The Determination of Humus. - ACS Publications

in the official method and suggested a modifica- tion of it. Cameron and Breazeale,3 two years later, determined the humus in six soils, using both th...
0 downloads 0 Views 787KB Size
A L W A Y , F I L E S AND P I N C K N E Y ON DETERMINATION OF H U M U S .

THE DETERMINATION O F HUMUS. By F. J. ALWAY,E. K. FILES,AND R. hl. PINCXI~EY.

Received -4pril 25, 1910.

The term humus is used in this country with two distinct meanings, the one including the whole of the more or less altered organic matter in the soil and the other only that portion of the black or brown material soluble in dilute alkalies. I n the present article, the term is used to s i h i f y only the part of the organic matter dissolved by ammonia after lime and magnesia have been previously removed by treatment with dilute hydrochloric acid. The results obtained by treating the same soil with 4 per cent. ammonia solution differ from IO to 230 per cent., according t o whether the Hilgard method or that of the Association of Official .Agricultural Chemists, which is commonly spoken of as the “official” method, be employed. The most important work on humus has been done by Hilgard, but his data cannot be compared with those obtained by the 4. 0. A. C. method. While but a comparatively small amount of data based upon the use of the latter has been published, it is still commonly accepted both in America and in England’ as a reliable method and is the one in common use. Humus as defined above is the matidre noire of Grandeau and the different methods which have been proposed for its determination are modifications of his original method. * HilgardP3to prevent puddling of the soil, placed it in a filter of paper and then covered it with a disc of filter paper, before the preliminary treatment with dilute hydrochloric acid and distilled water. The soil was then treated with 4 per cent. ammonia water until the filtrate came through colorless. Huston and McBride4 made a critical study of the Grandeau method as used by Hilgard and as a result recommended the method, which soon after, with a few minor modifications, was adopted by the Association of Official Agricultural Chemists. They determined the humus in eight different soils, .four of which were of a peaty nature, both by the Hilgard (called by them the Grandeau) method and by their own. They found the results obtained by the latter to be from 2 0 to 117 per cent. higher than those obtained by the former. Assuming that this increase was due to the more complete extraction of humus by their method, they regarded the method of Hilgard as unreliable. What they had, in fact, done was to incorporate a large amount of clay with the humus. Upon ignition, after drying, this clay lost water, thus increasing the total loss in weight on ignition, which they, like Grandeau and Hilgard, considered to represent the humus. Both the latter, Hall: The Soil. 2nd Edition, p. 165 (1908). 2 Traite’ d’ Analyse des Mafisres agricoles, 3rd Edition, I , p . 150 (1897). 3 Bull. 38, Bur. of Chem., U.S . Dept. of Agri., p. 80 (1893). 4 Bull. 38, Bur. of Chem., p. 84 (1893). 1

0

317

however, obtained ammoniacal filtrates practically free of clay, this being retained by the soil through which the ammonia had to pass. To the entire failure of Huston and McRride to take cognizance of the greatly increased amount of so-called “humus ash” obtained by their method is due their erroneous conclusions as to the reliability of the Hilgard method, About the same time Snyder1 published the description of a modification of the Grandeau method, in many respects similar to that of Huston and McBride. Frear,Z in 1901, pointed out the source of error in the official method and suggested a modification of it. Cameron and B r e a ~ e a l e ,two ~ years later, determined tke humus in six soils, using both the official method and the modification suggested by Frear, viz., the use of a Pasteur-Chamberland filter. They obtained only from one-half to one-twentieth as much humus by the latter as by the former method. They called attention to the fact that it was not certain that some of the humus had not been absorbed and thus removed from the solution and that “humus determinations are tedious, especially if recourse must be had to the filter.” Peter and Averitt,l referring to the unreliability of the official method, suggested that the results so obtained be corrected by deducting one-tenth of the weight of the so-called “humus ash.” Mooers and Hampton‘ introduced another modification of the Huston-McBride method intended to remove the suspended mineral matter. Using five soils, they compared the results obtained by their method with those obtained by the official method, by the Cameron and Breazeale modification, and with the Peter and Averitt corrected results. They concluded that while the official method gave results far too high, the Cameron-Breazeale method gave results from 33 to IOO per cent. too low, and that a cohection of 14 per cent. gave more nearly correct results than that of I O per cent. Notwithstanding the above criticisms of the Huston and McBride method which have appeared, it is still retained as the method of the Association of Official Agricultural Chemists.8 Description of Soils.-In the soils used, we had representatives of the arid IX, XI1 and XIII, of the semi-arid 111, IV, V, VI, VI11 and XI, and of the distinctly humid soils, VI1 and X. The other samples belong to the eastern portion of the transition region of the Great Plains. No. I. -4 composite of surface six-inch samples from a field of upland loess near the Nebraska Experiment Station farm. The field had been in cultivation for about forty years, with heavy applica-

