Geology of Titanium-Mineral Deposit Force1991

Downloaded from specialpapers.gsapubs.
org on March 16, 2016

Geological Society of America
Special Paper 259
1991
ABSTRACT
More than 90 percent of the titanium minerals currently produced come from
magmatic ilmenite deposits and from young shoreline placer deposits. This means that
the two geologic processes most directly responsible for economic titanium-mineral
deposits are (1) the accumulation of dense oxide-rich liquids immiscible in cooling
magmas of ferrodioritic to gabbroic composition, and (2) the interference between
deposition and entrainment in the enrichment of dense minerals on the upper swash
zones of beaches (and removal of some concentrates to eolian environments). Both
processes are essentially mechanical; i.e., chemical remobilization of titanium does not
form its major ore deposits.
Both processes also require precursor conditions that ensure that titanium is pres-
ent predominantly in the form of oxide minerals. In magmatic deposits, these are physi-
cal and chemical conditions that favor titanium-oxide over titanium-silicate minerals. In
sedimentary deposits, these conditions are a combination of proper source rocks, weath-
ering history, and sedimentary conduits, all necessary to permit the supply of favorable
minerals and prevent their dilution with unfavorable ones.
Some titanium-mineral production currently comes from fluvial placer deposits
(Gbangbama, Sierra Leone) and from deeply weathered alkalic pyroxenites (Tapira,
Brazil). In addition, several other deposit types could well become economic in the near
future: (1) rutile from eclogites, (2) rutile from contact-metasomatic zones of alkalic
anorthosites, (3) perovskite from alkalic pyroxenites, and (4) rutile byproduct from
porphyry Cu-Mo deposits; detrital titanium-mineral deposits could be exploited (5) on
continental shelves, (6) in Pleistocene glaciolacustrine deltas, or (7) in older, semiindu-
rated beach deposits. If young shoreline placers are depleted, these other deposit types
may become important.
1
Downloaded from specialpapers.gsapubs.org on March 16, 2016
Chapter 10.
Introduction
The impact of titanium-mineral geology on industry and trade
Titanium minerals are currently mined from four quite dis- Board (1983) has published an extensive review of titanium
similar types of deposits. To a great extent, the structure of the metal use. Titanium is one of the few metals the use of which is
titanium industry reflects the mineralogy of the different deposit likely to increase in the near future (U.S. Bureau of Mines, 1986).
types and subtypes, and world trade patterns for titanium prod- The titanium industry is an infant compared with long-
ucts reflect deposit-type distribution. This chapter explores the established industries using many other minerals and metals.
relation between geology and industry; for this purpose the eco- Large-volume pigment uses date from the late 1930s, the metal
nomic geology of titanium is introduced in nongenetic terms. uses from the late 1950s. The history of large commercial markets
for titanium metal has been short but irregular, making normal
USES AND VALUE OF TITANIUM MINERALS market cycles look placid by comparison; fitful U.S. government
acquisitions of military aircraft have partially governed the
Titanium metal and TiC>2 pigment are the two main prod- industry.
ucts made from titanium minerals, and on them large industries Worldwide, titanium-mineral mining in 1987 produced
depend. First by far in terms of volume produced is microcrystal- about 5.8 million metric tons of titanium-mineral concentrates
line TiC>2 for white pigment. Because of the extremely high re- (Lynd, 1988), worth approximately US $915 million. Titanium
fractive index of Ti0 2 as rutile (2.6 to 2.9, or higher than dioxide pigment and metallic titanium, made from titanium min-
diamond), it is the chief opacifying pigment used in paint and erals, in the same year were worth about US $4,200 million and
other products such as plastics and paper, not only for white color US $700 million, respectively. Total employment in these
but for quite a range of colors. It has supplanted lead-based titanium-mineral mining and conversion industries is more than
pigments in many of these roles. Titanium dioxide pigment com- 20,000 workers. Past this point, it is impossible to track the
monly forms more than 20 percent by weight of some paints. The increasing value of titanium products as they become finished
pigment industry consumes more than 90 percent of all titanium goods in complex industries.
minerals mined.
