Physical properties of Hawaii soils with special reference to the colloidal fraction

MISSING IMAGE

Material Information

Title:
Physical properties of Hawaii soils with special reference to the colloidal fraction
Series Title:
Hawaii. Agricultural Experiment Station. Bulletin ;
Physical Description:
45 p. : ill. ; 23 cm.
Language:
English
Creator:
Richter, Charles, 1892-
Publisher:
U.S. G.P.O.
Place of Publication:
Washington, D.C
Publication Date:

Subjects

Subjects / Keywords:
Soils -- Hawaii   ( lcsh )
Colloids   ( lcsh )
Soil physics -- Hawaii   ( lcsh )
Genre:
federal government publication   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Bibliography:
Includes bibliographical references (p. 41-42).
Statement of Responsibility:
by Charles Richter.
General Note:
Cover title.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029611789
oclc - 16324502
Classification:
lcc - S52 .E1 no. 51-63
System ID:
AA00014569:00001


This item is only available as the following downloads:


Full Text



































'INTRODUCTION
The -origin, formation, and.composition of tropical Hawaii soils
i es-ft them with physical and chemical properties which vary widely
Li"- de quite unlike those of the soils of the mainland of the United
j( ae. In-somea cases they also differ considerably from those of
:ilbe of other tropical countries.
; hei; large amounts of fine material, organic and inorganic, with a
^henical composition typical of lateritic decomposition, cause condi-
' ::ios which are radically different from those of other soils and which
a omii:ietimes extreme in nature. Moisture relationships especially
.e. ibt.sual behavior in many of the soils.
I~t :;,been known that the colloidal fraction of the soil plays
ae..Z: !!! I ::. le-in the soil complex phenomena. The investigations
iof if 1 past decade have definitely established the fact that practically
.':di theisntial properties of soils are directly or indirectly linked with
.':e ioi. d ai se; and that such important processes as flocculation,
Sdrainmage, plasticity, moisture-holding capacity, shrinkage, fixation and
ava:bi~ihty of plant food elements, hydrogen-ion concentration, and
Sbase exchange are functions of the colloidal part of the soil.
I The soils -f Hawaii, because of the above-mentioned differences in
i in'::;i 'formation, and composition, may be logically expected to
1:1iiia: bit a diierdnee in the nature of the colloidal complex as com-
Ipar d iith soils of other countries.
I .........1-.1 1


. ..;;;;;iiiiiil::. .;;... ....... .. ...













near the seashore at or near sea level is the modifying effect .
stone in the form of coral beds noticeable in soils adapted to sj


FIGURE 1.-Undecomposed lava flow (a-a formation)


FIGURE 2.-Field of volcanic ash showing the end of a recent lava flow (pahoeioe
formation) .
ture. Both the nature of the original parent material peculiar ti
volcanic countries and the soil-formation processes typical of tropical
or subtropical latitudes are responsible for the particular types d1
soil formed.
Kelley and his coworkers (21)' and McGeorge (24), in former
publications of the station, give information at some length cos

1 Italic numbers in parentheses refer to Literature Cited, p. 41.






































rise, among others, to the well-known lava types of.a-a
(l (Figs. 1, 2, and 3.)
.. the. chemical composition of the original parent lava is
oiJJiii" lto a certain extent for the composition of the resultant
]il,, i eatherin prdeesses, often referred to as "laterization," play
i:n cipal r61e. Most of the known weathering agencies enter
l;; ':" sil formation to a greater or less degree, but the more impor-
I" tt are temperature, rainfall, and humidity of the atmosphere.
I:::::Under tropical temperatures decomposition and oxidation of the (
S mineral constituents are rapid, and water which is present in very
I- ge amounts is constantly exerting its hydrolizing and solvent
: action. As a result, the basic elements, through action of the dis-
Ssolved carbonic dioxide in water, are leached out as carbonates and
bicarbonates. The alkaline carbonates also react with the silica,
ormn:i'*iming soluble alkali silicates which are leached out. Conse-
i quently, of the remaining constituents, the sesquioxides of iron and
alumina dominate the mineral part of the soil. The resulting soils
are usually heavy in character, the clay fraction consisting of
hdrted 6 oxides of iron and alumina and double silicates of iron and


i." ... .. ... .















poses. Later on, the definition of the term was concerned e
with the processes of formation. Campbell (11, pp. 1-f -1}::
rates the processes of tropical rock decomposition into two
alteration and weathering--depending on the state of saturati
water of the decomposing rock material. Attempts have been"
to define laterites on the basis of chemical composition. Fi
(18, pp. 559-566) proposes that the term "laterite" be rest
to soils containing 90 to 100 per cent lateritic constituents (
of iron, aluminum, titanium, and manganese more or less hy
that soils containing 50 to 90 per cent lateritic constituents be
"siliceous laterites"; that those containing 25 to 50 per cen
these constituents be called "lateritic soil"; and that soils ha
less than 25 per cent receive no designation on this-basis. ...
Martin and Doyne (25, p. 546) proposed that the classification.
laterite and lateritic soils be made on the basis of the compoit
of the clay fraction. They suggest that soils of which' the
alumina ratio in the clay fraction falls below 2 be called late
and that when this ratio falls below 1.33 the soil be described.
laterite.
It should not be inferred from the foregoing that all the si
of Hawaii are of the same average composition and exhibit simil
physical characteristics. Even though the parent material fra
different sources exhibits marked similarity in composition, d,
ferences in age and variations in weathering agencies are so la
as to bring about a wide variety of decomposition products. Ta
perature changes may range from tropical heat to that approach
Temperate Zone conditions, and rainfall may vary from a few ineb
to several hundred inches a year. Examination of the Appendi
(p. 43), giving certain climatic conditions and corresponding ad
with some of their physical and chemical properties, show that tt
soil types of the Hawaiian Islands vary widely. Kelley and H
coworkers (21) and McGeorge (24) show that soil types exist .1
Hawaii which are not laterites nor even lateritic in any of t
senses mentioned above. The soils of Hawaii in large part ma
however, be classified as laterites or lateritic.
Organic matter modifies the soil complex to no small degree aU
varies considerably both as to amount and state of decompositia
In very humid regions where the organic material decomposes und
partly anaerobic conditions, the rate of decomposition is slow ai
there is likely to be an accumulation of partly decomposed organ
matter. In this case large amounts of organic matter are preset
not only in the clay or colloidal fraction, but in the different te
tural subdivisions of the silt fractions as well. In the less humn
areas, or where excess moisture is rapidly removed by good drain
age, the organic matter is present in smaller amounts and dec
position is more thorough. From a biological point of view the


































n it sometimes in larger quantities in Hawaii soils. Manganese is
| 0a' :::found in certain areas, 4 to 9 per cent, and titanium is present to the
extept of 2 to 4 per cent. Kelley and his coworkers (21, p. 7) report
ai : instance in which the soil contains as high as 35 per cent titanium
S'dioxi:de (TiO.). The amount of organic matter in Hawaii soils
is about two to three times as high as in mainland soils.
In view of these facts the soils of the Hawaiian Islands should
be i expected to exhibit physical characteristics different from those
Sof soils of the mainland of the United States.
DESCRIPTION OF SOILS USED IN THE INVESTIGATION
S In order to obtain the necessary material for the investigation
herein reported, trips were made to various parts of the islands
where typical and unusual soil types exist. The larger number of
i samples were obtained from localities widely differing in climate,
elevation, and the like. The description and physical character-
l istics of these soils are given in tabulated form in the Appendix
:(p. 8). Selection of soil samples was based on the chemical and
physical differences of the various soils, including soil reaction, per-
i entage of organic matter, mechanical analysis, water-holding ca-
pacity, hygroscopic moisture, and color, and on geographical loca-
1 ti origin, and climatic condition. Of the 82 samples obtained,
.7 came from Oahu, 25 from Maui, 10 from Kauai, an 40 from Ha-

!i.iiiiip. ii :: iih.. .






























whereas in certain very arid districts the annual mean teii
is around 800. Agricultural areas extend to elevations zvt
4,000 feet. Temperature conditions are governed chit ft
tion, humidity, and the prevailing air current.
Examination of certain physical and chemical data ol i th
shows the great variety of soils that can be collected in thell
Islands; however, on the basis of mechanical analysis, n"
soils are clays and clay loams. Because of differences iu -i
of weathering and oxidation, and percentage of organKic '
soils show all gradations of color from almost pure y elai
all the different shades of brown and red to gray and bladC.
2 gives the color analysis of 21 typical and unusual soil t
pressed on the basis of the proportion of four basic colors as a*
by the American Soil Survey Association.


TABmi 2.-Basic colors in Hawoaii soils


Soil type b


Lanikai clay-...------..
Salt Lake Crater clay....
Peninsula clay loam..-----
Olaa sandy clay loam......
----...do----------.........-------..........----
Volcano sandy loam.......
Kona clay------------
..--..do--.-..----.------........
-o clay-----------
Kohala clay-----.--...
Waimea clay loam--.......
Honokaa clay---------
Hakalau clay----------
Hamakuapoko clay------
.....--do---------------------
Haiku clay.------------
.....do....----...............--
Honokohau clay.-------
--...do.........-------
Lahaina clay.-..---.....
Kealia clay -----.......
-----do--------...............-------------


Depth


Proportion of basic color


White


Black


I-~ I


Inches
0-10
0-12
(I)
0-12
0-10
0-12
0-12
0-8
0-12
0-12
0-12
0-12
0-12
12-24
0-12
12-20
(a)
(E)
0-12
0-4
12-24


Per cent
3
4
7
0
1
6
2
0
0
0
0
0
0
0
0
0
14
0
9
2
5


Per cen
55
72
81
63
85
81
59
76
80
80
09
66
65
65
57
72
41
39
59
81
55


Yellow


Per cent.
23
13
6
18
8
6
a
20
12
9
10
16
16
11
11
17
10
35
15
14
9
20


Red


Per cent
19
6

9
19
12
11
10
15
18
24
24
26
18
10
'46
18
8
20


Cuter


Light brown.
Gray.
Light gray.
Light brown.
Graoi
Brown.
Grayish brown.
Brown.
Do.
Light brown.
Light yellowish
Brownish red.
Do.
Light brown.
Brown.
Yellow.
Light red.
Light gray.
-Gray.
Brownish gray.


*Using the Munsell color method of measuring the percentage of the four basic colors as applied i
Hutton-Rice system. HUTTON, J. G. (chairman), REPORT OF SOIL COLOR STANDARDS COMMITTEE .
Soil Survey Assoc. Rpt. Ann, Meeting 7, Bul. 8:53-56a. 1927.
b The word "type" here has not the same meaning as in a soil survey or soil classification. Lani i
for example, simply means clay from Lanikai.
* Subsoil.


'The table contains descriptions of 78 soils only.


Sam-
ple
No.


