Tải bản đầy đủ - 0trang
II. The Nature of Laterite
S. SIVARAJASINGHAM ET AL.
and tortuous tubular cavities either empty, filled, or partially filled with
a greyish-white clay passing into an ochreous, reddish and yellowish
brown dust; or with a lilac-tinted litheomargic earth. The sides of the
cavities are usually ferruginous and often of a deep brown or chocolate
color; though generally not more than a line or two in thickness, their
laminar structure may frequently be distinguished by the naked eye. . . .
The interior of the cavities has usually a smooth polished superficie, but
sometimes mammillary, and stalactiform on a minute scale . . . . The
surface masses of the softer kinds present a variegated appearance. The
clay and lithomarge exhibit lively colored patches of yellow, lilac, and
white, intersected by a network of red, purple, or brown. The softness
of this rock is such that it may be cut with a spade; hardening by
exposure to the sun and air, like the laterite of Malabar.” (Omissions are
by the present authors.) Vesicular laterite may be soft or of varying
hardness and commonly has earthy material in the cavities. It usually
occurs near the surface.
Cellular slag-like laterite is a scoriaceous mass. The many empty
cavities are separated by ferruginow material similar in appearance to
that which separates the earthy substance in vesicular laterite. Cellular
laterite is usually dark colored and may have a dull or lustrous surface.
It is of varying hardness and is brittle, being usually easily shattered
when struck a sharp blow with a hammer. According to Fox (1936),
cellular laterite is formed by removal of kaolin and other earthy material
from the cavities in vesicular laterite when the latter is exposed to
erosion and leaching at the surface. Falconer (1911), from his observations in nothern Nigeria gave a similar explanation, though he avoided
using the term laterite for “surface ironstone.”
Nodular laterite consists of individual concretions, pisolites or other
crudely round masses, usually the size of a pea but commonly larger or
smaller; it is generally ferruginous. The nodules may occur as a superficial covering or as a component in one or more horizons in the soil,
varying in concentration from low or insignificant amounts to very high
amounts. The nodules vary in hardness; some can be readily cut by a
knife but most are hard and brittle.
When the nodules of a layer are cemented together, hard “pisolitic”
or “concrete-like” laterite is formed. It occurs mainly at or near the
surface. The individual nodules may either be joined directly to one
another or be discrete entities in a cementing matrix of similar, but
usually less ferruginous, material.
Recent studies by Alexander and Cady ( 1962) present enlightening
detail on the physical arrangement of discrete components. Though
various specimens exhibit a great variety of micromorphological features,
certain structures are common to many, but not necessarily all, varieties.
Commonly under magnification in thin sections, tiny bodies ranging from
perfect spheres to oblong rounded forms may be seen embedded in a
matrix of fine particles; the matrix may be either very dense or spongelike. The rounded bodies may be individual units or, commonly, may be
aggregates of smaller spherical units closely packed. Such rounded bodies
may be widely spaced or closely packed in the matrix. Their boundaries
may be smooth and definite or irregular and indefinite in various specimens. The matrix may be unorganized, may have a gridlike rectangular
or reticulate network of oriented material, or may be largely oriented. Oriented material commonly lines pores and may appear as skins on the nodules. Crystalline oriented material is common as pseudomorphs after primary minerals, as porefillings, and as discrete bodies ranging from barely
visible units to relatively large homogeneous masses. Rock structure may
be preserved or may be entirely absent. Quartz particles may be
included, and in some specimens weatherable minerals encased in a
protective covering of weathered material have been observed.
Materials identified in the field as laterite have a wide range of
chemical characteristics. A prominent feature common to all laterites,
nevertheless, is a high content of either iron or aluminum or both
relative to other constituents (Alexander et al., 1956). This is clearly
illustrated by the following analyses, which are thought to be typical
examples (Table I ) .
Bases are almost completely absent. Combined silica is generally
Chemical Composition ( % ) of Selected Laterites
H,O (loss on ignition)
Site: 1, Coolgardie, Australia (Simpson, 1912). 2, Satara, Bombay, India
(Warth and Warth, 1903). 3, Bagru Hill, Bihar, India (Fox, 1936). 4, Cheruvannur,
India; Buchanan’s original site (Fox, 1936). 5, Djougou, Dahomey; laterite on
granite (Alexander and Cady, 1962).
0 ND, not determined.
S. SIVAFUJASINGHAM ET AL.
low (sites 1, 2, and 3, Table I ) , but some varieties, such as the original
laterite of Buchanan (site 4, Table I ) , may have significant amounts.
