Growth of Present-Day Concepts, up to 1977
Klovan 1966) methods. None of these methods appears to be completely satisfactory for discriminating the various depositional environments, although
the method of factor analysis (Klovan 1966) and the
system of graphic analysis proposed byVisher (1965,
1969a) appear to be the most consistent. Most of this
work has been strictly empirical in nature.
Attempts to relate grain-size distribution to sediment tranport mechanics were made by Middleton
(1976) and Sagoe and Visher (1977). Middleton
(1976) showed that the size break between the traction and intermittent suspension subpopulations in
a river bed depends on the dominant shear velocity
of the flow, a fact of some potential paleohydraulic
significance.
Methods for the study of particle roundness,
sphericity, and shape (a more comprehensive description than sphericity) were provided by
Krumbein (1941) and Sneed and Folk (1958). More
recent developments were summarized by Pettijohn
et al. (1972). Roundness increases downstream due
to abrasion; mean and maximum grain size decrease
in the same direction. These parameters are therefore- useful indicators of paleoslope and have been
used in many studies of fluvial sediments too numerous to mention (see Pettijohn 1962, for an early
review). However, their use demands much field
measurement and data processing, and, as paleoslope indicators, they are less flexible tools than
paleocurrent analysis. An additional problem is that
of inherited textural characteristics - sand which has
been recycled from earlier deposits may retain size
and shape characteristics which obscure those imprinted by the later environment. For these reasons,
grain size and shape studies largely went out of favor
during the 1980s, and further developments are not
dealt with in detail in this book.
2.3.5 Bedforms and Paleocurrents
It was realized early in sedimentological studies
that the morphology of individual hydrodynamic
sedimentary structures, and their assemblage and
relative arrangement, might yield clues as to the
depositional environment under which they were
formed, and that their scale was somehow related to
the energy level during deposition. Some of the conclusions of Lyell, Sorby, and others along these lines
have been quoted above. Other early work included
that of Spurr (1894a,b) who discussed the strength
and constancy (directional consistency) of currents
required to produce ripple marks, and Darwin
19
(1883) who demonstrated the existence of separation
eddies in the lee of ripple crests (Fig. 2.5) using a
rotating flume. Cornish (1899) observed the formation of antidunes in natural systems and Hunt (1882)
studied the growth of ripples on tidal flats. Hobbs
(1906) observed cross-bedding with a consistent dip
orientation and used this as part of his evidence for
interpreting the enclosing strata as "torrential" in
origin (see Sect. 2.3.6).
Sorby carried out experimental work on the "effects of current on sand". He explained (in his major
1908 paper, published shortly after his death):
"Fifty-nine years ago, when I was living at Woodbourne, a
country-house on the eastside of Sheffield, there was at the
bottom of the small park a brook entirely under my control. In order to investigate a number of questions, I constructed a place for experiment with some self.registering
appliances. I could easily regulate and measure the depth
and velocity of the current within certain limits."
The work carried out here and elsewhere, plus
careful observations of sedimentary structures in
ancient rocks, led Sorby to divide sandstone into
four types:
"1. Thinly or thickly bedded rock, without ripples or drift
bedding� and showing little or no graining of the surface in the line of the current ... This could be explained
by supposing that the water was at considerable depth,
and the material mainly deposited from above, not
drifted along the bottom where the velocity of the current was much less than 6 inches per second.
2. Thinly bedded rock, with well marked graining in the
surface in the line of the current, indicating a mean
velocity up to about 6 inches per second, but showing
few or no ripple marks.
3. More or less thick masses of rock almost entirely made
up of ripple drift. This must have been when the
velocity of the current was something like a foot per
second ...
4. What I have called drift bedding in numerous published papers ... The velocity of the current is indicated
by the nature of the sand; and probably further experiments would enable us to learn the approximate depth,
which was probably small, since an increase of a very
few feet made so great a difference in the strength of the
current.�'
As pointed out by Allen (1963b), sedimentary
structures in types 2 to 4 would now be termed
primary current lineation, small-scale cross-lamination, and large-scale cross-stratification, respectively. The order in which Sorby described the
sandstone types was one of increasing flow velocitywhich was the beginning of the flow regime idea,
although, as Allen (1963b) stated, Sorby was mistaken in that type 2 is now known to form at higher
Klovan 1966) methods. None of these methods appears to be completely satisfactory for discriminating the various depositional environments, although
the method of factor analysis (Klovan 1966) and the
system of graphic analysis proposed byVisher (1965,
1969a) appear to be the most consistent. Most of this
work has been strictly empirical in nature.
Attempts to relate grain-size distribution to sediment tranport mechanics were made by Middleton
(1976) and Sagoe and Visher (1977). Middleton
(1976) showed that the size break between the traction and intermittent suspension subpopulations in
a river bed depends on the dominant shear velocity
of the flow, a fact of some potential paleohydraulic
significance.
Methods for the study of particle roundness,
sphericity, and shape (a more comprehensive description than sphericity) were provided by
Krumbein (1941) and Sneed and Folk (1958). More
recent developments were summarized by Pettijohn
et al. (1972). Roundness increases downstream due
to abrasion; mean and maximum grain size decrease
in the same direction. These parameters are therefore- useful indicators of paleoslope and have been
used in many studies of fluvial sediments too numerous to mention (see Pettijohn 1962, for an early
review). However, their use demands much field
measurement and data processing, and, as paleoslope indicators, they are less flexible tools than
paleocurrent analysis. An additional problem is that
of inherited textural characteristics - sand which has
been recycled from earlier deposits may retain size
and shape characteristics which obscure those imprinted by the later environment. For these reasons,
grain size and shape studies largely went out of favor
during the 1980s, and further developments are not
dealt with in detail in this book.
2.3.5 Bedforms and Paleocurrents
It was realized early in sedimentological studies
that the morphology of individual hydrodynamic
sedimentary structures, and their assemblage and
relative arrangement, might yield clues as to the
depositional environment under which they were
formed, and that their scale was somehow related to
the energy level during deposition. Some of the conclusions of Lyell, Sorby, and others along these lines
have been quoted above. Other early work included
that of Spurr (1894a,b) who discussed the strength
and constancy (directional consistency) of currents
required to produce ripple marks, and Darwin
19
(1883) who demonstrated the existence of separation
eddies in the lee of ripple crests (Fig. 2.5) using a
rotating flume. Cornish (1899) observed the formation of antidunes in natural systems and Hunt (1882)
studied the growth of ripples on tidal flats. Hobbs
(1906) observed cross-bedding with a consistent dip
orientation and used this as part of his evidence for
interpreting the enclosing strata as "torrential" in
origin (see Sect. 2.3.6).
Sorby carried out experimental work on the "effects of current on sand". He explained (in his major
1908 paper, published shortly after his death):
"Fifty-nine years ago, when I was living at Woodbourne, a
country-house on the eastside of Sheffield, there was at the
bottom of the small park a brook entirely under my control. In order to investigate a number of questions, I constructed a place for experiment with some self.registering
appliances. I could easily regulate and measure the depth
and velocity of the current within certain limits."
The work carried out here and elsewhere, plus
careful observations of sedimentary structures in
ancient rocks, led Sorby to divide sandstone into
four types:
"1. Thinly or thickly bedded rock, without ripples or drift
bedding� and showing little or no graining of the surface in the line of the current ... This could be explained
by supposing that the water was at considerable depth,
and the material mainly deposited from above, not
drifted along the bottom where the velocity of the current was much less than 6 inches per second.
2. Thinly bedded rock, with well marked graining in the
surface in the line of the current, indicating a mean
velocity up to about 6 inches per second, but showing
few or no ripple marks.
3. More or less thick masses of rock almost entirely made
up of ripple drift. This must have been when the
velocity of the current was something like a foot per
second ...
4. What I have called drift bedding in numerous published papers ... The velocity of the current is indicated
by the nature of the sand; and probably further experiments would enable us to learn the approximate depth,
which was probably small, since an increase of a very
few feet made so great a difference in the strength of the
current.�'
As pointed out by Allen (1963b), sedimentary
structures in types 2 to 4 would now be termed
primary current lineation, small-scale cross-lamination, and large-scale cross-stratification, respectively. The order in which Sorby described the
sandstone types was one of increasing flow velocitywhich was the beginning of the flow regime idea,
although, as Allen (1963b) stated, Sorby was mistaken in that type 2 is now known to form at higher