SI


1 1





































Bobinson and Holmes (31, p. 25) show that the chemical compo-
... siion of the soil colloids is often indicated by their color and note
: that the soil and the corresponding colloid have the same color,
i differing only in intensity. Hence, there is a certain relationship
S between the color and the chemical composition of a soil. This
i has been observed in most instances with Hawaii soils also-gray,
dark-brown, or black soils and colloids having a higher, and yellow
and red soils and colloids a lower, silica-sesquioxide ratio.
For instance, in Table 9, giving the silica-sesquioxide ratio for the
different fractions of some important Hawaii soil types, samples
Ns. 291, 292, 428, and 474 are dark or gray organic soils, whereas
S Nos. 392, 448, and 547 are red, yellow, and brown, respectively.
i Oi Organic-matter content ranges from almost none in certain barren
infertile soils to as much as 20 per cent in some mountain soils where
on account of high rainfall, poor drainage, and dense native vegeta-
tion exi prior to cultivation, there is an accumulation of organic
material. Figure 4 shows the type of native vegetation that causes
S an accumulation of humus in certain mountain soils where high rain-
fall and slow drainage do not permit rapid oxidation of organic

















.LI JLU.LJ J J.U.L tjU J. .JL,.i. U LL L A.,L L L O t UCMIL.U JJ 1J f tEL cl L KU Ij l.i L
and nature of the soils follow.
Soil No. 2 is a clay from Lanikai, windward Oahu, near th
at sea level. It closely "resembles the adobe type of soil and isi
vial in origin, the fine clayey material having been washed*
from the surrounding mountain slopes. This is a very plt
sticky light-brown sodium clay of low organic content. Vege
does not do well in this soil on account of the deflocculated conkl
of the clay,
Soil No. 6 is a clay from Salt Lake Crater, Moanalua, near
lulu, where the Twin Salt Lakes are located. Originally a lak
tom of saline nature, this area was reclaimed for agrieultur.3.|
draining and covering with fertile loam from other localities. :
ever, agriculture (sugarcane production) had to be abandoned i
cause of the high alkalinity and high clay content of the soil. iii:
is a very sticky and plastic gray clay of low organic content i0
high specific gravity.
Soil No. 9 is a clay loam from near the'Pearl City Peninsula, OSi
very near the ocean (Pearl Harbor), at sea level. This is a hg
gray, pulverulent light clay, of the type used in rice culture .
black when moist on account of the large content of humus -BO
partly decomposed organic matter accumulating under anaeroi
conditions. The soil is very acid in reaction and has the lowet
specific gravity on record in the station laboratory. The clay is i
a perfect state of flocculation, due to the flocculating effect of Uj
large amount of electrolytes present.
Soil No. 12 is a sandy clay loam from field 4-5, section F, in ti~
Mountainview section of Olaa plantation, Hawaii. Residual :
origin, this light-brown mountain soil of low clay content contain
large amounts of undecomposed and partly decomposed orgazi
matter existing under semianaerobic conditions. The soils of' t
district (Nos. 10 to 14, inclusive) are characterized by high moistua
absorbing and moisture-holding capacity. Forming under excess"i
rainfall where drainage is poor, these soils are always water-logge4
The surface soils are dominated by a large amount of organic m|l
ter, whereas the subsoils are light-drab or light-brown mottle
puttylike clays. The lava from which they have been forme.
according to Burgess (10, p. 58), is at a depth of 2 to 3 feet. Sorn
of these soils will hold four to five times their weight in water. I
the station laboratory soil No. 11 had a moisture content of 411 pe
cent and soil No. 13 a moisture content of 500 per cent. (See alt
Shaw, 33, p. J20.) Agricultural operations are carried on witi









































iw."ll ":i:D BIfS UUJSJ% L LJ. L --L. JL LI *JL I UALJ.JL.L JL LL UJLJL.O L L J U J. IJ YT .
;i Soil No. 33 is a brown clay loam from an experimental edible-
Seanna field of the station at Waimea, 4 miles east of Kamuela. This
i soil is typical of a large part of the Waimea plains-a rather cool,
Shamid, and wind-swept region at an elevation of about 2,700 feet.
The soil is of high organic content, fine uniform physical structure,
and good productivity.
Soil No. 37 is a clay from the upper end of field 2, Honokaa plan-
tation, about 100 feet from the main road. This light-brown acid
Soil is typical of the mauka (upper) fields of the upper Hamakua
District. The high organic content of the soil here often causes an
acid condition, which is being remedied by the application of lime.
SThe soil is of a good capillarity and good moisture-holding capacity.
Soil No. 42 is a clay from field 8 of the Honohina section of Haka-
ilau plantation, Hawaii. This is a light yellowish-brown surface
S soil of considerable plasticity from the low elevations of the Hilo
Coast where the annual precipitation is upward of 200 inches.
!' Soil No. 48 is a clay from the Haleakal substation, Makawao,
!Maui, on the slopes of Haleakala, at an elevation of 2,160 feet.
I::8::.. 1;;1 ---"

..i ;g: ...iii : .:.i .. .. .. ,












under cultivation for some time, having been planted with k 7
sified crops.
Soil No. 49 is a clay and the subsoil of soil No. 48, taken.
depth of 24 inches. It has a high moisture-holding capacity i:-
low organic-matter content.
Soil No. 56 is a clay from the Haiku district, Maui, Kaupa
section of the Haiku Pineapple. Co. The soil is typical of the
apple soils of this district. It is a light-brown acid surface so
good fertility notwithstanding the fact that the clay fraction is
high. It has a specific gravity of 3.24.













tion. Note stratification (soils Nos. 65 and 66) in central part of bolder
o 2 I a o o o. .







FIGURE 5.--Soil formation in situ. Large, exposed bowlder undergoing deeompoc- ,,r
tion. Note stratification (soils Nos. 65 and 66) in central part of bowlder
Soil No. 57 is a clay and the subsoil of soil No. 56, taken at a delp ':o:li
of 20 inches. In all characteristics it is similar to soil No. 56.
Soil No. 65 is a clay from the side of an exposed bank at the e ai:d
of Honokohau Gulch, Maui, near the sea, at an elevation of 100 fe et
This is a yellow, inorganic, very colloidal subsoil of consideraIiblei|
depth resulting from the decomposition of the parent rock in situi
The soil appears as a clearly defined stratum and is conspicuous l ifi
visible from a long distance. The width of the stratum is only about
2 feet. The soil is low in organic matter and is slightly alkalina iATi
reaction, indicating an accumulation of basic elements. Weatheri4ilg
takes place under very hot and dry conditions. In moisture-holdingi
capacity the soil is high. .3.
Soil No. 66 is a clay from the same exposed bank at the end ofi
Honokohau Gulch, Maui, from which soil No. 65 was taken. Thiu.i i
subsoil, many feet deep and of red color, covers soil No. 65 as ,ai
separate and clearly defined layer. In physical characteristics, eixr
cept color, it is similar to soil No. 65. Both soils are extremeiltI
colloidal in nature. (Fig. 5.)


































product. When weathering is chiefly of a physical
te -:rerlting soil is usually coarser in texture than when
tio-~ifused by chemical changes, such as occur in
limtesn. Moreover, the resulting soil is usually more evenly
in the different fractions when weathering is chiefly physi-
A.t The difference in solubility is also a factor when textural dif-
.i.c. are considered. Quartz, on account of its low solubility,
longer in the coarser fractions, whereas minerals of a less
Sli: tant nature such as feldspars, hornblende, and augite, con-
I ta~te gradually in the finer fractions. It is not surprising, there-
.i... to earn, in the case of Hawaii soils, the parent material of
hic consists largely of the last three minerals mentioned, that the
iier separates, such as the fine silts and clay, are present in larger
i i amounts than in most of the mainland soils.
.ii Th....... e soils of Hawaii are on the whole heavier than those of the
....mainland. of the United States, inasmuch as they contain larger
percentages of the fine fractions, especially clay. Most soils belong
ii: the clay and clay-loam types, these designations being in accord-
rsac with the United States Department of Agriculture Bureau of
Chemistry and Soils classification on the basis of mechanical analy-
The Appendix gives the mechanical analyses of 78 Hawaii
as determined by means of the rapid hydrometer method of
Donyoneos ( 6).




a ... ........ "....:....
























5 sandy loam, and 2 silt loam.
In the figures given in the Appendix, the limits in the dii
range for the different grades of particles do not' correspo


MO aM eenvr


,
.,. i-
.;l.....





; I;


'11


*1

iiiih iii
... :* '::.:il


: ii:!!!:, iiiU~


.. .: .. -. **** (





.:i'. i.. i:i l l

f" "" :;;
... '.: iiii
::.a.m
, : "" < i


zaso Ote'd
FIGURE 6.-Diagram showing the distribution of 71 Hawaii soils into the principal::
soil classes on the basis of mechanical analysis '


those used in the United States Department of Agriculture Bure
of Chemistry and Soils (sand particles from 2 to 0.05 millimetel
in diameter, the silt particles from 0.05 to 0.005 millimeters, a
the clay particles less than 0.005 millimeter in diameter), but t|
different textural subdivisions are taken in the meaning of Bouy3s
cos (5, p. 320; 6, p. 234; 7, p. 475). He states that the rapid hydrqW
eter method includes a part of the fine silt fractions in the clay .-i
colloid fraction and that a part of the silt fraction is included i
the fine sand.
The soil classes obtained by the rapid hydrometer method wei
checked in the case of-21 soils with the method of repeated sample.
at a certain depth in order to obtain a comparison. The soil classic
cation obtained was the same by both methods for 14 of the 21 soil
In the other 7 slightly different classification was obtained due t


'a
'A
I.
'A

























































Mechanical subdivision
.. .,. .. '-3 -'"" '


Sand
(0.02
milli-
meter)


Slt (0.01
to 0.02
milli-
meter)


Fine silt
(0.005 to
0.01 mill-
meter)


Very fine
silt (0.002
to 0.005
milli-
meter)


Colloida
(0.002 mil-
limeter)


SI II---- I I


Per ast
25.2
15.7
34.0
65.0
06.5
4L 5

54.O
34.0
I.2
a.7
12.8
10.2
42.
28.2
683.0


Per cent
8.0
3.9
12.
120
&0
7.2
10.1
15.0
.9
1&0
9.1
8.0
t.8
9.2
9.4

104
6.9


Per emt
8.7
&.4
10.5
7.0
5.4
2.6
12. 1
ill
11,0
1L6
9.0
10.7
8.7
206
&0o
1LS
10.
11.6
7.6
13.6
8LB
8 6
.8


Per cent
5.8
7.4
16. 0
11.0
&6
.0
15&4
12.0
15. 6
5.0
1. 2
1& 8
7.2
18.2
14.4
7.7
&.7
14.9
15.8
5.2


Per cent
51.5
66.6
27.0
150
9.5
11.8
23.9
15.
49.4
17.0
23.1
32.5
30.1
21.3
a. o
40.1
35.1
3006
48.9
25.8
28.1


Organic Reaction
matter pH


823
8.73
4.80
5.87
5.77
6.50
6.55
7.17
7.40
6.70
6.36
6.50
5.97
5.77
4.75
5.03
7.42
7.25
7.77
7.72
7.74


Per aet
3.94
&389
18.52
15.75
2L&52
8&49
13.36
18.1
L 15
10.95
1L566
0.07
6.67
2.34
7.10
6.27
&o9
2.32
1.22
7.70
1.91


-;" .-


Soil classes


Clay.
Do.
Clay loam.
Sandy clay loam.
Do.
Bandy loam.
Clay.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Sandy clay loam.


i M, a ultS: analyuls by a modifal pipette method.
.l usl; So" asesI ordin to a nation given by R. O. Davis and H. H. Bennett (15).


.. .....: ,
!! i ':. iii. iii~ii~ li.:i;F;;;i ... ..ii :: : .


'iii No.l


= 4------
t: ---------
:iii: ...;Bi. ....



-::. ,, ......



ML0-------
----------
i.::. .........
;i........




r! :i 3..........
L ..... ..


