-
Economic Geology Vol. 78, 1981, pp. 077-093
Copper Mineralization and Magmatic Hydrothermal Brines in the
Rio Pisco Section of the Peruvian Coastal Batholith
R. A. AGAR
Geological Survey Department, Ministry of Mines, P.O. Box R.W.
135, Lusaka, Zambia
Abstract
Cu-Fe-Mo mineralization in the Rio Pisco section of the Peruvian
Coastal batholith is spatially associated with the Linga superunit,
a suite of monzonitic rocks intruded into Albian volcanics.
Petrochemical studies of this superunit indicate emplacement of a
differentiation series at a subvolcanic level in the crust. The
Cu-Fe-Mo mineralization is located principally in the Albian
volcanic envelope and is essentially a low-grade porphyry copper
type, although the grade is enhanced where structural controls on
the movement of the ore-bearing fluids produced more sulfide-rich
vein- and manto-type deposits. Alteration patterns associated with
both the mineralization and the Linga superunit suggest a close,
predominantly magmatic control on the nature of the hydrothermal
fluids. Fluid inclusion studies of quartz from the Linga superunit
support this and indicate that emplacement of magmas and
mineralization took place at a depth of approximately 8 km. The
characteristics of the Linga porphyry copper are compared to those
of other such deposits and used to suggest a possible telescoping
of geometry of the Andean model of Lowell and Guilbert (1970).
Thus, in magmatic hydro- thermal systems like the Linga, the deeper
parts of the model are effectively brought nearer the surface.
Introduction
THIS paper describes three types of copper minera]- ization in
the Coastal batholith of southern Peru, from the Rio Pisco section
(Fig. 1). The three types of mineralization, porphyry, vein or
fissure filling, and manto, are all associated with a suite of
monzonitic rocks. This relationship, the nature of the deposits and
their hydrothermal alteration, as well as fluid inclu- sion
studies, are used to demonstrate a magmatic source for both the
metals and the hydrothermal fluids.
Geologic Setting The Peruvian Coastal batholith extends for
more
than 1,500 km from the border with Ecuador south- ward to the
Chilean border (Fig. 1). The batholith ranges in age from Upper
Cretaceous to Lower Ter- tiary and for most of its length intrudes
Mesozoic and Lower Tertiary rocks (Pitcher, 1978). The batholith
north of Lima has been studied extensively and the publications of
Pitcher (1978) and Cobbing et al. (in press) provide more than
adequate summaries. Re- cent work shows the batholith to be
variable along its length and to consist of three segments, the
Trujillo, Lima, and Arequipa (Fig. 1; Cobbing et al., 1977). Each
segment is characterized by its own particular intrusions which can
be grouped into superunits, and each superunit comprises a suite of
units which are closely related in space, time, chemistry and
petrol- ogy (Cobbing et al., 1977).
The Rio Pisco section is in the Arequipa segment
of the batholith. Although the Lima and Arequipa intrusions are
different, the Rio Pisco section shows the modes and levels of
emplacement of these two segments of the batholith to be the same
(Agar, 1978). Both were emplaced at subvolcanic levels and both
exhibit brittle fracturing, cauldron subsidence, and gaseous
entrainment phenomena (cf. Bussell et al., 1976).
The intrusions of the Arequipa segment as exposed in the Rio
Pisco section can be grouped into four superunits together with one
other unit of limited extent (Table 1). The gabbros are very
similar to those of the Lima segment, previously described by Regan
(1976), Bussell (1975), and Mullan and Bussell (1977). They are
variable texturally, mineralogica]ly, and chemically even on
outcrop scale. This is due to late- stage net veining by related
dioritic and tonalitic magmas accompanied by associated
hybridization and amphibolitization (Regan, 1976).
The Incahuasi superunit is a suite of diorites and granodiorites
and, like the gabbros, has no associated mineralization. The Linga
monzonitic rocks and the Tiabaya tonalite-adamellite superunit,
both have as- sociated mineralization (Fig. 2) though only the for-
mer is considered here. The youngest unit is the Char- acas
monzogranite which has only a limited outcrop and is of relatively
minor importance.
The envelope of the batholith in the Pisco valley consists of a
series of sedimentary and volcanic rocks ranging in age from
uppermost Jurassic to Lower Tertiary. Some gabbros are deformed
along with their
677
-
678 R.A. AGAR
Km 0 300 Km
RIO PISCO SECTION
XArcquipa ( /
/
FIG. 1. The location of the Rio Pisco section in the segmented
Peruvian Coastal batholith.
folded Albian envelope, whereas others clearly trun- cate late
Albian fold axes. The Lower Tertiary vol- canic rocks of the area
are intruded by outlying stocks of the Tiabaya adamellite but lie
unconformably over older members of the batholith. Thus, as in the
Lima segment, igneous activity commenced in the Albian and
continued at least into the early Tertiary (Wilson, 1975, Bussell
et al., 1976).
The Linga Superunit The Linga superunit is a suite of monzonitic
rocks
ranging from monzogabbro to monzogranite (Table 1, Fig. $) which
crop out together on the western margin of the batholith (Figs. 2
and 4). The rocks are intermediate to fine grained and show good
igneous textures. In the monzogabbro and monzodiorites, cu- mulose
textures are well preserved whereas grano- phyric textures
predominate in the monzogranitic units.
Early workers in the Arequipa segment of the
batholith considered the Linga superunit to be a suite of hybrid
rocks produced either by the assimilation of early gabbros by later
granitic magmas (Stewart, 1968) or by the potash metasomatism of
early gabbros by volatile fluids emanating from later granitic mag-
mas (Hudson, 1974). However, the presence of the above igneous
textures makes such an origin unlikely. Furthermore, petrologically
the modal abundances of hornblende and plagioclase, when plotted
against those for quartz, reveal a near linear regression which
corresponds to the order of emplacement and hints, in the case of
plagioclase, at the differentiation of three separate rhythms, the
Humay, Auquish, and Rinconada, each of which represents a discreet
batch of magma derived from a common differentiating parent (Fig.
5). A quartz-alkali feldspar-plagioclase triangular plot shows the
same rhythms (Fig. 13). While the major oxide chemistry of the
Linga su- perunit fails to reveal these separate rhythms, all show
a strong linear relationship when plotted on a Larsen variation
diagram lending further weight to the ar- gument for a primary
magmatic origin for the Linga superunit (Fig. 6).
