The moisture equivalents of soils

MISSING IMAGE

Material Information

Title:
The moisture equivalents of soils
Series Title:
United States. Bureau of Soils. Bulletin
Physical Description:
23 p. : pl., tables, diagr. ; 23 cm.
Language:
English
Creator:
Briggs, Lyman J ( Lyman James ), 1874-1963
McLane, John Wallace ( jt. author )
Publisher:
Govt. print. off.
Place of Publication:
Washington
Publication Date:

Subjects

Subjects / Keywords:
Soil moisture   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Additional Physical Form:
Also available in electronic format.
Statement of Responsibility:
By Lyman J. Briggs and John W. McLane.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029606089
oclc - 29750378
lccn - agr07002172
System ID:
AA00025979:00001

Full Text

1 41A lastred. B eptember'9 1907.
ARV'ENT OF AGRICULTURE,
B ~~A#. OF SGIS--:-BULLETIN No. 45.
-MILTON W]KITNEY, Chief.






-T-UlE-'EQUIVALENTS

'OF' S,-0Ls
IV










U.S. DEPOSITORY
..g" BY,

,-4lA .L -BRIGGS ANI JOHNq W., Mc'LANE.
t 400



r rJ







WASHNTON:




oovlnxxawr PRITIMMG OMCEO.
1907v
144
-xi

l,;










































iiAi Ncil W.,Dousii, in charge of Alkali hand Reclamat
JAYA. BoNSErnL, in charge of Soil Survey.












i i i i i i i i i = i iii
4 NQ~il Icn :




I rseirll of the Uniled States CovernmenL
Issued September 9, 19007.
U. S. DEPARTMENT OF AGRICULTURE,
BUREAU OF SOILS-BULLETIN No. 45.
MILTON WHITNEY, Chief.






THE MOISTURE EQUIVALENTS

OF SOILS.





BY

LYMAN J. BRIGGS AND JOHN W. McLANE.















WASHINGTON:
GOVERNMENT PRINTING OFFICE.
1907.






















LETTER OF TRANSMITTAL.


U. S. DEPARTMENT OF AGRICULTURE,
BUREAU OF SOILS,
Washington, D. C., May 8, 1907.
SIR: I respectfully transmit herewith a technical report, entitled
"The Moisture Equivalents of Soils," prepared by Messrs. Lyman J.
Briggs and John W. McLane, formerly of the Physical Laboratory
of this Bureau. This paper covers work done during their connection
with this office, and I recommend that it be published as Bulletin
No. 45 of the Bureau of Soils.
Respectfully, MILTON 'WHITNEY,
Chief of Bureau.
Hlon. JAMES WILISON,
Secretary of Agriculture.
2














CONTENTS.

Page.
Introduction -------------------------------------------- -------- 5
Moisture equivalents ----------------- 5
Description of centrifugal machine used for determining moisture equiv-
alents ----------------------------- 6
Effect of duration of test on the moisture equivalent----------------- 8
Effect of initial water content on the moisture equivalent------------ 9
Effect of speed on the moisture equivalent --------------------- 10
Accuracy of the determinations of the moisture equivalent ------------- 13
Moisture equivalents of typical soils ---------- 14
Relation of mechanical composition to the moisture equivalent---------- 17
Norfolk and Portsmouth series .-------- _------------------ 19
Marshall series ------------------------------ 20
Moisture equivalent coefficients of Cecil, Hagerstown, Miami, and Ver-
non series ------- ----------------- ------- ------------ 21
Summary --------------------- -------------------------------22





ILLUSTRAT 0 N S.


PLATE.
Page.
PLATE I. Fig. 1.-Centrifugal machine used for determining moisture
equivalents. Fig. 2.-Head of centrifugal machine, showing
cylindrical cups with perforated bottoms for holding the
moist samples of soil ----------_ ----------------- 16

TEXT FIGURE.

FIG. 1. Relation of the moisture equivalent to the force employed -------_ 12
3



















Digitized by the Internet Archive
in 2013













http://archive.org/details/eequivale00brig














THE MOISTURE EQUIVALENTS OF SOILS.


INTRODUCTION.

It is important in the comparative study of soils that the mechan-
ical composition, which forms the basis of most comparisons of this
kind, shall be supplemented by the quantitative measurement of other
characteristics, such as the moisture retentivity and the rate of
capillary movement of water under standard conditions.
The measurement of the water-holding capacity under the action of
gravity does not furnish a satisfactory method for comparing soils
in a quantitative way, as such determinations are dependent upon the
percentage of interstitial space in the soil, which in its turn depends
upon the way in which the soil has been manipulated. It occurred
to the authors that a satisfactory basis of comparison might, however,
be established by determining the amount of water which different
soils are capable of retaining when the moisture in the soil is sub-
jected to a constant measured force, sufficient in maIgnitude to remo'e
the water held in the larger capillary spaces. This may be easily
done by placing the moist soils in suitable perforated cups and sub-
jecting the soil moisture to centrifugal force. The magnitude of the
force employed can be readily calculated by determining the radius
of rotation and the speed.

MOISTURE EQUIVALENTS.
The percentage of water retained by a soil, when the moisture con-
tent is reduced by means of a constant centrifugal force until it is
brought into a state of capillary equilibrium with the applied force,
we propose to designate by the term moisture equivalent. In a series
of samples in which the moist soils are all in equilibrium with the
same force, it logically follows that if these soils are placed in contact,
each soil containing an amount of water equal to its moisture equiva-
lent and the packing of each soil being the same as in the centrifugal
machine, the soils will be in capillary equilibrium and no mo\vement
of water from one soil to another will take place. This conclusion
has been verified experimentally. The packing to which each soil
5





6 THE MOISTURE EQUIVALENTS OF SOILS.

is subjected in the centrifugal machine seems as nearly definite and
reproducible as it is possible to obtain, since each element of mass at
the same distance from the axis is subjected to the same centrifugal
force. When a layer of soil. of uniform thickness is employed, the
conditions are practically the same for each soil examined. Conse-
quently the determination of the moisture equivalents of a series of
soils shows -the amount of moisture each will retain when all are
packed in a uniform manner and are subjected to the same centrifugal
force.

DESCRIPTION OF CENTRIFUGAL MACHINE USED FOR DETERMINING
MOISTURE EQUIVALENTS.

In designing apparatus for the determination of the moisture
equivalents of soils the first requisite is a shaft with bearings capable
of standing high shaft velocities without undue heating. To save
expense various commercial machines were considered with a view
to their adaptation to this purpose, and a grinding machine was
finally secured as most nearly fulfilling the desired condition. The
bearings did not, however, prove satisfactory, and it was only after
special bronze bearings had been substituted and the shaft had been
carefully ground that a speed of 5,000 revolutions per minute could
be maintained without unsafe heating. The steam turbine, with the
centrifugal apparatus mounted on, or directly connected to, the tur-
bine shaft, is the ideal arrangement for high-speed machines of this
class.
The machine as finally used is shown in Plate I, fig. 1, with the
centrifugal head in position. The main shaft passes through the
cylindrical head, which is secured in position by being clamped
between two heavy flanges by means of a nut threaded on the shaft.
The other end of the shaft was designed to take a centrifugal head of
another form.a In the illustration, however, this end of the shaft
simply carries a small worm which engages a gear wheel having 100
teeth. A small spur on the side of the gear wheel momentarily
closes a circuit once each revolution, actuating a sounder in the
engine room, and in this way the speed of the machine could be deter-
mined and kept nearly uniform by regulating the speed of the engine.
The construction of the centrifugal head is shown in Plate I, fig. 2.
a Such a machine is capable of serving a double purpose, since with another
forin of centrifugal leand it can he used as a means of extracting and collecting
from i moist soil a portlio of its soil solution. The composition and concentra-
tion of the solution thus removed would appear to be identical with that in the
soil from which the plant derives its mineral food, and the method seems a
promising one for investigating the relation of the composition and concentra-
tion of the water-soluble constituents to the productivity of the soil.