The product ranked second by volume, though perhaps not DISTRIBUTION OF TITANIUM AND TITANIUM
in importance, is titanium metal. The high strength-to-weight MINERALS IN THE EARTH'S CRUST
ratios and resistance to corrosion and high temperature of tita-
nium metal and titanium-based alloys make them important in- Titanium is sufficiently abundant in the earth's crust (0.86
gredients in many industries. Most important is the aircraft percent Ti, or 1.4 percent Ti02, according to Turekian, 1977)
industry, where the use of titanium has been growing for more that it is customarily listed with the major elements. Table 1 lists
than 30 years, to the point that commercial airliners of the current the Ti02 contents of some important rocks; note that many
generation can contain 30 percent titanium by weight. Another common rocks have Ti0 2 contents of more than 1 percent. Tita-
trend during the same period has been a diversification of nium is thus fundamentally unlike many mineral commodities
titanium-metal use in other industries. Many industries take ad- that are present in only trace quantities in common rocks.
vantage of titanium's corrosion resistance, as in heat exchangers Titanium is present in rocks as oxide and silicate minerals.
and desalinization plants. The U.S. National Materials Advisory The minerals that account for most of the titanium in rocks are
3
4 E. R. Force
TABLE 1. TYPICAL TITANIUM CONTENTS AND Economic minerals
TITANIUM PARTITIONING OF SOME COMMON ROCKS*
The currently economic titanium oxide minerals are rutile,
Ti0 2 TÌO2 in anatase, and ilmenite. Rutile, with a theoretical composition of
Rock types (%) oxide minerals
pure Ti02 (Table 2), is the most valuable, currently at about US
(as % of
$600 or more per metric ton. Two polymorphs of rutile—anatase
total Ti0 2 )
and brookite—have the same theoretical composition, but they
commonly contain detrimental chemical impurities. Anatase
Igneous rocks
concentrates are just beginning to come onto the market, and if of
Ultramafic 0.2-0.8 1-4
Mafic 0.9-2.7 50
suitable composition, they may also command a high price. A
Felsic 0.2-0.8 3-30 value for perovskite has not yet been established; it will depend
Alkalic 0.1-3.3 3-50 on the efficiency of the process eventually used to convert it to
Charnockitic 0.4-1.6 50-95 Ti02- Titaniferous magnetite in the strict sense has no value in the
Anorthositic 0.1-0.5 50-95 titanium industry at present. Silicate minerals having high tita-
Metamorphic rockst nium contents, such as sphene, have no economic value either.
Gneiss 0.6 5-100
Ilmenite, the most important source of titanium products,
Schist and phyllite 0.6 1-70
has a complicated story both geologically and economically.
Amphibolite 1.4 15-70
Unweathered ilmenite is commonly intergrown with iron oxide
Serpentinite 0.0 n.a.
minerals and thus contains less Ti02 than its theoretical composi-
Eclogite 1.0-6.0 50-90
Sedimentary rocks
tion indicates (Table 2). On the other hand, weathering leaches
Sandstone 0.2-0.6 10-100 iron from ilmenite, resulting in poorly crystalline mineral grains
Shale 0.6-0.7 ? residually enriched in Ti02- The term ilmenite,
as used in the
Limestone 0.1-0.2 ? titanium-mineral industry, commonly covers the entire range
from unweathered ilmenite with Ti02 contents below 50 percent
"Modified from Force, 1976a. to altered ilmenite containing more than 60 percent Ti02- When
t S e e Chapter 2 for variation of partitioning with grade. the Ti02 content of altered ilmenite exceeds about 70 percent, it
is commonly referred to as leucoxene.
Price per metric ton varies
over this range from as much as US $500 for "leucoxene,"
listed in Table 2. The partitioning of titanium between oxide and

silicate phases varies greatly among different rock types (Table 1). TABLE 2. COMPOSITION OF
The economic geology of titanium begins by focusing on rock SOME COMMON TITANIUM MINERALS
suites that contain titanium largely in oxide-mineral form.
Mineral Theoretical H O 2 content
INTERPLAY OF GEOLOGIC AND formula (%)
ECONOMIC FACTORS
Oxides
Only the titanium-bearing oxide minerals have economic Rutile Ti0 2 >95
Anatase TI02 >95
value. All the oxide minerals containing more than about 25
Brookite* Ti0 2 >95
percent Ti02 have some present or potential economic value, and
Ilmenite FeTi0 3 52t
no silicate minerals are valuable regardless of Ti02 content. Thus
Perovskite* CaTi0 3 59^
the economic geology of titanium is the geology of titanium oxide Magnetite Fe304 0-15
minerals.