’ J . A . C . S., 16, 210 (1894).

Bull. 69, Bur. of Chem., p. 40 (1902). J . A . C . S . , 26, 29 (1904). Keritucky Bull., 126, 63 (1906). 5 J .A . C. S., 30, 805 (1908). Bull. 107 (Revised), Bur. of Chem.. p. 19 (1908).

3



3'8

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y .

tions of barnyard manure during the last twelve. Normal annual precipitation = 29.36 ins. No. 11. A composite of surface six-inch samples from a virgin prairie near Broken Bow, Nebr. The soil is loess though more sandy than No. I. Normal annual precipitation = 25.35 ins. No. 111. The surface four inches of soil from a virgin prairie near Madrid, in western Nebraska. A sandy loam residual soil. Normal annual precipitation is about 18 ins. No. IV. A composite of many samples taken to a depth of twelve inches from both cultivated and prairie fields near Moose Jaw, Sask., Canada. The soil is a heavy lacustral clay derived from the drift material and deposited in glacial Lake Saskatchewan. Normal annual precipitation is about 15.6 ins. No. V. The surface foot from the virgin prairie near Lethbridge, Alberta, Canada. -4 clay loam, probably similar in origin to No. IV. Normal annual precipitation is about 14 ins. No. VI. Surface six inches from virgin prairie near Solano, northeastern New Mexico. ,4 loam of residual origin. Altitude 6000 feet. Normal annual precipitation is about 16 ins. No. VII. Surface soil from near Cloudcroft in the Sacramento Mountains, New Mexico. A black loam. A dense growth of coniferous trees covers the mountains at the altitude where the sample was secured, zliz., 8,900 feet. Normal annual precipitation = 25.14 ins. No. VIII. Surface foot of soil from Parsons' dry land orchard near Parker, Colorado. Normal annual precipitation is about I 7 ins. No. I X . Surface six ins. from recently broken grass land in Sulphur Spring Valley, Arizona, eighteen miles north of Douglas, which lies on the Mexican boundary. Soil of alluvial origin. Normal annual precipitation = I O . 5 ins.

July, 1910

a residual soil derived from basalt. Normal annual precipitation = 14.2 ins. No. X I I . Surface six inches from long cultivated field east of Fresno, California. A red loam which had never been irrigated. Normal annual precipitation = 9.00 ins. No. X I I I . Surface six inches of sandy alluvium near Delano, California. The field had been under '' dry-farming " about 2 0 years, being non-irrigable, Normal annual precipitation = 6.07 ins. 1433. From the same prairie as 111. 902 to 917. Similar to I, but taken to a depth of 12 inches. Comparison of Results Obtained by Different Methods.-With all the methods the soils were treated similarly previous to the application of ammonia, the lime and magnesia being washed out with I per cent. hydrochloric acid and the acid removed by washing with distilled water. It was found economical of time, as most of the soils were of fine texture, to use an ordinary filter in a funnel without suction, rather than a Gooch crucible with a n asbestos filter. In all determinations a 4 per cent. solution of ammonia was used. Freshly distilled ammonia was used because a n ammonia solution allowed to stand in bottles of ordinary glass soon dissolves enough to seriously affect the amcunt of residue remaining after the ignition-the so-called humus ash. The humus extract, by whichever method obtained, was placed in a platinum dish, evaporated on a steam bath and dried at 110' C. Numerous comparisons showed that the same results were obtained as by drying at 100' C., and the time of drying, instead of being 12 to 24 hours, was reduced to 2 to 4 hours. Table I gives the humus and the humus ash as obtained by the different methods, the percentages given in the tables being the average of two, and in most cases of three determinations :

TABLEI .-COMPARISON OF RESULTSOBTAINED BY DIFFERENT METHODS. Soil No.