DESCRIPTION OF CENTRIFUGAL MACHINE. 7

The cylinder, which is of brass, is 24.7 cm. outside diameter, and the
thickness of the cylindrical wall is 1.3 cm. Care was taken to secure
a uniform, homogeneous casting, and both the cylinder and cover
were. accurately finished inside and out, which gave a perfectly bal-
anced system. The cylinder carries in its interior eight brass
cylindrical cups with fine gauze bottoms covered with thin disks of
filter paper, in which the soils under investigation are placed. Each
of these cups rests upon the flat surface of a cylindrical segment
fitted to the inside of the centrifugal head. These segments are
equally spaced about the interior of the cylinder and are held in
position by screws. To keep the cups in position when the machine
is at rest a brass diaphragm .having apertures for each cup is fitted
into the cylinder, as shown in the illustration. The cover is pro-
vided with a slightly tapered flange which fits snugly over the open
end of the centrifugal head. The moisture removed from the sam-
ples works out through the joint between the head and the cover.
Since the whole outer surface of the cylinder is accurately turned
and smoothly finished, the air friction is reduced to a minimum.
While it is possible to examine eight samples of soil at one time
with the machine as described, this procedure is apt to throw the
machine out of balance, for the samples will in general lose different
amounts of water during the test. Since duplicate determinations
were desirable, four samples were run at one time. The cups con-
taining the same sample were brought to equal weight and placed on
diametrically opposite sides of the cylinder, which kept the machine
in balance throughout the experiment.
This machine was normally driven at 5,000 revolutions per minute
by means of a steam engine running at 300 revolutions. A belt from
the fly wheel of the engine drove a shaft, to which the centrifugal
machine was belted, at 2,000 revolutions per minute. The centrif-
ugal machine and the pillow blocks carrying the end of the slhaft
opposite the machine were mounted on the same cement foundation,
which rested upon a bed of sand and was free from the floor and
walls of the building. This effectually prevented the machine from
producing any serious vibration or jarring of the il ilding.
It is an essential condition in the successful operation of all high-
speed centrifugal machinery that the rotating system shall be as
nearly perfectly Ialanced as possible. Machines wlhich are not ill
perfect balance vibrate hadly when the period of rotation corresponld
to the natural period of vibration of the machine. This speed is
sometimes spoken of by imanufacturers of centrifugal a)pparatus as
the critical speed," above or below which the machine operates
much more quietly.





8 THE MOISTURE EQUIVALENTS OF SOILS.

The centrifugal force f,a expressed in dynes, acting on an element
of water, dim, is
f=4 nT n 2 r. dm
where n is the number of revolutions of the apparatus per second and
r is the distance of the element of water from the axis. Since the
bottoms of diametrically opposite cups, when in position in the cylin-
der, were 21.5 cm. apart, we would have for a layer of soil 5 mm. in
thickness, a mean radius of rotation of 10.5 cm. The centrifugal
acceleration, i. e., the force acting on unit mass of water, correspond-
ing to each of the different velocities employed in the following exper-
iments, is given below (Table I). We can also express the centrifugal
accelerations in terms of the acceleration of gravity as a unit, since
the, acceleration of gravity is equal to 980 dynes. To avoid confusion,
the calculations in the table are given in terms of unit volume (1
cubic centimeter) of soil moisture, which is assumed to have unit
mass. The second column gives the force in dynes, and the third
column gives the force in terms of the force of gravity as a unit.

TABLE I.-Centrifugal force acting on runit mass of soil moisture at the various
speeds employcd.

Revolu- Force, Force,
tionsper dynes per grams per
minute. c. c. c. c.

2,700 F39 > 10 857
3,000 1,036 1,057
3,200 1,179 1,203
4,100 1,936 1,975
4,200 2,031 2,073
4,300 2,129 2,174
5, 000 2,878 2,937
5,500 3,483 3,554


EFFECT OF DURATION OF TEST ON TIE MOISTURE EQUIVALENT.

In order to determine the time required to bring the moisture con-
tent of the samples into capillary equilibrium with the centrifugal
force applied, a series of preliminary experiments (varying in dura-
tion from 15 to 60 minutes) was made upon several soil types. The
results are given in Table II. The first column of the table gives the

a Sonie English writers object to the term centrifugal force." The term
has, however, so thoroughly established itself in this country and is so generally
unlderstood that it appears advisab)le to use it here. There is, of course, no
actual force exerted in a centrifugal machine which tends to pull the water
out of the soil. What actually occurs in the mlachine is just the opposite; that
is to say, the soil is pulled away from the water. There is an acceleration of
the soil directed inwardn toward the axis, which leaves the water behind. The
result is, of course, the same as if the water were subject to an equal accelera-
tion outward.





EFFECT OF INITIAL WATER CONTENT. 9

type of soil used, the second column gives the percentage of moisture
remaining in the soil at the end of the test, and the third column gives
the time during which the moisture in the sample was subjected to
centrifugal force. It will be noted that there is a slight reduction in
the moisture equivalent as the time increases in nearly every case,
although this effect is small after a period of thirty minutes.
These determinations were made upon a thicker layer of soil than
was employed in the final experiments, so a somewhat longer time
was necessary to attain approximate equilibrium. Thirty minutes
was accordingly decided upon for the duration of the tests in the final
experiments. In some extremely heavy soils the moisture content
might be slightly reduced by. extending this period, since the distri-
bution of moisture in a sample would probably reach a state of com-
plete equilibrium only after a considerable interval had elapsed.
Errors due to other causes, however, such as the fluctuation in the
-speed of the machine, did not appear to us to justify an attempt for
further accuracy in this direction.

TABLE II.-Effect of duration of test upon moisture equivalent.

Soil typ Moisture Duration Speed per Soil t Moisture Duration Speed per
Soi ype after test. of test. minute. type. aftertest.1 of test. minute.

Per cent. Minutes. Revs. Per cent. Minutes. Revs.
Leonardtown loam. 14.9 15 4,800 9248 Hagerstown
14.6 15 4,800 loam............. 17.3 30 5,000
13.7 30 4,800 17.2 45 5,000
13.1 30 4,800 16.2 45 5,000
13.3 45 4,800 9037 Hagerstown
13.4 45 4,800 clay............. 22.2 30 5,000
13.0 t0 4,800 21.9 30 5,000
12.4 60 4,800 21.7 45 5,000
10212 Hagerstown 21.4 45 5,000
silt loam......... 16.0 30 5,000 8589 Norfolk fine
15.7 30 5,000 sand............. 3.1 30 5,000
15.5 45 5,000 3.4 30 5.000
15.2 45 5,000 3.1 45 5,000
9248 Hagerstown 3.3 45 5.000
loam............. 17.2 30 5,000


EFFECT OF INITIAL WATER CONTENT ON THE MOISTURE EQUIVALENT.