I!
'I























I o .i




50

0 o ___ Ij r ___



0.3 3.7 2.0 2.3














DA _ETERMM.)
200













SHawaii s
50.IM R M









011
20 goo











any intermediate point may be obtained by interpolation. The
"equivalent diameter" is used instead of "diameter" for obyi
LOGARTHM OF EQUIVALENT DIAMETER i
S53 3.7 2.0 2.3



0.reasons, and the logarithm of the equivalent diameter is plotted



stead of the equivalent diameter itself to avoid crowding the p
E A AlE TER (NM I)



Figure 7 Mec shows graphically these summation curves of the fraction of ls
this manner smooth curves are obtained, making interpolation
any intermediate point may be obtained by interpolation. The
"equivalent diameter" is used instead of "diameter" for o
reasons, and the logarithm of the equivalent diameter is plotted
stead of the equivalent diameter itself to avoid crowding the .
Figure 7 shows graphically these summation curves for six soixa,. ::|
this manner smooth curves are obtained, making interpolation 0






























irou materials varying results are obtained when different liquids
.f.e used in the determinations of true specific gravity. In these
" i^,determinations water and toluol were used, but in applying the fig-
us .to Stokes' law the specific gravity as determined in water was
ieed. The determinations were made as follows:
The soil material was ground to pass a 60-mesh sieve and was
driad in an electric oven at 1100 C. for at least 18 hours. A 50 cubic
i l~R meter soil pycnometer was half filled with recently boiled dis-
il: ed water and weighed. From 6 to 8 grams of the oven-dried soil
Mi:was added, the whole weighed and then boiled for five minutes to
-expel any air occluded in the interstices of the soil particles. When
i: the pycnometer had cooled and the soil particles had settled, leaving
lte supernatant liquid clear, it was filled with water to the point
io'overflow. The pycnometer and contents were brought to the
i. temperature at which the determinations were made (30), the ground-
g~mii stopper was inserted, the liquid which had overflowed was
relll li and weighing was quickly done. Duplicate determina-
l'ions made in this manner usually checked within 0.05.
In several instances the true specific gravity was determined,
ii:!::~ o being used. In most cases the agreement was good, the aver-
a being 2.88 in toluol as compared with 2.8 in water, where both
deter^minations were made. In determining the apparent specific
;I,: riy or volume weight, the same soil material was used as in the
specific gravity determinations. A tared glass cylinder grad-
I'iiittted to 50 cubic centimeters was filled with the soil to the upper
mark, dropped 4 to 6 inches a number of times until compactin
!'!ws no long1 noticeable when the volume was noted and the soil
with the cynder weighed to 0.1 gram.



L. ..... .....
"................. .....
















True Appar- True Appar-
ent nt ....
Soil No. specific spec Porosity Soil No. specific spe
gravity gravity av gravity

Per cent B
2------------------ 283 1.25 55.8 48............ 3.36 1.82
6-....---------... 2.98 1.32 55.7 49-...----..--..---- 2.70 L 06
9----------------- 2.19 .66 69.9 560---------....--- 3.24 1.32
12-.---------------- 2.53 .93 63.2 57----------------- 3. 31 1.25
15 ---------------- 2.31 .98 57.6 65--------- 2----86 L04
17....----------. ... 2.79 1.19 57.3 66--.----...--........ 8 .94
21 --------.------ 2.57 1. 04 59. 5 69....- --------- 2. 90 1.17
23.....-----------. 2.52 1.04 58.7 76----......-..--. 2.92. 1.09
31_ ------------ 2.79 1.16 58. 4 77...----------. 2. 96 1. 14 .
33.....----------. 2.68 .93 65.3
37. -------------- 2.52 .96 61.9 Average.... 2.80 L 09
42----------------- 2.84 1.04 63.4
42.................. 2.84 1. O3.4.... ..

As was anticipated, large variations were obtained. Soils withM
larger percentages of organic matter gave on the whole much
figures than soils of low organic content. The lowest figure record |i
was 2.19 for soil No. 9-a highly organic, light-gray soil contaim.n
18.5 per cent organic matter; whereas the highest figure recorded w
3.36 for soil No. 48-a brownish-red, ferruginous clay of 6.67 per ceO
organic content. The average for all the soils was 2.8 despite the fha
that the series contained several abnormally organic soils whichii
lowered the average. In the case of soils the organic content of which
was below 10 per cent, the specific gravity averaged 2.96, whereas f "" h
soils the organic content of which was over 10 per cent the speci.f!
gravity averaged 2.47.
In determining the volume weight or apparent specific gravity, i "
is realized that the figures are only approximate and that they do n ....
represent actual field conditions in so far as structural relationshipSi ,'
are concerned. The high percentage of organic matter in Hawa;jil;
soils is one of the main determining factors and is often the cause of :I
considerable variation. On the whole, the apparent specific gravity
of Hawaii soils is much lower than that of the mainland soils
partly for the above-mentioned reason and partly on account of t l
finer texture. For the whole series it averaged 1.1, the values fluctuatM ,:
ing between 0.66 in soil No. 9, a highly organic kind, to 1.32 in soi Il
No. 6, an inorganic clay resembling adobe. -;,
The percentage of pore space calculated from the specific gravities :ii..
indicates the fine texture of the Hawaii soils. As is known, coarse ....
soils are much heavier than fine soils, because of the nature of thei!
contact of the particles and their disparity in resisting compaction.:
The method employed in determining the apparent specific gravity,:a
when disturbed soil is used, gives lower results for that constant than: -
does either the iron-cylinder or the paraffin-immersion method. On ::I
mainland soils the apparent specific gravity determined by the iron-:j
cylinder and the paraffin-immersion methods gave results 0.4 higher -:i























R AS'l J EIMp.LU J.IL bLO IflW UtC IvM U.ILUr, i7RLU CI.LLuLlUl,. AiU a IJ. LiU VUI I LJUO
aistres and their relationships to plant nutrition. It is well
that the different moisture constants vary between very wide
for different soil types and sometimes vary considerably even
oils belong g to the same type. This is- true especially of
soils which, by virtue of their peculiar chemical composi-
.i.d physical characteristics, show great variation in their mois-
II at.i. onAships.
Obtain a more or less true picture of the moisture relationship
~ iin Hawaii soils of different types, a number of soils were
for their moisture-holding and moisture-adsorbing prop-
.. sties. The various physical and chemical properties of these soils
ich affect their moisture relations varied between wide limits.
m-The data are given in Tables 1, 2, 3, 4, and 5, and in the Appendix.
...he moisture constants determined for these soils were the hygro-
soipic and capillary moistures, the moisture-holding capacity, and
CIt moisture -iqivalent. It should be stated that notwithstanding
i.l::the: use of standard methods and of precautions to secure uniformity
miin conditions surrounding the determinations, the results do not rep-
S.resent actual field conditions for obvious reasons. In the first place
Ithiirte textural and structural conditions of the soil as they exist in the
. .field .:.have been disturbed, and their moisture characteristics have
..l :thus been altered. This is an unavoidable source of error in most
i.ideterminations, and in order to obviate it attempts are made to sub-
jel t all soils approximately to the same degree of alteration by apply-
i.g uniform preliminary treatment in the laboratory. In the second
plhie, in most colloidal soils upon air drying a change seems to take
iici e in the colloidal complex which alters the moisture-adsorbing
iwer to a considerable degree. This alteration may take place to
rent degrees depending- on tlie nature of the colloidal phase and
Ithe extent the soil has been subjected to drying.
Some soil investigators believe that the moisture properties are
lll:iiialtered to a certain extent immediately upon the removal of the soil
...... romm the ground to a container. As an extreme example to illustrate
i ::: point: Certain soils from the Olaa district, Hawaii, having
been formed under extreme rainfall and poor drainage conditions.
l :-are almost constantly in a water-saturated condition. As a result,
52031--31-----


iEi Ei :!! ........ .... : ": .! i !........ : .... 1:..













water-a fact indicating the irreversible nature of the reactions tl
ing place in the colloidal phase. This fact also indicates that
soils in situ have never been deprived of their moisture to :
appreciable degree.
The hygroscopic-moisture content was determined by exposing ..
powdered soil to an atmosphere saturated with water vapor a
3.3 per cent sulphuric acid at 300 C. for six days, when the gn
in weight was determined by drying to constant weight. The a
rious factors entering into this determination, such as effect of ti...
vapor pressure, temperature, degree of vacuum, and size of partietsi
have been thoroughly investigated by Middleton (26, p. 455), whoq
findings were observed in these determinations. The amount of
hygroscopic water which the soil holds under ordinary air-dry con.,
editions was also determined. The soil was brought to a constant :.
weight by drying it at room temperature, 28 to 320 C., at a relative|.|:
humidity of approximately 50 per cent, and the moisture content was.ii
determined by drying in the oven at 1100 for 18 hours or longer.;I,'
The percentage of moisture was calculated on an oven-dry basic :
both in this and in the other moisture-constant determinations. i
The capillary moisture was determined by filling a tared glass:;
tube, 1 inch in diameter by 15 inches in length and fitted with
copper-gauze bottom, with air-dried soil to the 12-inch mark, and : '
dropping the tube repeatedly 4 to 6 inches until conpacting was.
no longer noticeable, when the weight was recorded. The base of the 3
tube was then placed in a pan of water, until the water rose to the
surface of the soil. Then the tube was removed from the water,
placed on a blotting pad to remove excess moisture, and weighed. '
The time required for the surface to become moist was also recorded.
The moisture-holding capacity was determined on the same sample
on which capillary moisture was determined. The tube holding the
soil plus capillary moisture was placed in a cylinder of water the
level of which was the same as that of the soil, and was allowed to
remain until the soil had become thoroughly saturated. At tthis
point the tube was removed, drained of excess water, and weighed.
The moisture equivalent was determined according to the specific
cations of Briggs and McLane (8, p. 140). Approximately 30 grams .
of soil on an oven-dried basis was placed in the centrifuge cups the
wire-gauze bottom of which was first covered with a sheet of filter
paper. The cups of soil were then thoroughly moistened by placing ii
them in a shallow pan containing water to a depth of one-fourth to ::
one-half of an inch. The cups were allowed to stand overnight in '.i
the pan during which time they were suitably covered to protect
against evaporation.
The time of centrifuging was 30 minutes at a rate of 2,440 revolu- :
tions per minute at the end of which the moisture retained was i
determined by drying to constant weight at 1100 C.
The various moisture constants obtained are given in Table 5 and,
are shown graphically in Figure 8.






















































FO/SZ /YO.
SFiGura 8.-Relation between various moisture constants


TABza 5.--Moistre relationship in Haueaii soils





.... moisture mosture rise capacity lent
per cent
II :SO
..... ....W..
,I^_----------------- ------ ______________ --------_-------_--------_------- _-------

Per cent Per cent 'Per cent Days Per cent Per cent
i!.ii-,-- ------- .6.1 13.3 31.8 16 43.9 39.6
i. ------- --------10.8 20.8 34.3 14 48.1 48.5
------------------------ 10.0 77.7 8 119.6 67.2
----- ------------------- 20.5 21. 6 35.4 5 74.1 67.5
------- -------- -------- -------- 14.0 ---------- ---------- ---------- ---------- 42.7
3.1 --------- ---------- ---------- --------- 21.
-- ----- ------ 19.3 23.7 41.6 4 66.0 53.4
its-.... 15.8 --------- 45.1 4 63.5 46.1
---------- -w-- :-:--------- ^. I7
-:-- -15.7 ----------- 49.4 1 79.8 61.8

-------------------------------112 1. 6 42.3 1 60.0 2.0
--. -5.5 9.7 36.1 1.8 50.6 37.7
S- 21.1 -------- 53.2 1.8 89.6 71.3
------- --------- 3.8 9.3 29.1 4.8 46.9 30.9
-- .-------3.1 12.0 31.4 2.9 45.5 27.4
----.. .----. ......... .. ---- 15.5 29.6 54.2 27 90.0 6L2
.- -.. .... ----9.9 25.6 52.5 3 73.5 50.4
-.--.....I--------- -------- .2 18.8 30.6 36 42.9 40.7
SI.-: --..- -------5.8 14.8 42.65 2 63.6 6.9
:...:.......--- --------. 41.9 .8 5.7 31.1

:-A-rasb-,------- .10.7 18.4 430.........- 64.7 4.7


ll".:',,';::i" ... ...,

















the amount of moisture adsorbed or held depends on the surfa e
posed, and for this reason soils containing large amounts of:.wi
70 .
.- ----"--- ...----I ,,,-| .7

mAaP oc'r o V7' 0 *, 4W V /C ,J.4Y7


'4

1I4


2 / 2/ 73/ 7 672 656 67 S 5" 3 .9 7W
SO/Z. NO. : .
FIGURE 9.-Relation between per cent colloids, per cent organic matter, and per cent ::::
water vapor adsorbed
matter or colloids adsorb or hold iuch more moisture than soils con:..l;i
training small amounts of such matter. The same relationship exi=*sA::'
between organic matter and moisture held or adsorbed. Further- 1
more, since usually all the organic matter is a part of the colloidalh':
fraction on the basis of particle size, it may be said with a good :e-,
gree of accuracy that the moisture properties of soils depend on the.: ::.i
colloidal fraction--organic and inorganic. i
To permit study of the relationship between organic-matter con:Il
tent, colloidal content as determined by mechanical analysis, andl':
moisture properties of soils, these three factors were plotted in Figuii|
9. The surprising feature of this diagram is that the hygroscopiq g
moisture not only does not follow the colloid content but in the largerti
number of cases there is an inverse relationship. However, the hygro-:.....
scopic moisture conforms more or less to the per cent organic matter, :
































Per cent Per cet Per cent Per cent
S m ........................................ 25 26-32 19.4-29.1 0
lasn)--------- .--------------------------. 87 70 45. 9-55. 9 5-25
-..------------ 181 180 106. 5-253. 6 25-50

S;i;sa rcha( i7yi, p1. 158).

