138
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
growth of the succeeding crop. There is thus presented a problem in crop rotation. If the influence of a certain kind of plant on the formation of nitrates, after its removal, holds for soils, other than the one used in this experiment, something has been learned. That other, and under some conditions. influences more potent than nitrate formation obtain is also indicated by this experiment. SUMMARY
Plats of land planted t o certain crops in 1 9 1 0 were kept bare of vegetation during the early part of the growing season of 1911. Determinations of nitrates, in the soil of these plats showed a distinct and characteristic relation of the several plants to the nitrate content of the soil in the year following that in which the plants were grown. Maize was the only crop following which the nitrates in the previously planted soil were higher than in the unplanted soil. Potato soil was the next highest in nitrates, and oat soil contained least nitrates. Millet planted on these plats July 1st was markedly influenced b y previous crops, but the luxuriance of growth was inversely proportional t o the nitrate content of the plats. The beneficial influence of a crop on a succeeding one was not, in this case, t o be attributed to the favorable influence exerted on nitrification, but this would doubtless differ with different soils, and if this effect of certain plants on nitrification in the following year should be of general application, the influence of a crop on nitrification may be a n important factor in crop rotation. DEPARTMENT OF SOILTECHNOLOGY CORNELL UNIVERSITY ITHACA, NEWYORK
A COMPARISON OF SOME QUALITATIVE AND QUANTITATIVE METHODS FOR CARBONATES IN SOILS By E. W. GAITAER Received Aug. 28, 1912
I n connection with certain chemical investigations of soils, which are being conducted in the laboratory of the Chemical Department of this Station, it was found necessary t o determine the carbon dioxide existing as carbonates in the soils. After making a number of determinations by the official method,I it was found t h a t there was no correlation between the reaction of the soils with litmus paper or the Veitch qualitative test and the CO, obtained by this method. If i t were assumed t h a t the CO, was derived from carbonates and it were calculated t o CaCO,, the average amount for acid surface soils would be 3 2 4 4 pounds CaCO, per z,ooo,ooo pounds of soil and that for acid subsoils would be 1036 pounds per 2 , 0 0 0 , 0 0 0 pounds of soil, which would be a sufficient amount of this material present t o render any of these soils alkaline or t o indicate that there was a sufficient amount of this compound in all of these soils for all practical purposes. I n many cases, where the soil reddened blue litmus, more CO, was obtained than from those which turned red litmus blue. These f
Bull. 107, Bureau of Chemistry, U. S. I). A.
Feb., 1 9 1 3