I.

Method. Hilgard. 2.22 Huston-McBride.. 3.33 Cameron-Breazeale. 2 .06 Mooers-Hampton. 2.07 10 per cent. correction for ash.. 2.55 14 per cent. correction for ash.. .......... 2.23

............................... ...................... ..................... ...................... ..........

....... ._

Hilgard.. ..................... Huston-McBride.. Cameron-Breazeale. Mooers-Hampton.

...................... ..................... ......................

0.44 7.83 0.16 0.71

11. 1.64 2.45 1.62 1.63 2.07 1.88 0.67 3.80 0.32 0.27

111. IV. v. VI. Humus. Per cent. 1.01 1.61 0.85 1.61 2.33 3.39 2.32 3.62 1.01 1.62 1.60 0.91 1.60 0.92 1.03 1.56 2.62 2.21 2.00 1.24 1.86 2.22 1.74 1.81 Humus ash. Per cent. 0.76 0.84 0.47 0.67 3.32 10.03 11.77 10.64 0.43 0.61 0.56 0.51 0.29 0.42 0.27 0.39

No. X. Surface foot from a bog near Fremont, Nebraska. The soil is so little above the level of the Platte River that i t is saturated at all times of the year. No. XI. A composite of surface six-inch samples from long cultivated fields near Pendleton, Oregon. The soil is commonly known as c'volcanic ash," but is

.

VII.

VIII.

IX.

x.

XI.

XII.

XIII.

7.31 8.34 7.53 7.47 7.51 7.20

0.71 2.70 0.81 0.79 1.40 0.89

0.57 1.85 0.65 0.63 1.15 0.87

9.53 9.93 9.55 9.56 9.41 9.21

0.49

0.90 0.45 0.50 0.68 0.59

0.41 0.70 0.37 0.36 0.28 0.11

0.45 1.10 0.38 0.43 0.87 0.78

2.72 8.32 1.55 0.55

0.35 12.95 0.41 0.12

0.67 7.03 0.44 0.20

1.35 5.17 1.28 0.71

U.29 2.20 0.39 0.13

0.30 4.23 0.43 0.08

0.20 2.30 0.26 0.20

The results for humus obtained by the Hilgard method in no case agree with those obtained by the Huston-McBride method, but in all cases they agree with those obtained by the two methods in which the clay is removed from the humus extract, viz., the Cameron-Breazeale and the Mooers-Hampton. I n case of only Nos. VI1 and X does the humus found

A L W A Y , F I L E S A N D P I N C K N E Y ON D E T E R M I N A T I O N OF H U M U S .