Table 6 indicates that the moisture-adsorbing power of humus is
.a the average about two and one-half times as great as that of pure
Cqity; the average water-holding capacity of humus about three and
i pnerhird times that of clay. This explains the fact that certain soils
h in organic matter, such as many of the Hawaii soils, show
gh'" adsorptivity-despite the fact that their clay or colloidal fraction
i, relatively Tow. On,the other hand, it is easily possible for a soil
.with a high colloidal fraction to show a relatively low adsorptivity.
:i As the organic-matter content of the soil increases the hygroscopicity
rises rapidly.
That this disparity in hygroscopicity between the various soil con-
: tituents may easily obscure the parallelism between amount of clay
or colloids as determined in a mechanical analysis and the hygro-
Iecopicity of the soil is illustrated- by Lyon and Buckman (23) in the
ease of mainland soils. Table 7 shows the hygroscopic coefficient com-
pared with other soil factors.

7TArg 7-The hygroscopic coefficient compared wjith certain other soil factors

Hygro-
Soil Bme Cna uym
eo Clay ignition H
client

Per cent Per cent Per cent Per cent
Dii.kiii 'ilty dM y loam (a -------- --- ------------- 3.80 12.9 5.08 1.26
SDoakfrktaltyy loam tobl) -------------..----------- --.77 20.0 3.05 .20
OIY Cdy 0 loam (sa -------)------------------------- 18.90 20.1 14.54 4.34
V umnn.el day (gubuolO)--.----------------------------- 17.40 74.5 .7 .4

I i Ly anmd Duekman (0).
..........a:'.: ::. : ... ..... .1.
,,. Vergelm esela y .( ...... .................17. 40 74. 5 .. 79 4 .



x m :: x












a higher hygroscopic. coeficient" than thi Vergennes clay, t*l
fraction of which is 74.5 per cent. A comparison of .the .
subsoil with the Clyde clay shows that they both possess th.e.
amount of clay, yet the Clyde soil has a hygroscopic coefficieiftt
three times as large as that of the Dunkirk soil, due to the
in humus.
In the above-cited example the magnitude of the humus co
ranged from 0.2 per cent to 4.34 per cent, small figures, relate
speaking, as compared with the percentage of organic matter|
Hawaii soils. If such disparity is possible with soils contaiiu
from 0.2 to 4.34 per cent of organic matter, it is logical to ss
that much larger differences are possible with soils containing fr
two to three times that amount of organic matter. This actually
the case with Hawaii soils. The wide limits between which .tb&i
organic-matter content of Hawaii soils varies explains the wid
limits obtained for the moisture constants themselves.
The moisture-equivalent figures vary between 21.9 and 71.3 l
cent, the average of the whole series being 46.7 per cent. Consider7
able variations were obtained for this moisture constant even within : i
the same soil type. For instance, soils Nos. 57 and 9, which are .: i-:
classified as clay and clay loam, on the basis of mechanical analysisilMl
had moisture equivalents of 27.4 and 67.2 per cent, respectively. In:
this case the large difference in the respective organic-matter content:
(6.27 and 18.52 per cent) may explain the disparity in the moisture:: ,;
equivalent. However, in the case of soils Nos. 2 and 49, the moisture: :
equivalents of which were 39.6 and 71.3 per cent, respectively, thI!
difference in organic-matter content does not explain this disparity :
(organic matter 3.94 and 2.34 per cent). This case is all the more :
remarkable because soil No. 2, with the lower moisture equivalent, in:iI
a clay with 51.5 per cent colloids, whereas soil No. 49, with a moisture' a
equivalent almost twice as great as that of soil No. 2, is a clay with ;
only 21.3 per cent colloids.
These discrepancies may be explained by the considerable differ, :il
ences that may exist in the chemical composition of the inorganime':
colloids, and by the amount and nature of the coarser-than-clay frac- .
tions, such as silts and fine silts. As will be pointed out later (p. 25Y),:;:,;
the chemical composition, especially the silica-sesquioxide ratio, may :]!
vary very considerably in the different fractions of the same soil.
Similar discrepancies in moisture equivalents are noted in thel :
figures published by Bennett and Allison (4, pp. 51, 2S, 308) in the ":.
case of Cuba soils. For instance, Alto Cedro clay No. 32828, with:::
a clay content of 52.1 per cent, gave a moisture equivalent of 75.2 i
per cent, whereas Nipe clay No. 32830, with a clay content of 44.5
per cent, gave a moisture equivalent of 23.7 per cent. Both soils:i
are very low in organic matter. In this instance the low moisture-m:,l'::,
retaining capacity of the Nipe clay is attributed by the authors t
certain physical peculiarities, notably to large pore space. ;:


'WE








di ce&1ln&ici9. i calculsted. by dividing the moisture
by the Briggs and'Shants factor 1.84 (9, p. 68), the figure
Sbe. obtained for -the whole series. This average figure
Iel with experimentally obtained data. Hartung (19919. 16),
with pineapples, noted that no root growth took place when'
content fell below 22-6 per cent. Feeble efforts at root
mettook place between 22.5 and 97.0 per cent moisture

Aen (17, p. 157) determined experimentally the wilting coeffi-
for three. different soils on which pineapple plants, were -grow-
"and obtained 23.6, 25.7, and 27.8; correlating these figures with
,osture equivalent, which was also determined, he calculated the
which he found, to be 1.451 1.4, and 1.46-fig res much lower
i he-*Briggs and,.Shantz factor. ,Although tl factor 1.84, is
'roinmately applicable for an over-all average, the factor. appar-
yvaries considerably and must be determined experimentally
for* different Soil types. However, the critical. moisture point of
A& an"i soils may safely be said to be about 25 per .cent. This
agrees well with results obtained in tests with Porto Rico
Cuba sugareanes soils. These soils, as is known, are somewhat
similar to Hawaii Soils being red, laterites, and lateritic. Crawley,
a qutedby Bennett and Allison (4, P. 307) observed that certain
I- o Rico clay-land cane suffered where the soil contained 23.5
'-4T cent,. and the subsoil 24.3.per cent of moisture. He was also
of the opinion that where the moisture in Cuban red soil is
reduced below 25 per cent, cane will suffer."
RELATION OF ORGANIC AND INORGANIC PROPERTIES TO
ADSORPTION
ORGANIC PHASE
It has already been stated that in most cases the amount of organic
-matter presetaf i-the determining factor in the moisture relation-
ships in Hawaii sol.It has been suggested as a possibility that
all the organic matter does not exist in Ile clay or colloidal fraction
of ,a mechanical analysis, and that a part of the organic matter, on
account of the size of the particles, is present in the subdivisions of
silt. Logically, it, may be supposed that the soil organic matter
undergoing decomposition passes through different gradations in
particle size before assuming colloidal dimensions. In these investi-
gations all -particles smaller than 0.002 millimeter in equivalent
diameter axe considered as colloidal. With the realization that this
limit is an arbitrary- one, it is reasonable to suppose that the organic
matter shows adsorptive power before reaching this arbitrary limit.
it may be argued that organic material existing in this coarser state
Us simply colloidal aggregates resisting dispersion, but for practical
considerations they are particles of silt dimensions exerting their
power of adopinin tjmt subdivision. Further, the organic matter
is often sttdto exist chiefly as coating around particles of colloidal
dimensions, but it is reasonablet suppose that it also exists as coat-
Ing aroundl the. fine silt particles Moreover, coarser aorticles of






















1AjLUULjLJLbjAaA1 j CuiL&*LIUiAJ OLDJ. 5 WLPJ LP I J W JLJU L JC J. jSAm AUD &Dfl..mplUX ris .*j ip
tion of the colloidal and noncolloidal fractions in these sois
', i
TABLE 8.-Comparison of loss on ignition of colloidoa an4 noacoiW'W l
in HTawaii soils

Boil Colloids N
Soil n o. .
Loss on Loss on A
ignition Amo ignitionAmo

Per cent Perct Per cent Pt cent i.
6........................... ................ 10.0 66. 10.6 33.4
9----------------------------------------- 30.0 27.0 19.7 73.0
12.---.....-- ...................---------------..--- 32.8 15.0 43.9 85.0
31.........------- --------...... --- ---..--...---...-- 16.3 49.3 17.5 5i. 7
42 -------------------------------------. 25.0 32.5 24.6 67.,
48 ------.-------.--------------------- .- 13.0 30.1 15.3 69.9
65 -----------------------------------,-- ---9.4 35.1 9.0 64.9
66---------------------------------------------- 9.6 30.6 8.6 60.4
76 -----------------------.--------------- 14.3 25.8 19. 74.2
77 -------------------------------------- 18.9 28.1 18.8 71. 9

Table 8 shows that a considerable portion of the organic matEg
is often obtained in the noncolloidal fraction. In most instances 1i
loss on ignition of the nonoolloidal fraction is close to that oi
colloidal fraction and in some instances exceeds it. In the case;
soil No. 9, a highly organic type, a much higher percentage of orag
material is indicated in the noncolloidal fraction than in the colloidd
INORGANIC PHASE
When the percentage of the organic matter is not the determi-- i
factor in soil-moisture properties, such as water-vapor adsorpt ,
the texture of the soil partly determines that property. Struct~ r
character plays a role also, but under laboratory conditions
factor is eliminated. That the chemical composition of the ti
ganic phase plays a significant role has been indicated already whO
it was shown that two soils equally low in organic-matter cont
may have nearly the same percentage of clay fraction, and yet c'
soil may have more than three times as great a moisture equivalent
as the other (see p. 22). Of course, this phenomenon maybe note
not only in the case of moisture equivalents, but also with othh
moisture constants, such as hygroscopic moisture, or for that matte
with other physical properties, such as heat of wetting and adsorp
tion of ions. That one may expect such behavior may be seen front
the work of other investigators.






























oi easm of silica at least particles much larger than clay or colloidal
ns give considerable water-vapor adsorption. It may be
assumed that other substances behave similarly.
as quoted by Alexander (1, p. 451), showed that aluminum
Sde, manganese oxide, and ferric oxide gave widely differing
for heat evolved when wetted with water. The heat evolved
aluminum oxide was about ten times that of ferric oxide (1.16
i0.12 C. rise, respectively). According to Wollny, as quoted by
ngn (65, p. 62), the rise in temperature when quartz sand,
inum silicate, and hydrated ferric oxide were moistened with
tr was 0.10, 0.83, and 6.6, respectively.
. In view of these facts, it is easy to see that difference in chemical
eirnposition of the various fine fractions of any one soil considerably
fets their moisture-adsorption properties, especially where large
genes are involved.
ITo show that such large differences may exist, the silica-sesquioxide
io was calculated for the different fractions of nine typical
Hawaii soils. These figures are given in Table 9.

riu 9.-The iaO4/Fe.Os+AlsO. (mol) ratio in the different fractions of some
Ii; important Hawaii soil types

Soil fractions
Soil No. Coarse
I' Clay Fine silt Silt Fine sand Cose

--------------------- ------- 1.25 1.32 0.07 0. .-....---
--------17-2 1.88es 1.47 1.28 1. 0
--------------1.W 2.00 1. 8 2.54 231
----------------1.28 1.34 1.23 1.18 1.12
-- -------------------- ----1.08 .96 .66 .85 .64
,. .. ..... ............ ~ .62 1.41 1.67 2.01 1.78
----- ---------.-------------------33------- .63 .51 .7 1.13
------------------------------------------- 1.59 2.02 1. 1.67 1.
.---..------------ 1.17 .84 .2 .22 .30

i ires wee eainhared fram the data published by McGeorge (24, p.8).

ii-.