variations led to the conclusion that this method was unreliable as a measure of soil carbonates. While considering this point, attention was directed by Mr. J. W. Ames, Chief of the Department, t o a method devised by the late F. S. Marr, Carnegie Research Scholar, Rothamsted, England, and published in the Journal of Agricultural Science, Vol. 111, Part 2 , pp. 155-160, for determining CO, as carbonates in soils, in which the soil was boiled with very dilute acid under reduced pressure. It is claimed t h a t under these conditions no organic matter is decomposed and t h a t the CO, obtained is derived from carbonates only. Some of Marr's work was repeated and his results were confirmed. He found that CO, was evolved from acid soils even when boiled with distilled water a t I O O O C. and that the amount was increased upon the addition of a mineral acid, while if boiled a t 50' C. under reduced pressure with 2 cc. of conc. HC1 per I O O cc. of water, the results were either negative or only a very small amount of CO, was evolved. These results were considered to be within the limits of experimental error. The double titration method of Brown and Escomb with the Amos apparatus' was used by Marr, while a slight modification of the same method and a n apparatus designed b y the writer2 was used in this work.3 I n checking Marr's work, four soils were chosen, one acid, the other three giving from slight t o strong alkaline reactions to litmus and progressive amounts of CO, with weak acid under reduced pressure. Comparative results under different treatments follow. TABLEI-PER CENT. OF coz BY DIFFERENT METHODS O F TREATMENT Per cent. Per cent. COz when Per cent. CO1 when boiled at COS when boiled at 50' C. for 100' C. 30 minPer cent. boiled at with 20 utes under COz when 50' C. unboiled at der recc. of 1-1 reduced loODC. duced presHC1 and pressure Reaction sure with 80 cc. of with diswith 2 cc. of soil to Lab. tilled distilled HaO for HC1 per litmus] No. water water 30 mins. 100 cc. paper 2 X hours 2 % hours 4715 0.025 0.00 0.190 0.00 Acid 0.00 0.205 0.01 S1. alk. 4722 0.010 4724 0.015 0.00 0.175 0.03 Alk. 4731 0.265 0.04 4.700 4.62 St. alk. ABBREVIATIONS USEDIN THE PRECEDING AND FOLLOWING TABLES. A. = acid, SI. = slightly, S t . = strongly, Alk. = alkaline. Mi. = Miami, V. = Volusia, D . Dunkirk, Wab. = Wabash and B. or Blk. = Black. Gr. = Gravel, Sd. = Sand.Si. = Silt, L. = Loam, and C. = Clay. V. 0. W. = Virgin open woodland, V. P. Virgin pasture, M. = Meadow, P. = Pasture. V. F. R. = Virgin fence row, V. M. G. Virgin "Maple grove," R. = Road side, C = Cultivated.
.
-
f
-
From these results,it is seen that the soil which reddened blue litmus paper gave no CO, when boiled under reduced pressure either with distilled water or with dilute HC1, and that none of the other soils gave CO, with distilled water under reduced pressuret except 4731, which contained a very large amoun, of carbonates, while'the acid soil gave more CO, when 1 Jour. 2
Agv. Sci., 1905, 1-322.
IS JOURNAL. Aug., 1912.
3 The apparatus permits the use of a srnalIer amount of 4 per cent. NaOH for absorption of Con and makes it convenient t o titrate all of the solution rather than an aliquot, thus reducing the analytical error.
I
Feb. , p 913 ~
T H E J O U R N A L OF I N D U S T R I A L AhTD E N G I N E E R I N G C H E M I S T R Y .
