Qgwntazla ma. ELECTROLYSIS OF HUMUS SOLUTIONS. BY J. B. RATHER, ASSISTANT CHEMIsT. The term “humus” is used in this article to signify .that portion of the organic matter of the soil dissolved by ammonia after the removal of lime and magnesia by Washing with dilute hydrochloric acid. Grandeau’s1 work on the ammonia-soluble organic matter of the soil is‘ the basis of the method of the Association of Cfiicial Agricultural Chemists" for the estimation of humus. Grandeau mixed ten grams of soil, freed from lime and magnesia by Washing With acid, With coarse sand and placed them in a small funnel, at the bottom of Which Were- fragments of porcelain. This Was moistened with ammonia and allowed to digest for several hours. The ammonia Was then displacedWith Water and ‘the filtrate evaporated and dried to constant Weight, ignited and Weighed again. The loss on ignition is termed humus and the- residue in the dish, “humus ash.” Grandeau’s method Was used by Hilgard3 With modifications. Hil- gard placed the soil on a paper filter and covered it With a disc of filter paper, then Washed it With dilute hydrochloric acid and then with dis-- tilled Water. The soil Was then extracted With 4 per cent ammonia until the extract Was colorless. i Huston and MCBTldQ4 further modified the Grandeau method and their modification is substantially the official method of the Association of Ctficial Agricultural Chemists. In this method 10 grams of the sam- ple are placed in a gooch crucible and extracted With 1 per cent hydro- b chloric acid to remove the lime, and then Washed With Water to remove the acid. The entire contents of the crucible are Washed into a glass- stoppered cylinder" With 500 c.c. of 4 per cent ammonium hydroxide and allowed to remain With occasional shaking for twenty-four hours. The soil is then allowed to settle twelve hours. The supernatant liquid is filtered, and an aliquot evaporated, dried at 100° C., and Weighed. It is ignited, and ‘weighed again The loss on Weight on ignition is termed humus. Considerable clay is brought into suspension With this method. This clay on ignition loses water, which causes an error in the humus determination. In one soil used by the Writer nearly a third of the sample Was held in suspension after the soil had been allowed to settle a Week. Filtration does not remove the clay and continued settling Will not remove it. Recent Work (5, 6, '7 and 8) has proved this method to he entirely unreliable for certain soils on account of the clay present. SnvderQ proposed a method which differs little“ from that of Huston and McBride. In this method the soil is treated With successive por- tions of acid in a tlask, and Washed With Water in the same Way. It is- then treated With successive portions of ammonia, made up to volume, and an aliquot evaporated, dried and weighed, ignited and Weighed again. As early as 1901” it Was pointed out that the official method W38- unreliable on account otf the clay, which lost itsWater of combination on ignition, Which Was thus calculated as humus. Cameron and Bifeazeale,“ in 1904, used the Pasteur filter to remove the suspended clay, although 4 Texas AGRICULTURAL EXPERIMENT STATIONS they did not propose the method as quantitative. 'l‘his method modified‘ . has given fairly satisfactory results, but is long and tedious. Peter and Averitt" proposed a correction by subtracting from the humus 10 per cent of the so-ealled “humus ash.” They admitted that this correction Was uncertain, but it is certainly better than none. Fraps _ and Hamner“ have shown that the average correction would be more nearly correct, if placed at 8 per cent. Mooers and Hampton7 have introduced a modification of the ofiicial method to remove tl1e clay. “The ammoniacal humus extract, contain- ing clay in suspension, is evaporated to dryness over a steam bath, by which means the clay is flocculated