:..:.........;...















greatest difficulty is had in obtaining a perfect dispersion..: 'i
faculty is especially noticeable with certain highly organic
with soils the colloidal complex of which contains in the a
state certain bases, such as calcium, which have a flocculatingJ
upon the soil suspension. Several suggestions have been made i
past few years by various workers both as to the mechanical
to be employed and as to the chemical treatment the soil :
receive to insure as nearly perfect dispersion as possible.
light of recent studies on the exchangeable bases, much valuable
nation has been obtained regarding the r61e of various bases in
dispersion and flocculation of clays. For instance, it is knownit:
sodium clay is much more readily dispersed than ammoniuwr a
potassium clays. It would appear logical, therefore, to employ
alkaline solution of a sodium salt to obtain maximum dispersion. i
Hawaii soils contain a larger amount of-organic matter thn :4t
most soils of the mainland of the United States, and in certain ar"...
especially in the lowlands, highly calcareous soils exist. In addis
the soils are on the average heavier, containing upward of 50
cent clay and fine silt. In view of these facts, an investigation
made to learn what manner of mechanical dispersion.gives the
efficient results and what chemicals, if any are usdd, bring abtE1
complete deflocculation.
In the studies undertaken in the station laboratory, the work '",
dispersion was therefore divided into two parts, one comparing J.:":
ferent mechanical means and their relative efficiency when the cheE
cal treatment is constant, and the other comparing the efficiency fl::-il
various chemical treatments when the mechanical method is const '
In measuring the degree of dispersion, Bouyoucos' hydrometer :i
used: (1) Because under the same conditions (concentration al i
temperature) it gives comparable figures; (2) because mechanical
analysis performed with the aid of it compared favorably with otf -::: i-
standard methods; (3) because duplications carried out under idea-. :
tical conditions gave good agreements; and (4) because of it i'
rapidity. :.i
The five soils, varying widely in nature, selected for the work rep .
resented certain Hawaii soil types and were in most part typical
of the larger soil areas. A description of these soils is given (p. 5) :.i
MECHANICAL ASPECT .:
Three methods widely used in mechanical soil analysis were trie4lf:.
(1) Rubbing in a mortar with a rubber pestle under water contaii:i ,i:
ing alkali (KOH) for varying lengths of time; (2) shaking in -siA
end-over-end shaker in a machine revolving at the rate of about ol .
revolution per second; and (3) using an electrical stirrer. The first.
method has been used by certain investigators, notably Engre ani
































































Solid matter dispersed in-


0
minutes


Grama
7.4
8.5
5.7
5.4
0.2


15
minutes


Grami
20.2
21.7
11.5
14.2
23.7


30
minutes


Grams
21.3
21.6
13. 2
17.3
28.1


45
minutes


Grams
24.9
25.3
14.4
19.4
30.9


USING END-OVER-END SHAKER


Solid matter dispersed in-
E ....Soi N o.
0 hours 6 heurs 12 hours 18 house 24 hours


------------------------------------------- 7.4 S.5 0L.7 U31 32.7
-- .. ....--- ..... 7.4 30.5 30.7 33.1 32.7
---------------------------------------------- 8.5 37.5 38.7 41.1 41.7
---- -----_----_ ----------------. 5.7 19.0 20.2 21. 21.2
.................----- -------......... 4 24.0 25.7 24.6 25.2
...------.............- -. ---.. .-- &6.2 35.0 6.2 381 8.7


S ..::..: .. ... ... ..... .
r ie~i~ii~....; i :iiiiii:. ;..i:...ii........ iiiiiii.i.i.. i + ,SM-


SoA No.


60
minutes


Grams
27.3
27.3
15.0
20.5
31.9


e -----------------------------------------------
-n-----------------------------------------























The method of dispersing with an electrical stirrer is t
to be far superior to the methods of rubbing in mortar or
in an end-over-end shaker. Dispersing in an electricalstit
45 minutes was found to be equivalent in efficiency to sharing
end-over-end shaker for 24 hours. Rubbing in a mortar
least efficient dispersion of the three methods studied. In m
work, therefore, the electrical stirrer was used. The fig
show that maximum dispersion was obtained by this method
stirring for 60 minutes. The relative efficiency of these
devices is shown graphically in Figure 10,. which gives in ad&
to the individual dispersion curves, a composite average c

CHEMICAL ASPECT

Various chemicals tried, supplying a hydroxyl ion to stabihnim::*A^
dispersed colloidal particles, included ammonia (1 pe" cent soutin)
potassium hydroxide and sodium hydroxide (5 cubic centimVi: '.l
normal solution per liter of suspension), and sodium carbon.
(0.2 per cent). Table 11 gives the magnitude of dispersion obtuaii I
with the aid of these deflocculents. The mechanical means employ~
was always the same, that is, stirring in an electrical stirrer for :'
minutes. This method gave the maximum dispersion in the f::tiM
part of the study. Bouyoucos' hydrometer was used to determi
the degree of dispersion in the same manner as described in i tn::
foregoing.

TABLE 11.-Effect of various chemical treatments on the degree of dIsporsfa e
[Results expressed in grams per liter of solid matter in suspension after 15 mitnutes.. :i:
..... .:: ,,' ....
Solid matter in suspension when using-
Soil N o. .. : ki--,;_i i
H0O NHIOH KOH NaOH NasC ti;,

Grams Grams Grams Groams Gram.ll:
2-.....---.------------------------------ 34.0 27.9 34.0 36.4 It
6. -.----.--.------------------------ ---- 43.7 37.2 42.5 44.9 k.
37...-------------------.------------- 10.4 25.6 26.0 28.6 I
42.....---- ------.----------------------- 25.6 28.3 30.0 32. 8. &i .A:
48..s ..s. .. ... 4. ....... 38. 8 38. 4 39. 5 41. 3,,
48---------------------------------------------------38.8 38.4 39.5 I1.
H E:::: :'





F.











































4 0 r -- -- -- -- ......... .

ii *.,4 0- --- -T W Ti- ,,.,O.urs).
40 stirri-n! LJ


.::: h......... .


i6 /1 1/ 2 6 /Z /
6 .. *1 ,Minutes
o10 t2 JO 40 50 60 70 0 10 20 30 40 50 60 70 80
'"i n" a 1.---Graphical comparison of the efficiency of three mechanical dispersing
devices

t did not increase the dispersion obtained in water alone. In
case of soil No. 48 (acid), the use of ammonia did not increase
he dispersion over that obtained with water.3
:The relative efficiency of the various chemicals is shown graphi-
s1y in Figure 11 for the different soils, as well as the composite
iry. for the averages of all soils.
s.:The effect of increasing the concentration of a deflocculent
(potassium hydroxide) was studied on a typical Hawaii clay. The

*W:Sice these determinations were carried out, other investigators have shown that lithium
ieimaate causes an even better dispersion than does sodium carbonate.


MR;;I: :























Soi/ No.2


Soil No. 6


- S:
..: :..


15 minutes 15 minutes
FIGURE 11.-Graphical comparison of the efficiency of various chemicals in dispersion


TABLE 12.-Effect of increase in concentration of alkali (KOH) wpon degree-of
dispersion of a typical Hawaii clay

[Results expressed in grams per liter of solid matter in suspension]


Solid matter in suspension after settling for- i

Amount of KOH, cubic centimeters
0 min- 15 min- 30 min- 45 min- 60 min- 75 min-.
utes utes utes utes utes utes


Grams Grams Grams Grams Grams Grams
5 --- ------... ------.....-.. --....-.---. 40.3 28.3 26.3 25.3 24.3 2 .5
10...--.........----.............. ........ 39.9 27.4 26.4 25.4 24.9 23.4
15 -......--......---...... ----.-----.--. 39.5 28.0 26.5 25.5 24.0 23.6 f
20---..--------. ...-----....---..-.--..... 39.1 28.6 26.6 25. 1 24.6 23..
25---.-----------.............--.--..-- .. 39.7 28.2 26.7 25.2 24.2 22.7
..'--.--
"ii 4


U)


40

I..


S30


U)


* j
.. ..i
;:





























plate and kept there until the evolution of oxygen was no longer
Soti.ceable. During heating, the beaker was covered with a watch
glass to prevent loss by spattering. After the mixture cooled, 250
biic centimeters of acid solution was added, and the whole was
,stirred and allowed to stand overnight. During the day the mix-
tre' was stirred several times. The following morning the mixture
wiii.as poured into a porous filtering dish, and the acid was removed
.iy suction- Threfiltrate was tested for calcium with ammonium
o:: late solution. The soil was repeatedly leached with the acid
solution until the filtrate gave no test for calcium. Thereupon the
i: zcess acid was removed by washing with distilled water until the
.flttrate gave no test for chlorides. The soil was then transferred
Cto a dispersing cup; the alkali solution was added, and the soil was
diersed for one hour. After the usual settling period, the degree
of dispersion was read by means of the hydrometer.
Table 13 shows the effect of the treatment with hydrogen peroxide
I:and the removal of calcium by hydrochloric acid upon the degree of
dispersion. Column 2 gives the degree of dispersion when neither
ijth organic matter was affected nor the calcium removed, and the
: l was dispersed with the aid of potassium hydroxide alone.
olumn 8 gives the dispersion with treatment by hydrogen peroxide
i :ut without removal of the calcium. Columns 4 and 5 give the dis-
Spesion with the calcium removed by fifth-normal and twentieth-
orml hydrochloric acid, respectively, but without the hydrogen
i ieroxide treatment prior to the acid treatment. Column 6 shows
ithe degree of dispersion after both hydrogen peroxide and acid
Iieatments. Column 7 shows the percentage loss due to superoxol
i' treatment of the soil.