m T REDDEN BLUELITMUS PAPER. (SEE no. I) P e r cent. COz P e r cent. COz by Per cent. Litmus test Humus by Marr method official method' Condition2 Lab. No Veitch test 1 Acid 1.so Acid 4688 0 0.0242 C. 0 0.47 2 5081 0.0308 C. S1. Alk. S1. Acid P. 0 1.53 3 4687 0.0308 V. SI. Acid C. Acid 0 0.91 4 4709 0.0341 Acid C. 0 Acid 1.62 5 4677 0.0352 0 C. 1.18 6 468 1 0.0440 C. 4680 S1. Acid 0 7 0.0460 1.24 4711 0 V . F . R. 8 0.0462 1.47 S1. Alk. 0 C. 4734 9 0.0462 1.01 M. 465 1 0.0484 10 Acid 0 Acid 1.56 M. 0 11 4652 0.0484 1.79 0 C. 12 5668 0.0484 3.03 S1. Alk. S1. Acid 0 C. 13 4673 0.0484 1.20 0 C. Acid Acid 14 47 19 0.0506 1.61 0 V. M. G. S1. Acid 15 0.0506 1.52 4650 0 C. Neutral 0,0550 S1. Alk. 16 4727 1.55 Acid 0 C. Acid 17 4655 0.0572 3.54 0 C. 18 0.0572 4663 1.61 0 C. 4690 19 1.13 0.0572 Sl. Alk. 0.0594 0 C. 4640 20 1.88 C. 21 1.50 0.0594 S1. Acid 0 4671 0 C. Acid 22 4684 1.57 0.0594 Acid 0 .... C. 23 4642 0.0616 0 0.0616 C. 24 4643 1.93 0 C. 25 4661 0.0616 S1. Acid 0 2.04 C. 26 4704 0.0660 0 C. 1.21 27 4699 0.0680 Acid 0.0022 v. 0. w. 4660 0.0682 51. Acid 1.35 28 0 C. 2.40 4639 29 0.0682 Acid 0 M. 4662 30 0.0682 2.12 C. 4641 31 0.0682 0 1.74 0 C. 5279 32 0.0704 1.25 S1. Acid 0 C. 1 .50 4675 33 0.0704 C. 0 2.31 4695 34 0.0715 4665 51. Alk. Acid 0 C. 1.34 35 0.0726 0 S1. Acid M. 1.79 4732 36 0.0748 M. 0 4647 37 0.0770 Acid Acid 1.38 C. 0 2.09 4654 38 0.0770 C. 0 1.93 4658 39 0.0770 Alk. 0.0792 Acid 0 M. 1.62 4679 40 4733 41 0.0814 S1. Alk. 51. Acid 0 C. 2.49 0 V. F. R. 2.17 4674 42 0.0814 Acid 4715 0 v. 0. w. 2.02 43 0.0836 44 0 V. 0. w. 1.28 4736 0.0902 45 0 C. 4670 0.0902 Acid 1.82 0 46 C. 1.57 0.0946 4648 0 V . 0 . w. 47 4703 0.0957 V. S1. Acid 2.63 0 M. 48 Acid 2.02 4644 0.0990 0 v. 0. w. 4657 0.1012 S1. Alk. 49 2.40 0.0132 R. V. F. R. 50 4721 0.1034 S1. Acid Acid 1.44 0 51 4683 0.1078 v. 0. w. Acid 1.61 v. P. 52 5283 0.1100 0 2.21 0 53 4685 3.44 0.1100 C. 0 54 V. 0. w. 4646 1.71 0.1166 Sl, Acid 0 v. 0. w. 55 4706 3.39 0.1188 v. 0. w. 56 47 18 0.0022 Acid 3.24 0.1320 57 0 4737 3.15 0.1518 Alk. C. Neutral Average 0.0714 per cent. COz = 3244 pounds CaCOa per acre, 7 inches b y official method.
I39
TABLE 11. PART 14URFACE SOILS 0-6 INCH=
Type1 Mi. Sd Mi. Si. L. Waverly Clay Mi. Sd. V. Si. L. Mi. C. L. Sand Mi. C. L. D. Gr. Sd. L. D. Gravel D. Sd. L. V. Si. L. Mi. C. L. D. Sd. Gr. L. Sandy L. D. clay Mi. C. L. Mi. Gr. L. Mi. clay Wab. L. Mi. C. L.
....
TABLB11, PART 2-sUBSOILS
Lab. No. 4686 47 10 4689 4745 4720 4700 5280 4645 4664 469 1 4672 4649 4682 4676 4659 4666 Average, 0.0228 per
Veitch test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 cent. COZ =
Acid
6-36 INCHESTHAT REDDENBLUE LITMUS PAPER.
Subsoil P e r cent. COz Per cent. COz b y of surface Litmus test by Marr method official method Soil No. Acid S1. Acid Acid S1. Acid Acid
0.0022 0.0088 0.0088 S1. Alk. 0.0110 Acid 0.0132 0.0132 0.0198 0.0220 0.0220 0.0220 0.0242 0.0264 0.0264 Neut. 0.0319 Acid 0.0330 0.0792 1036 pounds CaCOa per acre, 7 inches b y official method. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
53 4 1 67 14 27 32 48 18 19 21 46 6 33 39 35
D. Sandy L. Blk. Sd. L. D. clay D. Sd. L. Mi. C. L. Sd. L. Mi. C. L. D. Si. L. V. Si. L. Clyde clay Mi. St. L. Mi. C. L. V. Si. L. D. Gr. D. Sd. L. v. s1. L. Blk. Sd. L. V. Si. L. D. clay Clay L. D. Gr. L. D. Gr. Sd. L. Mi. Si. L. V. Si. L. D. clay
M. C. L. Waverly clay V. Si. L. Clyde clay Wabash loam Clyde clay
(SEE FIG. 11)
Condition
C. C. C.
C. C. C. C. M. C. C. C. C. C. C. C. C.
Type Waverly clay Sand Sand Silt loam Clay D. clay D. Si. L. V. Si. L. Clay loam Mi. Gr. L. Wab. loam D. Gr. Sd. L. Clay loam V. Si. L. D. Sd. L. Mi. Si. L.
