1.485 Potassium was applied at the rate of 90 lbs. K2O per acre. The check plot as shown in Table XVIII has produced more hay than the plot fertilized with potassium chloride. The difference is small but is not an unusual observation. The sulphate of potas- sium has produced practically no increased yield during this period. A large number of analyses have been made to determine whether the yield affects the composition of the crop. Crops pro- ducing both small and large yields were studied to determine whether the amount of potassium removed was directly propor- tional to the weight of the crop. With the almost unlimited and inexhaustible amount of potassium existing in the soil this becomes a very important consideration. Samples of hay have 08 Original Communications: Eighth International [vol. been analyzed from the plots already discussed located on the different types of soils and the amount of K2O removed per acre calculated. These data follow: Table XIX No. Yield per acre Per cent. K2O Lbs. K2O 1 945 1.80 17 2 1875 1.77 32 S 2344 1.67 30 4 2818 1.87 52 5 3088 1.54 48 6 3268 1.85 60 7 3616 1.61 58 8 4504 1.51 68 9 5174 1.86 96 10 5412 1.73 94 11 6010 1.62 97 12 6374 1.81 115 It is evident from Table XIX that the percentage of potassium is practically the same for all samples of hay from the unfer- tilized plots regardless of the yield. While slight variations were found the general tendency has been for the potassium to vary directly with the amount of hay. These results are shown graphically in Fig. 1. In the figure the amount of potassium is shown along the line AB and the yield of hay along the line AC. The line AD repre- sents the relation between the yield of hay and quantity of KtO contained. The range represented in this figure is wide and in the case of the heavier crops the yield is near a maximum. The analysis of samples of hay grown on plots which have received potassium fertilizers does not show that they contain a higher per cent, of potassium than hay grown on unfertilized plots. In eases where the yield has been increased by the addition of nitrate of soda and acid phosphate the composition is again fairly con- xv] Congress of Applied Chemistry stant showing that the soils must have supplied relative amounts of potassium. D B .110 .too T^i. 1 X / £ .80 y / 5 / '51 en / + « +» .60 O X"^ a. «f- X\ o .IfO «/\ T3 J r / 3 .zo / o y Ol. / /" / 1000 1 ICOO 3O0O ifOOO 5000 -J 60OO 1 1 t — ^ Pounds of Hay Figure 1 In the light of the information received from the percolation experiments and from the observations made in field work it appears to be logical for the yield of the nitrate of soda plots to be equal to those which have been treated with nitrate of soda in connection with potassium salts. From the percolation experi- ments it is evident that where nitrate of soda is used, the results are produced by the action of the nitrate of soda itself and also by the action of the nitrate of soda on the potassium in the soil. The effects of nitrate of soda, as observed should therefore be much the same as would be obtained by the use of both nitrate of soda and potassium salts. In this connection the percolation experiments indicate that common salt might in some ways 70 Original Communications : Eighth International [vol. replace potassium salts in fertilizers for some soils. The same might be said of acid phosphate. Field observations show that a large number of our soils con- tain sufficient soluble potassium to produce maximum yields of grass. This appears to be true whether the potassium is in its natural condition or whether it is affected by salts which have a tendency to increase its solubility. From a practical point of view little encouragement can be given to the grass grower to use potassium fertilizers on these soils. The soils furnish sufficient potassium to produce very large crops and hundreds of years would be required to diminish the supply materially. Field observations show that practically all of the potassium in fertilizers produces no beneficial results and that in no case have we found that potassium has produced a profitable increase whether used singly or in combination with other fer- tilizers. Summary It has been pointed out that : 1. A large amount of potassium fertilizer is not used profitably at the present time. 2. New Hampshire soils are rich in potassium and naturally adapted to the production of hay. 3. The soil potassium is present in clay and in mineral form. 4. The soils remove large quantities of potassium from solution under both laboratory and field conditions. 5. When potassium phosphate reacts with the soils no new soluble salts appear in solution. 6. When other potassium salts react with the soil new bases do appear in solution. 7. Excepting phosphoric, the solubility of the common acid radicals is not affected by the action of the soil. 8. The effect of such salts as sodium chloride, sodium nitrate, sodium carbonate and acid phosphate is to greatly increase the solubility of the soil potassium. 9. The reaction between these salts and the soil is chemical. xv] Congress of Applied Chemistry 71 10. Calcium carbonate, calcium sulphate and calcium oxide have practically no effect on the solubility of soil potas- sium. 11. The feldspar minerals have a definite solubility in water. This solubility is affected by lime and the common salts found in fertilizers. The effect of these is modified by the presence of clay. 12. Field observations show that potassium fertilizers do not pro- duce increased yields of grass, particularly on clay soils. In some combinations they are more effective on the sandy soils, but not profitably so. 13. In many cases nitrate of soda alone produces yields as good as are obtained with a combination of nitrate of soda and potassium salts. This may be due to the effect of the nitrate of soda on the soil potassium. 14. The composition of the hay shows that when no potassium fertilizers are used the soil affords plenty of potassium for the growth of the crop. This is true for large yields. 15. From a practical point of view little profit can be expected from the use of potassium fertilizers for the production of hay.