Solid matter in s t p ...t tNl

soil No. KOH Superoxol NIS HC N/H1 l i
on plu NWs plus i 3
alone KQHi KOH E
....... ... ....
Grams Grames 'ram l.a...: i lli
------------ 34.0 7 349 5.6
........................................ 42.5 43.3 44.6 ,* 0 44..04
37............................------------ 26.0 26.1 304 6 430L.4 i
42---.............-----.--------------...... 30 0.4 31 38.68 3.2
48.----....---.... ----.------------..... 39.5 41.2 41,1 41. 1 4
I Figures were corrected for loss due to superoxol treatment, assuming that all the les was bodp
fraction that stayed in suspension after thejsettling period.
In a study of Table 13, several points are noteworthy:
Treatment with hydrogen peroxide solution apparently do
increase the degree of dispersion even in the case of soils .co.
from 9 to 11 per cent organic matter. A special determinate
the effect of superoxol on a highly organic soil was made to
ment this table. Soil No. 9, containing 18.5 per cent organic ...
was treated in the manner described, and the dispersion was
mined. In this case the soil treated with hydrogen peroxide
dispersion of 30.5 per cent as compared with 29.5 per cent i
the treatment. According to these findings, the treatment of organs.
matter with hydrogen peroxide is not necessary except, perhaps i
soils containing upward of 15 per cent organic matter.
The use of acid to remove calcium is necessary to obtain
dispersion. On an average, 2.6 per cent more solid matter is
in suspension in the case. of acid-treated soils than in untreated.' ...
after the dispersion has been completed.
Twentieth-normal hydrochloric.acid was just as effective in bt:
removal of calcium as was the fifth normal. As a matter of f
of 21 soils treated in this manner, only in one instance-soil No. 7 .
was the twentieth-normal acid not sufficiently strong to remove
calcium. In this case fifth-normal acid was used.
These results are in good agreement with those obtained by othbi::r
investigators, notably Charlton (12, p. 9), and Puri and Aai::'il
(28, p. 15), working with tropical lateritic clays in India.::i
ESTIMATION OF COLLOIDS BY THE WATER-VAPOR ADSORPTION i
METHOD
It has long been realized that certain soil properties are due largely,
if not wholly, to the colloidal fraction of the soil. To such propertiesiii
belong heat of wetting, base exchange, and adsorption. This fa
suggests the possibility of estimating the amount of colloids by ...I
termining one of these properties in the soil and also in a representa-s34ii :
tive portion of the colloidal fraction of the soil that has been e ii.
traced. The ratio of the two figures obtained would give the:
percentage of colloidal material in the soil. 7
Several assumptions had to be made that such methods would yieli "
even approximately correct figures: (1) It had to be assumed that:;
the extracted colloid is representative of the whole colloidal complex i",!
.. : .i; iiii


S;i








































gram anquois were repeateidy snaken witn dcstllea water m an
i-rver-end shaker for four hours and allowed to settle according
iStoes' law for the length of time required to leave only particles
afler than 0.002 millimeter in suspension, The suspension was
as siphoned off, the collected fractions were mixed, and the col-
ds in suspension were passed through Pasteur-Chamberland filters.
is separated colloids were dried at room temperature and passed
rough a 60-mesh screen. A fairly large amount of soil colloids was
gained, asis-s1s6*n in Table 14.


Proportion
of total
colloids to
colloids
heated
(a/6)


Soil No.


three successive dispersions


Colloids
extracted
(a)


Total
colloids
(b)


Proportion
of total
colloids to
colloids
extracted
(alb)


S-...... 40a 82 5L5 79. 2 42--....-. 7.16 3256 22.0
E 1la6 6 .6 75.3 48---...-.. 10. 04 301 33 4
a- 2 27.0 0.1 49. 1---.- 2.76 2L3 13.O
3.6 15.0 24.2 56 -------- 22.76 48.0 47.4
.. 9.36 23.9 39.2 57 .- 23.24 49.1 47.3
S L88 15.5 121 65-.----------- 1.88 51 45.2
to o.i 49.3 10.5 66.------- 1214 30.6 30 7
S.......6 017.0 &.6 76..---.-- --, 9.94 25.8 L385
4L... 2 L 1 200 77- -......-..& 21 0a2


IThis method of procedure permitted presumably little change in
UsM adsorptivity of the soil colloids during treatment.
iAlthough all the soils, with the exception of soil No. 69, contained
4y a small amount of soluble salts, the samples for determination
wre washed with water to prevent the salts from influencing the


14.-Percentage of total colloids extracted by


.:


Colloids
eatracted
(a)


Total
colloids
(6)


.__


...... ... .. ..
.: ,:: ..: "*..... "".."
,...:: .. .. ... ....























their corresponding soils was always determined togt" i;r
desiccator was evacuated by a vacuum pump to about .284.i!i
mercury-about 50 millimeters absolute pressure--placed in
mostat which was regulated to 300 C., and kept there for s"si
At the end of that time adsorption was considered a complete
dishes were immediately covered, weighed, and dried at 1~5
18 hours, and then weighed again. Table 15 gives the adsorp
obtained and the percentage of colloids calculated therefrom
results are shown graphically in Figure 12.


OM COLLOI/P


2 6 1 /2 2/ 3/ 37 2 94Z 5 6 7 A
.SO/. /vO.
FIGURE 12.-Water vapor adsorbed by soil and extracted

TABLE 15.-Water vapor adsorption in Hawaii soils and colo
of dolloidal content


iV

I.
a :: ':;i: .....





(."- ... inwi .".il;:,,
^y:. "9 r::: ...... ::,:;;
..I.: ::::: ::: ::
.. ,E.
.:. ':.;1 i

colloid N *:'

." as' a ". ""
4d~s as a ::.c-
I ::''


Soil No.


2.----......
8 ---------
12-........
21------........
31..------
37...- --
42.------
48-------


Adsorp-
tion per
gram of
soil (a)


Gram
0. 133
.208
.216
.237
.178
.226
.186
.097


Adsorp-
tion per
gram of
colloid
(b)


Gram
0. 236
.271
.221
.240
.219
.208
.183
.122


Com-
puted
percent-
age of
colloids
in soil


-aX100


Per cent
56.4
76.7
97.7
98. 7
81.3
108. 6
101.6
79.5


Amount
of
colloids
by
pipette
method


Per cent
51.5
66. 6
15.0
23.9
49.3
23.1
32. 5
30.1


Soil No.







56....----..
57--------
65-------
66-------
69-------
76-------
Average. -


Adsorp-
tion per
gram of
soil (a)


Gram
0.093
.120
.296
.2586
.188
.148


.184


Adsorp-
tion per
gram of
coliold
(6)


Gram
0.124
.138
.331
.266
.260
.223


Com-
puted
percent-
age of
oolloids
in soil






Per centl
750O
87.0

2S.1
72.3
a'l.


----.iN
S .E. .. .:
,s i
as.";~


. '.


.217


A I I i ......... 1.































M~lrE e appIled LI MUMiS case LM al lMeU SoIlS sincitn UU LluULLiUauIIS II1
mie. were rather considerable. The percentage of colloids com-
ed from the adsorptivities gives excessive results, in two instances
h figure being above 100. Onl in two instances is there a sem-
kite of agreement with the colloids estimated in the mechanical

' dis cssingithese, very divergent results, certain relationships
iich were brought out earlier in this bulletin (pp. 23, 25) are
iled to the reader's attention. It was pointed out that the amount
Pd distribution of organic matter and the chemical composition of
e inorganic phase in the various soil fractions seem to be of such
Iatire n many Hawaii soils as to preclude the use of the water-
tpor adsorption method for computing the percentage of colloids.
A. figures obtained indicate that the above-referred-to assumptions
mae i connection with this method are not permissible for many
wai soil types containing high percentages of organic matter.
h figures given in Table 8 indicate that a large part of the organic
mater may be present in the silt, and even in the sand fractions.
e as separate particles or as aggregates resisting dispersion, ex-
iing high adsorptive properties. Furthermore, it is an estab-
hled fact that particles show colloidal properties, even to a lesser
flint perhaps, long before the arbitrary limit made for colloids
SEay is reached.
Mais (14, p. 876) finds that the dispersion of soils in the course of
.e.S.nidal analysis may be so incomplete that the silt fraction will
ntain 25 to 97 per cent colloids in the form of aggregates, and that
ean the sand fraction may contain 2 to 25 per cent colloids in the

...........

L ::m .:.: :















content, as determined by the water-vapor 'adsorption meth
quently runs 10 or 15 per cent in excess of the actual cday
"probably because of exceptional physical properties of
material.' This fact is all the more remarkable because t
fraction of mechanical analysis has an upper 'limit of.
millimeter diameter, whereas the upper limit of colloids, as
preted in the water-vapor adsorption method, is much
ably below 0.001 millimeter. Therefore, the percentage of
should be considerably less than that of clay. Furthermo |
organic matter of the Cuba soils referred to is very low in
son with Hawaii soils, most of them ranging, between 1 and .
cent. Much larger discrepancies may logically be expected
soils having two or three times this amount of organic matter. '
Steele (33, p. 31), referring to the colloids of Ohio soil pr.;.
remarks that "there was a fairly good correlation betweenM
Robinson, water-absorption, and heat-of-wetting methods, ex I
in soils high in organic matter."
Alway (, p. 246) believes that "it does not appear yet satis l
torily established that the ability of soils to absorb water vapor :iJs iir
reliable measure of their colloid content."
Joseph and Snow (20, p. 119), working with heavy Sudan s :i1i2!
state that-- *
There is no connection between the proportion of very fine material (a71
below 0.5p) and other important soil properties. The proportion of the .: :.
fine material determined in this way [viz, sedimentation] would not, there~ t
afford any indication of the colloidd" properties of the soil. .
The moisture-adsorbing capacity of two different fractions ofeli:
colloidal phase was determined in the case of two soils and compar:i::
with that of the soil from which they were isolated. One fracHi..
was that part of the colloidal phase which could be isolated fromt:ii.
suspension by means of supercentrifuging at the rate of 86,000
revolutions per minute, whereas the other part was that fra~cti.iii.
which could not be separated from suspension by this treatment, ':k:
had to be filtered through Pasteur-Chamberland filters and contcen-ei;;i.
treated on the surface of the filtering tubes. This latter fraction i :::
sometimes referred to as "ultra-clay." This ultra-clay amounted
sometimes to 1 to 1.5 per cent of the soil. Table 16 shows the hygP-:;
scopicity of these two fractions as compared with that of the parentt;
soil in the case of two highly organic, highly hygroscopic soils. ;.:i3
In both instances the fine fraction of the colloidal phase or ultra :-
clay showed less adsorption than the coarse-colloidal fraction, in&fdi g!|
eating the probable presence of the highly adsorptive organic mat&iia
rial in larger amounts in the coarse-colloidal fraction and in smallerA
amounts in the fine-colloidal fraction. Moreover, in the case of soi
No. 12 the adsorptivity of the original soil is higher than that of'
,..- -4!














Water vapor adsarbed per gram of

d ,i ". so N o. m tar

p, ..:.._ m__
matter _________ _,...
solc.. .-.Fine

...... racti
---- ----------. nas am s ea .
-----:::m .-. ---.: .... ----" ---- -- -- as. e. e a e

6 ........ionclusion, therefore, it seems inadvisable to attempt to trans-
* *pical prp"yw properties into particle size with soils of high organic
fl each as are found in humid tropical regions. The writer is
lA opinion that the determination of the various fractions in the
lnical analysis and the determination of the magnitude of a
ufint physical property, such as water-vapor adsorption car-
Ml "ut on a soil, will adequately describe the quantitative colloidal
tiesof that soil, even though a part of the colloidal.property
dti for some reason, to the fine silt fraction for instance. If to
me determinations is added the chemical analysis of the clay frac-
a, an approximate qualitative picture may be obtained of the frac-
a chiefly responsible for these colloidal properties. It appears,
ever, that with a highly heterogeneous substance, such as soil,
mesing a number of compounds both organic and inorganic in
isly varying proportions not only in different soils but sometimes
the diffent ions of the same soil, the expression of a certain
pal property in terms of particle size may sometimes become

EPARISON OF THE HYDROMETER AND A MODIFIED PIPETTE
METHOD OF MECHANICAL ANALYSIS
"ouyoucos (5, 6, 7) proposed the use of a hydrometer for meas-
g soil-suspension concentrations. In an early publication (6, p.
P) on the subject, he proposed a rapid method for the determina-
a of the sand, silt, and colloid fraction of soil. In this method the
rormeter reading is taken of the soil suspension after 1 and 15
~ntesof settling time, and from these readings the three fractions
i computed. The author pointed out that in this method the col-
4 fraction includes a part of the fine silt fraction. In his subse-
at. article (7, p. 986), Bouyoucos proposed the use of the hydrom-
~ for a detailed mechanical analysis in which any number of class
sons might be made. Hydrometer readings are taken at differ-
intervals according to Stokes' law of settling particles and
ai the readings the particular fraction in question is calculated.
*i method and its different modifications are too well known to
Describing here in detail, hence only a ~sum6 is given.