so that during extraction with 4 per cent ammonia it can be retained by anlordinary filter paper. Two evap- crations are necessary’, as a rule, in order to get a clear filtrate, in which the humus is determined as usual.” They compared their method ‘with that of Huston and McBride, that of (Jameron and Breazeale, and that of. Peter and Averitt. They concluded that the Huston and McBride method ‘gave results far too high. The Pameron-Breazeale method gave low results, and a 14 per cent correction was thought better than 10 per cent. tllhe method of Mooers and Hampton is open to the objection that some of the humus might be decomposed by the continued baking necessary to flocculate the clay. The method also requires several days to complete a determination. i ‘Stoddan 1*‘ precipitated the clay with ammonium sulphate and acid- ified the alkaline solution of humus. He filtered the resulting precipi- tate on a gooch, evaporated, dried and iveighed. Fraps and Hamner‘ have shown that this method gives low results, the average recovery being 64L per cent. Buthelot and Andre“ found that one-halt of the carbon of the soil, soluble in dilute alkalis, was not precipitated on the addition of acid- Several comparisons of these methods have been published. Alway’ et al.5 and Leavetts find the htooers and Hampton method satisfactory, but it has been shown by Fraps and Hamner that in many soils of the Southwest the method does not remove all the clay. The Pasteur filter as used by the latter tailed to give satisfactory results. Some of the clay was removed, but a part of the humus failed. to go through the filter. hlooers and Hampton failed to get good results with the Pas- teur filter. In a recent bulletin“ of the ’l‘exas Experiment Station, Fraps and Hamner mentioned an attempt to use electrolysis tor the removal of the suspended clay. It was to ascertain if the electric current could be used for this purpose and to study its effect on the humus that the work here presented was undertaken. i Cushman and Hulobarrllt have shown that feldspars, when slimed with xvater, can be removed from suspension by means of the electric current. It is a Well-known fact that most colloids migrate toward the anode upon electrolysis. Since humus is a colloidal, it is possible that both the clay and the humus would be affected by the current. EFFECT OF ELECTROLYSIS ON THE CLAY. Six soils high in clay and low in humus were selected. The solutions of humus were prepared according to the method of the Association of ELEOTROLYSIS or HUMUS SOLUTIONS 5 Official Agricultural Chemists. However, instead of filtering, the solu- tions were decanted, allowed to settle a week and then decanted again. They were shaken thoroughly before each aliquot was removed. The following four methods were tested on these solutions: (a) Humus and ash were determined in 100 cubic centimeters by evaporation, drying to constant weight and subsequent ignition. (b) Humus and ash were determined in 100 c.c. according to the method of Mlooers and Hampton, the solution being twice evaporated to dryness, and taken up in 4. per cent ammonia, filtered, evaporated, dried, weighed, ignited "and weighed again. (c) Humus and ash were determined by the electrolytic method adopted after a number of preliminary experiments to study the condi-- tions xvhich should prevail in this work. One hundred and thirty-five cubic centimeters were placed in an electrolyticcell and a current of .05 amperes (728 volts) was passed for sixteen hours. The solution was decanted through a filter into a dry flask and 100 c.c. taken for deter- » mination of humus and ash by (lirect evaporation and ignition. The electrolytic cell used was a 200 c.c. cylinder. The electrodes were platinum and were attached to platinum wires sealed in glass tubes. The tubes were supported by means of a perforated tin plate which