This is probably largely in the form of kaolin, which has been found in
recent work by Alexander and Cady (1!362) to be the principal or only
identifiable silicate clay mineral in samples from Africa. Alumina may
be the principal sesquioxide (site 3, Table I ) , but more commonly iron
oxide (site 1, Table I ) or iron oxide and alumina together (sites 2, 4,
and 5, Table I ) are the major constituents. Combined water, determined
by loss on ignition, is appreciable but is generally higher in aluminous
than in ferruginous varieties, as is shown in Table I. Titanium is also
common in significant amounts in most varieties and may be a major
constituent (site 3, Table I ) . Vanadium and chromium are found, but
rarely in appreciable quantities.
Quartz may be absent or present in only limited amounts, but on
rocks high in quartz it is commonly a significant or major component, as
on the granite of site 5 of Table I, for which petrographic studies showed
that much of the total silica was contributed by quartz. Quartz is also
common in laterite over nonquartzose rock, where it appears to be
derived mainly from wind-blown or detrital material from outside sources.
Ten samples of detrital laterite from various parts of India had an average
of ,2074 quartz (Warth and Warth, 1903). Pendleton and Sharasuvana
(1912, p. 10) have emphasized that differences in amount of quartz
commonly contribute to major variation in SiO, among samples, even
in the same profile.
The impoverishment in combined silica and bases and concentration
of sesquioxides during weathering and laterite formation on a dolerite
is illustrated in Table 11. The “primary laterite” of Harrison is not to be
confused with ‘laterite” as used in this review. It was a weathered earthy
product that lay between the surficial hard laterite crust and the unweathered dolerite rock. Major differences in proportions of iron and
aluminum and in amount of combined water between weathered material
and laterite presumed to have formed in similar material are common, as
in Table 11, but it is rarely possible to be certain that the laterite crust has
indeed formed in material like that of the underlying weathered product.
No consistent relationship seems to exist between the relative amounts
of silica, iron, and alumina and the degree to which the physical properties of laterite are developed. The shortcoming of any chemical classification was shown by Fox (1936) from the analyses of laterite samples
from Buchanan’s original sites (site 4, Table I ) . These would have been
called ‘lateritic lithomarge” in Fermor’s ( 1911) classifkation because of
the high content of combined silica, though the material was vermicular
and was being quarried for building purposes.
The analyses considered so far refer to bulk samples of massive
laterite without distinction between segregated nodular material and the
matrix. The nodular material is, however, found either to be similar in
composition to the matrix or to contain less combined silica and more
Chemical Composition ( ”/o ) of Dolerite, “Primary Laterite,” and Associated Laterite
Ironstone at Eagle Mountain, British Guianaa
H,O (loss on ignition)
From Harrison (1933).
Selected Chemical Constituents (%) of Nodules of Laterite at 5 Sites and of the
Matrix in Which the Nodules Were Embedded at One of Them
a Site: 1, Natal; ferruginous nodules (Beater, 1940). 2, Welimada, Ceylon;
aluminous and silicious nodules (Joachim and Kandiah, 1941). 3, Peradeniya,
Ceylon; manganiferous nodules (Joachim and Kandiah, 1941). 4, Hambantota,
Ceylon; titaniferous nodules (Joachim and Kandiah, 1941), 5, Congo; soft nodules
and matrix (Alexander and Cady, 1962).
ferric oxide. Site 5 of Table 111 illustrates the latter in soft nodules of a
ground-water laterite of the Congo. Prescott and Pendleton (1952, p. 21)
believed that nodules usually contain less free alumina than the more
massive forms and that they are low in manganese.
S. SIVARAJASINGHAM ET
Nodules studied by Alexander et al. (1956) were high in sesquioxides
and low in silica. Other workers have, however, reported appreciable
contents of both quartz and combined silica (Joachim and Kandiah, 1941;
Waegemans, 1954). The data in Table 111 illustrate the wide range of
silica, alumina, iron, titanium, and manganese in nodules from different
places. The work of Bennett and Allison (1928) also reveals variations in
the composition of nodules in different soils.
C. M W ~ L O G I C ACHARAC~ERISTLCS
Chemical analysis alone is not sufficient to reveal the nature and
origin of laterite ( Harrison, 1910; Campbell, 1917). Laterites having
similar physical properties, such as hardness or morphology, may differ
greatly in chemical composition, and, conversely, laterites having similar
chemical compositions may have greatly different physical properties.