" .... .... .... .." ... ."
iii ::::i: :: ,2 : iii::;,, :, :.. ..; .. .. .-...... .......... ... .. ...... .
















percentage of. the fraction in question. The results by this
are said to check with those of the heat-of-wetting method of~
mining soil colloids.
In the study of Hawaii "soil colloids in the station labor'
this method was used as a routine analysis to classify a larger a
ber of soil samples taken from the different parts of the f
cipal islands. Because of.this fact it became desirable to d
(1) how closely the rapid method proposed by Bouyoucos (
sisting of two readings only) agrees with the. pipette methodW ,
(2) how closely a detailed mechanical analysis, as performed :ii
the aid of a hydrometer, compares with a pipette method of ana
In checking the rapid method of hydrometer analysis no
eration was taken of the rather wide variations in specific g.r ~ it
of Hawaii soils. The readings were taken always after sett
for 1 and 15 minutes, regardless of the specific gravity of the: po'
in question. That this sometimes introduces appreciable er r .i.e
known from the fact that the specific gravity of Hawaii soilf wa& .
found to range from 2.17 to 3.36. In the detailed hydrometer analy~i
sis, however, the correct settling period was calculated from thei:iiii
specific gravity of the soil and from the viscosity of water at' tE i-'"s
temperature of the measurement. The depth in the hydromriti.er *
method was considered as 32.5 centimeters, in accordance w','.ith"'i:
Bouyoucos' calculations, the Bouyoucos hydrometer cylinder used Q:
registering the average density of a soil-suspension column 8-2.:::i
centimeters high.
The method with which the hydrometer method was compared iiln
this investigation had the same principle as the original pipette
method of Robinson (SO, p. 311), that is, sampling a soil suspension .:
at a constant depth at calculated intervals and determining the con- .:i:S
centration. It differs from it in that sampling was not done by
means of a pipette, but by a stopcock fitted in the wall of the cylin'
der. In this method the samples were taken from a 100-milimeter ::i
depth, corresponding to about an 8-hour settling period for thei
clay fraction. The exact time of settling was calculated from ti.hei i
specific gravity of the soil and the temperature. Puri, Amin ( l'
p. 5), and Robinson (30, p. 318) show that closely agreeing figaur#l. ..
were obtained from sampling at different depths at the corre-ij|
spending calculated settling time. Temperature conditions were:
observed by placing the soil suspensions in an electrically controlle jdi
thermostat, regulated at 300 C. with an accuracy of plus or minus' ."4
one-half degree. The cylinders were kept in the thermostat at
this temperature throughout the experiment, the door being openedii:K
only to permit taking a reading or for making a sampling. i
I






























iiiiiancal analysis by Mechanical analysis by pipette method
....hydrometer

g ;.liB. Colloids Very fine Fine silt Bilt (0.01 Sand
<(0.0 o000 (0.005 to to 0.02 >(0.02
Colloids Silt Band <(0.002 o (005 to to 0.02 >(0.02
milliu 0.01 milli- milli- milli-
meter) metS) meter) meter) meter)
meter)

Per cet Prcent Pe Per ent Percent Per cent Per ce7n Per cen Per ent
...... 61. 17 b 2a 51.5 5 8 8.7 8.8 25.2
S 80 13 7 66. 6 7.4 6.4 3.9 15.7
---...... 30 40 30 27.0 16.0 10.5 12.5 34.0
....... 2 16 61 15.0 11.0 7.0 12.0 55.0
-..--.....25 17 58 9.5 8.6 .5.4 8.0 68.5
--------.. 19 21 60 11.8 9.0 2.5 7.2 69.5
..------------ --- 44 24 32 23.9 15.4 12.1 10.1 38.5
.........-----. 30 32 15. 5 12. 0 12.0 15.0 45.5
,-.----..... 72 17 11 49.4 15.6 11.6 6.9 16.5
---30 33 37 17.0 5.0 0 15.0 54.0
----------...39 17 44 23.1 16.2 10.7 9.1 40.9
------ ----. 5 17 30 32.5 16.8 8.7 8.0 34.0
r ------ -. 68 31 1 30.1 23.4 20.6 14.7 11.2
------ 3 22 46 21.3 7.2 6.0 9.8 55.7
I,-.4---8.. 76 21 3 48. 0 18. 2 11. 8 9.2 12. 8
-...-.------ 76 21 3 49.1 14.4 10.9 9.4 16.2
48 31 21 35.1 7.7 11.6 18.9 26.7
-.----------- 44 20 36 30.6 8.7 7.6 10.9 42.2
------------------.... 70 10 20 48.9 14.9 13.6 8.4 14.2
------------ 26 30 25.8 15.8 86 21.6 28.2
-------- 43 22 35 28.1 5.2 6.8 6.9 53.0
8m--.- 3. 2L.2 29.5 31.4 121 9.6 10.8 36.1


Table 17 shows a fair agreement between the two methods. Tak-
g the over-all averages it is seen that the colloids indicated by
Shydrometer method included all the very fine silt fraction and
the fine silt fraction of the pipette method. Since the sand
c n was obtained in both instances by subtraction, it is not
y comparable, yet the disparity exhibited is not great.
detailed hydrometer method was also compared with the
tt method. Oly the colloidal and the three silt fractions were
Srrmied the sand fraction having been obtained by subtraction.
data obtainedare given in Table 18.

ll Tis method of sampling was checked against the original method of sampling by
iham of a pipette, and the agreement between them was found to be so close as to permit
HraniDrg the results with considerable accuracy.
















Percent Per cona Pat cen Per eJ'N .
2 Hydrometer .....----...--.....---.......... 49.3 8.0 7.0 '7.0 :rI
SPipette.......----- ..----........... 51.5 5.8 87 B
6 Hydrometer-......----------..---- 63.3 12.0 6.0 0
6 Pipette------.. ---....... ------- 66.6 7.4 6.4 3.9
37 JHydrometer...........-----------. 23.3 12.0 10.0 8.0
SPipette--.--------------- ------ 23.1 16. 2 10.7 9.1
42 Hydrometer----------------------- 28.3 18.0 10.0 6.0
4 Pipette---...--- -------............. .32.6 16. 8 7 &8 '
48 JHydrometer-----. ...........------- 30.0 18.0 17.0 23.0
4Pipette----------------------- 30.1 23.4 20.6 14.7
SHydrometer- -. --------------- 59. 9 6------.--------.-.----.--.
4 Pipette------------------------ 59. ------------------

The figures given in Table 18 show that the agreement in thliu,
is much closer than in the case of the rapid hydrometer m
The over-all averages for the five soils, Nos. 2, 6 37, 42, and 48,q i
pare for the different fractions as follows: Colloids, 38.8 and 40.7pq
cent; very fine silt, 13.6 and 13.9 per cent; fine silt, 10 and 11
cent; silt, 9.8 and 8.9 per cent; and sand, 27.8 and 25.4 per cent. :In
each case the first figure given was by the hydrometer method a:t
the second by the pipette method,

SUMMARY..
Data relative to the origin, formation, and composition of tropictai
Hawaii soils are given.
Physical characteristics of 21 soils which were selected for the::ii
investigation are described.
Mechanical analyses of a few typical soils, with special referetm c
to the fine fraction, are given. Variations in apparent and real se;;
cific gravities are pointed out. Hygroscopic and capillary moistriil
moisture-holding capacity, and moisture equivalent of these soilss;
were determined and correlated with the texture and organic-matter. :
content. The amount and distribution of organic matter is a:,:!i
determining factor in moisture relationships.
A considerable portion of the organic material is often in the non-
colloidal fraction. With soils the organic content of which is'W
small, the textural conditions and the chemical composition are thej:
determining factors. Sometimes large differences exist in the silica
sesquioxide ratio of the different fractions in any one soil.
Experimental results on the dispersing qualities of typical Ha-::
waii soils are given. For mechanical study the electrical stirrer:
was found to be far superior to either of the methods of shaking or !
rubbing. For chemical treatment prior to dispersion it was found::i
that in most cases cold twentieth-normal hydrochloric acid is satis-"!
factory for the removal of calcium, that the destruction of organicqt,
matter with hydrogen peroxide is not necessary, and that sodmi:'um i
carbonate is the best agent to aid deflocculation. i
i
aE:::E~







































(6)-


17)


1928.


THE COLLOIDAL CONTENT OF SOInS. Soil Sci. 23: 319-330, illus.

MAKING MECHANICAL ANALYSES OF SOILS IN FIFTEEN MINUTES. Soil
SeL 25: 473-480.


S" .1928. THE HYDROMETER METHOD FOR MAKING A VERY DETAILED MECHANICAL
ANALYSIS OF SOILS. Soil Sci. 26: 233-238.
(8) BE.!;S, L. J.,fand MOLANE, J. W.
..: 1910. MOISTURN EQUIVALENT DETERMINATIONS AND THEIR APPLICATION.
Amer. Soc. Agron. Proc. 2: 138-147, illus.
0(gB) and SHANTZ, H. L.
1912. THE WILTING COrZYICIENT FOI DIFFERENT PLANTS AND ITS INDIRECT
IDraIN ATION. U. S. Dept. Agr., Bur. Plant Indus. Bul. 230,
83 p., illus.
10) Buansss, P. S.
1017. A STUDY OF THE PRINCIPAL PLANTATION SOIL TYPES AS FOUND ON THE
ISLAND .OF HAWAII. Hawaii. Sugar Planters' Assoc. Expt. Sta.
SBul. 45, 100 p., illus.
31) OCAnPus, J. M.
1917. LATIrr: ITS O eIGIN, STRUCTURE, AND MINERALS. Mining Mag.
[London] 17: 67-77, 120-128, 171-179, 220-229, illus.
CHABLTON, J.
- 1927. THE MECHANICAL ANALYSIS OF TROPICAL SOILS. Agr. Research Inst.,
Pusa, Bul. 172, 9 p.
B1a) us, W.
i 1915. LAVAS OW HAWAII AND THERB RELATIONS. U. S. Geol. Survey Prof.
S Paper 88, 97 p., illus.
) DAvis, R. 0. E.
1925. COLLOIDAL D rE NATION IN MECHANICAL ANALYSIS. Jour. Amer.
Soc. Agron. 17: 275-279.
---) and BmNxr r, H. H.
S 1927. GOUPING Or SOILS ON THU BASIS OF MECHANICAL ANALYSIS. U. S.
Dept. Agr. Ire. 419, 15 p.