140
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
TABLE11, P A R T 3-SURFACE SOILS 0-6 INCHESTHAT TURNLITMUS BLUE. (SEE FIG. 111) Per cent. COz b y Per cent. COz by Per cent. Lab. No. Veitch test Litmus test Marr method official method Difference humus Condition 4708 58 Acid S1. Alk. 0.0011 0.1122 0.1111 2.88 v. 0. w. 4667 59 Alk. 0.0022 0.0726 0.0704 1.19 V. 0. w. 4747 60 S1. Alk. V. S1. Alk. 0.0033 0.0440 0.0407 0.97 C. 4722 Alk. Alk. 0.0044 61 0.0902 0.0858. 1.72 C. 4740 62 Neut. 0.0044 0.0704 0.0660 1.27 River T. 4746 S1. Alk. Alk. 0.0066 63 0.0462 0.0396 1.43 V. F. R. 4725 S1. Alk. 0.0088 64 0.0968 0.0088 1 .80 C. 4701 65 Alk. 0.0110 0.1034 0.0924 3.35 C. 4724 Alk. S1. Alk. 0.0132 66 0.0770 0.0638 1.79 v. s. Y. 4744 Alk. 0.0132 0.0660 67 0.0528 1.31 C. 5085 68 0.0143 0.0528 0.0385 ..., C. 4739 69 S1. Alk. 0.0281 0.1496 0.1215 3.49 V. F.R. 4730 70 Alk. 0.0484 0.1232 0.0748 1.93 V.B. G. . 4741 71 0.0506 0.2168 0.1662 6.48 Alluvial 475 1 72 0.0506 0.0814 0.0308 0.98 V. 4712 73 0.0539 0.1210 0.0671 1.96 C. 4749 S1. Alk. 0.0682 0.1210 74 0 .OS28 1.13 C. 5083 75 Alk. 0.0792 0.1540 0.0748 1.34 v. 0. w. 5281 76 0.1782 0.0836 1 .50 0.0946 C. 5087 77 0.2530 0.3630 0.1100 3.26 Blk. Swale 4698 78 S1. Alk. Neut. 0.5632 0.6809 0.1177 1.88 V. F. R. 47 14 Alk. Alk. 0.5720 79 0.6600 0.0880 1.23 V. F.R. 4743 80 1.5220 1.8830 0.3610 6.56 V. Alluv. Average difference, 0.0912 per cent. COz = 4144 pounds CaC03 per acre, 7 inches b y official method. 6-36 INCHES THAT TURNLITMUSP A P E R BLUE. (SEE FIO. Iv) Per cent. COz by P e r cent. COz by Subsoil of surface ' Litmus teat Marr method official method Difference Soil No. Condition Neut. 0 0.0572 0,0572 64 C. SL Alk. 0 0.0132 0.0132 16 C. 0 0.0176 0.0176 2 C. V. S1. Alk. 0 0.0286 0,0286 68 C. Neut. 0.0022 0.0253 0,0231 , 26 C. Alk. 0.0066 0.0462 0.0396 61 C. S1. Alk. 0.0088 0.1012 0.0924 57 C. Alk. 0.0088 0.0396 0.0308 17 C. S1. Alk. 0.0088 0.0242 0.0154 65 C. 0.0088 0.0506 0.0418 55 v. 0. w. Neut. 0.0110 0.0682 0.0572 C. Alk. 0.0770 0.1331 0.0562 34 C. 0.3652 0.4048 0.0396 71 Alluv. 0.5522 0.6050 0 .OS28 60 C. 2.0680 2.1010 0.0330 70 V. B. G. 2.0900 2.1010 0.0110 73 C. C. 51. Alk. 2.1030 2.1650 0.0620 74 Alk. 7.9090 8.0960 0.1870 77 Blk. Swale 2167 pounds CaCO3 per acre, 7 inches by official method.