A second series of experiments was made to determine whether the
percentage of water used in moistening a soil had any influence
upon the moisture equivalent, providing, of course, that the amount
of water used was always greater than the soil could retain at the
speed employed. It was thought that perhaps something akin to
puddling might take place when relatively large amounts of water
were used. Experiments were male with several differenit moistiure
contents in the case of each soil examined. One sample of each soil
was saturated and another was puddled as much as possible by
thorough mixing so as to accentuate the influence of the preliminary
treatment of the soil upon the moisture equivalent. It will be noted
from a comparison of the results, which are given in Table III,
490--No. 4."--07 1M- 2





10 THE MOISTURE EQUIVALENTS OF SOILS.

that the initial moisture content in the case of the Leonardtown
loam had very little influence upon the per cent of moisture remain-
ing after the test. The same is true of the subsoil of the Norfolk
sand. In the case of the subsoil of the Cecil clay, which is a heavy
soil, the results show a gradual increase in the value of the moisture
equivalent as the initial water content is increased, and the effect of
puddling in this soil is marked, the percentage of water retained by
the soil being increased from 32 to 39 per cent by the puddling
process.
Therefore, in preparing the samples for a determination of the
moisture equivalents, care was taken not to saturate the samples,
and to stir the soil only enough to insure the distribution of the
water throughout the sample.

TABLE III.-Effect of initial water content on the moisture equivalent of soils.

Approxi- Moisture Approxi- Moisture -
Soil t mate initial content il t mate initial content
Soil type. moisture after Soil type. moisture after
content. test. content. test.

Per cent. Per cent. Per cent. Per cent.
Leonardtown loam (good).. 15 12.0 Norfolk sand (subsoil)..... 10 7.4
15 12.1 10 7.3
20 11.5 15 7.0
20 10.9 15 7.1
25 11.7 20 7.0
25 11.8 20 7.0
Puddled. 14.0 Puddled. 7.4
Puddled. 11.7 Puddled. 7.3
15 12.4 Cecil clay (subsoil)........ S5 .2.9
15 12.6 Z5 c3.1
20 11.4 O) 24.1
20 11.3 _0 34.3
25 11.8 43 25.0
25 11.7 43 24.8
Puddled. 11.9 Puddled. S9.2
Puddled. 12.1 Puddled. 39.0


EFFECT OF SPEED ON THE MOISTURE EQUIVALENT.

A third series of measurements was made to determine for different
soils the relation between the centrifugal force employed and the
amount of moisture retained by the soils. If any definite and simple
relation exists, then it would be possible to make the moisture equiva-
lent determinations at any suitable speed and reduce the results to
some standard condition, such as, for example, the retentivity of the
soil when subjected to a force of 1,000 times the force of gravity.
Such a relation would also enable us to calculate the amount of
moisture which would become available when the pulling force is
increased by a known amount.
The results of the experiments are given in Tables IV and V. In
Table IV the different speeds employed are given at the heads of the
columns, with duplicate determinations of the corresponding moisture
equivalents below. The same arrangement exists in Table V, except
that the means of the duplicate determinations are given. It will






EFFECT OF SPEED ON THE MOISTURE EQUIVALENT. 11

be seen in Table IV that doubling the speed, which quadruples the
centrifugal force, reduces the moisture content of the Dunesand only
from 3.0 to 2.6 per cent. In the case of this sand the capillary spaces
are relatively large in size and few in number, so that it seems prob-
able that the lowest speed employed is sufficient to reduce the mois-
ture content mainly to the form of a thin film on the surface of the
soil grains, so that the effect of quadrupling the force is small. In
all the other soils there is a well-defined and progressive decrease in
the moisture content as the speed increases. The results of Table IV
indicate that the Leonardtown loam, good, and the Leonardtown
loam, poor, are distinctly different in their relation to moisture."
Sample No. 4983, in Table V,-is seen also to differ from the other soils
of the series, giving up its moisture much more readily with increas-
ing speeds.
TABLE IV.-Effect of speed on moisture equivalent.

Revolutions per minute.
Soil type.
2,700. 3,000. 3,200. 4,100. 4,200. 5.500.

P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct.
New Mexico dunesand .... 3.0 3.0 2.9 2.7 2.9 3.0 2.9 2.7 2.6 2.6 2.7 2.6
Sassafras loam, good..... 18.5 18.7 17.8 16.1 15.5 15.0 15.0 15.0 14.1 14.0 12.4 12.7
Leonardtown loam, good.. 18. 1 18. 0 16. 5 16. 5 15.0 15. 4 13. 4 13. 8 14. 1 14. 3 12. 1 12. 1
Leonardtown loam, poor.. 12.0 12.0 10.1 10.5 9.4 9.9 8.0 7.5 7.4 7.4 6.3 6.3

TABLE V.-Effect of speed upon the moisture equivalent.

Speed-revolutions per
minute.
No. Type. minute.
4,300. 5,000. 5.500.
Per cf. Per ct. Per ct.
7233 Hagerstown stony loam ........................................ 16.9 15.0 14.6
6534 Hagerstown sandy loam ....................................... 12.2 11.0 9.9
4952 Hagerstown loam............................................. 21.8 20.8 18.4
9248 Hagerstown loam.............................................. 18.3 17. 5 16. 5
10212 Hagerstown silt loam.......................................... 17.9 16. 5 15. 8
4983 Hagerstown shale loam......................................... 31.6 25.0 17.3
4962 Hagerstown clay loam.......................................... 24.1 23.4 21.5
4966 Hagerstown clay loam......................................... 26. 0 25.5 24.8


A simpler way of comparing these results is to be found in the
graphical arrangement shown in figure 1, for which we are indebted
to Mr. Buckingham, of this laboratory. In this diagram the mois-
ture equivalents are plotted as ordinates, while the abscissas are pro-
portional to the reciprocal of the centrifugal force. If all of the
moisture could be thrown out of the soil by an infinite centrifugal
force, the moisture equivalent would be zero when the centrifugal
force is infinite, i. e., when the reciprocal of the centrifugal force
(plotted as abscissas) is zero. In this case all of the graphs would
pass through the origin. However, between the limiting values of
a See Hulletin No. 22, Bureau of Soils, p. 34.





12 THE MOISTURE EQUIVALENTS OF SOILS.

the centrifugal force employed in these experiments it will be seen
that in nearly every case a linear relation exists between the moisture
equivalent and the reciprocal of centrifugal force. These linear
graphs when produced do not pass through the origin, but, on the
other hand, the slope of these graphs is in nearly every case the same.







2 498
30
:28
426J ^~496
24 "-4962


020 /







2. DUNESAND
18 122 A/-RA64





14


66

S---- DUNESAND


0 234 56 7 8910 12 1 14 15
RECIPROCAL OF CENTR/IFUGAL FORCE
FVG. 1.-Relation of the mo'ture equivalent to the force employed.