319

by the Huston-McBride method approach that ob- dents obtained turbid solutions he had them make tained by the Hilgard, being less th.an 15 per cent. use of a Pasteur-Chamberland filter before evaporapractice which, as shown below, does not higher, while in all other cases it is from 50 to 280 tion-a per cent. higher. In the case of all except Nos. VI1 introduce any error. Y T SHSE HILCARD and 2;) a correction of I O per cent. of the humus ash TABLEII.-A~GREEIIENT O F DIFFERENTD E T E R ~ ~ I N A T BI O MZTHOD. gives results still much too high. When this correcDifference. Per cent. tion is raised to 14 per cent. the results agree fairly we1 in some cases, but are too high in some and too Soil Nu. I. 11. 111. Av. I Iumus A.;h low in others. When the data from individual de2.19 2.22 2.21 3.3 ... Humus.. , . . 2 . 2 6 I. Ash.. . . 0.61 0.43 0.27 0.44 . . . . 126 terminations are considered, it is evident that in many Humus. . . . . 1 . 6 9 1.60 . . . . 1.64 6.0 ... 11. cases the agreement of the averages is due to low re5 Ash . . . . . 0.68 0.65 . . . . 0 . 6 7 . . . . sults being counterbalanced by high ones. 1.00 1.01 7 .O 1.05 Humus . . _ .. 0 . 9 8 111. 4 0.77 0.76 0.76 A s h . . . . .. . , 0 . 7 4 .... The so-called “humus ash” shows still more variant 1.60 1.61 1. 0 1.62 Humus ... . . 1 . 6 0 I\-. results than does the humus. The amount found 16 0.91 0.84 0.78 Ash . . . .. . . 0 . 8 3 .... by the Huston-McBride method varies from 4 to 38 4 .. v. Humus.. . . . 1 58 1 6 4 . . . . 1 6 1 41 . 0.47 0.55 A s h . ,. . . . . 0 . 3 9 .... times as much as that found by the Hilgard. In . . . . 0.85 4.0 0.86 1-1. Humus . . . . 0 . 8 3 general, the higher the ratio of ash to actual humus, 11 .. 0.67 0.63 Ash ... . . . . . 0 . 7 0 .... the greater is the error in the amount of humus found 7.32 7.33 7.31 0.0 Humus ... . . 7 . 3 0 VII. 21 2.44 2.72 2.77 .... Ash .... , . . . 2 . 9 5 by the Huston-McBride method. In the majority 0.72 0.71 .. 0.73 Humus .... . 0 . 6 7 9.0 VIII. of cases less ash is found by the Cameron-Breazeale 81 0.26 0.35 0.47 Ash.. . . . , . 0 . 3 1 .... method and still less by the Mooers-Hampton than 0.57 0.57 0.59 0.56 5 .o Humus . . . IX. 3 0.77 0.67 0.67 Ash ... . . . . 0 . 5 8 .... by the Hilgard. There is no definite relation between 2.0 9.53 9.53 9.60 Humus. . . . . 9 . 4 6 X. the humus and the ash, the latter varying by IOO 4 Ash ... . . 1 . 3 8 1 . 3 3 ..... 1 . 3 5 . . . . per cent. in the case of duplicate determinations in 0.50 0.49 0.49 2 .o .. Humus.. . . . 0 . 4 9 XI. 10 0.30 0.29 0.30 Ash ... . . . . . 0 . 2 7 .... which the results for humus agree. 0.37 0.41 19.0 0.41 Humus. .. . 0 . 4 4 XII. The Hilgard Method.-The importance of a thorough 0.24 0.30 54 0.30 .... -4sh ... . . . . 0 . 3 7 examination of the method which has been used by 0.47 0.45 9.0 0.43 Humus ... . . 0 . 4 4 XIII. 0.16 0.20 56 0.19 .... Ash ... . . . 0.25 Hilgard is due less t o the probability that it will be used extensively in the future than to the fact that Table I1 shows the agreement of duplicate deterupon it depends the value of so much of his work minations by the Hilgard method, where extreme upon humus. For this reason it is important to con- care had been taken to not only wash out all the humus, sider just how he, himself, has used the method. but also to avoid any of the clay being carried over I n practice he has continued the treatment with the edge of the filter. Further, these determinations ammonia until the washings were colorless or so nearly were made by those who had had much experience so that the undissolved humus would make no ap- with the method and they, accordingly, are more conpreciable difference in the results. This is indicated cordant than the results obtained in the ordinary in his letter to Huston’ in April, 1892. Although it course of analysis. The variation between different is not specifically stated in the various publications determinations of the same soil is indicated in the describing his method,2 it should be self-evident last two columns of the table, in which is shown the that when the soil is being washed with ammonia relation of the difference to the lowest individual in order to dissolve out the matihe noire, the wash- determination. For the humus this varies from o ing should be continued until the ammonia no longer to 19 per cent. with An average of j . 5 , the greatest brings away any “black matter” as indicated by the difference being shown by a soil poor in humus. The color. After one has familiarized himself with the humus ash shows a variation of from 3 to 1 2 6 per method he can judge from the intensity of the color cent., with an average of 33.2. There is no relation of the washings whether enough humus remains to between the agreement of duplicate determinations cause an appreciable difference in the results. That of humus and that of the corresponding duplicates ten days or even longer may be required to complete of ash. the extraction in the case of some soils we have found As the results in the above table are not represento be true, but this does not a t all invalidate the ac- tative of the results of ordinary analyses, the followcuracy of the method. Hilgard, himself, regularly ing two summaries are given, they having been obreported the amount of humus ash. He obtained tained without reference to testing the reliability of the filtrates clear, before transferring them to the the method. In one series of 30 soils with a humus platinum dishes for evaporation. In a private com- content varying from 0 . 5 4 to I . 53 per cent., -the munication, in 1907, he stated that where his stu- maximum difference between either one of the duplicates 1 Bull. 38, Bur. of Chem., p. 84 (1893). and the average of the two was I j per cent. of the 2 I b l d , P. 80;J A C. S..16, 34 (1894), Circular N o . 6, Univ. of Calilower result, while the average difference was 3 . 5 forqia (1903). --_A_--

,

, ,

, ,,

,

, , ,

,

, ,

,

, , ,

,,,

,

,

,

,

T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .

320

July, 1910

per cent. In another series of I8 soils containing cases very slight and in most cases too slight t o be from 2 . 8 0 to 3 . 5 0 per cent. of humus the differ- detected. ences were 8 and 4 per cent., respectively. All the resid;es, except those of Nos. I1 and IX, Whei-e the ammonia solution was applied contin- were, after the above exhaustive extraction, transuously to the prepared soils until the washings were ferred to flasks and treated with 4 per cent. ammonia colorless, the extraction was in some cases incom- solution, etc., by the Mooers-Hampton method, plete, but when the soil near the end of the operation as though they had been fresh soils. This gave any was allowed to stand some time between successive of the particles of humus not previously reached by washings, the extraction could be carried to comple- the ammonia an opportunity to dissolve and also tion. Samples of the 13 soils, in sets of four, five subjected the whole of the remaining organic mator six each, were subjected to the continuous applica- ter to 36 hours' contact with fresh ammonia. The tion of 4 per cent. ammonia solution until the wash- averages of the determinations are shown in the last ings became colorless. Then the receivers were re- two columns of Table 111. Only Nos. VI1 and X placed by others and ammonia added to the residues gave any humus in the extraction. Before the large a t wide intervals, until after standing 24 hours without amount of suspended clay was removed by the evapwashing, an application of ammonia gave a colorless oration and subsequent filtration it was not possible filtrate. Determinations were made of humus and to decide from the color whether any humus had been ash obtained both previous and subsequent to the dissolved by this last treatment. The increase for change of receivers. Even where the second filtrate No. VI1 was only I per cent. and for No. X only 2 obtained by the continued washing with ammonia per cent. To determine what is the effect of filtration through was colorless, it was evaporated in a platinum dish a Chamberland-Pasteur filter upon the Hilgard humus and the latter dried, weighed, ignited and weighed extract, we made use of the humus extracts obtained again. The approximate average lengths of time reby continuously applying ammonia to the soils until quired for the different soils is shown in Table 111. the washings were colorless. Accordingly the results The time required to obtain a colorless filtrate by are not to be compared with those in Table I, but continuous washing varied from 3 to go hours. In with those in the third and fourth columns of Table the case of soil No. X, 19 days' additional treatment 111. In this case equal portions of the extracts was required in order to remove the last trace of from two or three samples of each soil were combined the humus. to form a composite which was passed through a TABLEIII.-SUCCESSIVE EXTRACTIONS OF THE SAME SOILS. filter prepared as follows: A new ChamberlandThird extraction. Pasteur filter, or one that had been burned out in First extraction. Second extraction. Mooers-Hampton Hilgard method. Hilgard method. method. the furnace after it had last been used, was washed .-. Humus. Ash. Humus. Ash. with 4 per cent. ammonia, as described below, until H u q u s . Ash. Per Per Per Time. Per Per Time. Per IOO cc. of the filtered ammonia solution, when evapcent. cent. cent. Hrs. cent. cent. Days. cent. orated in a platinum dish, left no residue. The comI. 10 2.12 0.41 5 0.10 0.03 0.00 0.00 11. 5 1.56 0.65 4 0.08 0.02 ... . . ... posite extract was then passed through this filter 111. 5 1.01 0.76 1 0.00 0.00 0.00 0.00 and three aliquot portions used for the humus deterIV. 8 1.52 0.82 5 0.09 0.02 0.01 0.00 V. 8 1.54 0.46 5 0.07 0.01 0.00 0.00 minations. The results of the three determinations, VI. 5 0.85 0.67 1 0.00 0.00 0.00 0.00 as was to be expected, agreed very closely with one VII. 40 6.71 2.59 7 0.60 0.13 0.09 0.02 VIII. 4 0.64 0.34 5 0..07 0.01 0.00 0.00 another. The averages compared with those of the IX. 3 0.55 0.67 5 0.02 0.00 .... . . . . unfiltered extracts are shown in Table IV. While in X. 90 8.99 1.25 19 6.54 0.12 0.17 0.02 XI. 5 0.49 0.29 3 0.00 0.00 0.00 0.00 two cases there is a distinct loss of ash due to filtra7 - - 7