It is important at this point to examine the nature and
distribution of quartz in the Linga superunit. In all units it
occurs graphically intergrown with ortho- clase, and only in rocks
which have undergone po- tassic or sericitic alteration is enlarged
or recrystal- lized quartz observed. Modal abundances of quartz
vary little within units and in the superunit as a whole increase
with decreasing age (Appendix I). These fac- tors, coupled with the
above-mentioned petrological and chemical trends, show the quartz
to be primary and related to the Linga magmas and not secondary or
hydrothermal in origin.
Each unit is emplaced in its own particular pluton. Contacts
are, in the main, steep, sharp and clear-cut. The main controls of
emplacement are cauldron sub- sidence, stoping and possibly uplift
of the roof. The collapsing of elongate blocks with subsequent em-
placement of magma around these blocks produced the characteristic
long, narrow plutons of the Humay rhythm (Fig. 4). The Auquish
rhythm, on the other hand, was emplaced around a central collapsed
block of gabbro to produce the Auquish ring complex (Fig. 4).
There is evidence in and around the Auquish ring complex of
gaseous entrainment in the form of col- lapse and intrusion
breccias (Fig. 11). This, together with the predominance of brittle
fracture phenomena during the emplacement of this superunit and the
intermediate- to fine-grained granophyric textures of its more acid
members, is taken to be indicative of
As used in this paper, rhythm refers to a series of
intrusions.
-
COPPER MINERALIZATION IN THE PERUVIAN COASTAL BATHOLITH 679
TABLE 1. Units of the Coastal Batholith in the Rio Piseo
Section
Youngest
Characas monzogranite Medium-grained, pink graphic granite with
less than 5 percent biotite and no other marlcs; plagioclase
(Ana0); quartz and perthitic orthoclase granophyric
Tiabaya superunit Tiabaya dikes Tiabaya adamellite Tiabaya
granodiorite Tiabaya tonalitc
All units except the dikes are coarse grained and leucocratic:
dikes are fine and darker; all are characterized by 8:1 prismatic
hornblendes and euhedral books of biotite; plagioclase (An_.);
orthoclase perthitic, quartz biastic; xenoliths are small, round,
and dioritic
Linga superunit Rinconada monzogranite Auquish monzogranite
Auquish prophyritic monzonite Auquish monzonite Humay monzonite
sheets Humay monzonite Humay quartz-monzodiorite Humay monzodiorite
Humay monzogabbro
Suite of variable grain size, medium to fine, characterized by
pale green laths of plagioclase (An), dark green anhedral
hornblende with pyroxene cores often distributed in clots, rare
biotite, low-quartz content (always less than 20%), and
salmon-pink, perthitic K- feldspars graphically intergrown with
quartz; volcanic xenoliths in swarms at contacts and isolated
dioritic xenoliths in the main
Incahuasi superunit Huaytara granodiorite Magocancha
granodiorite Incahuasi diorite
Medium to coarse grained with euhedral plagioclase (An) and
poikilitic hornblende and biotite; marlcs often occur in clots and
hornblende has pyroxene cores; orthoclase perthitic and quartz
biastic; xenoliths are usually small, dioritic, and occur in
trains
Patap superunit Gabbros A variety of olivine, pyroxene, and
hornblende gabbros; net veined and amphibolitized by
related diorites and tonalRes
Oldest
a subvolcanic or hypabyssal level of emplacement (cf. Bussell et
al., 1976). Copper-Iron-Molybdenum Mineralization in the
Rio Piseo Section
Cu-Fe-Mo mineralization in the Rio Pisco area has a clear
spatial relationship with the Linga superunit (Figs. 2 and 7),
although mines are almost universally located within the
superunit's volcanic envelope.
Three types of mineralization can be recognized; one of low
grade and two in which the grades are en- hanced by structural
controls on the migration of ore bearing fluids.
The low-grade ore is disseminated throughout the volcanic
envelope of each Linga unit with a geometry like that of a porphyry
copper deposit (Fig. 8A, type c). $ulfides of copper, iron, and
molybdenum are widespread close to the contact, but the grade is
gen-
-* - -- ' 2 Lingo Super- unit 2 Other Intrusive rocks K Volcanic
rocks Cu, Pb,Zn0Ag Mine workings Cu,Fe,Mo Nine workings
FIG. 2. The location of mine workings in the Rio Pisco
section.
-
680 lq. ,4. ,4G,41q
A P
Units of Hummy Rhythm ..... Trend of parent mmgmm Units of ^
uqulsh Rhythm ..... Trend of rhythm Rlnconada Ionzogranlte
FIc. $. A quartz (Q)-alkali feldspar (A)-plagioclase (P)-plot of
the Linga superunit. Numbers 1-5 refer to position of unit in em-
placement sequence of a particular rhythm (age 1-5, oldest to
youngest).
erally less than 0.2 percent Cu. Whenever the en- velope
consists of an earlier intrusion, the mineral- ization is
suppressed with only very narrow
chalcopyrite-pyrrhotite-molybdenite veinlets depos- ited along
joint surfaces.
Vein deposits located at a distance from the contact show
enhanced grade and are common throughout the volcanic envelope. The
veins are both parallel and perpendicular to the main Andean trend
and reflect the movement of hydrothermal fluids along joint and
fault planes. Grades of up to 8 percent copper can be found in
major veins, particularly in the heavily faulted Cinco Cruces
mining district (Fig. 7). Each vein shows a complex history of ore
deposition, often with as many as four or five episodes each
followed by disruption and brecciation of the ore and gangue
minerals from subsequent fault movements. This ep- isodic
mineralization is probably due to the successive intrusions of the
Linga units, which were genetically associated with resurgent
movements along the faults. Movement of the hydrothermal fluid and
the subse- quent deposition of ore occurred in a manner similar to
the "seismic pumping" mechanism of Sibson et al. (1972).
The principal ore minerals in these deposits are chalcopyrite,
pyrite, bornite, hematite, and molyb- denite, with quartz, calcite,
gypsum, azurite, and
malachite the principal gangue minerals. There is very little
supergene enrichment and mining opera- tions in the Cinco Cruces
area have now largely ceased. The only major mining operation in
this area, Mina Eliaria in Quebrada Rio Seco (Fig. 9), has now also
closed. Here, both of the above types of miner- alization can be
seen, but there is also a third type. A gabbro sill $00 m thick
intrudes volcanic rocks and both are in turn intruded by the
Rinconada Monzo- granite with a thin wedge of volcanic rocks
between the two intrusions (Fig. 9). Within this volcanic wedge,
close to the contact with the Rinconada mon- zogranite, is a
disseminated type of deposit which is typical of the area as a
whole. Mineralized veins ex- tend upward from this disseminated
deposit toward the gabbro sill, immediately beneath which are large
lenticular bodies of ore (Fig. t0). The gabbro is un- mineralized
except for minor traces of copper min- eralization along joint
planes, whereas in the volcanics above the sill, veins typical of
the area as a whole have been found (Hudson, 1968).
The volcanic rocks both above and below the sill are heavily
fractured and jointed; the gabbro is par- ticularly massive and
poorly jointed. The lenticular orebodies are located immediately
below the gabbro which acts as a ceiling for each lode (Fig. 10).
The gabbro is little altered and the ore-gabbro contact is sharp
and may be marked by a thin band of actinolite, although normally
massive chalcopyrite and pyrrho- rite are found immediately
adjacent to the gabbro. The centers of the lodes are always massive
chalco- pyrite and pyrrhotite but pass outward to zones where
actinolite and then pyrite become important (Fig. t0). The outer
margins of the lodes are marked by an assemblage of pyrite,
chalcopyrite, apatite, and ac- tinolite. Other minerals present
include bornite, mag- netite, and hematite. In addition to
actinolite and apatite, the gangue minerals include quartz and cal-
cite. The highest grades are found in the center of the lodes where
the ore may contain as much as 15 percent copper over distances of
up to 5 m.
The geometry of this deposit (i.e., both the zoning and
disposition of the ores) suggests that the hydro- thermal fluids
responsible for its deposition were ponded or trapped in the
heavily fractured, perme- able volcanic rocks by the overlying
massive, im- permeable gabbro sill. Philipps (1972) described a
similar situation where the impermeability of a thick mudstone unit
overlying heavily fractured gray- wackes caused the formation of
the mineralized fiats within the graywackes of the Van mine,
Montgomery, Mid-Wales.
The distribution of the mineralization in the area as a whole
might suggest an exhalative origin for the ores of each type of
deposit. For example, all min- eralization occurs in the Albian
volcanic envelope
-
COPPER MINERALIZA TION IN THE PER UVIAN COASTAL BA THOLITH
681
Drift Rinconodo Monzagranite
AUUISH RHYTHM
. A3 Monzagranite i:.:.,' A2 Porphyry fL A1 Ivlonzonite
H UbIAY RHYTHM
1H4 Ivlonzonit e Qtz Monza- H3 diorite !:.': H2 Ivlonzodiorit. e
m]- H1 Ivlonzogobbro
Pampa Cabezo De Toro
HUMAY
Cruces
tvvvv /VVVV
VV VVV
AUQUISH
VVVVV VVVVVV VVVVVV
VV
VVVVVV
Older Intruslw rocks F Volcanic rocks Pampa Chunchonga
0 5Kin
I J
IiI ' i
VVVVVV VVVVVV VVVVVV
VVVVVV 'VVVVVVV
FI(;. 4. The Linga superunit in the Rio Pisco section.
close to contacts with the Linga monzonites which may
conceivably have remobilized metals within this envelope during
emplacement.
Although there are no mines within the Linga su- perunit itself,
traces of copper mineralization occur along joint planes within the
various plutons of the superunit. Also, if the source of metals was
indeed within the volcanic envelope, one might expect to find
similar mineralization wherever the envelope is intruded by the
batholith. In the Rio Pisco area, this is not the case and only in
the immediate vicinity of the Linga superunit are any of the three
types of mineralization observed (Figs. 8 and 6).
Thus, in its simplest form, the Cu-Fe-Mo miner- alization in the
Rio Pisco area is a low-grade porphyry copper type associated with
the Linga superunit. The ore is disseminated principally in the
volcanic en-
velope. Ore grades were enhanced by resurgent movements along
faults coupled with episodic hy- drothermal activity and produced
the vein- and manto-type deposits, respectively, where structural
traps impeded the movement of the ore-bearing fluids and localized
the deposition of ores.
Hydrothermal Alteration Associated with the Linga Superunit
Associated with the above mineralization and with each unit of
the Linga superunit is a series of alter- ation zones which are
typical of those commonly as- sociated with porphyry coppers and
described by Lowell and Guilbert (1970) and by Sillitoe (1976).
Each unit of the Linga has its own particular series of alteration
zones within its own envelope, which may be volcanic, or an earlier
member of the Linga
-
682 l. A. AGAR
% a 25-
20- 15 - lO-
- 5
o-
b 60-
55-
50-
45- -o 35- o
: 30 - n 25-
20-
15
o
Quartz
Huma units Trend of parent magma Auqulsh units Trend of rhythm
Klnconada monzogranlte
FIG. 5. Modal abundances of hornblende (a) and plagioclase (b)
plotted against those of quartz of the Linga superunit.
superunit, or of the batholith as a whole. Three types of
alteration are found. They are potassic, sericitic, and propylitic,
and are usually arranged concentri- cally in zones around the
parent pluton with the po- tassic zone innermost and the propylitic
zone outer- most.
In the potassic zone, secondary biotite replaces hornblende and
other marlcs, occasionally in con- junction with chlorite;
K-feldspar replaces plagioclase and quartz crystals become
enlarged. Magnetite-il- menite intergrowths, common in fresh Linga
rocks, are altered to intergrowths of rutile-sphene-magne- rite,
not only in this zone but in the others also. Dis- seminated
chalcopyrite, pyrrhotite, and molybdenite are also present whenever
the envelope is volcanic. Where the envelope is intrusive (e.g., an
older Linga unit), the mineralization is suppressed and less evi-
dent, although the degree of alteration is apparently no different.
The secondary K-feldspar is much red- der in color than the
salmon-pink primary K-feldspar typical of fresh Linga rocks.
Therefore, the limits of this zone are reasonably recognizable in
the field.
The sericitic zone is dominated by sericite which
is secondary after both plagioclase and K-feldspar. The quartz
in the rock becomes enlarged, showing later secondary growth, and
chlorite replaces horn- blende and biotite. In terms of
mineralization, dis- seminated pyrite characterizes this zone,
although it also tends to be suppressed wherever the envelope is
not volcanic.
The sericite-quartz-chlorite assemblage of the ser- icitic zone
grades into the quartz-orthoclase-chlorite- epidote-calcite
assemblage of the propylitic zone. Ser- icitization of the
K-feldspar becomes negligible and that of plagioclase gives way
gradually to alteration to clinozoisite and calcite. Hornblende and
biotite are still replaced by chlorite, but epidote becomes more
common toward the propylitic zone. The quartz shows no apparent
change.
Quite frequently either the potassic or the sericitic zone is
missing. In the more basic Linga units, where the potash contents
are low (1-2%, Table 2), the as- sociated potassic zone is absent
or diminished. In con- trast, the widest zones of potassic
alteration are as- sociated with monzogranitic members of the
superunit in which the potash contents are much higher ($-4%, Table
2).
Alteration patterns in which the potassic zone is either absent
or diminished (Fig. 8B, type d) are con- sidered by Sillitoe (1976)
to be indicative of hydro- thermal systems in which the magmatic
contribution was minor compared to the meteoric. However, in this
case, there appears to be a direct relationship between the amount
of KeO in the parent magma and the extent of associated potassic
alteration (Table 2). Thus, it may be that the extent of potassic
alter- ation is related to the potash content of the parent
intrusion.
The sericitic alteration zone is also variable in ex- tent but
does not vary according to the chemistry of the parent magma.
Widths of only a few meters are typical and frequently the zone is
either absent or discontinuous. The restricted development of
sericitic alteration zones is common where the wall rock is of a
basic nature. This is due to the predominance of biotite over
sericite (Guilbert and Lowell, 1974), and may be the case here
where the wall rock is gabbroic. However, for the most part, the
wall rocks are neither gabbroic nor silica deficient. Sillitoe
(1976) believes that the absence of sericitic alteration (Fig. 8B,
type c) is more probably due to a relatively small meteoric
contribution to the hydrothermal fluids. Thus, the fluids in this
case would be largely of magmatic or- igin, a model which is
clearly more applicable here.
In detail, the pattern of alteration zones associated with a
particular intrusion can be complex. Fre- quently, more than one
Linga unit crops out in a small area and each has imposed its own
envelope of alteration zones. This may cause a great deal of
over-
-
COPPER MINERALIZATION IN THE PERUVIAN COASTAL BATHOLITH 685
0.12 0.10 0.08 0.06
18--
17--
16-
15--
14-
7-
6-
5--
4-
3-
2-
4-
3-
2-
1-
4-
11-
10-
9-
8-
7-
6-
5-
4-
S--
4-
3
2.
1-
0
-5
MnO 2
AI203
CaO
a20
Total Fe
MgO
I I I I o s lO 1s 20
Larsen Index
FIe. 6. Larsen variation diagrams for major elements in the
Linga superunit.
lap resulting in a mineralogically complex series of rocks from
the overprinting of alteration.
The tuffisites or hydrothermal breccias associated with the
Linga superunit (Fig. 11) are also typical of
porphyry coppers in general and have attendant min- eralization
and alteration. These tuffisitic bodies are located in and around
the units of the Auquish ring complex (Fig. 11). Usually, they
contain finely corn-
-
684 R.A. AGAR
i.::i Linga Super- unit. Older intrusiw rocks
Volcanic rocks Mine workings
Pampa Cabeza De Toro
Cinco
VVVVV VVVVVV
VV VV VVV VVVVVVVV
Cerros
VVVVVV
HUMAY
AUQUISH
Pampa
o 5 Km I I
FiG. 7. Cu-Fe-Mo mineralization mine workings in relation to the
Linga superunit.
minuted fragments of their host rock and are weakly mineralized
(mainly disseminated pyrite but minor amounts of chalcopyrite have
also been noted). Al- teration associated with these tuffisites is
mainly po- tassic, although argillic zones, in which plagioclase is
completely kaolinized, may be found locally.
Alteration patterns associated with the Linga su- perunit are
therefore generally typical of porphyry copper deposits, although
the variation.in the extents of both the potassic and sericitic
alteration zones shows a deviation from conventional models of por-
phyry coppers (Hollister, 1975). A greater than nor- mal magmatic
control of these phenomena is sug- gested by the positive
relationship between the development of potassic alteration and
potash content of the parent magma, and the restricted
development
of sericitic alteration regardless of both parent magma
chemistry and wall-rock composition.
Fluid Inclusion Studies
Quartz in all units of the Linga superunit contains three types
of fluid inclusions (Fig. 12). Type 3 in- clusions show a planar
relationship to one another and are considered to be secondary and
the result of an- nealing along cracks in the quartz (Yermakov,
1965; Roedder, 1967 and 1972). Types i and 2 both occur in
isolation and, as it is difficult to conceive of the annealing of a
crack to produce a single large inclu- sion, they are thought to be
primary (Roedder, 1967).
All three types of fluid inclusions may contain daughter
minerals (Fig. 12). These were identified wherever possible using
optical techniques (Table 3).
-
COPPER MINERALIZATION IN THE PERUVIAN COASTAL BATHOLITH 685
a b c
_J'"- Porp. hyritic Intrusive JEJOther Rocks rOCKS
'( Equigranular Intrusive Collapse Breccias rocks
'::..'. .{ Stock work g Disseminated Mineralizaion Alteration
Types
K KK Potassic sSs Sericitic pPp Propylitic F FF Fresh
(Affer Sitlilac 1976) FIG. 8. Intrusive situations and styles of
porphyry copper min-
eralization (A) and alteration (B).
Type I inclusions are primary, vapor rich and may have up to two
daughter minerals; halite and an opaque mineral. Type 2 inclusions
are also primary but have only a very small vapor bubble and may
have from four to eleven daughter minerals (Fig. 12). The number of
daughter minerals in these type 2 inclusions has a positive
relationship to the degree of differentiation of their respective
Linga units (Table 4). Type $ inclusions are secondary, also have a
small vapor bubble and have up to four daughter minerals (Fig.
12).
Microthermometric studies carried out on samples from each Linga
unit revealed that all type 1 and some type $ inclusions
homogenized completely; the former to a vapor and the latter to a
liquid. Type 2 and those type $ inclusions with opaque and bire-
fringent daughter minerals did not homogenize com- pletely due to
the presence of the above minerals which, being slow to dissolve,
would not have main- tained equilibrium with their surrounding
fluid. Sev- eral hours are normally needed for such minerals to
reach equilibrium (Yermakov, 1965). The results of the
microthermometric analyses are tabulated in Ap- pendix II.
Freezing data for type 1 inclusions revealed trapped fluid
salinities in the range from 8 to 15 equivalent weight percent NaC1
(Fig. 15). Similar data for two- phase type 3 inclusions gave
salinities of 14 to 20 equivalent weight percent NaCI. Heating data
for
type 2 and three-phase type $ inclusions gave salin- tries from
20 to 60 equivalent weight percent NaC1, although only very few of
the latter had salinities greater than 40 equivalent weight percent
NaC1 (Fig. 14).
The primary nature of type 1 and 2 fluid inclusions means that
both trapped fluids were present at the time of growth of the
quartz. The coexistence of two distinct saline fluids, one gaseous
and the other liquid, indicates that as the quartz crystallized,
boiling of its parent liquid was taking place (Roedder, 1972).
Knowing the salinities of both fluid phases makes it possible to
derive the vapor pressure at the time of boiling (Sourtrajan and
Kennedy, 1962). In the ease of the Linga superunit, the gaseous
phases have a maximum salinity of 15 equivalent weight percent NaC1
giving a minimum vapor pressure of 800 bars (Fig. 15). Similarly,
the minimum salinity of the liq- uid phase is 20 equivalent weight
percent NaCI giving a maximum vapor pressure of 900 bars (Fig.
15).
By definition, boiling can only take place when the vapor
pressure of a liquid is equal to the ambient pressure. In the ease
of the Linga magmas, the am- bient pressure may have been wholly or
in part due to the lithostatie pressure. Assuming that the ambient
pressure was entirely lithostatie and that a mean value of 0.28 kb
per km was the lithostatie pressure, the depth of boiling of the
Linga magmas would have been between 2.9 and 3.2 km. If the ambient
pressure at the time of boiling was only partially lithostatic,
then this depth range would be even shallower. Thus, boiling of the
Linga magmas as the quartz crystallized occurred at a maximum depth
of 3.2 km. The textural nature and distribution of quartz in the
Linga su- perunit point to a primary magmatic origin and deny
growth from secondary hydrothermal fluids (see above). The fluid
inclusion studies indicate that the late magmatic fluids from which
the quartz grew boiled during crystallization. This boiling
produced two distinct saline fluids. From the nature of the fluid
inclusion daughter minerals, the most saline of these fluids
contained both the alkali chlorides considered necessary for the
transport of ores (Barnes and Cza- manske, 1967) and the ores
themselves. Migration of this fluid out from each successive Linga
pluton would, on subsequent cooling and precipitation, have given
rise to porphyry-type mineralization and to fluids of lower
salinity such as those represented in the type 3 inclusions.
Summary and Conclusions The Cu-Fe-Mo mineralization of the Rio
Pisco area
is fundamentally a porphyry copper type which has been modified
locally to produce small deposits of a higher grade. A certain
amount of mineralization and attendant hydrothermal alteration took
place with the
-
686 R.A. AGAR
A
VVVVVV VVVVVVVV' VVVVVVVVVVV VVVVVVVVVVVV' VVVVVVVVVVV --
VVVVVVVVVVV
vMina liana v VVVV vvv
vv vv vv
vvvvv% vvvvNina S Martin vvvv VVVVVVVVVV VVVVVVVVVVV
VVVVVVVyVVV VVVVVVVVVVVV vvvvvv lVVVVVV VVVVVVVVVVVVVVV'
VVVVVVVVVVVVVVVV v Lomas Condor Grande vv
vv
vv
vv
vv
vv
vv
vv
vv
0 Km 1 I I
vvvvvvvvvvvvvvvvvvvvv Ouebrada fill VVVVVVVVVVVVVVVVVVVVVVV
VVVVVVVVVVVVVVVVVVVVVVV
V V VV V.&.&,&&.. V V VV V VV V V V V ' ," , ",
m, ,,-, , , V V V V V V V V&--AI&I&I &i &i
&i &i& V V V V v v/y&--,-,-,-,'--A'-- A- v v v
Gabbro
B Volcanic rocks : "'"'..'":. Ore
NW-SE
Section Across Lomas Condor Grande
FI;. 9. The geology of the Quebrada Rio Seco district.
intrusion of each successive Linga unit. Each unit appears to
have produced its own hydrothermal fluids to form their respective
deposits. An external source for the metals is discounted in view
of the presence of mineralization within the superunit itself and
the lack of similar mineralization wherever else the same volcanic
rocks are intruded by other members of the batholith.
From the distribution of alteration patterns asso- ciated with
both the Linga superunit and the min- eralization, the restricted
development of sericitic alteration is taken to indicate the
predominance of magmatic rather than meteoric fluids in the hydro-
thermal system at this level (Sillitoe, 1976). The pos- itive
relationship between the development of the potassic zone and the
K.O content of the parent in-
trusion further suggests a close magmatic control of the
hydrothermal fluids.
Both minerals and chloride salts are abundant in primary fluid
inclusions within quartz of all the Linga units. More importantly,
the positive relationship be- tween the number of daughter minerals
(hence the complexity of the fluid) within these inclusions and the
degree of differentiation of the Linga unit makes an external or
meteoric source impracticable. The idea of the hydrothermal fluids
being drawn into the pluton by convective circulation of ground
water as proposed by Phillips (1978) cannot be applied here.
In such a model, the hydrothermal fluids for each of the nine
Linga units would have been derived ini- tially from fresh ground
waters of more or less the same salinity. The observed increase in
salinity with
-
COPPER MINERALIZA TION IN THE PER UVIAN COASTAL BATHOLITH
687
T 5m
GABBRO
VOLCANIC ROCKS
'v V
;py-p) py-pyrr- oct
py-oct-py .. .... )y-op
vein deposits .py-
Disseminoted ,py- ))
Limd of )otossic
!
\'," RINCONADA 1,, MONZOGRANITE
FIG. 10. Schematic section through the Mina Eliana ore deposit,
Quebrada Rio Seco. Cpy, chalcopyrite; pyrr, pyrrhotite; py, pyrite;
act, actinolite; ap, apatite; mo, molybdenite; bo, bornitc.
time could have been brought about by the same meteoric fluid
leaching a greater amount of salts and metals from later
differentiates purely as a function of the greater concentration of
such ions in the more differentiated and fractionated rocks.
However, if
TABLE 2. Potassic Alteration Zones in Relation to the K20
Content of Linga Magmas
Linga unit
Maximum width of potassic K20 content
alteration zone of Linga unit
H, Humay monzogabbro 10 1.19% H2, Humay monzodiorite 100 2.38%
Ha, Humay quartz monzodiorite unknown 2.05% Ha, Humay monzonite 600
3.79% A, Auquish monzonite 200 2.07% A2, Auquish porphyritic
monzonite 900 4.08% As, Auquish monzogranite 1,500 unknown
Abbreviations: H = Humay rhythm, A = Auquish rhythm. Sub- script
numbers 1-4 refer to position of unit in emplacement se- quence of
a particular rhythm (age 1-5, oldest to youngest)
this were the case, the fluid inclusions preserved would not be
primary and their host rock would not be fresh and unaltered. The
increase in salinity might also be brought about if the same
hydrothermal fluids were in continual circulation throughout the
em- placement of the Linga superunit, thus undergoing boiling and
further concentration a total of nine times. In this instance,
however, one might reasonably expect to find secondary fluid
inclusions within the more basic Linga units to show a higher
salinity than the primary fluids trapped in the same rocks. This is
not the case. Thus, it is more likely that the observed
hydrothermal brines are directly of magmatic origin.
Furthermore, the primary igneous texture dis- played by quartz
in all the units, the uniformity of modal abundance of quartz in
each particular unit, the clear relationships of beth modal
abundance of quartz and silica content of the Linga units to time,
and the total lack of textural evidence for secondary growth of
quartz demonstrate conclusively a mag- matic origin for the Linga
quartz. It follows, there- fore, that the fluid from which the
quartz grew was
-
688 R.A. AGAR
Auquish Monzon ire.::, ..... R Gabbro :::.'.. '.... ' ":' '" '"
' '' '" ' "' "" M ine work ngs :')':':" '? Envelope ':. '.?.ii.
.? /-
TABLE . Daughter Minerals from Fluid Inclusions of the Linga
Superunit
Daughter mineral Physical and optical properties
Halite
Sylvite
Hematite
x ..:.: :i :i' x :::::i:
Calcite
Unknown A
FIc. 11. The Auquish rhythm and its associated breccias.
Unknown B
Unknown C
Cubic, isotropic, refractive index close to that of quartz
Cubic, isotropic, refractive index less than quartz, dissolves
at a lower temperature than halite
Hexagonal platelets, red to opaque; does not dissolve
Prismatic, colorless, refractive index greater than quartz,
birefringent, length slow in orientation of least birefringence,
does not dissolve
Rhombic, high relief, refractive index greater than quartz,
birefringent, did not dissolve
Small, amorphous, isotropic, did not dissolve
Narrow birefringent rod or speck, did not dissolve
Cubic opaque, probably a sulfide, did not dissolve
Unknown D Cubic opaque, smaller than C, did not dissolve
Three other unknowns are isotropic and occur as small specks;
two other unknowns occur as birefringent specks; and one unknown is
an amorphous opaque
magmatic. This fluid boiled during crystallization of the quartz
and gave rise to a low-salinity gaseous phase and a high-salinity
liquid phase. These mag- matically derived hydrothermal brines,
already ore- bearing, were, on migration out from the parent in-
trusions, responsible for the Linga mineralization.
The fluid inclusion studies confirm the subvolcanic level of
emplacement indicated by the textures of the superunit and its
associated brittle fracture and gas- eous entrainment phenomena. A
depth of around 13
km compares closely to that obtained by Roedder (1971) for the
porphyry deposit at Bingham. The al- teration halo (Fig. 16b) of
the Linga system, however, is more akin to that of deeper deposits
such as Wood- stock (Fig. 16a). Although depth determinations for
the Woodstock and neighboring deposits are unreli- able, they are
still considered to be representative of the deeper roots of
porphyry copper systems (Hollis- ter et al., 1974). It may be,
therefore, that magmatic hydrothermal systems in porphyry coppers
(such as
o 1O.m TYPE DISTRIBUTION NUMBER OF DAUG. HTE
MINERALS
I Gos_rch Omame$ In isolation 0 - 2 lmy mies In 5ion 4-11
5Ollnlty 5ecle5 Plon6r 0 - 4
FIG. 12. Fluid inclusions in the Linga superunit. V, vapor; L,
liquid; H, halite; S, sylvite; A, anhydrite; He, hematite; I,
ores.
-
COPPER MINERALIZATION IN THE PERUVIAN COASTAL BATHOLITH 689
4. Daughter Minerals in Primary Fluid Inclusions in Relation to
the Age of Respective Linga Units
Maximum number of daughter
Linga unit minerals
H, Humay monzogabbro 5 H2, Humay monzodiorite 6 Ha, Humay quartz
monzodiorite 6 Ha, Humay monzonite 8 Hs, Humay monzonite sheets 8
A, Auquish monzonite 8 A2, Auquish porphyritic monzonite 10 As,
Auquish monzogranite 11
Taken from ten inclusions per section with sections from eight
rocks per unit (i.e., 80 separate inclusions per unit)
the Linga deposit) produce a telescoping effect in their
patterns of mineralization and alteration, effec- tively bringing
the deeper parts of the porphyry cop- per system nearer to the
surface (Fig. 16).
Hollister (1975) defines two fundamental models of porphyry
copper deposits, each with its own dis- tinct petrography and
alteration halo. One such model, the diorite model is typical of
island-arc en- vironments and is characterized the by-product gold
and by the lack of sericitic alteration. The Andean model of Lowell
and Gullbert (1970) on the other hand is characterized by the
by-product molybdenum and the presence of all three alteration
zones. In spite of the restricted development of the sericitic zone
in the Linga porphyry copper deposit, it is, particularly with its
associated molybdenum, more akin to the Andean model of Lowell and
Gullbert which is now widely considered to be typical of
destructive con- tinental margins (Mitchell and Garson, 1976).
Acknowledgments This work is part of a larger research project
within
the Peruvian Coastal Batholith Research Programme
SOLUTION NaCI & SOLUTION
ICE & NaC 12H20 -30 ! I ! I I I I J I I 1
0 5 10 15 20 25 30 60 65 100 Wt% NaCI -
FIG. 13. Salinity ranges of two-phase inclusions from freezing
data. A-B, range of type 1 inclusions; C-D, range of type $ inclu-
sions (two-phase).
HO
FIc. 14. Salinities of type 2 and $ inclusions with two or more
daughter minerals from heating data. O, type 2 inclusions; I, type
$ inclusions with more than one daughter mineral; ..... , differ-
entiation trend of hydrothermal fluids.
carried out jointly by the University of Liverpool (England),
Institute of Geological Sciences (London), and Instituto Geologico
y Mineria (Peru). The re- search project itself was based on the
regional map- ping of the Coastal batholith in the Rio Pisco area
and was financed by the British Ministry for Overseas Development.
The author would like to thank the staff of Cobre Sur S. A. (Eliana
and Auquish mines) for their hospitality and permission to review
their
iaaa
t
500
Q_
liquid
....... Critical curv
" "..:"::..' ::::::::! 1 0 i I I I I I II I.'1 :f.'.,.v.v..
I
0.001 0.01 0.1 1 5 10 50 100
NaCI (Wt. %)
J-J Gas compositions, type 1 inclusions "':'"..:i Liquid .... 2
,,
FIG. 15. Coexisting gas and liquid compositions relative to
pres- sure (after Sourirajan and Kennedy, 1962).
-
\ \\ / // G,ngham (Peru) \ I /., ., V /
+ + ++ { Iwfoundlend) FRESH Cotheort I I ( Noine ) euATz
HONZONIT + '++ ariner +++++++ (Nova Scotia) +++++++
FIG. 16. Schematic representation of porphyry copper alteration
patterns relative to depth showing (a) meteoric hydrothermal
systems and (b) the Linga magmatic system.
workings and to publish the resulting ideas. Finally, thanks are
due to colleagues at the University of Liv- erpool and to Drs. R.
Prasad and A. Gunatilaka for their encouragement and their comments
on the manuscript. June 1, 1979; August 23, 1980
REFERENCES
Agar, R. A., 1978, The Peruvian Coastal batholith, its
monzonitic rocks and their related mineralization: Unpub. Ph.D.
thesis, Liv- erpool Univ., 261 p.
Barnes, H. L., and Czamanske, G. K., 1967, Solubilities and
trans- port of ore minerals, in Barnes, H. L., ed., Geochemistry of
hydrothermal ore deposits: New York, Holt, Rhinehart and Win- ston,
Inc., p. 286--388.
Bussell, M. A., 1975, The structural evolution of the Coastal
batho- lith in the provinces of Ancash and Lima, Peru: Unpub. Ph.D.
thesis, Liverpool Univ., 875 p.
Bussell, M. A., Pitcher, W. S., and Wilson, P. A., 1976, Ring
com- plexes of the Peruvian Coastal batholith; a long standing sub-
volcanic regime; Canadian Jour. Earth Sci., v. 13, p.
1020-1030.
Cobbing, E. J., Pitcher, W. S., and Taylor, W. P., 1977,
Segments and super units in the Coastal batholith of Peru: Jour.
Geology, v. 85, p. 625-631.
Cobbing, E. J., Baldock, J., McCourt, W. J., Pitcher, W. S.,
Taylor, W. P., and Wilson, J. J., in press, The geology of the west
cor- dillera of North Peru: London, Inst. Geol. Sci. Overseas Div.,
Mern.
Guilbert, J. M., and Lowell, J. D., 1974, Variations in zoning
pat- terns in porphyry ore deposits: CIM Trans., v. 77, p.
105-115.
Hollister, V. F., 1975, An apprisal of the nature and source of
porphyry copper deposits: Minerals Sci. Eng., v. 7, p. 225-233.
Hollister, V. F., Potter, R. R., and Barker, A. L., 1974,
Porphyry- type deposits of the Appalachian Orogen: ECON. GEOL., v.
69, p. 618-630.
Hudson, C., 1968, Mina Eliana (Rio Seco), estudio geologico:
Tesis Bach., Univ. Nac. de ingeneria, Lima, 15 p.
1974, Metallogenesis as related to crustal evolution in south-
west and central Peru: Unpub. Ph.D. thesis, Liverpool Univ., 484
p.
Lowell, J. D., and Guilbert, J. M., 1970, Lateral and vertical
al- teration-mineralization zoning in porphyry ore deposits: ECON.
GEOL., v. 65, p. 373-408.
Mitchell, A. H. G., and Garson, N. S., 1976, Mineralization at
plate boundaries: Minerals Sci. Eng., v. 8, p. 129-169.
Mullan, H. S., and Bussell, M. A., 1977, The basic rock series
in batholithic associations: Geology Mag., v. 114, p. 337-359.
Phillips, W. J., 1972, Hydraulic fracturing and mineralization:
Geol. Soc. London Jour., v. 128, p. 337-359.
-- 1973, Mechanical effects of retrograde boiling and its prob-
able importance in the formation of some porphyry ore deposits:
Inst. Mining Metallurgy Trans., v. 821, sec. B, p. B90-B98.
Pitcher, W. S., 1978, The anatomy of a batholith: Geol. Soc.
London Jour., v. 135, p. 157-180.
Regan, P. F., 1976, Mafic plutonic rocks of the Coastal Andean
batholith, Peru: Unpub. Ph.D. thesis, Liverpool Univ., 352 p.
Roedrier, E, 1967, Fluid inclusions as samples of ore fluids, in
Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New
York, Holt, Rhinehart and Winston, Inc., p. 515-574.
-- 1971, Fluid inclusion studies on the porphyry-type ore de-
posits at Bingham, Utah, Butte, Montana, and Climax, Colorado:
ECON. GEOL., v. 66, p. 98-120.
-- 1972, Composition of fluid inclusions: U.S. Geol. Survey
Prof. Paper 440-JJ, 164 p.
Sibson, R. H., Moore, J. M., and Rankin, A. H., 1975, Seismic
pumping: A hydrothermal transport mechanism: Geol. Soc. Lon- don
Jour., v. 131, p. 685-659.
Sillitoe, R. H., 1976, A reconnaissance of the Mexican porphyry
copper belt: Inst. Mining Metallurgy Trans., v. 85, p.
B170-B189.
Sourirajan, S., and Kennedy, G. C., 1962, The system H20-NaCI at
elevated temperatures and pressures: Am. Jour. Sci., v. 260, p.
115-141.
Stewart, J. w., 1968, in Garcia, W., Geologia de los
cuadrangulos de Mollendo y la Joya: Peru, Servicio Geologia
Mineria, Bol. 19, 98p.
Wilson, P. A., 1975, K-At age studies in Peru with particular
ref- erence to the Coastal batholith: Unpub. Ph.D. thesis,
Liverpool Univ., 299 p.
Yermakow, N. P., 1965, Research on the nature of mineral forming
solutions with special reference to data from fluid inclusions:
Oxford, London and New York, Pergamon Press, 743 p.
-
COPPER MINERALIZATION IN THE PERUVIAN COASTAL BATHOLITH 691
APPENDIX I
Modal Abundances of Plagioelase (P), Quartz (Q), Orthoelase (K),
Hornblende (H), Pyroxene (Px), Biotite (B), Ores (O), and Accessory
Minerals (A) in the Linga superunit
Specimen Unit P Q K H Px B O A
49846 Hi 60.4 5.9 10.1 18.9 tr tr 4.7 49889 H 58.8 5.5 18.0 28.8
tr tr 8.9 49898 H 68.1 6.7 6.7 20.2 tr 1.8 1.5 49890 Ha 58.9 5.9
11.6 22.1 2.8 tr 4.8 51878 Ha 57.9 5.7 18.9 18.4 tr 4.0 51879 Ha
59.4 6.8 17.5 12.2 tr 4.5 49860 Ha 48.2 11.4 17.8 14.7 tr 4.5 8.1
tr 49870 Ha 59.2 12.7 11.7 9.0 tr 4.2 2.9 tr 51880 Ha 54.2 11.0
18.7 17.2 4.2 2.6 tr 51844 H4 44.7 18.0 22.9 15.8 tr 2.1 2.1 51858
H4 45.9 13.8 20.2 18.5 tr 0.9 tr 51870 H4 89.6 18.5 80.7 12.7 tr
2.2 1.8 51816 H5 84.9 18.0 80.6 11.8 2.0 2.7 51842 H5 89.2 17.8
26.8 18.6 1.4 1.7 51858 Hs 87.5 17.6 80.9 10.8 2.1 1.6 49448 A 51.8
16.2 16.9 11.4 tr 8.4 tr 49458 A 47.6 15.9 18.0 14.6 1.5 2.4 tr
51812 A 48.8 15.9 16.7 20.1 2.8 1.5 tr 51872 A2 29.2 26.4 81.1 9.4
0.5 1.9 1.5 49445 As 22.2 29.0 40.4 0.7 5.0 1.9 0.8 49455 As 18.8
82.2 46.0 tr 1.6 1.4 tr 51729 As 10.8 85.8 50.4 tr 0.8 2.0 1.0
51856 R 20.6 83.7 34.2 4.0 7.4 tr tr
Tr = trace Unit abbreviations: H = Humay rhythm, A -- Auquish
rhythm, R = Rineonada monzogranite. Subscript number refers to
position
of unit in order of emplaeement of rhythm (age 1-5, oldest to
youngest) (see Fig. 4)
APPENDIX II
Microthermometric Data for the Fluid Inclusions of the Linga
Superunit
Unit Type D Ho V L Ha Sy F
Hi
H2
585+ 535+ 585+ 585+ 580+ 580+ 277.1 de 126.6 869.4 490.9 549.6
210.6 221.0
550+ 550+ 550+ 550+ 560+ 560+ 560+ 560+
150.9 154.5 168.6
182.5
182.2 126.6 869.4 490.9 549.6 192.1 180.6
? 560+ 150.9 154.5 168.6 182.5
585+ -5.1 585+ -8.6 535+ 585+ 580+ - 11.2 580+
55O+ 55O+ 55O+ 55O+ 560+ 560+
210.6 221.0
560+ 560+
560+ 195.5
-15.1
-12.4
-10.6
-14.8 -16.2
-
692 a.A. AGAa
APPENDIX II--(Continued)
Unit Type D Ho V L Ha Sy F
H3
H4
H5
Ai
A2
3 0 192.3 192.3 $ 0 266.6 266.6 3 0 457.9 457.9
1 0 487.2 1 0 518.9 2 5 550+ 309.6 2 5 510+ 346.9 2 5 510. Odc
248.1 3 0 145.0 145.0 3 0 192.1 192.1 3 0 206.9 206.9 3 0 249.0
249.0 3 0 271.2 271.2 3 1 168.9 119.2 $ 1 182.6 123.8
1 0 506.7 2 6 550dc 317.5 $ 0 101.2 101.2 $ 0 108.4 108.4 3 0
147.4 147.4 3 0 149.7 149.7 3 0 151.5 151.5 3 1 179.4 127.7 3 1
217.3 197.4 $ 1 343.1 175.1 3 1 550+ 238.5 $ 2 440dc 182.8
1 0 565+ 1 0 565+ 2 4 550+ 284.9 2 5 550+ 215.5 3 0 158.9 158.9
$ 0 163.1 163.1 3 1 165.1 162.4 3 1 171.3 137.2 8 1 208.7 194.5
1 0 460.2 1 0 526.0 1 0 55O+ 1 0 55O+ 2 4 560+ 333.7 2 5 560+
511.3 2 5 560+ 283.7 $ 0 170.0 170.0 3 0 420.1 420.1 $ 0 446.2
446.2 3 1 272.5 272.5 3 1 307.4 307.4 3 1 312.8 312.8
533.3 560+ 560+ 471.1dc 239.9 193.2 193.2 201.1 201.1 209.6
209.6 345.0 345.0 370.3 370.3 428.8 428.8 221.4 212.1 280.8
216.5
487.2 518.9
506.7
565+ 565+
460.2 526.0 55O+ 55O+
533.3 560+ 560+
55O+ 417.8 510+
168.9 182.6
344.4
179.4 217.3 848.1 550+ 44O+
55O+ 55O+
165.1 171.3 208.7
560+ 560+ 560+
137.6 185.9 242.1
471.1+
221.4 230.8
264.7 323.7 238.3
248.4
225.0
242.5 258.9
323.6 190.4 234.1
257.8
-16.4
-9.5
-17.0 -13.7
-7.8
-46.2
-27.8
-9.5
-16.7
-13.1
-14.1
-6.9
-14.5 -12.2 -15.2
-
COPPER MINERALIZATION IN THE PERUVIAN COASTAL BATHOLITH 6913
APPENDIX II--(Continued)
Unit Type D Ho V L Ha Sy F
A3
$ 1 252.4 190.8 252.4 $ 1 259.9 215.$ 259.9 $ 1 262.7 262.7
192.9
2 7 518+dc 277.2 518+ 2 5 470+dc $27.0 470+ 2 5 555+ $52.$ 554.8
2 5 550+ 420.2 329.8 $ 0 140.5 140.5 $ 0 150.2 150.2 $ 0 172.0
172.0 3 0 182.8 182.8 $ 0 189.1 189.1 3 0 325.7 325.7 3 0 335.8
335.8 3 1 320.9 320.9 269.8 3 1 327.4 327.4 254.6 3 1 329.2 329.2
246.0 3 1 550+ 481.8 550+
518+ 266.2 468.0 187.7
-14.5 -15.$ -14.9
Abbreviations: D -- number of daughter minerals, Ho =
homogenization temperature, V = temperature at which vapor
disappears, L = temperature at which liquid disappears, Ha = halite
dissolution temperature, Sy -- sylvite dissolution temperature, F
-- freezing temperature, dc = decrepitation temperature; all
temperatures in C.