Consequently we may represenllt within certain limits the relation
between the moisture equivalent m and the centrifugal force f as
follows:
7n = meo + -

where W77,, is the intercept on the axis.of ordinates of the prolonga-
tion of the linear part of the curve, and a is the slope. T'he constant





ACCURACY OF'THE DETERMINATIONS. 13

a may be computed from the data given in Table IV (the dunesand
excepted), from which we get a=6,100, expressed in terms of the
force of gravity. When a soil is in equilibrium with a force f, the
amount of moisture which is liberated when the centrifugal force
is increased from f, to f2 is
1
m1 m, = a (-f)

or the moisture set free is proportional to the difference of the re-
ciprocals of the two centrifugal forces, and is independent of the
initial moisture content. For example, if f,=1,000, f2=2,000,
a=6,100, then m, m,=3.05, or 3 per cent of water would be set free
when the centrifugal force is increased from 1,000 to 2.000 times the
force of gravity.
It must be distinctly understood that this relation has been tested
for only a few soils, and that it can hold only between certain limits.
It is evident from the diagram that the linear relation must fail
at very high velocities. In fact, samples 4962 and 4952 show a
marked tendency in this direction 'at the highest speed employed.
At very low speeds, also, the linear relation no longer holds. The
dunesand and sample 4983 are also exceptions, so far as a uniform
value for the coefficient a is concerned. But even with such limita-
tions this relation, if it is found to be applicable to other soils, will
prove of value in adjusting moisture equivalents to a common basis,
and in determining the amount of water available when the force
is increased by a definite amount. The fact that the amount of
water liberated from a soil in equilibrium with a given force, when
the force is increased by a definite amount, is, between certain limits,
independent of the initial moisture content, gives us a new point of
view regarding the availability of the moisture of different soils.
These experiments have not been carried sufficiently far to reduce
the moisture content of the soils to a point corresponding to drought
conditions, but it appears doubtful from these experiments whether
the linear relation would hold at speeds very much higher than these
employed. The subject is worthy of extended investigation, espe-
cially at high velocities.

ACCURACY OF THE DETERMINATIONS OF THE MOISTURE EQUIVALENT.

The accuracy of the determinations of the moisture equivalents
can best be judged from Table VI, in which is given a series of deter-
minations made on different dates. It will be seen that a good agree-
ment is generally obtained from different runs, the moisture equiva-
lents usually agreeing within 0.5 per cent of moisture. Samples 8969
and 9332 show, however, a steady decrease in the moisture equivalents
for successive determinations. The samples of these soils were small,






14 THE MOISTURE EQUIVALENTS OF SOILS.

so that it was necessary to save the soil used in a determination of the
moisture equivalent, for use in subsequent measurements. It is,
therefore, possible that the repeated oven drying, which a part of the
sample received, changed the character of the soil material sufficiently
to produce the diminution of the moisture equivalent observed.

TA.BLE VI.-Duplicate moisture equivalent determinations on different dates.
[5,000 revolutions per minute.]

Sam- First Second Sam- First Second
deter- deter- deter- deter-
ple Type of.soil. mina- mna- ple Type of sail. mina- mina-
No. tion. tion. No. tion. tion.

Per ct. Per ct. Per ct. Per t.
9248 Hagerstown loam: 8503 Miami silt loam-Contd.
July 11............... 17.3 17.2 July 20 ............... 18.3 18.5
Aug. 8.................. 17.8 17.9 July 21............... 18.6 ........
Aug. 10 .............. 17.6 17.3 5014 Miami clay:
4962 Hagerstown clay loam: July 21............... 17.7 17.9
Aug. 8............... 23.4 23.5 Aug. 7 ............... 18.9 18.7
Aug. 11............... 23.4 23.4 8398 Marshall sandy loam:
49C6 Hagerstown clay loam: July 18............... 13.0 13.5
July 12............... 24.9 24.9 July 22............ .. 13.3 13.6
Aug. 8.............. 26.5 26.2 9493 Marshall clay loam:
Aug. 11............... 25.6 25,3 July 19 .............. 23.3 22.9
8969 Miami stony loam: July 21............... 23.4 ........
July 18............... 17.7 15.9 9332 Oswego loam:
July 22............... 15.1 15.3 July 31............... 20.7 20.7
Aug. 5................ 14.1 14.5 Aug. 7............... 19.8 19.7
5008 Miami gravelly loam: Aug. 9................ 18.0 18.1
July 14............. 13.2 13.3 6917 Sedgwick clay loam:
July '0 .............. 13.5 13.6 July 28................ 21.3 22.2
5006 Miami loam: July 31................. 22.1 22.1
July 20............... 17.7 17.7 6632 Delavan silt loam:
Aug. 7................ 18.0 18.0 Aug. 10............... 26.8 26.8
Aug. 9................ 17.5 17.8 Aug. 2................ 24.1 23.5
8506 Mianfi silt loam: i Aug. 4................ 26.3 26.3
July 15............... 18.3 18.7


MOISTURE EQUIVALENTS OF TYPICAL SOILS.

With the experiments described in the foregoing pages as a basis,
determinations were made under uniform conditions of the moisture
equivalents of representative samples of each of the important soil
types established by the Bureau. These determinations are given
in Table VII, arranged in the form of correlated series, so far as
these series have been established.
The determinations were all made at a speed of 5,000 revolutions
per minute, which would correspond to a centrifugal acceleration
*equal to 2,940 times that of gravity. The determination of the
speed was not accurate to more than 1 per cent, which means a
variation of 2 per cent in the centrifugal force, since the speed-
enters as the square il the equation. The centrifugal force employed
in these measurements would then be

/=(29401 -0) g

where r represents the force of gravity.
The amount of :oil used covered the bottom of the cylindrical
cup (internal diameter 4.7 cm.) to a depth of approximately 5 mm.






MOISTURE EQUIVALENTS OF TYPICAL SOILS. 15

When large amounts of soil were used, it was found in some of the
more retentive soils that there was a tendency for a part of the
water to accumulate upon the top of the soil instead of passing
through the soil and escaping through the perforated bottom of
the cup. No trouble was experienced in this way when the layer
of soil did not exceed 0.5 cm. in thickness.
At the end of the run a moisture determination was made imme-
diately of each sample of soil used in the test, the whole of the
sample being employed for this purpose. The moist samples were
quickly weighed and then dried for eight hours in an oven main-
tained at a temperature of 1080 C., and then weighed again.
Such a temperature in an oven is easily maintained by the use of
toluol in the so-called water oven. If the apparatus is provided
with a good Allihn return condenser, very little toluol escapes, and
the apparatus works very satisfactorily.
In Table VII the first column following the type names gives the
moisture equivalent; the second the percentage of organic matter,
determined by the chromic acid combustion method; the remaining
columns give the mechanical composition. The limits of the diam-
eters of the particles in each group are given in italics under the
group number.

TABLE VII.-Moisture equivalents and mechanical composition of typical soils.
------------------------------------________________________________________ ____________
. (1) (2) (3) (4) (5) () (7)
C-f
No. Type. .