In addition, different titanium oxide minerals are best suited ilicates
for different industrial recovery processes. The market price of a Sphene CaTiSi0 5 41 +
titanium-oxide mineral concentrate is a function of its Ti02 con- Melanitic garnet* Ca3Fe2S¡30-| 2 0-17
tent and its suitability for a given process. Prices of such concen- Biotite K 2 (Mg,Fe) 4 (Fe,AI) 2 0-6
trates vary an entire order of magnitude. Si 6 AI 2 O 2 0 (OH, F) 4
Therefore, mineralogy is a more important factor in the Calcic amphiboles (Na,K)Ca 2 (Mg,Fe,AI) 5 0-10
economic geology of titanium than it is for most other mineral Si6AI2022(0H,F)2

Augite Ca(Mg,Fe)(Si,AI) 2 0 6 0-9
commodities for which chemical enrichment is the most impor-
tant factor. Examples will be given in this volume (Chapter 9) of
'Phases restricted to unusual rock types,
important titanium-mineral ores that contain less than the aver- t Stoichiometric value.
age crustal abundance of titanium.
Geology of titanium-mineral deposits 5
TABLE 3. TYPES OF TITANIUM-MINERAL DEPOSITS, THEIR ECONOMIC SIGNIFICANCE,
AND THEIR TITANIUM MINERALS
Class Type Typical Importance* Example

Mineralogy
1. Metamorphic a. Eclogite Rutile B Piampaludo, Italy

b. Aluminosilicate Rutile E Evergreen, Colorado
c. Ultramafic contact Rutile E Dinning, Maryland
2. Igneous a. Magmatic ilmenite Ilmenite A Allard Lake, Canada;
Roseland, Virginia (in part)
b. Anorthosite-margin Rutile, ilmenite C Roseland, Virginia (in part)
c. Albitite (kragerite) Rutile E Krager/S,Norway
d. Alkalic Perovskite, C Powderhorn, Colorado;
Nb-rutile, Magnet Cove, Arkansas
Nb-brookite
3. Hydrothermal Porphyry
Rutile C Bingham, Utah
4. Sedimentary a. Fluvial Ilmenite, rutile A Gbangbama, Sierra Leone
b. Glaciolacustrine Ilmenite C Port Leyden, New York
c. Shoreline (and Ilmenite, altered A Richards Bay, South Africa;
coastal eolian) ilmenite, rutile Stradbroke Island, Australia;
Trail Ridge, Florida (in part)
5. Weathered a. Alkalic parent rock Anatase B Tapira, Brazil
b. Mafic parent rock Ilmenite D Roseland, Virginia (in part)
c. Placer parent Altered ilmenite, A Trail Ridge, Florida (in part)
"leucoxene"
*A, of great present importance; B, of probable great importance in near future; C, of possible
importance; D, of moderate present importance; E, of minor present importance at world scale.
through about US $70 for slightly altered ilmenite containing Currently known shoreline placer deposits will be exhausted
about 54 percent TiC>2, to ilmenite with lesser Ti02 contents that in about 30 years, with several exceptions (Garnar, 1978; Shep-
is generally not traded (i.e., it is mined only by companies that herd, 1986; Fantel and others, 1986). The future importance of
consume it in their own pigment plants and smelters). Thus, the shoreline placer deposits is in question, and thus the economic
alteration state of ilmenite is of great economic importance. In the geology of titanium minerals is in flux. New types of deposits
ilmenite mining industry, the term grade commonly refers to the would supply their own characteristic suite of titanium minerals
TiC>2 content of ilmenite concentrates rather than to the amount (Tables 3, 4).
of ilmenite in ore.