Feb., 1913
Type Clyde clay Mi. Si. L. Silt loam Sand Clay Silt L. Sd. C. L. Sand Sand Silt loam Clay loam Clay Clay loam Wabash loam D. clay Clay loam Silt loam Silt loam
silt Clyde clay Clyde clay D. clay Wabash loam
TABLE 11, P A R T 4-sUBSOIW Lab. No. Veitch test 4726 1 S1. Alk. 4728 2 5082 3 Alk. 5088 4 4705 5 S1. Alk. 4723 6 4738 7 Alk. 4656 8 S1. Alk. 4702 9 10 4707 4717 11 Alk. 4697 12 4742 13 S1. Alk. 4748 14 Alk. 4731 15 4713 16 4750 17 5086 18 Average, 0.0477 per cent. COz
..
-
boiled at r o o o with distilled water than either 4722 or 4724, and that this was considerably increased for all soils on boiling a t r o o o with HCI. Sample No. 4731, which contained a very large amount of CO,, gave nearly the same with acid under both conditions, viz., 0.08 per cent. higher for the strong acid at r o o o C., which difference is not far from the average amount obtained from acid soils b y boiling at 100' with HC1, as is seen from Table 11. Repeated experiments with other soils gave the same variations as are above shown. Only soils containing large quantities of carbonates gave any CO, when boiled with water alone under reduced pressure. After this preliminary work was done i t was decided t o make the determination of CO, by both methods and compare these results with the litmus paper and Veitch qualitative tests. For this work 114 samples of soil were used, representing eighteen. different types, varying from light sands t o heavy clays and from 0 . 4 7 per cent. t o 6.56 per cent. of humus by the modified Grandeau method; 8 0 were surface soils, 2 I of which were virgin soils from timber lands,, fence rows or pastures; 34 were subsoils below the cultivated surface soils.
Type Sd. C. L. Sd. L. Mi. Si. L. Clay L. Blk. Sd. L. Sand Clyde clay D. Clay Sand Clyde clay Wabash L. Mi. Blk. C. L. Wabash L. Silt L. Clay L. Clay L. Silt L. Clyde clay
The litmus paper test was made by moistening the soil with boiled distilled water and making it into a stiff paste, a strip of blue and a strip of red azolitmin1 paper were placed about a n inch apart in a petri dish, covered with a piece of filter paper and the wet soil spread over the filter paper, other strips of red and blue litmus paper were folded and inserted into openings in the soils which were pressed against the folded paper so that when the paper was unfolded i t presented a clean surface that had not been in direct contact with the soil. A third set of strips were placed on top of the soil and pressed firmly against it so the contact would be perfect; the cover was then placed on the dish and allowed t o remain over night; the comparison was made with a moistened strip of. neutral litmus paper. None of the paper was handled with the fingers. The results were recorded in order of the degree of color change produced in the paper. There was very little difference in degree of color produced by the same soil for any of the methods of applying the paper, and it seemed t h a t one method was as good as the other so long as the soil remained covered and the con1
The best grade of azolitmin paper was used.
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
Feb., 1913
tact was allowed t o continue for a sufficient time. The Veitch qualitative test was made b y shaking IO grams of soil with I O O cc. of cold distilled water, free from CO,, and allowing t o settle over night or until sufficiently clear t o pipette off 5 0 cc., t o which was added a few drops of phenolphthalein; i t was then boiled t o about 15 cc. in a 150 cc. Jena beaker and the degree of color recorded, a s very slightly alkaline,
p
1
Surface Soils /Lo? R e d d e n B l u e L i f m u s F'aper
I
t h a t for all of these soils there is a wide variation between the two sets of results. I n many cases acid soils give more CO, b y the official method than alkaline soils, especially is this true of surface soils high in organic matter. I n the case of subsoils there is a somewhat closer agreement. For soils containing large amounts of carbonates the differences are not very great for such quantities, but are of about the same magnitude as those obtained from soils giving no CO, by the M a n method. TABLE111-LABOUTORYNOS.OF SOILSIN WHICH VARIATIONS OCCURRED BETWEEN THE ~UALITATNE T ~ T AND S COZ BY MARR METHOD'
C-6 0-6 6-36 C-6 6-36 0-6 C-6 0-6 C-6 6-36 6-36 14-24 C-6 0-6 0-6
qf d
7.9
70
20
15
33.
30
J L o S~a m p l e s .
YO
f.?
$0
5~-
a Y
6
S
N
/Z
IY
16
h& dSQ?IZf
/CS
FIG. I1
[
............ 4726 4727 4728 4732 4733 4734 4737 4745 5082 5088 4640 4657
467 1 4673 4687
.... ....
4728
4718 4721
4718 4721
........ ........ ............
4726 4727
................ . . . . . . . . . . . . . . . . ................ .... 4737 .... ............
5082 5088
............ ............
................ ................
................
. . . . . . . . 4660 4660 . . . . . . . . . . . . . . . . ............
. . . . . . . . . . . .
............
....
4708
0.0022 0.0132 0 0 0
0 0 0 0 0 0 0
0 0 0 0.0022 0
0 0 0
0.0011.
For other data on these samples see Table 11.
slightly alkaline and alkaline. If no pink color appeared, i t was recorded as acid. All these tests were checked b y two analysts. These reactions were tabulated and compared with the CO, results obtained by the two methods as found in Table 11: From this table it is seen t h a t there is a very marked agreement between the reacticn of these soils and the CO, obtained by the Marr method, while there is little or no agreement with the official method.
2
0-6 0-6 0-6 0-6 0-6
1
FIG.
............
0-6
A /LA
A
0
141
A comparison of the CO, b y the two methods shows
Table 111 shows the variations which occurred between the litmus and Veitch tests and the CO, from these soils b y the Marr method. Samples 4718-4721 and 4660 were the only soils which were acid t o litmus paper t h a t gave any CO, a t all (these results were checked and four determinations gave the same amount of CO,). Sample 4718 is a virgin timbered soil overlying a n alkaline subsoil containing CaCO,; 4721 is a virgin soil from a fence row on a roadside and may have been contaminated with CaCO, in the road dust; 4660 was a virgin timbered soil; 4708, which was acid b y the Veitch test and alkaline t o litmus, was a virgin timbered soil, no subsoil being taken. With the exception of these four virgin soils, it is observed t h a t all of the soils, which reddened blue litmus paper, failed t o givc any CO, a t all. Only three soils t h a t react alkaline t o litmus paper fail t o give CO, and these are all subsoils which are only slightly alkaline. This alkalinity, as pointed out by Cameron,I may be due t o the presence of certain minerals which give alkaline solutions by hydrolysis, independent of the presence of carbonates. Of the 17 soils t h a t were alkaline b y the Veitch test and gave no CO,, six were alkaline or neutral t o litmus paper, the remaining eleven reddened blue lit1
Bureau of Soils, Bull. SO.