I 0 t o d '4 J d

P.ct P. c. P. ct. P. tP. ct.t P P. ct. P. ct. P. ct.
9107 Norfolk coarse sand.................. 4.6 0.9 9.6 35.3 19.1 16.4 7.2 7.3 4.8
8522 Norfolk sand..........................1 8.6 1.0 2.0 8.5 19.0 52.6 7.1 5.8 4.7
8589 Norfolk fine sind..................... 3.8 0.8 0.0 0.7 13.7 09.0 5.4 7.9 3.0
5665 Norfolk sandy loam................... 6.5 0.7 4.3 12.7 18.3 33.1 6.6 20.0 5.6
8706 Norfolk fine sandy loam............. 6.8 1.3 0.0 0.5 4.8 54.6 13.4 1. 1 8.5
8682 Norfolklo: m....................... 7.7 1.9 0.3 8.9 1..2 16.7 5.4 42.5 9.7
5493 Norfolk silt lr m................... 11.1 1.2 0.0 1.4 2.8 8.9 6.7 68.3 12.0
8530 Portsmouth sa d.................... 7.1 3.0 1.9 16.8 26.0 31.7 5.5 6.7 11.3
8537 Portsmouth sndy loam.............1. 6 10.9 0.8 4.9 7.3 31.7 13.0 23.6 18.7
10666 Portsmouth fine sandy loam......... 11.8 1.0 0.3 1.5 2.0 45. 4 15.1 2. 2 12. 6
8022 Portsmouth loam................... 10.6 2.7 0.0 03 0.3 14.7 24.5 43.5 16.8
7848 Orangehurg sndyloam............ 5.7 1.7 9.5 20.8 13.3 22.7 15.4 6.4 11.6
85S4 Orangeburg f.re sandy loam......... 8.2 0.4 0.0 0.2 2.2 69.6 8.3 16.3 3.3
8339 Oran-e'mr-gclvy (cly loam)......... 21.4 0.5 5.2 4.5 4.9 9.3 23.4 24.8 28.S
100,7 Houston silt loam ................... 13.7 1.7 1.0 1.2 1.0 5.8 15.1 63.4 12.3
10574 Houston blbck clay loam............. 82.4 3.7 0.3 0.6 0.8 12.1 17.1 42.5 26.6
78t7 HIoustonclay (sllty)................. 2.5 1.3 0.5 1.7 2.8 13.3 9.1 45.9 26N.4
10083 Houston black clay.................. 88.2 1.4 0.5 0.9 0.6 2.2 6.1 56. 33.4
- 7753 Vernon s.nd......................... .6 0.5 0.0 7.4 25.0 45.3 13.4 5.4 3.6
10118 Vernon fine sand.................... 4.8 0.2 0 0 .1 0.2 49. 42.1 5.9 2.1
7757 Vernon sandy loam......... ...... 10. 0.8 0.2 8.8 17.9 25.6 22.8 13.2 11.3
7737 Vernon fine sandy loam............. 2. 3 0.8 0.0 0.7 0.5 2.6 47.0 42.3 6i. 7
7749 Ver:ion silt loam..................... 19.8 0.9 0.1 0.1 0.0 0.3 16.4 72.2 10.7
7713 Vernon clay (loam) ....... ..... 23.1 1.1 0.5 0.2 1.3 3.6 27.1 51.1 15.9
6442 Yazoo loam .......................... 1.9 1.3 0.3 0.9 0.7 3.7 19.9 (4.1 10.1
6430 Yazooclay......................... 81.7 1.4 0.0 0.1 0.2 1.6 4.0 63.2 31.1
5741 Cecil sand................... ...... .1 0.8 7.6 17.7 13.6 25.0 12.0 17.4 6.1
5327 Cecil sandy loam.................... 7.7 1.2 14.7 15.6 9.8 21.3 13.1 17.8 7.9
7185 Cecil loam ........................... 16.6 1.9 1.9 5.1 4.8 15.3 9.6 39.6 23.6
4988 Cecil mica loam...................... 12.2 1.9 7.8 4.3 2.2 14.4 30.9 33.3 7.8
5719 Cecil silt loam ........................ 8.8 0.6 2.6 1.7 0.7 3.4 6.1 74.2 11.8
5727 Cecilclay (clay loam)...............I 21.2 2.0 1.4 2.7 2.0 7.1 10.1 44.8 31.6







16 THE MOISTURE EQUIVALENTS OF SOILS.


TABLE VII.-Moisture equivalents and mechanical composition of typical soils-
Continued.


i (1) (2) (3) (4) (5) (6) (7)


No. Type.