Recovery processes and relation to geology
Deposit types
The chloride and sulfate processes are the two recovery
Titanium minerals are mined from hard crystalline rocks, processes used in the titanium mineral industry. The chloride
weathered rocks, and unconsolidated sediments. Table 3 shows process converts titanium minerals to titanium tetrachloride and
the great diversity of deposit types and the relations among them. thence to either TiC>2 pigment or to titanium metal. This process
These deposit types are described in detail in following chapters. is the more recent of the two and is preferred because it is less
At present, shoreline placer deposits supply more than half polluting. However, it requires a high-TiC>2 feed (and certain
the titanium minerals mined worldwide. These deposits supply trace-element maxima). Initially, only rutile and leucoxene (TiC>2
rutile and variably altered ilmenite. Most of the remainder is >70 percent) could be used in the chloride process, but some
supplied by magmatic ilmenite deposits from rocks of the companies have learned to use feeds of altered ilmenite contain-
anorthosite-ferrodiorite suite, which supply unaltered ilmenite. A ing about 60 percent TiC>2. Accordingly, plants using the chloride
fluvial placer deposit produces rutile, and a deposit in the weath- process are supplied predominantly with concentrates from
ered mantle on alkalic igneous rock has started producing weathered shoreline placer deposits. Chloride plants are also fed
anatase. by placer rutile deposits of fluvial origin and could possibly use
6 E. R. Force
TABLE 4. WORLD PRODUCTION, RESERVES, AND IDENTIFIED suitable for the chloride process or, if processed via sulfate, greatly
RESOURCES OF TITANIUM MINERALS reduce the volume of effluent. In the second subprocess, "syn-
(in thousand metric tons of contained TÌO2*) thetic rutile" is produced from ilmenite by a number of methods;
generally, feeds containing about 55 percent TiC>2 from slightly
1987 Reserves§ Identified
weathered placer deposits are used. The price of synthetic rutile,
Country Production* Resources**
like that of rutile, is about US $600 per metric ton. Synthetic
rutile is used in the chloride process. In fact, one producer is
Australia 1,250 27,000 131,000
150,000
making titanium metal from ilmenite via synthetic rutile. In the
Brazil 30 54,200
Canada 890 24,000 81,000 near future, the smelting and synthetic-rutile subprocesses will
China 80 28,500* 38,000* handle increasing proportions of lower-TiC^ oxide feeds and will
Finland 1,400 4,000 allow such material to cross over into the chloride process.
India 100 32,700 79,000
Italy ? 9,000 The meaning of a titanium-mineral resource
Malaysia 240 ? ?
Mexico 3,000
Because the economic value of a titanium-mineral deposit
Mozambique 2,000
depends on complex mineralogic factors, cutoff grades cannot be
New Zealand 47,000
specified with simple chemical values. Orderly comparison of
Norway 550 29,000 89,000
world resources of titanium minerals (Fantel and others, 1986;
Sierra Leone 110 1,800 2,000
Lynd, 1988; Towner and others, 1988) has necessitated a defini-
South Africa 680 37,700 58,000
Sri Lanka 80 4,300 5,000
tion of an economic resource of titanium minerals that excludes
U.S.A. 360 10,600 103,000 titanium enrichments that are of no present economic interest.
U.S.S.R. 210 8,000* 16,000* The definition most commonly used at present is the test of
economic relevance by Force and Lynd (1984):
Total 4,580 259,200 817,000
* = none reported; ? = unknown. Only the titanium oxide minerals rutile and its polymorphs [anatase
*Lynd, 1988. and brookite], altered ilmenite, ilmenite, and perovskite, which are
§ Modified from Lynd, 1985. known or thought to have some economic value, are included. . . .Ex-
"Including reserves. Fantel and others, 1986; Table 5 of this paper; cluded from resources are titanium minerals of finer grain size than 20
and my information. Aim (0.02 mm), on the grounds that they cannot presently be separated.
^Reserve and resource figures are difficult to calculate for these Where ilmenite is known to be present as separable grains intergrown
countries because the Force and Lynd (1984) resource definitions do with magnetite, resources of the ilmenite are included. Where insepara-
not apply. Figures listed by Towner and others (1988) are much larger. ble intergrowths of magnetite and ilmenite together contain 25 percent or
more of TiC>2, resource figures are also included on the grounds that this
material could be smelted into high-TiC>2 slag. . . .Our figures include
only deposits containing at least 1 percent ilmenite or 0.1 percent rutile
anatase from weathered alkalic deposits or rutile from metamor-
or linear combinations thereof in unconsolidated deposits, or 10 percent
phic deposits. ilmenite or perovskite or 1 percent rutile in hard rocks. Lower grade
The sulfate process digests titanium minerals in sulfuric acid resources are included if titanium minerals could be produced as by-
enroute to recrystallization as TiC>2 pigment. Effluents from this products of other minerals already being mined in the same deposits. . . .
process are powerful pollutants unless they are neutralized. High-
TiC>2 feed is not necessary in this process; indeed, the higher-TiC^ Resource distribution and trade networks
feeds such as rutile and leucoxene are unreactive in it. Accord-
ingly, plants using the sulfate process are fed by ilmenites contain- All but about 1 percent of 1987 world titanium-mineral
ing 45 to 60 percent TiC>2. The trace elements acceptable in the production came from eleven countries (Table 4): Australia,
sulfate process are also different from those acceptable in the Canada, South Africa, Norway, the United States, Malaysia, the
chloride process. Most of the ilmenite used in the sulfate process U.S.S.R., Sierra Leone, India, China, and Sri Lanka, in order of
comes from deposits in crystalline rocks and is unaltered. Some total Ti02 content of concentrates. Brazil will probably join this
also comes from shoreline placers that are little weathered. list in the near future.
Two subprocesses are used to convert a low-TiC^ feed into Resource distribution is also uneven (Fig. 1; Table 4). Brazil
a high-Ti02 feed for the two main recovery processes. First, in has the largest identified resource in terms of contained Ti02,
the smelting subprocess, low-TiC^ feeds of appropriate composi- followed by Australia. The twelve countries mentioned plus New
tion (i.e., low Ca content but as little as 30 percent TiC>2) are Zealand dominate the resource picture. Resource figures for the
smelted to a high-TiC^ slag plus pig iron. This slag has a much U.S.S.R. and China are not adequately known, and the resources
higher market price than ilmenite (about US $275 per metric ton of Madagascar and a few other countries may be far greater than
for 80 percent Ti0 2 slag from Canada and US $300 for 85 are currently recognized.
percent TiC>2 slag from South Africa). Some of these slags are Thus a lively trade is inevitable between countries mining
Number Deposit (and type")
Metamorphlc deposits (Chapter 2)

1. PiampaJudo, Italy (1a)
2. Sunntjord region, Norway (1a)
3. Shubino Village. USSR (1 a)
4. Evergreen. Colorado (1b)
5. White Mountain. Calllomla (1b)
Anorthoslte-ferrodlorlte deposits
(Chapter 3)
6. Sanlord Lake, New York (2a)
7. Allard Lake, Canada (2a)
8. Tellnes, Norway (2a)
9. R06eland, Virginia (2a. 2b. 5b)
10. San Gabriel Range, California (2a)
11. Laramie Range, Wyoming (2a)
12. Duluth Complex, Minnesota (2a)
£
8
13. Piuma Hidalgo, Mexico (2b)
14. Kragero, Norway (between
deposits 8 and 18) (2c)
Alkallc Igneous deposits (Chapter 4)
15. Powderhorn, Colorado (2d)
16. Taplra, Salltre, and Cataiao. Brazil
5"
(2d. 5a) s
17. Magnet Cove, Arkansas (2d) S'
18. KodaJ. Norway (2d) 3
Other igneous deposits (Chapter 5)
19. Bingham. Utah (3)
Nonmarlne placer deposits (Chapter 8) s
20. Gbangbama. Sierra Leone (4a)
21. Pott Leyden, New York (4b) 2
Quaternary shoreline placer deposits
(Chapter 9)
f
22. Jacksonville district. Florida
(4c. 5c) 1
23. Eastern Australia districtt (4c) E?
24. Geographe Bay, Australia (4c)
25. Eneabba, Australia (4c)
26. Travancore coast, India (4c)
27. Richards Bay. South Alrlca (4c)
Pre-Quatemary shoreline placer deposits
(Chapter 10)
28. Lakehurst. New Jersey (4c, 5c)
29. McNairy Sand. Tennessee (4c, 5c)
30. Cretaceous sands, western U.S.
<4c)
31. Karoo basin. South Africa (4c)
"From Table 3.
tTwo numbers show extent.
Figure 1. W o r l d map showing location and type of titanium-oxide mineral deposits described in this
volume. A more detailed m a p of U.S. deposits is contained in Force and Lynd (1984). Currently
important producers shown solid.
—1
8 E. R. Force
TABLE 5. IDENTIFIED TITANIUM-MINERAL RESOURCES OF THE UNITED STATES,
INCLUDING RESERVES*
(In thousand metric tons of contained TIO2)
State District or Typet Rutile + Altered Low-TlC^ Perovskite

description polymorphs ilmenite ilmenite
Alabama Sand and gravel 4 100

Arizona Porphyry copper ore 3 4,000
Arizona Yuma County 1b 200
Arkansas Magnet Cove 2d 200
California San Gabriel Mountains 2a 4,800
California lone 4a 600
California White Mountain 1b 300
Colorado Powderhorn 2d 20,000
Colorado Evergreen 1b 200
Florida Old beach sand 4c(5c) 1,700 9,700
Florida Phosphate 4 200
Georgia Old beach sand 4c 500 2,400
Georgia Silica sand 4 100 200
Idaho Latah County clay 5 1,300
Maryland Harford County 1c 700
Minnesota Duluth Complex 2a 900
Mississippi Ship Island 4c 100
New Jersey Lakehurst 4c 10,100
New Mexico Cretaceous sandstones 4c 700
New York Sanford Lake 2a 8,600
New York Port Leyden 4b 6,300
North Carolina Yadkin Valley 2(a?) 200
North Carolina Other 4c 400
Oklahoma Wichita Mountains 4a 3,900
Oregon Salem bauxite 5 1,800
South Carolina Hilton Head 4c 100 300
South Carolina Charleston 4c 100 1,100
Tennessee Cretaceous sand 4c 1,300 8,400
Utah Bingham 3 4,000
Virginia Roseland-Piney River 2a,b(5b) 1,000 5,500
Virginia Willis Mountain kyanite 1b 300
Washington Spokane clays 5 400
Wyoming Cretaceous sandstones 4c 500
Totals by deposit types 1b 1,000

1c 700
2a 20,000
2b 1,000
2d 200 20,000
3 8,000
4 100 500
4a 600 3,900
4b 6,300
4c 3,700 32,400 1,300
5 3,500
Total by mineral 14,700 33,500 35,000 20,000
Total all minerals 103,200
'Modified from Force and Lynd (1984); none reported or negligible.

tFrom Table 3.
titanium minerals and those producing and/or consuming TiC>2 TABLE 6. BYPRODUCT RELATIONS AMONG TITANIUM MINERALS
pigments and titanium metal. Prominent among producers hav- AND OTHER MINERAL COMMODITIES*
ing few titanium-mineral resources are Japan, the United King-
dom, and West Germany. The United States is also heavily Currently Potentially
Recovered Recoverable
dependent on imported titanium minerals; although U.S.
titanium-mineral resources are considerable (Table 5), they are Porphyry Cu, Mo deposits
overshadowed by the country's needs. The United States imports Sand and gravel deposits
Titanium as Fluvial tin deposits
not only titanium minerals but also titanium metal. byproduct Silica sand deposits Silica sand deposits
Countries that do not produce sufficient amounts of a badly (recovered in small part) Aluminosilicate rock deposits
needed mineral may refer to that mineral as strategic. In the Palabora-type deposits Bauxite (some) deposits
United States, titanium minerals are considered among the top Nepheline syenite deposits Feldspar deposits from anor-
thosite
ten strategic mineral commodities, partly because of the need for (recovered in USSR only)
Detrital phosphorite deposits
metallic titanium for the aircraft industry. Internationally, tita-
Sandstone uranium deposits
nium is considered sufficiently strategic that it was chosen for
Mafic V, Pt, Cr, Ni-Cu
study in the International Strategic Minerals Inventory (Towner deposits
and others, 1988).
Byproduct-coproduct relations Titanium From shoreline placer From alkalic deposits:

as major deposits: Niobium
Titanium minerals are produced as byproducts of a number product§ Zircon Rare earths
of other mineral commodities, most notably tin from alluvial Aluminosilicates
mining in southeast Asia. Titanium-mineral resources, not recov- Monazite From eclogite deposits:
ered, are present in a large number of other types of active From anorthosite- Garnet
deposits (Table 6). Potentially the most important is rutile from ferrodiorite deposits:
porphyry-type deposits, discussed in Chapter 5. Magnetite
Vanadium
Conversely, byproducts of titanium-mineral mining are also
Apatite
numerous; these vary considerably among types of deposits
(Table 6). Zircon from shoreline-complex placer deposits is
'Modified from Force, 1976c, 1980a.
worthy of special mention, as it attains coproduct status in some
"(Titanium minerals as byproducts of established mining operations of
districts. other commodities.
§Other minerals as byproducts of titanium-mineral mining.
Chapter 2.
Occurrence and deposits of titanium oxide minerals in metamorphic rocks
Titanium oxide minerals formed by metamorphic processes and the Fe/Mg ratio, which governs reactions of the type:
play three roles in the economic geology of titanium. First, some
extreme metamorphic processes actually form deposits of eco- pyroxene + rutile = ilmenite + quartz + MgO. (2-2)
nomic interest. Such deposits account for less than 2 percent of
the identified resources of the United States, but the eclogite type Figure 2 shows the relation of these variables and of metamor-
of metamorphic deposit may become more important elsewhere phic grade to titanium mineralogy.
in the future. Second, titanium oxide minerals from metamorphic Volatile constituents may affect partitioning of titanium
rocks are the predominant source of those minerals in placer among oxide minerals and between oxides and silicates. Volatiles
deposits. Third, the massif anorthosite-ferrodiorite suite of igne- of importance include oxygen, carbon dioxide, and sulfur. Oxy-
ous rocks that contain important titanium-mineral resources is gen controls reactions of the type:
characteristically emplaced under particular metamorphic condi-
tions. These igneous rocks in turn form a subordinate source of ilmenite + 0 2 = magnetite + rutile (2-3)
detrital titanium oxide minerals in placer deposits. Because of
these roles of metamorphism in influencing igneous deposits and and sulfur controls reactions of the type:
controlling placer deposits, titanium will be traced through
metamorphic processes with some diligence. ilmenite + S 2 = pyrite + rutile. (2-4)
Titanium contents of rocks generally remain approximately
constant during metamorphism. In fact, titanium is so unusually Similar reactions can also be written to show liberation of tita-
immobile in many subsolidus geologic processes that calculations nium from silicates. CO2 enters into the reaction:
of the amount of change in other elements are commonly normal-
ized to titanium, assuming it to be conserved. The partitioning of sphene + C0 2 = rutile + calcite + quartz. (2-5)
titanium between oxide and silicate phases, however, varies
VARIATION IN PARTITIONING WITH
markedly among metamorphic facies. For example, 0.6 percent
METAMORPHIC GRADE
TiC>2 in some ordinary metamorphic rocks (Table 1) may be
present as 1.5 percent sphene at low metamorphic grades or as For convenience of discussion, metamorphism is divided
1.2 percent ilmenite or 0.6 percent rutile at high grades. into higher and lower grades between the upper and lower am-
phibolite facies. For rocks of high-pressure facies series, it is di-
VARIATION IN PARTITIONING WITH vided between the higher and lower blueschist facies.
ROCK COMPOSITION
Lower-grade metamorphic rocks
Titanium partitioning between oxide and silicate phases is
also a function of rock composition. The compositional variables Most rocks at lower metamorphic grades contain the bulk of
of greatest importance appear to be the Al/Ca ratio, which gov- their titanium in silicates. Sphene is the most important carrier of
erns reactions of the type: titanium in many such rocks, and biotite and hornblende are also
important carriers in some rocks. Figure 2 shows the relation of
sphene + A1203 = rutile + plagioclase (2-1) sphene stability to rock composition and metamorphic grade.
11
12 KR. Force
Increasing temperature and pressure
Greenschist facies A m p h i b o l i t e facies

Granulite Eclogite
Biotite and Kyanite and facies
^ C h l o r i t e zone facies
garnet zones sillimanite zones
Aluminous Rutile + Ilmenite Ilmenite Ilmenite + Rutile
rocks magnetite rutile
(sillimanite-
kyanite- (Rutile)+ Ilmenite +
staurolite) sphene sphene Rutile Rutile Rutile
Sphene Sphene + Ilmenite Ilmenite +
" N o r m a l " rocks
ilmenite rutile
(chlorite-biotite-
muscovite-
garnet)
Sphene Rutile Rutile
Sphene + Sphene + Ilmenite Rutile
Calcic rocks
ilmenite ilmenite
(amphibole-
clinopyroxene-
calcite)
Sphene Sphene Rutile
FeO/MgO1/ Figure 2. Schematic relation of rutile, ilmenite, and sphene occurrence to composition and grade in
high metamorphic rocks. Circled minerals indicate deposits, currently subeconomic; n is chemical potential.