cov- ered the top of the cyflinder and the electrodes were adjusted by means of sections of rubber tubing fitting the glass tubes above the plate. The- anode was placed at the bottom of the cylinder and the current was regulated by moving the cathode, no resistance box being necessary. After sixteen hours the clay was found in a compact mass around the anode and the solutions were in all cases clear and nearly colorless,‘ showing that the humus was precipitated to a considerable extent, or else oxidized. When the upper electrode was made the anode. the pre- cipitation was not complete. On four of the determinations corrections. were made for volatile and nonvolatile solids found in the ammonia used. (d) The clay in 300 c.c. was flocculated with 0.3 gram and 0.6 gram ammonium chloride and an aliquot subjected to electrolysis as described above, another aliquot being used for the determination of clay by direct evaporation and ignition. ResuZts.-—'l‘he results by these methods as regards ash are shown in Table 1. A very large amount of clay remained in suspension in the case of soil No. 893. ’l"'his amount of clay would lead to a serious error if the official method (a) were used to oletermine humus. A large part of the clay was removed by the method of Mooers and Hampton. In only two cases, however, was it reduced to less than 1 per cent. Electrolysis removed more suspended clay than the Mooers and Hamp- ton method. The results, with one exception (823) are fairly uniform and average less than “half those of the Mooers and Hampton method. Elec- trolysis following the partial removal of clay by ammonium chloride, in the proportion of 1 gram per liter, removed some clay but was not as effective as electrolysis in the absence of this salt. When the clay was first precipitated with 2 grams of ammonium chloride per liter, there was no clay removed by subsequent electrolysis. The amount of humus “ash” left by these t-Wo methods is greater than by electrolysis alone. 1t appears that the current is a more efficient pre- 6 - TEXAS AGRICULTURAL EXPERIMENT STATIONS cipitant for clay than 1 gram per liter of ammonium "chloride, but efficient than 2 grams ammonium chloride per liter. I Cushman and Hubbard“ have shown that feldspars in suspension '_ water are decomposed by the electric current. It is probable that portion of the ash remaining after ignition consists of substances 0th than clay. TABLE NO. 1. y PERCENTAGE or HUMUS ASH IN SOILS, ESTIMATED BY DIFFERENT METHOD Precipatron of clay with ammou-i . A . A ium chloride. t» s ST. a - a ‘ . "g 3; a g2 g g 1 gram per liter 2 grams per Illa ‘J a o fi-cl " z Kind of Soil. g5 g _U~;_§ v.9 _ _ " b 3g s = ‘6 Z5. '5 "‘ 3 53a s v 5 °° s a z» s’ z- z- E. To E 5.2 2 ‘=13 . 2 - 2 o ‘s 3v» e Q a 3 a fi as o =2 2 E .2 .2 :2 A o 2 8 3 m a <1 m (a) (b) (C) (dl (B) (d) 114 Travis gravelly loam . . . . . . . . . . . . . . . . . . . .. 8 .80 0 .41 0. _823 Orangeburg fine sandy loar 28 .20 1.23 0. 982 Cameron c1ay..;.. .... .. 4.59 1.78 0. 993 Orangeburg cla .... .. 2.66 0.41 0. 978 Lufkm clay.“ ....... 6.69 1.12 0. 1203 Houston c ay.. 5 .36 1.98 0. Average .............. .................... .. 9 .38 1.16 O. Results on Humnasx-Jllhe humus determinations made as already de- scribed are presented in Table 2. The results by the Mooers and Hamp- ton method average one-half of those of the A. O. A. C. method. This. is probably due to the Water lost on ignition of the clay. This is in accord with the Work of others 5, 6, '7' and 8), who have shown that ‘ the otficial method is unreliable when the solutions contain much suswi pended clay, Soil 823 contains very little humus (0.86 per cent) but’ gives 5.65 per cent by the A. O. A. C. method, which shows that this? method is misleading for such soils. a .T'he Mooers and Hampton method does not remove the clay com-i pletely. When the results are corrected