I 9



P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct. P. ct.
5904 Penn sandy loam..................... 9.3 0.9 2.2 11.6 13.4 27.6 11.8 21.0 12.1
5898 Pennloam ........................... 17. 1.3: 2.6 7.6 7:0 14.3 7.8 37.4 22.8
7703 Pennclay (clay loam)................ 17. 1.3 1.4 2.7 2.1 5.0 14.0 42.6 32.2
7240 DekaLb stony loam................... 12.8 2.5 3.4 11.1 9.0 17.7 10.8 33.4 14.6
10025 Dekalb sandy loam .................. 11.7 1.3 0.5 11.0 26.1 24.2 4.3 17.8 16.1
10216 Dekalb fine sandy loam............. 12.8 1.0 0.4 1.4 8.8 44.2. 8.1 25.0 11.9
10154 Dekalbclay loam.................... 21.6 2.3 0.4 0.7 2.1 6.6. 14.8 48.6 26.9
7233 Hagerstown stony loam............. 15.0 1.6 2.7 4.9 3.6 7.3 14.1 45.1 22.1
6534 Hagerstown sandy loam............. 11.0 1.0 0.4 5.0 8.8 23.4 20.2 33.0 9.4
4952 Hagerstownloam .................. 20.8 1.4 2.1 3.4 3.4 6.8 12.4 57.2 14.9
9248 Hagerstown loam................... 17.5 1.0 1.0 1.8 1.4 7.4 18.6 49.9 19.7
10212 Hagerstown silt loam .............. 16.5 1.0 0.3 0.6 1.5 9.1 10.5 60.3 17.4
4983 Hagerstown shale loam.............. 25.0 2.6 13.1 5.1 2.0 3.8 7.9 55.5 11.7
4962 Hagerstown clay loam............... 23.4 2.2 1.0 2.7 2.6 6.7 10.6 61.7 14.3
4966 Hagerstown clay loam.............. 25.5 1.3 1.9 1.8 1.2 2.5 6.3 69.6 17.0
9037 Hagerstown clay.................... 23.0 1.1 0.8 1.8 1.1 2.5 4.1 54.9 34.8
6024 Clarksville stony loam............... 13.7 0.3 1.5 2.1 1.2 3.2 8.0 67.9 16.5
6006 Clarksville loam .................... 18.5 1.3 0.1 0.1 0.4 17.1 23.6 39.4 19.0
6018 Clarksville silt loam ................. 15.1 1.2 1.2 1.4 0.9 2.9 5.6 73.2 14.8
6008 Clarksville clay loam................ 2).8 1.9 0.2 0.5 0.7 3.7 3.9 71.1 20.0
10184 Clarksvilleclay..................... 31.7 2.3 0.4 0.9 0.5 1.2 4.5 54.0 37.7
6092 Miami sand.......................... 4.3 1.9 0.5 9.1 38.1 41.3 1.9 5.2 3.7
6696 Miami fine sand.................... 4.7 1.1 0.3 17.4 26.0 35.6 10.3 5.9 4.5
9606 Miami sandy loam.................. 7.0 2.1 3.4 11.8 15.0 33.2 14.7 14.4 7.3
9626 Miami fine sandy loam.............. 13.3 1.7 1.6 5.1 7.8 25.4 15.2 32.3 12.4
8969 Miami stonv loam.................... 15.4 4.7 2.1 7.3 5.5 7.4 16.5 44.8 16.5
5008 Miami gravellyloam................. 13.4 0.8 5.6 11.4 9.5 18.0 15.9 26.1 13.2
5006 Miami loam ....................... 17.7 2.2 1.6 10.7 10.7 21.7 19.8 22.0 13.3
8506 Miami silt loam...................... 18.5 0.9 0.3 1.2 0.8 1.2 6.4 79.6 16.7
5014 Miami clay loam ..................... 18.3 1.2 0.7 1.6 1.8 4.2 11.2 68.4 12.0
9505 Marshall sand........................ 7.2 1.7 2.5 16.7 19.1 34.3 6.8 10.6 10.2
8804 Marshall fine sand.................... 5.4 1.3 0.6 2.2 7.7 57.8 21.5 4.2 5.9
8398 Marshall sandy loam................. 13.4 3.5 4.6 14.0 12.7 22.5 10.1 24.0 12.1
8418 Marshall gravelly loam .............. 19.7 7.8 7.3 20.2 10.8 8.7 5.8 28.9 18.0
9458 Marshall stony loam................. 30.2 3.9 2.3 8.7 10.4 28.7 15.1 22.8 12.1
8788 Marshall loam....................... 18.8 2.1 1.5 6.4 8.0 19.5 16.3 23.3 24.7
8728 Marshall silt loam................... 26.9 2.1 0.2 0.9 0.5 0.6 3.7 76.3 17.7
9493 Marshall clay loam................... 23.8 5.1 0.2 4.1 4.9 8.2 4.3 52.8 25.4
8453 Marshall clay ....................... 36.4 5.4 0.2 1.7 1.7 4.3 2.0 49.3 40.8
8664 Sioux sand ......................... 4.9 2.4 0.4 14.2 30.7 43.5 1.1 5.9 4.4
8670 Sioux sandy loam................... 13.8 3.3 0.6 13.6 16.1 20.4 9.7 28.1 12.5
9535 Sioux fine sandy loam................ 21.4 2.6 1.2 4.5 4.9 16.9 21.9 38.6 11.8
9174 Sioux clay........................... S.4 5.2 0.0 0.3 0.8 5.8 4.5 32.1 56.6
9288 Dunkirk gravelly loam................ 7.9 2.7 17.4 18.1 12.6 15.7 9.2 17.4 9.6
9296 Dunkirk sandy loam ............... 10.5 2.2 0.8 3.4 3.9 42.7 26.1 13.0 9.8
9286 Dunkirk clay........................ 37.8 3.7 0.9 1.8 1.2 5.3 5.9 31.6 53.1
9334 Oswego fine sandy loam............... 9.7 1.3 0.1 0.9 1.1 26.1 31.2 29.3 11.1
9332 Oswego loam........................ 19.4 2.7 0.7 1.8 0.8 3.6 15.8 61.1 16.2
9340 Oswego silt loam..................... 11.8 1.2 0.3 0.4 0.3 1.9 9.6 74.1 13.2
6921 Sedgwick sandy loam................ 7.9 1.3 0.3 8.9 18.6 32.3 18.6 14.0 7.0
6909 Sedgwick loam....................... 15.4 2.3 0.3 5.3 9.6 12.2 11.7 51.4 9.4
6917 Sedgwick clay loam.................. 21.9 1.0 0.1 1.5 2.3 5.3 10.4 66.0 14.2
6919 Sedgwick black clay loam............ 15.2 1.3 0.3 1.0 1.3 8.3 31.2 48.6 9.1
7978 Memphis silt loam.................... 12.9 1.2 0.3 0.8 0.6 0.7 2.9 83.8 10.6
GG32 Delavan silt loam.................. 25.6 2.9 0.3 0.9 0.7 1.0 8.9 69.4 18.8
7008 Marion silt loam ..................... 14.7 1.7 0.1 1.0 0.7 0.7 3.0 82.9 11.5
7009 Marion silt loam...................... 14.7 0.7 1.1 1.8 1.0 1.1 3.5 79.8 11.8
7385 Waverly silt loam.................... 24.4 2.0 0.7 2.6 1.1 2.7 7.6 62.9 22.2
4445 Maricopa gravelly loni.............. 8.9 0.2 11.3 11.0 12.0 15 0 28.4 10.6 5.7
4494 Maricopa sandy loam............... 18.8 3.0 Tr. 2.3 15.0 27.7 28.2 13.5 13.0
8134 Maricopa silt loam .................. 36.2 1.1 0.0 0.1 0.1 0.9 2.1 73.0 23.8
4486 Maricopa clay ....................... 19.1 0.4 1.3 4.0 10.2 19.1 23.1 19.4 22.8
4887 Fresno sand......................r.. 8.0 0.4 0.9 15.8 31.3 22.6 18.1 8.2 3.0
4(83 Fresnosandylomin..................... 5.3 0.5 Tr. 1.0 3.9 23.5 31.7 32.9 7.0
5081 Fresno flne sandy loam............... 12. 0.6 Tr. 3.0 3.8 16.9 35.4 36.3 4.7
5790 Yakima sand ....................... 5.4 0.3 0.3 1.5 9.6 38.9 40.6 5.8 3.1
7ti90 Yakimni fine sand(.................... 10.0 0.8 0.2 2.2 5.8 24.8 22.6 37.8 6.4
536;4 Yakima sandy loam ................. 13.4 0.7 0.9 3.3 7.3 23.0 22.9 33.2 9.2
7509 Yakimn silt loam.................... 23.6 2.3 0.5 0.8 0.6 1.8 7.2 81.3 7.5
9136 Yakiima loam ...................... 16.2 1.7 Tr. 0.5 0.6 13.0 35.0 34.0 16.6
5325 Cecilclay (subsoil).................. 86.2 0.3 0.9 4.0 4.1 9.4 6.0 16.2 59.8
57W2 Iredell clay loam (subsoil) ........... 46.5 0.4 1.4 1.6 1.7 8.9 &5 17.8 59.8






Bul. 45, Bureau of Soils, U. S. Dept. of Agriculture. PLATE I.







V -



















FIG. 1.-CENTRIFUGAL MACHINE USED FOR DETERMINING MOISTURE EQUIVALENTS.

























FIG. 2.-HEAD OF CENTRIFUGAL MACHINE, SHOWING CYLINDRICAL CUPS WITH PER-
FORATED BOTTOMS FOR HOLDING THE MOIST SAMPLES OF SOIL.





SRELATION OF MECHANICAL COMPOSITION. 17

RELATION OF MECHANICAL COMPOSITION TO THE MOISTURE EQUIVALENT.