142
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
mus. Some of these gave no alkaline reaction to phenolphthalein when the soil itself was boiled with water and the clear extract tested. By inspecting Figs. I and 11, in which the percentage of CO, by the official method is arranged in ascending series, i t is readily seen t h a t there is no regularity in the variations of the amount of CO, obtained by the two methods. By eliminating the three samples, 4718-4721 and 4660 (to which attention has been directed), the curve representing results obtained by the Marr method becomes a straight line for all soils t h a t redden blue litmus paper, whether surface or subsoils. For alkaline soils the results by the Marr method
Feb., 1913
T H E M E T H O D F O R CO, I N S O I L S
The changes made in the methods used by Marr are in the amount of 4 per cent. NaOH used, the titration of all of the solution instead of a n aliquot, and the use of a different apparatus. Repeated tests using c . P. carbonates and samples of soil justified these changes. The method a s used in this laboratory is as follows: P u t from z t o zo grams of soil (depending upon the amount of carbonates present) into any suitable flask or bottle of about 2 5 0 cc. capacity t h a t will stand a vacuum of 7 0 cm. of mercury, add 80 cc. of C0,-free distilled water and mix thoroughly, connect t o distilling apparatus1 and start vacuum pump. When
-I Y
Z
+
L
#
&&*Z
. f P 3 Y J -
FIG.IV
FIG.111
are arranged in ascending series. Figs. I11 and IV show that there is still no regularity in the variations of the two methods until the amount of CO, approaches 0.3 per cent., from which point the differences are not so noticeable, but are still of about the same magnitude as for soils giving no CO, a t all. As a rule, the differences between the two methods are greatest for virgin soils or soils containing large amounts of organic matter. While no pot or field work has been done on these soils, observations made a t the time of sampling and the agricultyral history of the sections from which these samples were taken, point to the fact that there is a close agreement between their carbonate content a n d the productivity of the soils, especially as indicated b y the clover yields.
the air has been removed from the apparatus and a vacuum of 65 t o 7 0 cm. is obtained, run in 2 0 cc. of a solution of HC1 ( z cc. HC1, 1 . 1 9 sp. gr. t o 18 cc. of water) through separatory funnel. Boil for 30 minutes; the CO, evolved is absorbed in NaOH made from Na (zj cc. 4 per cent. NaOH and sufficient C0,-free water t o cover the broken glass rod in the absorbing tower). The bottom of the distilling flask or bottle should be about 3 / , inches above the gauze protecting i t from the free flame. If the liquid is thrown into the condensing tube the flame should be lowered. Relieve vacuum and draw out the absorbing liquid from the tower into a 400 cc. beaker or Erlenmeyer flask, wash out tower with 2 5 0 cc. of C0,-free distilled water, add about I cc. of phenolphthalein and run in normal HC1 until red color begins t o fade; finish this titration with N / z o HC1. When this end point is reached, add 2 drops of methyl orange solution ( I THISJOURNAL, 4, 611-612
Feb., 1 9 1 3
T H E JOUR!Y.4L OF I.\-DU.STRIAL
gram per 1000) and fill burette with ? \ i / 2 0 HC1. Take no account of A; or N / 2 0 acid used in the previous titration. Run in N / 2 0 HC1 until the lemon color of the alkaline methyl orange just darkens to slight orange,' take reading of N / z o acid and subtract blank, which should be about 4 cc. I cc. iV/20 HC1 = 0 . 0 0 2 2 gram CO,, 0 . 0 0 2 8 gram CaO or 0 . 0 0 5 gram CaCO,. CONCLUSIONS
which must of necessity attach t o an essential and very elusive element of plant food. While some plants may prefer t o take up nitrogen as ammonia there can be but little question as yet t h a t most plants prefer the nitrate, and we have further evidence that in the absence of such nitrate material the enforced absorption of ammonia by plants is apt to result in certain manifestations of disease due to the disturbance of their physiological mechanism. This, therefore, together with the fact which Hilgard has pointed out,' that uvthurnified organic matter, in his definition of the term, will not nitrify, would appear to invest the humus, determination in soils by some method similar to that of Grandeaul with primary importance in the chemical analysis of soils. Equal importance will, of course, attach to the nitrogen determination in humus since from what has been said above it is that, and not the total soil nitrogen, which may be depended upon t o yield through nitrification an available supply of nitrogen to plants. The general acceptance of the views above given has led t o the prosecution of extensive studies looking toward the perfection of the method for determining humus in soils and the reader is referred for a discussion thereof t o the recent work of Always and his associates and t o that of Kelley and M ~ G e o r g e . ~But while much work has been done on the method of determining humus, but little or no systematic work has been done on the determination of nitrogen in the humus since it has always been assumed t h a t the latest modifications of the Kjeldahl or Gunning method were as reliable for determining nitrogen in the humus as they were for determining that element in other substances. To obtain a more reliable and uniform method for this determination we have carried out a series of investigations in which some of the best known methods for the determination of nitrogen in humus were compared and the results of which we describe below.