Petrographic studies of thin sections ( Harrison, 1910, 1933), adsorption
of dyes (Hardy and Rodrigues, 1939), differential thermal analysis
(Humbert, 1948; Bonifas, 1959), and X-ray analysis (Alexander et al.,
1956; Bonifas, 1959) have been used to supplement chemical determinations.
Free alumina is mostly in the form of gibbsite (A1203.3H20),as
boehmite ( A1203.H20), or as an amorphous hydrated form which has
been called cliachite and a variety of other names (Hanlon, 1944;
Palache d aE., 1944). Iron is found in the form of goethite (FeO-OH),
hematite ( FeeOa),and as amorphous oxides or unidentifiable coatings on
other minerals (Alexander et al., 1956). Free silica is mostly inherited
quartz (Alexander et al., 1956),though Harrison ( 1933, p. 40) reported
Photomicrographs illustrating features of weathering and laterite formation
A. Weathering diorite, North Carolina. Crossed Nicols. The lath-like forms are
gibbsite pseudomorphs after feldspar. Some dark areas are allophane and some are
€3. Soft laterite from granite, Nigeria. Crossed Nicols. The light areas of the
crystal aggregate (upper left) and of the filled channel (lower right) are gibbsite
formed upon weathering of kaolinite. Dark areas are iron-impregnated clays and
iron oxides, which are isotropic or have a very low birefringence.
C. Hard laterite from granite, Nigeria. Plain light. The dark areas are impregnated with iron by local redistribution from the light areas. The higher population of quartz grains (white areas) in the part that has lost iron indicates that these
parts have collapsed.
D. Hard laterite from granite, Kigeria. Crossed Nicols. The yellow parts are
crystalline goethite which forms a continuous network, especially on the walls of
the small channel at the left. The dark streak through the center is a former channel
filled with fine-grained hematite. White spots are quartz grains.
This Page Intentionally Left Blank
secondary quartz in laterites presumed to have been derived from basic
igneous rocks containing only small amounts of quartz. Chalcedony and
opaline forms have been observed also. Presence of colloidal silica has
been suspected. Besides quartz, other resistant minerals like magnetite,
zircon, sphene, anatase, and ilmenite have been found.
Combined silica is known to occur in kaolin and comparable silicate
clay minerals that can be identified but is present in amorphous or
subcrystalline forms as well. Montmorillonite and illite types of clays
have not been identified in significant amounts.
The distribution and form of constituents within the laterite has
special significance. Recent studies by the authors have shown that the
finely divided matrix is commonly largely unoriented and unidentifiable
by petrographic techniques, though identsable portions range from none
to most of the material. Grains and large patches of oriented aggregates
ranging from kaolin slightly stained with iron oxide through kaolin
highly impregnated with iron oxide to almost pure goethite or hematite
are among the more prominent identifiable constituents. Variable optical
density within the matrix is related to the degree of iron impregnation
(Plate 1, C and D). Hardness of the mass appears to be related to the
degree of crystalinity and continuity of the crystalline phase of the
impregnating iron, which is largeIy in the form of goethite. Tiny spherical
bodies comparable to incipient nodules of centripetal enrichment described by Bryan (1952) are largely unoriented earthy material having
a higher degree of iron impregnation than the surrounding matrix. Some
have films of crystalline goethite on the surface or as concentric shells
within the body. Other spherical bodies embedded in the matrix may
be concretionary or pisolitic forms of gibbsite, boehmite, goethite, or
hematite. Gridlike networks of oriented materials are found in the matrix
of many specimens and are composed principally of goethite (Alexander
et al., 1956) or hematite ( Sivarajasingham, 1961).
Gibbsite pseudomorphs after feldspar (Plate 1A) and goethite pseudomorphs after ferromagnesian minerals are common constituents. These
appear to be most abundant in young laterites whereas concretionary or
pisolitic nodular forms of the same minerals are more conspicuous in
older varieties (Alexander et al., 1956). Gibbstite and boehmite are
found commonly as fillings in cracks and pores, both in the matrix and
in nodules. Pores and cracks may also be lined with oriented kaolin
having varying degrees of iron impregnation, or even with oriented
goethite (Plate 1D). In hard laterite crusts, hematite may be found as
pore linings, as bands, or as discrete masses (Plate 1D). In varieties
derived from quartzose material, quartz grains are generally distributed
randomly through the matrix (Plate 1C) and earthy nodular bodies.
S. SIVARAJASINGHAM ET AL.
Quartz, apparently derived from outside sources, may also be present in
laterite over nonquartzose rock (Alexander et aZ., 1956).