XII. XIII.

64 4

7

0.34 0.36

0.23 0.15

7 11

7--7

0.07 0.09

0.07 0.05

0.00 0.00

0.01 0.03

In the case of the soils other than Nos. VI1 and X I the additional amounts of humus obtained by the continued extraction were not greater than the differences between duplicate determinations. The individual determinations made a t the change of receivers agreed almost as well as those indicated in Table 11, where all had been carried to exhaustion. The effect of the continued extraction, however, was to lessen the differences, that member of a set of three which showed the smallest amount of humus a t the change in receivers unusually giving the most in the second extraction. The increase in ash was in all

TABLEIV.-EFFECT

OF THE

CHAMBERLAND-PASTEUR FILTERUPON HUXUS SOLUTION.

Before filtration. ___A__-

Soil No. I. 11.

111. IV. V. VI. VII.

VIII. IX. X. XI. XII. XIII.

Humus. Per cent.

Ash. Per cent.

2.12 1.56 1.01 1.52 1.54 0.85 6.71 0.64 0.55 8.99 0.49 0.34 0.36

0.41 0.65 0.76 0.82 0.46 0.67 2.59 0.34 0.67 1.23 0.29 0.23 0.15

.

----

After filtration. Humus. Per cent.

Ash. Per cent.

2.07 1.56 0.99 1.55 1.53 0.83 6.68 0.65 0.55 8.94 0.48 0.32 0.30

0.44 0.68 0.70 0.39 0.49 0.52 1.75 0.34 0.66 1.51 0.31 0.15 0.21

-----

LOSS.

Humus. Per cent.

0.05 0.00 0.02 -0.03 0.91 0.02 0.03 -0.01 0.00 0.05 0.01 0.02 0.06

Ash. Per cent. -0.03 -0.03 0.06 0.43 -0.03 0.15 0.84 0.00

0.01 -0.28 -0.02 0.08 -0.06

A L W A Y , FILES A N D P I N C K N E Y ON DETERMINATION OF H U M U S .



tion, the changes in humus are not greater than the experimental error. Accordingly, a ChamberlandPasteur filter so prepared may be considered to have no effect upon the amount of humus, although it lowers the ash in many cases. The Hzcston-McBride or ‘‘ Official” Method.-The analyses were conducted according to the directions in the “ Official and Provisional Methods of Analysis of the Association of Official Agricultural Chemists,”l except that instead of using a Gooch crucible for the preliminary treatment of the soils with acid and dist h e d water, we used ordinary filters in funnels provided with protecting discs of filter paper, and instead of glass-stoppered cylinders we in most! cases employed glass-stoppered, graduated flasks. The use of Gooch filters had not been found economical of time when working with many samples, and numerous comparisons of the cylinders with flasks had shown that where the soils stood in contact with the ammonia solution for 36 hours the shape of the vessel made no difference in the results. With peaty soils, however, there might be an advantage in using cylinders. The differences between duplicates were greater in the case of the humus and less in that of the ash than those found in the Hilgard method. The maximum difference in humus between duplicates was 73 per cent. of the lower while that in ash was 7 0 per cent., the average difference being 17 and 40 per cent., respectively, against 5 . 5 and 33. o in the Hilgard method. The differences between the amounts of humus obtained by the two methods are great and bear no definite relation to the humus obtained by either method. The success of the attempt to make satisfactory corrections for the data from the Huston-McBride method would be dependent upon the percentage of ash being stated. Even with the ash given, the factor to be used would vary with different soils and to a considerable extent with the same soil. Thus, from Table I it will be seen that for soil No. I11 a correction equal to 40 per cent. of the ash is necessary, while for X one of only 8 per cent. is necessary. As pointed out above, even where 14 per cent. seems satisfactory for the average of two or three determinations, this is often due to the wide variation in ash between the different determinations. Thus, in the case of No. VIII, the average for the three determinations is 0 . 8 9 per cent. of humus, which is not so far from the 0 . 7 1 per cent. found by the Hilgard method. This fairly satisfactory average, however, is obtained from three determinations which gave I 41, 0 . 7 2 and 0 . 5 3 per cent., respectively. As Huston and McBride have stated, soils extracted by the Hilgard method give a n additional loss on ignition (called by them hzcmw) when treated by their method. Ten samples of No. 1433, ex1 BuU.

101, Bur. of Chem.. p. 19 (1908).

321

tracted by the Hilgard method, were subjected to treatment by the Huston-McBride method. The results ranged from 0.37 to 0 . 6 2 per cent. of humus and from I .44 to 2 . O I per cent. of ash. Five other soils were similarly treated. Although the ammonia was not colored by freshly dissolved humus, the large amount of clay held in suspension caused a loss on ignition, after the second extraction, about equal to that after the first, as shown in Table V: TABLE

TREATMENT

O F RESIDUES FROM HILGARD METHOD BY HUSTON-MCBRIDE METHOD. First extraction. Second extraction. Hilgard. Huston-McBride.

____&__

Soil S o . 902 911 915 916 917 1433

-

--

_ A -

Humus. Per cent.

Ash. Per cent.

Humus. Per cent.

Ash. Per cent.

2 .68 1.24 1 .so 1.91 1.88 0.82

0.44 0.26 0.35 0.62 0.41 0.25

1.25 1.83 1.76 2.07 1 .92 0.37

5.77 10.51 10.66 12.06 11.21 1.44

to 0.62

to 2.01

The Cameron-Breazeale 3lethod.--We used the apparatus devised by the U. S. Bureau of Soils for the Chamberland-Pasteur filter, forcing the solution through the filter by means of compressed air. On the pressure cylinders, which were silver-plated, the plating was so thin that it was necessary to coat the inside of the cylinder with paraffine to prevent the solution of copper. The humus extracts were prepared by the official method and a composite sample of the extracts was made by combining equal volumes. The Chamberland-Pasteur filter was washed with distilled water, after which 4 per cent. ammonia solution was forced through until a portion of the filtrate, on evaporation, gave no residue. Then 300 cc. of the composite humus extract was forced through the filter and the filter washed by forcing through 4 per cent. ammonia solution until the washings were colorless. Filtrate and washings were made up to a definite volume and aliquot portions used for the three determinations in the case of each soil. The averages of the three, which were in all cases concordant, are given in Table I. The percentages of humus agree with those found by the Hilgard method, but the ash is in most cases much lower. While this method, when carried out as just described, gives satisfactory results, i t is so much more tedious than either the Hilgard or the Mooers-Hampton method that it is not to be recommended. Unless the above precaution be observed, the results are not to be relied upon. The Mooers-Hamfiton Method.-The percentages of humus obtained by this method agreed within the limits of experimental error with those obtained by the preceding method, but the ash was in most cases much lower. The duplicate determinations of humus agreed well, the percentage difference varying from I to 1 5 per cent. with an average of 5 . 7 . As in the

THE JOURNAL OF INDUSTRIAL AND ENGINEERING CHEMISTRY.

322

July, 1910

case of the Hilgard method, the ash showed a much greater variation, ranging from o to 164 per cent. with an average of 41. The Snyder Method.-As this method appears to have been used by no one except the author of it, and as it offers no particular advantages over the Hilgard and the Mooers-Hampton methods, we gave i t a careful test with only a single sample of soil, one from the loess of eastern Nebraska. The results were as follows:

In all but one case more humus was dissolved by 4 per cent. ammonia solution than by 2 per cent. Practically the same amount was dissolved by 8 per cent. as by 4 per cent.; more by 16 per cent., and in most cases, still more by 28 per cent. It is evident from the table that slight variations in the strength of the ammonia solution above or below 4 per cent. are not of much consequence. The amount of ash varied independently of the strength of the ammonia.

SNYDER METHOD.