The boiling of soils with mineral acids a t 100' C. (as pointed out b y Cameron, Marr and others) decomposes organic matter and liberates CO, which is not present as carbonates and therefore the official method is of no value as a measure of carbonates in soils. That if soils are boiled under reduced pressure a t 50' C. with very dilute mineral acids no organic matter is broken up and the CO, evolved is all derived from carbonates. That the litmus paper test, if properly made, is the best qualitative test we have for the presence or absence of native carbonates in soils from humid regions. That some soils may give alkaline aqueous solutions, not due t o the presence of carbonates but t o the hydrolysis of certain minerals which exist in these soils. That, although the reddening of blue litmus paper b y soils may be due t o the absorption of the base from the hydrolyzed litmus salt, the presence of native carljonates in a soil may either prevent this selective absorption or cause a n interchange of bases t o take place. That if a native carbonate, capable of being decomposed by weak HC1 a t low temperatures, is present in a soil it is indicathd by the bluing of red litmus paper when brought in contact with the moist soil. That if there are no alkalies or basic materials, in a soil, which are capable of giving alkaline solutions, the absence of such substances will be indicated b y the reddening of blue litmus paper. That if these conclusions are correct, as is strongly indicated by the data already presented, there exists HUMUS E M P L O Y E D I N THE EXPERIMENTS in soils which redden blue litmus paper a condition I n order to experiment with widely different kinds favorable to the formation of acids or acid salts which would unite with the base absorbed from hydrolyzed of humus we extracted soils of low, good, and very litmus and fail t o return another base in its stead, high humus content in accordance with the method thus producing the reddening of the indicator, even outlined in Circ. 6 , of the California Agricultural though it yields no H ions to a water solution. I t is Experiment Station. The soils employed for this therefore possible that such a soil may be capable purpose were as follows: of producing a physiological effect similar to that proN o . 1. Light sandy soil from walnut orchard, Anaheim, California. duced by stronger acids which do yield H ions t o Humus content 0 . 5 5 per cent. No. 2. Silty clay loam derived from the State of Washington. Humus aqueous solutions. content 8 .89 per cent. LABORATORY OF
THE
CHElllCAL
DEPARTMEW OF THE OHIO AGR.EXPT.STI.
No. 3. Tule soil from island in the Sacramento River. organic matter. Humus content 2 8 . 7 per cent. EXPERIMENTS
A CONTRIBUTION TO OUR METHODS OF DETERMINING NITROGEN I N HUMUS B Y CHARLESB . LIPMAZAND H . F. PRESSEY Received Sept. 30, 1912
So far as investigations on the subject have gone t o date, nitrogen as nitrates still retains the importance It will be well for those not familiar with this titration t o practice on solutions containing a little Na2C03. It will be necessary for each analyst t o establish and adhere strictly t o a constant end point for both indicators.
OS
THE
DETERMINATIOX
Searly all
O F NITROGEN
I N V A R I O U S H U M U S SOLUTIONS
Portions of fifty cubic centimeters of humus extract were boiled with one gram of magnesium oxide in 5 0 0 cc. Jena Kjeldahl digestion flasks until no more ammonia was given off as indicated by a test with 1 2
3 4
"Soils," Jfacmillan 8; Co., 1910. p. 359. Circ. 6, California -4gr. Expt. Station. Bzrll. 115, Nebraska rlgr. Expt. Station. B d l . 33, Hawaii Agr. Expt. Station.