for water in the clay by thf Peter and Averitt method, they are considerably lower in some eases‘ and average about one-tenth of 1 per cent less, on account of the clay‘ ‘present. .3; The humus left after removal of. the clay by electrolysis was appars ently higher than before. But it was evident to the eye that humus. had been precipitated to some extent or else oxidized, for the solution_ were nearly colorless, in most cases. An examination showed that} nitrates were present after electrolysis, though absent from the originali solutions. Apparently the current had produced nitric nitrogen, Nitrates were determined colorimetricallyf by the phenolsulphuric acid? method, a blank on the original solution being run with each determina-ij tion and the results are given in the "tablié. - d When the nitrogen in nitrates is calculated as ammonium nitrate, and; subtracted from the apparent humus found after electrolysis, the “humus: in solution is found to be less in all cases than the humus by the Mooers"; and Hampton, method. On an average, 0.31.1. per cent humus disap-i peared from. the solution i11 electrolysis, either by precipitation or electrolysis. a ELECTROLYSIS or HUMUS SoLirrIoNs _ '7 Cushnian and Hubbard, as mentioned on a preceding page, have shown that potassium hydroxide is formed when feldspars are decom- posed by the electric current. Hence it is possible that there is some potassium nitrate in the alkaline humus solutions after electrolysis, but most of the nitric nitrogen must be combined With ammonia. TABLE NO. 2. PERCENTAGE OF HUMUS ESTIMATED BY DIFFERENT METHODS. i; a Mooers and is g .52 T's é gE a g’; Hampton (evap- Q- 5 5L3 2 a Q5 ‘fig p g3 ration and solu- ‘$2.4: an: _ gang s; 75E Q i _ v7; tion 2 times.) g3 53.3 g iii w as; g Kind of Soil. T3 é E g,’ ggz, Eu c: 5-35: *5 ‘E Q *5 l5 .9 C! o?) 8 m2 F: ‘a *5 5 _ Cor- B g g 5 515-3 E g E O Direct. rected. m a I z? i‘ g (3% (a) (b) (c) i 114 Travis gravelly loam .......................... .. 1.98 0 .79 0 .75 0 .87 0 .08 0 .46 0 .41 823 Orangeburg fine sandy loam .. 5 .65 0 .98 0 .86 1 .33 0.13 0 .74 0 .59 982 Cameron clay... ..... .. .. 1.59 1.44 1.26 2.16 0.20 1.14 1.02 993 Orangeburg cla 0.61 0.57 0.53 1.55 0.20 1.14 0.41 978 Lufkin clay . . . . . . , . . . . . .. 1.75 1.47 1.36 1.56 0.15 O .86 0 .70 1203 Houston clay... .......................... .. 1.14 1.13 0 .93 1 .87 0 .25 1.43 O .44 Average .................................... 210i 1.06) 0.95 166' 0.17 0.96 0.60 EFFECT OF ELEOTROLYSIS ON THE HUBIUS. The object of this work was to study the efiiect of electrolysis upon the organic matter in solution. Six soils were selected which gave solu- tions higher in humus and lower in clay than the soils used in the work just described. The humus solutions were prepared as previously de- scribed. Humus was determined by the methods following; (a) The humus was determined in 100 c.c. by the method of Mooers and Hampton (evaporation and solution). (b) Humus by electrolysis. Three hundred cubic centimeters were treated with 2 grams per liter of ammonium chloride and allowed to stand over night. The clay was then filtered off. One hundred cubic centimeters-were the-n dilute-d to 200 c.c. with water and the current passed through as described on a preceding page. The precipitated humus was filtered on a gooch crucible, washed with 1 per cent hydro- chloric acid and dried "to constant weight at 100° C. (c) Humus by precipitation with acid. One hundred cubic centi- meters of the solution freed from clay’ as described in (b) were acidified ivith hydrochloric acid, filtered on a gooch, washed with 1 per cent hydrochloric acid, and dried to constant weight at 100° C. Results 0f the W 0111.". -—The results are shown in Table 3. The “ash” remaining in the solution purified by the Mooers and Hampton method was considerable, and the corrected’ humus ranges from 