In order to determine -whether a definite relation could be found
between the mechanical compositioni of the soils and the moisture
equivalents, the data given in Table VII were examined by the
method of least squares. It was assumed that the sands, silt, clay,
and organic matter all helped to hold the moisture in the soil, and it
was further assumed that the retentive power of each constituent
was directly proportional to the amount of that constituent present.
The object was then to determine the moisture retentivities of the
various constituents. 'The observational equations, in accordance
.with the above assumption, were of the form
A, 1Z B+ Z2+ C, Z+3+D Z4,+ E1 Z5= 1,
A2 Z + B2 Z + C2, B3+ D2, + E2 -- =M, (1)

AnZ,+B,Z,+ CZ3 D,Z,+EZ,= M,,
in which
A =per cent, groups 1, 2, 3 (2-0.25 mm.),
B=per cent, groups 4, 5 (0.25-0.05 mm.),
C'=per cent, group 6 (0.05-0.005 mm.),
D=per cent, group 7 (0.005-0 mm.),
E=per cent, organic matter,
M=moisture equivalent,
and Z1, Z2, Z,, Z,, Z.,, were the unknown coefficients, whose values
we wished to determine.
Each soil examined gave an observational equation of this form,
amounting to 104 equations in all. Owing to the inherent difference.
in the character of the clay and of the organic matter as well as the
observational errors, these equations could not be exactly satisfied by
one set of values of the Z-coefficients. It was therefore necessary to
determine by the method of least squares the most probable values of
the Z-coefficients; that is, the best values that could be deduced from
the given observations. The most probable values of the coefficients
as determined from the 104 equations were:

Z.,= 0.002
Z,=0.130
Z4 = 0.622
Z. = 0.(;27
The equation expressing the relation between the mechanical cor mpo-
sition of a soil and its moisture equivalent, as determined fromll the
104 observational equations, would then be

0.022.1 + 0.002B1 + 0.130 + 0.622D + 0.627E = /.l. (2)
.at






18 THE MOISTURE EQUIVALENTS OF SOILS.

By substituting the above values of the Z-coefficients in each obser-
vational equation, we determined the difference between the observed
and calculated values of the moisture equivalent in each case. From
these residuals we calculated the probable error a of a single determi-
nation of the moisture equivalent by means of equation (2), which
was found to be
r=-- 3.1.

The equation expressing the relation between the moisture equiva-
lent and the mechanical composition for the 104 soils becomes, when
the probable error is included,

0.022A + 0.002B + 0.13C + 0. + 0.22 + 0.G27E = M + 3.1. (3)

According to this equation, equal quantities of organic matter and
clay have nearly the same effect on the moisture equivalent. For
the centrifugal force employed, the moisture equivalent, expressed
in per cent, is proportional to about 0.62 of the total percentage
of clay and organic matter together. The silt has only about one-
fifth the effect of the clay and the organic matter, the moisture
equivalent being proportional to 0.13 of the silt content. The coef-
ficients of the sand groups are both very small, as we should expect.
An anomaly arises, however, in the case of these two groups, the
coarser sand having apparently a greater influence on the retention of
moisture than the finer grades of sand. This results from the fact

a The term probable error is here used in its technical sense, and serves as
a measure of the accuracy of the observations, in this case, of both the moisture
equivalent and the mechanical composition. As here used, it includes also the
departures due to the lack of uniformity in the character of the clay and organic
matter in the different soils considered. The uncertainty resulting from this
lack of uniformity would be more properly expressed by the term probable
departure," since it is not due to the observational errors. Iowever, it is not possi-
ble in a series of observations of this kind to separate these two terms, so that
the terml probable error has been used in this inclusive sense.
The term probable error should not be interpreted to mean the most prob-
able error," or the most probable value of the actual error." In any series of
errors, tie probable error has such a value that the number of errors greater
than it is the same as the number of errors less than it. In other words, the
chances are even that an observation selected at random will have an error
greater than or less than the probable error.
It should be emphasized that the pronbble error is concerned with errors of
an accidental ch(aracter-errors which are as likely to lead to results too large
as to results too small. A single determination of the probable error can not
reveal lie existence of a constant error, nor show whether the assumption
made regarding tie relation between the olserved quantities is correct. Errors
of this kind must he detected by changing tie form of the observational
equation, and determining thle probable error for each form of equation assumed.
The formi giving the smallest probable error most nearly represents the relation
sought.






NORFOLK AND PORTSMOUTH SERIES. 19

that the soils vary greatly in character, so that the individual pecul-
iarities of the soils tend to mask the true values of coefficients that are
very small.
The slight influence of the sand groups is shown by the fact that a
change of 50 per cent in the amount of coarse sand amounts to a
change of but 1 per cent in the moisture equivalent. We can conse-
quently ignore the influence of the sands in an approximate equation,
whence equation (3) becomes,

0.13(' + 0.62 (D+E) = J1 3. (4

NORFOLK AND -PORTSMOUTH SERIES.

In the case of a well-defined series of soils in which the material
is mainly derived from the same source one would expect a more
definite relation between the mechanical composition and the mois-
ture equivalent than would be the case for soils derived from differ-
ent rocks under widely varying conditions. We have accordingly
determined by the method of the least squares the coefficients for the
Norfolk and Portsmouth series combined, which are quite similar in
character, save that the Portsmouth includes heavier soils and carries
more organic matter. The coefficients of the sands were taken equal
to zero, and for the other coefficients the following values were
obtained:
Z,= 0.042
Z4=0.590
Z5=0.528

The equation of the combined Norfolk and Portsmouth series is then

0.04C+0.59D+0.;)3E= J1 (5)

A comparison of the observed moisture equivalents with those cal-
culated with the aid of equation (4) is given in the following table:

TABLE VIII.-Con pari.son of ob.ciredi and calculated moisture cqirh'lents of
the Norfolk and PorNtsmouth .seric..

Sample 0 E Moisture equivalent.
Organic"
number. (0.05-0.0005). (0.005-0.0). Omatue, .,1 .. b
imatter. Calculated. Observed. Residutals.
9107 5 0.9 3.7 4.6 -0.9
9022 7 4; 1.0 4.4 3. 6 r S
8589 8 3 S 2.5 3. -1.3
iTl5 19 5 7 4.1 6.5. --2.4
8706 In 8 1.3 6. 2 6. 8 -0. t
8682 42 10 1. 9 7 7.7 + 1. 0
5493 (6 12 1.2 10. 6 11.1 0.5
K530 7 11 3.0 8.3 7.1 +1.2
10666 24 13 1.0 9.2 11. : -- 2. (
l022 44 17 2.7 13.3 10.6 +2.7 i
8537 24 19 10.9 18.O 1s. 6 -0.6 0






20 THE MOISTURE EQUIVALENTS OF SOILS.

From the residuals in Table VIII we can calculate the probable
error of a single determination, which we find to be

r= 1.09

which means that if the soils employed in the measurements were
typical, then the chances are even that the moisture equivalent of any
other typical soil can be determined from equation (5) within 1.1
of its absolute value. The characteristic equation of the combined
Norfolk and Portsmouth series then becomes

0.04C+ 0.59D +0.53E= 1.1 (6)

Comparing this equation with equation (4), we see that the ratio
of the probable errors is 1:2.7. In other words, equation (6) ex-
presses the relation between the mechanical composition and the
moisture equivalent for the Norfolk and Portsmouth series with an
accuracy 2.7 times that of the equation (4) for the 104 soils. The
coefficients for the Norfolk and Portsmouth series are all smaller than
in equation (4), the organic matter coefficient being reduced from
0.63 to 0.53, the coefficient for the clay group from 0.62 to 0.59, while
the greatest change occurs in the silt group, which is reduced from
0.13 to 0.04.
MARS ALL SERIES.