illus. Honolulu.
(20) JOSEPH, A. F., and SNow, O. W.
1929. THE DISPERSION AND MECHANICAL ANALYSIS O2 OFBAVY.
SOmL. Jour. Agr. Sc. [England] 19: [106]-120, llMas.
(21) KELLEY, W. P., MCGEofGE, W. .[T.], and THOMPSON, A. B.
1915. THE SOILS OF THE HAWAIIAN ISLANDS. Hawaii Agr.
Bul. 40, 35 p. :.
(22) LANG, R.
1915. VEMsUOH EINERB XAKTEN KLASSmIKpATIo DB BOiDEN IN
SCHE UND OIDLOGISOHEB HINSIGHT. Internal. Mitt.
5: [312]-346, illus.
(23) LYON, T. L., and BUCKMAN, H. O.
1929. THE NATURE AND PROPERTY OF SOILnS; A COLLG. TEXT 0 .
OLOGY, Rev. ed., 428 p., illus. New York.
(24) MCGEOBGE, W. T.
1917. coMPOSIONrr oF HAWAIIAN SOIL PARTIoLES. Hawaii Agr.
Sta. Bul. 42, 12 p.
(25) MARTIN, F. J., and DOYNE. H. C. iV,
1927. LATERITE AND IATERITIO SOILS IN SIERA LEON. Jour. Ag
[England] 17: [580]-547, illus, I..
(26) MIDDLETON, H. E. i
1927. THE ADSORPTION OF WATER VAPOR BY SOILS AND SOIL C. 00LiLiI8.
First Internatl. Cong. Soil Sci. Proc. and Papers. 1: 45.
(27) MITSCHaELICH, E. A. -n
1923. BODENKUNDE FUR LANND UND EBrSTWIrtl. Aufl. 4, neubearb.,sM,, i p,,-
illus. Berlin. ':' .~
(28) PUB, A. N., and AMIN, B. M. .
1928. A COMPARATIVE STUDY OF THE METHODS OF PREPARATION 0W T :i
SOIL FOR MECHANICAL ANALYSIS, WITH A NOTE ON THE FP
METHOD. Agr. Research Inst., Pusa, Bul. 175, 15 p., illus. :::::
(29) RATHEa J. B.
1918. AN ACCURATh LOSS-ON-IGNITION METHOD FOR THE DEWm NATIQN Oi
ORGANIC MATTER IN SOILS. Jour. Indus. and Engin. Ciet..
10: 439-442, illus.
(30) ROBINSON, G, W. ... !n .
1922. A NEW METHOD FOB THE MECHANICAL ANALYSIS OF BOILS AND OT fI
DISPERSIONS. Jour. Agr. Sci. [England] 12: [306]-321, ia .t
(31) ROBINSON, W. 0., and HOLMas, R. S.
1924. THE CHEMICAL COMPOSITION OF SOIL COLLOIDS. U. S. Dept. Agr. B1l.
1311, 42 p.
(32) STEINKOENIG, L. A., and Far, W. H.
1917. VARIATION IN THE CHEMICAL COMPOSITION OF SOILS. U. S. PL
Agr. Bul. 551,.16 p., illus.
(33) SHAW, C. F.
1925. TWO UNUSUAL COLLOIDAL SOILS. Soil Sci. 20: 419-423.
(34) STEELm, [J. G.]
1929. VARIATION IN THE COLLOIDS IN OHIO SOIL P1ROILE. Ohio Agr.
Sta. Bul. 481: 31.
(35) WARINGTON, R.
1900. LECTURES ON SOME OF THE PHYSICAL PROPERTIES OF SOIL. 21
Oxford. V










































Volcano, Hlawaii......
.....doo- ... -... .
Pahala, Hawaii-------.......

.do .........

.--do--::
.....do............
---do------- -- --

.....do.................
Wohala, Hawaii-"""-
...--do........----....
.....do.....--.--.---...
...-.do............-----.
Waimesa, Hawaii--....


Fern forest.......
Volcano ash........
Sugarcane---------...........
Coffee-...............
.....do.-.............
.....do...............
..-...do...-----............
--....--do.............--
Barren----..............
Sugarcane...........
----do...............
Barren- ---------
Pasture ............
Pineapples-----.....
----....do...............-
Canna..............


4,000
1, 700
1,550
1, 550
1,475
1,475
1,500
1,400
600
600
300
1, 00
750
1,950
2,700


130
130
50
80
80
80
80
70
90
40
40
50
65
60
60
50


0-12
0-2
0-12
0-12
12-24
0-8
12-18
12-18
12-18
0-12
12-24
72-84
0-2
0-12
0-12
0-12


Sandy loam--.----
--do.............----
Clay---m-.........
----.do-..--..; --
Sandy clay loam..
Clay..............
Clay loam-........
Sandy clay loam..
Silt loam..........
Clay------..........
--.-do.-------.----
--....do----.........
Clay loam----......
Clay..........-----
Sandy clay loam..
Clay loam-...----.


Gray ...........
Light gray......
Reddish brown.
Brown-..........
.....do---..........
Grayish brown..
Brownish gray.
Light brown-..
Red.-----.---...
Light brown..-..
Brown..--.....-
Yellowish brown
Reddish brown.
Brown---.......
-....do-.......-...
-...-do...........


See botnotes at end of table.


17 -- .-
MeS......: -
M........

21-..-

3----
-.-..-...
M5--....-.
26-......
37-.......
24--

37--.-
Sl ....


31....

33---
89........
38........


O 49
0
5.27
13.36
8.37
12 45
8.33
6.30
4.61
7.12
5.81
5.67
0.63
8.15
8387
10.95


U. 7L
6.50
7.25
. 40
0.55
6.16
7.17.
6.58
. 81
5. 13
7.33
7.02
4. 75
6.43
7.40
7.00
6.70


L@. A
.1
0
11. 8

3.0
15.8
18.6

11.4
7.4
7.8
16. 2
17.2
7.4
11.8
15. 7


. saw

.128



.008AM



.140
.085

















Sol o. District and island Vegetation El va rai tm- Depth Soil class ICor icmtpH aIieul
ature tr adSl Ca



Feet Inches OF. Inches etcn etcn etcn
iiliiiiiiiiiiiiiiiiiliiiiiiiiiiiiiilllli;iiii j ii liiiiiiiii iilirrriiiiiir iliiiliiiiiiiiiliiiiiiiii" iili8ii~ iiiiiiiiA B ii ll iiiiliiiiiiii iii:iiiiii i iiliiii iiiiiiiliii
s s :i~ii i iI; iiiii;;i~ l ii~' i :I n iiii i::III::::II'iil~i;i: IiIIiiI ii=iiiiiiiiii:iiiiiiiiiiiiiiiiiliiiiiiiii lil i iilRilili iIii = ==i :i; ::i





4---- aimea, Hawaii---- Canna------------ 2,700 50 65 16.-20 Clay lo ---- -iiii5ii@i@ii@ii3i





35 ------Ahaloa Homesteads, Pasture.------------- 2, 000 90 72 0-12 Sandy clay loam ----d ----- 29566 48 6 3 2 .54 t
Hawaii. t
..36...............d o ---.............................................................d yl.................. d o0 0 0 9 0 72 1:-: S a n d y lo.m ....... -- d o -
37 ----- onolcaa, Hawaii... Sugarcane ----------- 1400 80 74 0-12 Clay. -------- Lgtbow- 15 63 89 4 1' .8
38 ----- ----do ----------------- -----do --------------- 1400 80 74 12-24 Sandy clay loamssrw ----- 25 58 47 6 6 2 --- 0
39 ------ ----o----- --------- ---- do----------- 00 74 0-12 Clay -----;rii::iibs:;s~:l.-------- -----d" .,,: -- o 4 0 62 01 1. 4 07







..40................do-----------------.-----....--------------...0.........2.....a.............................................. ...... ....... ii ii i
":;~iii@', @@@@@@ii @@@@@@@ i@ii@@@@@i@@i i







41------ phoehoe, Hawaii- Grass------------ 0 100 74 0-8 Cla ---------







42 ----- akalau, Hawaii.------ Sugarcane ----------- 300 200 73 0-12 ---- do-------Lgt elwsh 90 .0.22 3 1 3 .9

43 ----- ----do ---------------- ----- o -------------- 300 200 73 12-24 Sandy clay loam-- R d ihbo n 58 .0 2. 22
44----------do.... ------------- -----do ------------ 1 25 200 73 0-12 S andy clay ---- ro n------ 1.9 4.1 1..4.5 3 M4
45 --------..... --- d --------- ---do ------------ 1:25 200 73 12-24 Sandy clay loam- e ds bo n .9 5.5 1. 9 1 1 ---
46 ----- ilo, Hawaii ..... ------- ---do ------------ 1,10 180 73 60-72 Sandy loam ---- -- d ----- 6.5 54.73 4 60 .9
.;"' ~ ~ ~ :iiiiiiiiii""ii= iiii""si' ~~i;;



















47 --- -- -- --do. --------------------do -------------- 1,2350 180 73 48460 Silt loam --- -- --d -- -- 5 3 0.64.8 -1
48 ----- amakuapoko, Maui. Corn -------------- 2, 000 50 73 0-12 Clay --------.Bonihre.-.66.59.551.1.6 8
--- -- -- do .---- --- --- --- -- .. doe-------------- 2, 000 50 73 12-2 ---- do -- -- --.o-- --.. 4 5. 7 2..4 2 32 : 2
50 ----- angi, M5aui ---------- Sugarcane ---------- 1, 000 go 75 0-12 Sandy clay loam- rw -----67 --- 84 5 4 2 9
53 ~ ..4 1 ------ --- o ---------- Pasture ------------- 250 100 75 12-16 Clay. -------- -- o -----..2 .2 04 262.2 .4
54 ----- aiku, Maui.. ------ Pineapples ---------- 90 70 73 0-12 ---- .do. ------- Lih raih 360 66 SA0 3 .7 .

55 ~ ~ .d ------- --- do -- ------- ---- do -------------- 900 70 73 12-20 ----- do -------- R d ih r w 6.3i.0.. 0 13 5 .4
-- ---do --- -- ------ --- do ------------- 1,250 so 73 0-12 ---- do. -- -- Lih r w 0 47..83 2 6 8
7--------do ------ do -- 1,250 80 73 12-20 do ------













----------do ----- Abandoned pineap- 1, 200 80 73 0-12 ----- do -
ple field.
59 ----- ----do.. --- -- ------ ---- .. do ------------- 1, 200 80 73 12-18 ----do --- -- ak r wn2 98 : .0...72.7 0
60 ----- amakuapoko, Maui- Pineapples .---------- 2, 400 50 73 0>-12 do.. 21 310.001
61 ----- ----do ---------------- -----do.. ------------ 2, 400 50 73 24-28 Sandy elay loam _D r bo n 45 6.3 1 8 51 Z20V
ft ------ aiako, Maui ........ Diversified ---------- 2, 600 65 73 0-12 -----do ------- rw ------ .4..72 1..5.7 23 .8
63 --- ---do.. --- --- --- --- -----do -------------- 2 600 65 738 12-22 Clay -- -- -- o-- -- .5..6.6..1 25 3 ,,. 7
-------- Maul Agricultural Co., Barren-------------- 000 60 7 0-12 -----do -
Maui
65 ----- onokohau, M aui. --- ----do -------------- 100 40 77 Subsoil. -----do ------ elo ------ .0..4.155 2
66----- ---do --- ---------do --------- 100 40 77 M any feet. --- Lih r 4d..-.
67.. ----- do ----------------do ---------- g 40 77 eep Yelow
69 ---- 1-IA--iatn -- ---- d -- ------- 0 5 770-1














































































































.iiis .









* ,,, :: iiiiii;: i































It










'* j
. t ...
















I


... .... ...;







.........
.;.::fl
"*.y






.iii i









o i::l
....a i




























































































































P//i: il:,, ;,


















































I I








":ii
-9-l














:il
i:



























ii
I-
:ii3
... a qi.:ii

..Ii:ii









UNIVERSITY OF FLORIDA
IIlllli llH IIIll
3 1262 08929 0677




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EOF2JDZ0F_MPQ9TZ INGEST_TIME 2013-11-02T01:52:16Z PACKAGE AA00014569_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES