as the sabkhas of the Persian Gulf, or associated with the
heavy metal deposits in the deep, hot-brine basins of the
Red Sea (Miller et al., 1966), and in Clayton Playa,
Nevada (Moiola and Glover, 1965).
By the time the solution has been reduced to about 10 %
of the original volume, most of the calcium sulfate has
been deposited, and halite starts to separate along with
gypsum (or anhydrite). When the volume of the solution
is less than 5 % of the original, polyhalite takes the place
of gypsum or anhydrite. Polyhalite separates together with
halite until the solution is saturated with respect to
Mg-bearing sulfates free from Ca and K. If the earlyformed salts remain in contact with residual liquids at all
stages and react freely with them, the final products of
crystallization of the bittern salts are:
0
to 30
C: epsomite + carnallite + bischofite + halite +
anhydrite
13
to 17.5
C: hexahydrite + carnallite + halite +
anhydrite
17.5
to 110
C: carnallite + bischofite + halite +
anhydrite
Usiglio’s experiment reproduced most of the evaporite
mineral sequence found at Stassfurt.
To provide the theoretical or physical chemical explanation of the evaporite sequence at Stassfurt, the Dutch
chemist Jacobus van’t Hoff dedicated most of his career
in the late 1890s and early 1900s to the study of the mineral equilibria of the Stassfurt salts. This was the beginning of experimental petrology. Van’t Hoff’s calculation
produced a sequence of evaporites that was remarkably
similar to the Stassfurt sequence. The main differences
were with the potash salts. An informative historical summary of van’t Hoff’s work is provided by Eugster (1971),
and an extensive description of the Zechstein is provided
by Borchert and Muir (1964).
To explain the sedimentology of the sequence of minerals found at Stassfurt, Carl Ochsenius expanded the earlier model of Bischoff for the formation of evaporite
minerals by evaporation of seawater and proposed the
barred-basin model whereby a lagoon or seaway was separated from the open ocean by a bar. Ochsenius suggested
that the Gulf of Kara Bogaz (Kara-Bogaz-Gol), which is
separated from the Caspian Sea by a bar, might serve as
an analog for a barred basin but on a smaller scale, both
in terms of area and depth, than the large Phanerozoic
evaporite basins.
Permian Castile and Salado formations
Another modification of the Bischoff–Ochsenius bar theory was proposed by R. H. King (1947) to explain the
thick (1,300–2,000 ft; 400–600 m) anhydrite deposits,
with minor halite, in the late Permian (Ochoan; 255 ka)
Castile Formation in the Delaware Basin of west Texas
and southeastern New Mexico. The Delaware Basin is
an intercratonic basin with an area of 25,980 km
2 and
was probably about 500 m deep before evaporite
deposition (Kirkland, 2003). King proposed a “reflux
model” whereby some brine was returned (“refluxed”) to
the open ocean but a continuous supply of seawater kept
the salinity within the anhydrite stability field.
Adams (1944) showed that the “minor” halite in the
Castile was not minor in the deeper, eastern part of the
basin where it mostly occurs as thick beds of halite
interlaminated with anhydrite. He produced the first comprehensive cross section of the basin based on borehole
lithologies and defined the boundary between the Castile
and the overlying Salado Formation. In the eastern part
of the basin, halite makes up about one third of the thickness but represents only one fiftieth of the time of deposition (Anderson, 2011). In the shallower, uplifted western
part of the basin, halite beds have been dissolved resulting
in beds of dissolution breccias. During the late Permian
when the Castile was deposited, the Delaware Basin was
an embayment within a region of extreme aridity in the
equatorial western margin of the supercontinent Pangaea
surrounded by the well-known Permian reef complex
(Guadalupian Capitan Formation) with saline groundwater seeping into the basin from the south and west from
the giant Panthalassa ocean (Anderson and Dean, 1995;
Anderson, 2011).
The Castile anhydrite consists mostly of seasonal laminae couplets of dark, organic-stained calcite and lightercolored anhydrite, the typical “banded anhydrite” of most
of the Castile (Figure 1) (Anderson et al., 1972). Each
light–dark couplet is interpreted as an annual increment
of sedimentation (a varve) (Anderson and Kirkland,
1966; Anderson et al., 1972). The Castile is divided into
four anhydrite members, I–IV (from oldest to youngest),
and three halite members, I–III (from oldest to youngest)
(Anderson et al., 1972). Each halite member represents
the culmination of a major salinity cycle. The halite units
consist of cm-thick halite layers and thin anhydrite laminae. At the top of a salinity cycle, the typical calcitebanded anhydrite grades into nodular anhydrite
(Figure 1). Dean et al. (1975) illustrate a progression from
typical laminated anhydrite to nodular anhydrite but still
with laminae of calcite that may be highly distorted to
even resemble nodular “chicken wire” anhydrite that is
common in supratidal evaporates (discussed below). The
laminae are remarkably uniform in composition, thickness, and delicate detail over distances of at least 113 km
(Anderson et al., 1972). The calcite–anhydrite varve couplets are repeated with remarkable regularity with more
than 209,000 calcite–anhydrite and anhydrite–halite
varve couplets in 450 m of stratigraphic section
(Anderson, 2011). The evidence supports the deep-basin,
deep-water interpretation especially for the typical
“banded anhydrite” that represents the preponderance of
time of deposition of the Castile.
The calcite–anhydrite varves of the Castile are very
similar to the dolomite–anhydrite varves (jahresringe) of
the Permian Zechstein Formation of Germany that were
correlated for distances of as much as 390 km by
Anderson and Kirkland (1966) using measurement by
MARINE EVAPORITES
429
heavy metal deposits in the deep, hot-brine basins of the
Red Sea (Miller et al., 1966), and in Clayton Playa,
Nevada (Moiola and Glover, 1965).
By the time the solution has been reduced to about 10 %
of the original volume, most of the calcium sulfate has
been deposited, and halite starts to separate along with
gypsum (or anhydrite). When the volume of the solution
is less than 5 % of the original, polyhalite takes the place
of gypsum or anhydrite. Polyhalite separates together with
halite until the solution is saturated with respect to
Mg-bearing sulfates free from Ca and K. If the earlyformed salts remain in contact with residual liquids at all
stages and react freely with them, the final products of
crystallization of the bittern salts are:
0
to 30
C: epsomite + carnallite + bischofite + halite +
anhydrite
13
to 17.5
C: hexahydrite + carnallite + halite +
anhydrite
17.5
to 110
C: carnallite + bischofite + halite +
anhydrite
Usiglio’s experiment reproduced most of the evaporite
mineral sequence found at Stassfurt.
To provide the theoretical or physical chemical explanation of the evaporite sequence at Stassfurt, the Dutch
chemist Jacobus van’t Hoff dedicated most of his career
in the late 1890s and early 1900s to the study of the mineral equilibria of the Stassfurt salts. This was the beginning of experimental petrology. Van’t Hoff’s calculation
produced a sequence of evaporites that was remarkably
similar to the Stassfurt sequence. The main differences
were with the potash salts. An informative historical summary of van’t Hoff’s work is provided by Eugster (1971),
and an extensive description of the Zechstein is provided
by Borchert and Muir (1964).
To explain the sedimentology of the sequence of minerals found at Stassfurt, Carl Ochsenius expanded the earlier model of Bischoff for the formation of evaporite
minerals by evaporation of seawater and proposed the
barred-basin model whereby a lagoon or seaway was separated from the open ocean by a bar. Ochsenius suggested
that the Gulf of Kara Bogaz (Kara-Bogaz-Gol), which is
separated from the Caspian Sea by a bar, might serve as
an analog for a barred basin but on a smaller scale, both
in terms of area and depth, than the large Phanerozoic
evaporite basins.
Permian Castile and Salado formations
Another modification of the Bischoff–Ochsenius bar theory was proposed by R. H. King (1947) to explain the
thick (1,300–2,000 ft; 400–600 m) anhydrite deposits,
with minor halite, in the late Permian (Ochoan; 255 ka)
Castile Formation in the Delaware Basin of west Texas
and southeastern New Mexico. The Delaware Basin is
an intercratonic basin with an area of 25,980 km
2 and
was probably about 500 m deep before evaporite
deposition (Kirkland, 2003). King proposed a “reflux
model” whereby some brine was returned (“refluxed”) to
the open ocean but a continuous supply of seawater kept
the salinity within the anhydrite stability field.
Adams (1944) showed that the “minor” halite in the
Castile was not minor in the deeper, eastern part of the
basin where it mostly occurs as thick beds of halite
interlaminated with anhydrite. He produced the first comprehensive cross section of the basin based on borehole
lithologies and defined the boundary between the Castile
and the overlying Salado Formation. In the eastern part
of the basin, halite makes up about one third of the thickness but represents only one fiftieth of the time of deposition (Anderson, 2011). In the shallower, uplifted western
part of the basin, halite beds have been dissolved resulting
in beds of dissolution breccias. During the late Permian
when the Castile was deposited, the Delaware Basin was
an embayment within a region of extreme aridity in the
equatorial western margin of the supercontinent Pangaea
surrounded by the well-known Permian reef complex
(Guadalupian Capitan Formation) with saline groundwater seeping into the basin from the south and west from
the giant Panthalassa ocean (Anderson and Dean, 1995;
Anderson, 2011).
The Castile anhydrite consists mostly of seasonal laminae couplets of dark, organic-stained calcite and lightercolored anhydrite, the typical “banded anhydrite” of most
of the Castile (Figure 1) (Anderson et al., 1972). Each
light–dark couplet is interpreted as an annual increment
of sedimentation (a varve) (Anderson and Kirkland,
1966; Anderson et al., 1972). The Castile is divided into
four anhydrite members, I–IV (from oldest to youngest),
and three halite members, I–III (from oldest to youngest)
(Anderson et al., 1972). Each halite member represents
the culmination of a major salinity cycle. The halite units
consist of cm-thick halite layers and thin anhydrite laminae. At the top of a salinity cycle, the typical calcitebanded anhydrite grades into nodular anhydrite
(Figure 1). Dean et al. (1975) illustrate a progression from
typical laminated anhydrite to nodular anhydrite but still
with laminae of calcite that may be highly distorted to
even resemble nodular “chicken wire” anhydrite that is
common in supratidal evaporates (discussed below). The
laminae are remarkably uniform in composition, thickness, and delicate detail over distances of at least 113 km
(Anderson et al., 1972). The calcite–anhydrite varve couplets are repeated with remarkable regularity with more
than 209,000 calcite–anhydrite and anhydrite–halite
varve couplets in 450 m of stratigraphic section
(Anderson, 2011). The evidence supports the deep-basin,
deep-water interpretation especially for the typical
“banded anhydrite” that represents the preponderance of
time of deposition of the Castile.
The calcite–anhydrite varves of the Castile are very
similar to the dolomite–anhydrite varves (jahresringe) of
the Permian Zechstein Formation of Germany that were
correlated for distances of as much as 390 km by
Anderson and Kirkland (1966) using measurement by
MARINE EVAPORITES
429