For the Marshall series, omitting 'samples 8418 and 9158, which
are said not to be typical, the following values of the Z coefficients
were obtained:
Z,= 0.20
Z4= 0.62
Z,=-0.11

The equation for the Marshall series is then

0.20+00.(21)- 0.11E=M (7)

A comparison of the observed moisture equivalents with those cal-
culated fromn equation (7) is given in Table IX.

TABLE IX.-Corlmparixon of the obhsrred (1and (ulcuilitd moisture cquiralents of
the ,lar ihall series.

CE Moisture equivalent.
Sam ornlic - ----
number. (0.05 0.005). (0.005- 0.0). tr.latd. Observed Resiuals.
atter. Caleifluted.1 bserved. Residuals.

9505 11 10 1.7 8.2 7.2 +1.0
8801 4 6 1.3 4. 1 5.4 -1.0
8398 24 12 3.5 11.8 13.4 -1.6
8788 23 25 2.1 19.8 18.8 +1.0
8728 76 ,J8 2.1 26.1 26.9 -0.8
9493 53 25 5. 1 25. 5 23.3 +2. 2
8453 49 41 5.4 34.6 36.4 -1.8





MARSHALL SERIES. 21

For this series the probable error in the calculation of the moisture
equivalent of a typical sample from its mechanical analysis is
? r= 1.04

which is slightly less than thc probable error for the Norfolk and
Portsmouth series.
The equation of the Marshall series then becomes

0.20C+0.62D-0.11E = + 1.0 (8)

MOISTURE EQUIVALENT COEFFICIENTS OF CECIL, HAGERSTOWN, MIAMI,
AND VERNON SERIES.

In Table X the coefficients given above are summarized, together
with the results of.a number of calculations made in different ways
upon other soil types. In the Cecil series, for example, the coeffi-
cients have been determined: (1) for the silt, and for the clay and
organic matter combined; (2) for the silt, clay, and organic matter
separated; (3) for the clay and organic matter, assuming that the
silt and coarser grades have no effect on the moisture equivalent.

TABLE X.-Summary of moisture equivalent coefficients.

OClavy Probable
Series. Silt. Clay. maO rganic error,
matter.matter. moisture
equivalent.

All............................................... 0.13 0.62 0.63 ........... 3.1
Norfolk and Portsmouth....................... .04 .59 .53 ........... 1.1
Marshall...................................... .20 .62 .11 ........... 1.0
Cecil ............................................ .03 ..................... 0.62 1.7
Do ........................................... .06 .39 3.6 ........... 1.2
Do................................ ... ......... .44 3.6 ........... .7
Hagerstown............. ....................... .35 .......... .......... .06 2.4
Do........... ........................... .25 .11 2.4 ........... 2.0
D o........................................... .......... .61 4.7 ........... 3.2
Miami......................................... .09 ......... .......... .69 2.2
Do...................... ..... ....... ................. 70 1.8 ........... 2.7
Vernon........................................... .17 .................... .74
Do.......................................... .......... 1.4 02 ........... 2.4


According to these results, the moisture equivalents of the Cecil
series could best be expressed in terms of the mechanical analysis by
considering only the clay and organic matter, since on this assumnp-
tion we obtain the smallest probable error. This illustrates the
danger of generalizing from a small number of observations in a
question of this kind, since no one would admit the effect of the silt
to be nil. It will be noticed, also, that in all other instances we get
the highest probable error on this assumption. The danger just
spoken of is especially well illustrated in the Ilagerstown series,
where the reductions would indicate that the'clay had less effect than
the silt in the retention of moisture. This relation is altogether
improbable, and undoubtedly would not appear in the reductions if





22 THE MOISTURE EQUIVALENTS OF SOILS.

a suitable number of samples were considered. With this precaution,
this method of reduction could be used to determine to what extent
one series differs from another in the relation of its silt, clay, and
organic matter to the retention of moisture. This would, however,
resolve itself into a special study of the characteristics of the differ-
ent series. The present reductions emphasize the individuality of
different soils.
SUMMARY.

In the comparative study of soils it is important to supplement the
mecranical analyses by quantitative measurements of other charac-
ters, especially those relating to the movement and retention of
moisture.
The present paper deals with a method of determining the amount
of water which different soils are capable of retaining when the soil
moisture is subjected to a constant measured force sufficient in magni-
tude to remove the moisture from the larger capillary spaces. The
method is as follows: The soils under investigation are first thor-
oughly moistened, and are then placed in the perforated cups of a
centrifugal machine, where they are subjected to a constant centrifu-
gal force until they cease to lose moisture. The percentage of water
remaining in the soil is then determined.
It is possible to reduce the moisture content of a soil in this way
so that it is no greater than the moisture content of the soil under
favorable field conditions. By this method, then, it is possible to
determine the retentive power of different soils for moisture when
acted upon by the same definite force, comparable in magnitude with
the pulling force to which the soil moisture is subjected in the field.
Furthermore, this method of comparing the relation of soils to mois-
ture avoids to a large extent, if not entirely, the errors due to differ-
ences in packing, since the soils are packed by centrifugal force,
which acts upon each individual particle. This is further safe-
guarded by the high speed employed, which is sufficient to remove
the moisture from any large capillary spaces that may possibly be
formed.
The method has the further advantage that the force employed
can be accurately determined by measuring the radius of rotation
and the speed of the machine. A wide range in the force used can
also be easily secured simply by changing the speed of the machine,
since the centrifugal force increases as the square of the speed.
The maximum percentage of moisture which a soil can retain
when in equilibrium with a definite force we have designated as the
moisture equivalent of that soil for the particular force employed.
It logically follows that a series of soils which have thus been brought





SUMMARY. 23

into equilibrium with the same force will be in capillary equilibrium
with one another when brought into contact, and that no capillary
movement of moisture will take place between them. In other words,
the Mioisture equivalents of a series of soils represent the moisture
contents which those soils must have in order to make it equally
difficult to remove a very small additional amount of moisture from
any of the soils. It is from this point of view that the determination
of the moisture equivalent becomes of especial importance in the
comparison of the moisture contents of different soils under growing
crops.
The moisture equivalents of over 100 samples of type soils have
been determined, employing for this purpose a centrifugal force
about 3,000 times the force of gravity. These moisture equivalents
vary from 3.6 per cent in the coarser sandy soils to 46.5 per cent in the
case of a heavy clay subsoil.
These observations were reduced by the method of least squares
to determine the influence of the sand, silt, and clay groups, and of
the organic matter, upon the retention of moisture. It was found
for the whole series that each per cent of clay or of organic matter in
the soil corresponded to a retention of 0.62 per cent of moisture when
the soil was subjected to a force 3,000 times that of gravity. Each per
cent of silt, under similar conditions, corresponded to a retention of
0.13 per cent of moisture, and the coarser grades show practically no
retentive action against this force. The probable error for these
coefficients was rather high, and better results were oblained for
smaller series of related soils, using a different set of coefficients.
It is interesting to note that the organic matter, for the force em-
ployed, has a retentive power no greater than the clay group.
In investigating the influence of the speed upon the moisture
equivalent it was found, for the series of soils examined, that between
certain limits the amount of moisture set free when the pulling force
is increased by a definite amount is the same for the different soils.
In other words, when this series of moist soils is in equilibrium with
a given force, and the force is then increased by a definite amount,
the amount of water set free is independent of the initial water
content. Within these limits, then, a sandy soil and a heavy soil
of this series part with equal amounts of moisture.

0




A'*



































































44