I . The Hilgard, Huston-McBride, Cameron-Breazeale and Mooers-Hampton methods for the determination of humus were compared, using arid, semiarid and humid soils. 2 . The Hilgard method, as used by its author, was found to give entirely reliable and satisfactory results. In the case of soils rich in humus and those of very fine texture the method is a t times very tedious. 3. The Huston-McBride or “official” method gives results which are entirely unreliable and which, in the case of most soils, are far too high. The errors incidental to the method are of such a nature that it seems impossible to apply to the results any satisfactory correction formula. 4. The humus extract obtained by the HustonMcBride method, after being passed through a properly prepared porcelain filter, gives results which are concordant with those obtained by the Hilgard method. 5 . The Mooers-Hampton modification of the Huston-McBride method gives results entirely concordant with those of the Hilgard method. For some soils it is preferable to the latter, requiring much less time. 6. Ammonia solutions of ‘different strengths were not found to have the same solvent power for humus. The differences for strengths between 2 and 8 per cent. were slight, but where solutions of from 16 to 28 per cent. were used ,considerably more humus was dissolved. 7 . The amount of so-called “humus ash” found in the case of a soil sample vanes from method to method and bears no relation to the strength of the ammonia used. .There is little or no agreement between duplicate determinations of the (‘humus ash.” From 4 to 38 times as much ash is obtained by the HustonMcBride method as is obtained by the other methods. It is probable that the greater portion of the “humus ash” obtained by any of the methods is not an essential part of the humus. 8. In reporting the humus the per cent. of “humus ash” should always be reported, as it serves as an indication of the reliability of the humus determination.

Humus. Ash

2.

1.

.................

2.44 1.56

.....................

2.38 1.58

Average. 2.41 1.57

MOOERS-HAMPTON METHOD. 1.

2.

Humus. ................. 2 . 2 1

Ash.. . . . . . . . . . . . . . . . . . . .

2.34 0.33

0.23

Average. 2.27 0.28

This indicates that the method can be used to obtain results comparable with those of the other reliable methods. It is, however, likely in unpractised hands to give too high results. The Influence of the Strength 0) the Ammonia Solution.-Huston and McBridel reported experiments made to determine the influence of the strength of the ammonia solution upon the amount of humus dissolved. They used the Hilgard method, as well as their own, with ammonia solutions of 2 , 4, 7 . 3 and 8 per cent. Their results, in general, indicated that the stronger the ammonia the more humus was dissolved. This was not an unvarying rule, however. . We have tried ammonia solutions of the strengths of 2 , 4, 8, 16 and 28 per cent. in the Mooers-Hampton method. The six soils used were similar to Nos. I to’ IX, but on account of our supply of Nos. I1 to I X being exhausted, similar, instead of identical soils were used. Two or three determinations were made in the case of each soil. The averages are shown in Table VI. Soils Nos. XIV,*XV, XVI, XVII and XVIII are similar to Nos. 11, VII, VIII, VI and 111, respectively, having been taken from the same localities and to the same depths: TABLEVI.-COMPARISON

OF RESULTSUSING AMMONIASOLUTIONS OF DIFFERENTSTRENGTHS.

Soil N o . I. XIV. X V . XVI. Strength of ammonia solution. Humus: Per 2 per cent . . . . . . . 2 . 0 1 1 . 7 2 6 . 5 3 0 . 5 3 4 p e r c e n t. . . . . . . 2 . 1 2 1 . 8 2 6 . 3 4 0 . 6 0 8percent 2.16 1.85 6.28 0.62 16 per cent.. . . . . . 2 . 3 4 2 . 0 6 6 . 6 2 0 . 6 7

.......

28percent . . . . . . .

2.34

2.14

6.63

0.71

Humus ash: 2percent . . . . . . . 4 per cent.. 8 per cent 16percent 2 8 p e r c e n t .......

..... ....... .......

1 L O C . Cif.

0.36 0.41 0.33 0.38 0.30

0.19 0.26 0.23 0.34 0.33

1.72 0.23 1.30 0.21 0 . 8 9 0.08 0.87 0.18 1.15 0.16

X V I I . XVIII.Average.

cent. 0.94 1.16 1.12 1.17 1.23

1.26 1.39 1.40 1.64 1.17

2.16 2.24 2.24 2.42 2.47

0.38 0.43 0.24 0.16 0.21

0.51 0.47 0.33 0.36 0.41

Per cent. 0.19 0.20 0.18 0.21 0.32

SUMMARY.

LABORATORY O F AGRICULTURAL CHEMISTRY, UNIVERSITYOF NEBRASKA. LINCOLN, NEBRASKA.