111. The Environment of Laterite
Laterite is widely distributed in Africa, Australia, Southeast Asia, and
South America; Prescott and Pendleton (1952) have given a comprehensive review of its geographical distribution. The observed distribution
of laterite relative to present environment, however, is not necessarily a
criterion of conditions under which laterite forms (Hallsworth and
Costin, 1953, p. 25; Pendleton, 1936, p. 107).
Pendleton (1941) considered that an effective rainfall capable of
supporting a forest is necessary and that laterite probably never develops
in a climate that would support a savannah-type climax vegetation.
Maignien (1958, p. 15) concluded that the Sudano-Guinea climate of
80 or more inches of rain per year with 2 to 4 months relatively dry is
optimum for mobilization, accumulation, and induration of iron in
laterite. Humbert ( 1948, pp. 281-282) concluded that continuously wet
conditions do not favor laterite formation. The occurrence of a period of
drought, which is observed in many areas where laterite occurs, was
though to be a requisite by Maclaren (1906). This was disputed by
Scrivenor ( 1909), however, who declared that laterite occurs in Malacca,
where there is no such alternation of seasons. Campbell (1917) and
Humhert (1948, p. 282) considered that a regular alternation of wet
and dry seasons is not necessary if there are periods of wetness and
dryness, even though irregularly distributed. Mohr and van Baren (1954,
pp. 69-71) have suggested, on the basis of effective rainfall and evapotranspiration, that a monthly rainfall not exceeding 60 mm. would constitute an adequately “dry” season to provide no excess of rain over
evapotranspiration in tropical regions. Simpson ( 1912) thought that
laterite was forming under the semiarid conditions of western Australia.
It appears clear that some minimum amount of water is necessary for
weathering, the removal of bases and combined silica, and segregation
of iron so evident in the resultant material. It also appears reasonable
that periods of drying should favor the crystallization of goethite or
similar minerals, which appear to be associated with hardening. Observations by the authors suggest that approximately equal wet and dry
seasons favor crust formation and that some degree of alternating wet
and dry season is probably essential for this process. It is not clear whether
the material may or may not be conditioned for this final segregation
and crystallization in permanently wet conditions. To some degree,
conflicting views may be the result of two confounding factors: (1) the
occurrence of laterite under climates that are unlike those under which
the laterite formed, and ( 2 ) differential effectiveness of climatic wetness
or dryness in combination with differences of topography, parent material, and time.
Laterite is found where temperatures are warm or are believed to
have been warm at the time of formation. This does not preclude the
possibility, however, that time, not temperature, is the controlling
factor. If the principal effect of high temperature is to accelerate rates of
reactions, the time interval necessary for laterite formation in temperate
regions may exceed the period of time that appropriate landscapes have
been stable. Similarly, observations of Oldham ( 1893), Joachim and Kandiah (1935), and Prescott and Pendleton (1952) that laterite has not
formed under the cool temperatures of highlands in the tropics may be
due less to cool temperature than to landscape instability. Nevertheless,
field observation generally agree with the hypothesis that warm temperatures favor the formation of laterite.
B . VEGETATION
A forest vegetation was considered necessary for laterite formation by
Glinka (1927, pp. 33-34). More detailed observations have shown that,
while laterite occurs in regions with a rain-forest vegetation, welldeveloped laterite is most commonly found under a low forest and that
hard surficial laterite is a very common feature of the open savannah
adjacent to the forest. Maignien (1958, p. 16) concluded that laterite is
most extensive and strongly expressed at the boundary of forest and
savannah. DHoore (1954,1957, p. 55) found iron mobilized to a greater
degree and to greater depth under tropical grasses than under forest.
Buchanan’s type of soft laterite is found in Ceylon in the forested areas
of the southwestern lowland.
Humbert (1948) suggested that laterite forms in a climate that has a
wet and a dry season, and his descriptions and illustrations indicate that
the laterite he observed was in an open savannah that was gradually
replacing forest, a common condition where a dry season is prominent.
The change from soft laterite to the hard form within a few years after
man has cleared the forests has been reported by Alexander and Cady
( 1962) in French Guinea and other parts of West Africa. Blackie’s (1949)
description of soils of Fiji and Aubert’s (1950) description of Dahomey
also confirm the hardening of laterite following a change from forest to
savannah caused by man’s activities. It would appear from the literature
that laterite is most extensive in areas of savannah but that laterite forms
under forest though its hardening is favored by lack of forest cover.