0.05 per cent to 0.17 per cent less than the uncorrected humus. The humus precipitated by acid was less in all cases than the cor- rected humus by the evaporation and solution method and-averages 8 Texas AUrRICULTURAL EXPERIMENT STATIONS about half of it. There does not appear to be any relation between the total humus and the humus precipitated by acid. The humus by electrolysis varies from 0.10 per cent to 0.27 per cent and averages 0.18 per cent. This is about one-fifth of the humus pre- cipitated by acid, and one-tenth of the total humus. In this case also there appears to be no relation between the amount as determined by the different methods. . The current alone precipitated twice as much humus as when ammo- nium chloride was present. (See Table 1.) In the latter case the solu- tions were only slightly lighter and the cathode was blackened by some substance yvhich could only be removed by burning. Ammonium sul- phate, ammonium nitrate and sodium chloride were used to precipitate the clay before electrolysis, but they also interfered with the precipita- tion of the humus by the current. The current was increased but the amount of humus precipitated did not appear to be greater and the solution remained dark colored. It appears that salts interfere with the precipitation of the humus by electrolysis. TABLE NO. 3. PERCENTAGE OF HUMUS PRECIPITATED BY ELECTROLYSIS AND BY HYDRO- CHLORIC ACID AFTER THE REMOVAL OF CLAY BY MEANS OF AMMONIUM CHLORIDE. g Mooers and Hampton Method (evaporation 79: “E g and solution 2 times.) . h 3 41 E, z ‘is I? i :1 "' d Q Kind of Soil. l Corrected g3 ‘g _ for water m m 8 5+5 B _ Ash. Humus. 1O per cent i ‘a’ 3 s I I clay. ,:: g 334 Houston 10am ........................ .. 1.66 1.82 1.65 1.33‘ 0.16 829 HOUStOH loam ................... .. 1.39 2.52 2.38 1.21 0.10 896 Norfolk fine sandy loam. O .84 1 .58 1 .50 0 .87 0 .26 845,Wabash silt 1oam.. 0 .49 2 .16 2.11 1.05 0.17 941lH0l1St0n 10am....- 0.55 1.25 1.19 0.60 0.11 1121 Hagensport loam 1.22 1.32 1.20 0 .32 0.27 Average... .. . 1.03 1.78 1.67 I 0.89 0.18 EFFECT 1N NEUTRAL SOLUTION. One hundred cubic centimeters of humus solution were evaporated in a 150 e.c. Jena beaker to a volume of about 25 e.c. and made up to 100 e.c. with water. The current was then passed as already described. The humus and clay were precipitated almost completely as before. Only a trace of nitrates (0.3 mg. per 100 e.c.) uras found after elec- t-rolysis, showing that the nitric nitrogen produced. is principally from the free ammonia. When it was found that the humus was precipi- tated with the clay, the work on neutral solutions were discontinued after the following experiment had been made. One hundred cubic centimeters were evaporated to e.c. and placed inside of a diffusion shell of parchment. ( C. S. & Schull.) The shell was placed in a beaker of Water, the cathode of the electrolytic apparatus was placed inside the shell and the anode outside. A current of .05 amperes (28 volts) was passed for sixteen-hours. The humus collected on the walls of ‘the cell, and the water outside the shell remained colorless. ELEOTROLYSIS or HUMUS SOLUTIONS 9 AC KNOWLEDGMENH‘. ‘Ems work was done under the supervision 0i Br. Y}. Q. llraps, ior Whose advice and criticism the xvriter Wishes t0 acknowledge his thanks. Thanks are also due Professor J. W. Kidd and Mr. N. C. Ha-mner for valuable silggestions. This article Was presented t0 the faculty 0f the Texas Agricultural and Mechanical College, as part 0f the requirements ifOT the degree 0t Master 0f Science. 4 per cent ammonia as in the method of Mooers and Hampton,_ev AN IMPROVEMENT 1N THE METHOD i ESTIMATING HUMUS IN SOILS. i As shown in the preceding article, the chief difficulty with the method for the estimation of humus is the presence of clay, which l’ water by ignition and thereby increases the apparent quantity of or ‘i’ matter which is present. * [2, It is well known that salts can coagulate clay and throw it ou suspension. Fraps and Hamner used non-volatile salts to precipi the clay, but call attention to the fact that the salt used might be ‘composed or otherwise lost on ignition. It occurred to the writer if a salt could be found that volatilized below 100° C., its use ‘would come this difficulty. Ammonium carbonate decomposes at 85° C., 6 was accordingly tested. ExperimcntaL-Txvelve soils, some low and some high in humus, W. compared by means of the methods described below. The soluti were prepared by the A. O. A. C. method referred to in the previ article. i i; 1. Humus and ‘ash were determined in 100 c.c. prescribed in i oflicial method. _ g 2. One hundred. c.c. were evaporated two times and taken up orated, and completed as usual. 3. Carbon dioxide, obtained by heating sodium bicarbonate 7_ washed with water, was passed through 130 c.c. of the humus in stoppered 200 c.c. cylinders for five minutes. The solutions were- lowed to settle over night and decanted through a filter, 100 c.c. ta and the determination completed as usual. ! 4. Carbon dioxide was passed through 100 c.c., as above descri for 3.5 minutes, the clay allowed to settle over night and the solutia decanted through a filter. The clay was then washed five times l. decantation with successive portions (20 c.c. each) of 4 per cent a monia containing 10 grams ammonium carbonate per liter. 5. Ammonium carbonate (tested purity) at the rate of 5 grams w.’ liter was added to the solutions, the precipitate allowed to settle f night, the clear supernatant liquid decanted through a filter, and-i aliquot taken for analysis as described in the oflicial method. f 6. One hundred c.c. of the clay was precipitated with ammoni ‘ carbonate as in 5, and washed with ammoniacal ammonium carbon solution as in 4. The solutions by the carbonate methods were in it caseslperfectly clear. In all but two cases the Mooers and Hampton so l i tions were not clear. ' ‘ i The ammonium carbonate used was Eimer and Amend “tested purl and was testeil as follows: One gram was placed on a tared Wat, IMPROVEMENT IN Mum-Ion FOR ESTIMATING HUMUs IN S-orLs 11 glass and kept in a steam oven at 100° C, for three hours. The resi- due Was unweighable. , » ' All determinations were made on the same solutions. T'he time of drying of the humus in all cases was three hours, as numerous tests by us and by others in_ this laboratory showed that length of time to be quite sufficient. Results 0f the lV0'r'k.»—'l‘he results are shown in Table 1. The “hu- mus” by the official methodl varied from 0.86 per cent (soil 993) to 6.20 per cent (soil 947). The “ash” by the oflicial method varied from 3.97 per cent (soil 993) to 33.45 per cent (soil 823) and averaged for the twelve soils 10.‘/30,per cent. y The humus by Mooers and Hamptorfs method2 varied from 0.58 to 5.77 per cent with an average of 1.69 per cent, about one-half of the A. O. A. C. average for humus. The “ash” by this method varies from 0.35 per cent (soil 993) to 3.94 per cent (soil 947) and averaged 1.37 per cent, about one-half of the average for ash by method 1. - 3i 12 TEXAS AGRICULTURAL EXPERIMENT STATIONS .............. .. wad awb awé mwd mad mp6 .52..........doHHoPCoo 28o .6» HQH . . . . . . . . . . . . . . I ZHHOHQUQMHOO QQQU MQQ .35 .8. H om. o H5. o H5. o ha. o 2H. o 85 3. o NH. H 5m ................ :83: Q53 2.22am HE £223 so wzow 8 owahoiw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. mo. H 8. m mm. m S» m 8. m Hw. m ..§ox@nH .2 sawwm 5.6a Ho 50b com 3m mH... o mm. o mm. H oo. H Hm. m wHd .............................. fiownsw Eu 535cm momH om. o 5. H mm. o mm. H ma. HH HQH. w ............................ zoumfisw 5&2 QHSESHH H2. w». o B. m HAW. m n». m mo. mH om. o iéxovam E52 .22.“ .3138 53H mom .35 pm. o m? o mm. o mm. o E. m. ow. o .................. ..ou35w 52o 5555650 m8 3. o wo. H Hm. H NH“. H 3. w m». H .............HHown:w >20 cQMQEQO N3 . . . . . . . . . . ‘ . . . . . . . . . . . . . S. o om. H 3. o mm. H 8. o mm. H .......... $8~H5m 5&2 Hésw 2E vHHoEoZ 8w mm. o S. c wH. H HH. H Hw. w o». H ................ .9535. 5&2 >=Q>Sw @355 0mm EN. o mm. H m». o 3.. H om. w mm. H .................... 50355 5&2 sofisofi Haw 2a w». o Hm. o ww. o 9.. mm 8. m .... H655 55H 35$ 2E wmfiwwzfiho Maw HN.O wm.O m®.O M$.®M WM.M . . . . . . . . . . . . . . . . . . . . . . . . . . . tEdOH %ZQ>S.HN 3555B fiaa sm< 525E 23H W585i 29w 555E HHw< 555E s3» 525m 54 £55m w. . 5w Ho HEUH Nq m5 dofiHEHfiH HHWHQEO m m -253 was Eamon 25E menses wHHEQB UHHd use? H25 @6032 WW -28 E5co5E< -58 5550554» @5525 c0280 oHzxoHH. 502.60 .1 M wmomemz ezmmmmmHQ Wm m3. 52mm Q24 355m hQ mwl _' l § Kind of Soil. l l l f3 l l l f, N0 _cor- 10 per lNo _cor- 10 per No cor-l 10 per 3 rection. cent. lrection. cent. rection.‘ cent. l I l i l " 114|Travis gravelly loam ....................... ................. .. 3.35 1.41 l 0.78 ‘ 0.69 l 0.58 l 0.56 323 Orangeburg fine sandy loam SUbSOll .............. 5 .00 1 .65 l 0 .86 ‘ 0 .81 0 .78 0 .76 329 Houston 10am surface ...................................... .. 1.85 1.40 l 1.54 1.47 l 1.33 1.30 330 Laredo gravelly 10am surface ............................ .. 1.70 1.22 l 1.11 l 0.99 0.90 0.87 895 Norfolk fine Sandy loam surface ...................... .. 1.2:; 1.21 l 1.32 l 1.25 l 1.20 1.14 978|Lufk1n clay 1oam....._. .......................................... .. 2.15 1.05 1.29 l 1.09 10.95 l 0.90 982 Cameron clay SUbSOll ...................................... .. 1.75 1.31 l 1.42 1.25 l 1.08 l 1.01 993 Orangeburg clay surface .................................... .. 0.86 0.46 l 0.58 l 0.54 . 0.45 l- 0.41 947 Soil from alfalfa field, North Dakota .............. .. 6 20 4,89 l 5.77 l‘ 5.38 l 5.07 l 4.99 941 Houston 10am surface ........................................ .. 3 14 1.94 l 1.25 l 1.19 l 1.07 l 1.05 1203 Houston clay SllbSOll. .......................................... .. 2 18 1.23 l 1.00 l 0.85 l 0.83 l 0.78 949 Soil from old field, Edgely, N. D ...................... .. 3 81 3.24 l3 .40 l3.15 ; 3.06 l 2.96 Average (12 soils), ................................... .. 2.71 1.75 l 1.59 l 1.55 l 1.44 l 1.39 Average (6 solls highest in humus)........ 3 .01 2.34 2.45 l 2.28 l 2.14 l 2.08 Average (6 SOllS lowest in humus) ....... .. 2.54 1.17 l 0.62 l 0.94 l 0.72 l 0.71 When a correction is applied to the results on the six soils on which all methods were used "for water iI1 the clay lost on ignition, the results by methods 3, 4. 5 and 6 are nearly the same. Washing increased the ash without noticeably increasing the humus, and carbon dioxide has no advantage over ammonium carbonate. 'I‘he washing‘, therefore, ap- pears to be unnecessary, as the use of ammonium carbonate is the easiest method, and reduces the clay to a Ininimum. it was the method finally adopted. In Table we correct the results for humus by subtracting 10 per cent of the ash, as suggested. by Peters and Averitt. The cor- rections reduce the amount of apparent humus obtained by} the official 14 Texas AGRICULTURAL EXPERIMENT STATIONS method enormously in some cases. “Humus” in soil 823 was reduced from 5 per cent to 1.65 per cent; with soil 114. the difference is almost as great, the humus being reduced from per cent to 1.41 per cent. The Mooers and Hampton results Were also reduced considerably by correction, averaging 0.14 per cent less. When. the results by ammonium carbonate Were corrected they were only slightly lower (0.05 and 0.06 per cent) than before. It is doubtful if a correction should properly g be made here because this ash is practically free from clay. lllhe average after correction is 1.75 (A. O. A. (l), and 1.55 (Mooers and Hampton) and 1.39 (ammonium carbonate). The results by the ammonium carbonate method were lower in all cases than by the Mooers and Hampton method and the corrected results average 0.16 per cent less. The corrections applied are purely arbitrary and represent aver- ages. Fraps and Hamner have shown that the amount of water in the clay varies from 8 per cent to 20 per cent. These diiferences are not great and would possibly disappear if the evaporation and solution (Mooers and Hampton method) ivere continued until the clay were, entirely removed. But the evaporations take considerable time and the ammonia is liable to absorb acid. fumes from the laboratory. In addi- A tion, the continued. baking is liable to oxidize ordecompose some of the constituents of the humus. There is possibility of precipitation of some of _the humus, for in clay precipitated by ammonium sulphate, Fraps and Hamner found carbon in varyfing amounts, a portion of which must be in organic com- bination. At the same time the clay by the Mooers and Hampton method contains humus. If enough ammonia is used to get all the humus, most of the clay comes with it, and vice versa. SUMMARY AND CONCLUSIONS. Electrolysis 0f Hit-watts Solutions. 1. Electrolysis removed most of the suspended clay from humus solutions and precipitate-d some of the humus. _ 2. Electrolysis removed more clay than the Mooers and Hampton method. 3. Electrolysis precipitated more clay than 1 gram per liter of am- monium chloride, but less than 2 grams per liter of the salt. 4c. After precipitating the clay with 1 gram per liter of ammonium chloride, the current used did not complete the precipitation of the clay. The presence of the salts interfere with the precipitation. 5. Nitric nitrogen was formed by the current in the presence of. free ammonia. (3. Only a small amount of humus was precipitated by the current in the presence of 1 gram per liter of ammonium chloride, much less than by hydrochloric acid. In the absence of ammonium chloride about a third. of. the humus was precipitated. '7'. All of the humus was not "precipitated by hydrochloric acid. 8. Humus and clay are precipitated from neutral solution by the current. 9. The electrolysis can not be used as a quantitative method for removal of clay or of estimation of humus. IMPROVEMENT IN hlnrnon FOR EsTrMArINo HUMUs IN SoiLs 15 Estimation of Harm/s. 1. Clay in humus solutions may be precipitated by ammonium car- bonate or carbon dioxide, and the precipitant disappears on evaporating and drying the residue. _ 2. Evaporation and solution does not remove the clay completely. 3. Precipitation of the clay with ammonium carbonate is more nearly complete than by evaporation and solution, and is a much shorter method. LITERATURE oifrnn, 1. "Grandeau Compt. Bend, 1872, page 988. 2. Bulletin 107 (revised) Bureau of Chemistry, United States De- partment of Agriculture. ' 3. Bulletin 38, Bureau of Chemistry, United States Department of Agriculture. » _ 4;. Huston and McBride, Journal of the American Chemical Soci- ety, 1894, page 4:1. 5. Alxvay, Files and Pickney, Bulletin 114, Nebraska Experiment Station. g . 6. Fraps and Hamner, Bulletin 129, Texas Experiment Station. i7. Mooers and Hampton, Journal of the American Chemical Soci- ety, 30, 800 (1908). I 8. Leavitt, Journal of Industrial and Engineering Chemistry 2, page- 269 (1910). ' 9. Snyder, Journal of the American Chemical Society 16, 210 (1894). 10. Frear, Bulletin 69, Bureau of Chemistry, United States Depart- ment of Agriculture. 11. Cameron and Breazeale, Journal of the American Chemical S0-- ciety 26, page 29 (1904). 12. Peter and Averitt, Kentucky Experiment Station, Bulletin 126, age 63. p 13. Berthelot and. Andre, Compt. Bend, 116, page666. 14. Cushman and Hubbard, Bulletin ~28, Office of Public Roads, United States Department of Agriculture. 15. Stoddard, Journal of Industrial and Engineering Chemistry 1, page 72.