FeO/MgO
low
In subgreenschist and lower greenschist (chlorite zone) fa- grade contains a little more than 1 percent TiC>2. Sphene con-
des, the predominant carrier of titanium in most rocks is appar- tinues as an important carrier of titanium in rocks of a wide range
ently sphene (Force, 1976b). Sphene as fine dusty aggregates is a of compositions and becomes coarser grained. Ilmenite coexisting
common and moderately abundant phase in these rocks, both with hematite is also common (Rumble, 1976; Nedelcu, 1986); it
felsic and mafic. Relict titanium oxide minerals commonly are in may form with plagioclase by the breakdown of sphene and
advanced stages of recrystallization; for example, ilmenite and muscovite.
rutile in placer concentrations metamorphosed at low grades are Retrograde reactions in originally high-grade rocks are, of
commonly recrystallized to sphene and magnetite (Goldsmith course, similar to reactions in low-grade rocks and can commonly
and Force, 1978, Fig. 3). Magnetite at this grade has low Ti02 be recognized by rims of sphene and biotite around titanium
contents (Abdullah and Atherton, 1964), and there is little tita- oxide minerals. Chemical microenvironments limited to single
nium in chlorite. Extremelyfine-grainedanatase may be inter- grains can be recognized in some retrograded rock textures; for
grown with sphene in some rocks (Rumble, 1976; Herz and example, rutile needles in chlorite may replace titaniferous biotite
Force, 1987). Rocks having very high Al/Ca ratios may contain grains.
fine rutile (cf., Zen, 1960). Those with high Al/Ca ratios, high
Fe/Mg ratios, and high oxygen fugacity may show rutile + mag- Rutile in altered lower-grade metamorphic rocks
netite or hematite assemblages (Meilke and Schreyer, 1972;
Rumble, 1973,1976). Metasomatic additions of constituents such as magnesium or
In upper greenschist (biotite zone)-facies rocks, biotite be- sulfur, or depletion in calcium, may permit the formation of rutile
comes a significant carrier of titanium. At this grade, biotite typi- (equations [2-1] to [2-4]). The introduction of magnesium into
cally contains a little more than 1 percent Ti02- Many rocks, "blackwall" metasomatic zones adjacent to serpentinites ties up
especially mafic ones, continue to contain a considerable share of available iron; rutile is therefore commonly found in such black-
their titanium as sphene. Ilmenite may form in place of rutile in walls (Chidester, 1962). Southwick (1968) and Herz and Valen-
the high-oxygen pelitic rocks by reaction (2-3). tine (1970) have described an unusual chlorite rock containing
In the lower amphibolite facies, hornblende becomes a car- coarse rutile, magnetite, and apatite that apparently formed in this
rier of titanium in mafic rocks. Hornblende at this metamorphic manner; it is discussed later in this chapter as a resource.
In a similar manner, rutile is found in the wall rocks of
metamorphosed massive sulfide deposits. Nesbitt and Kelly
(1980) show that around the Ducktown sulfide bodies of eastern
Tennessee, metamorphosed to staurolite grade, ilmenite in wall
rocks has responded to local high sulfur pressure in going to
pyrrhotite and rutile. However, the economic importance of this
occurrence type is thus far limited to prospecting for massive
sulfides.
Hydrothermal stripping of cations from metavolcanic rocks
and formation of pyrite have led to appreciable rutile contents
along with andalusite and topaz in several pyrophyllite deposits
of the slate belt in North Carolina and South Carolina (Schmidt,
1985; oral communication, 1986). These deposits occur in host
rocks representing the greenschist facies (even though kyanite is
locally present; Carpenter, 1982). Schmidt (1985) and Carpenter
and Allard (1982) reviewed a large number of related world
Figure 3. Diagram of rutile distribution in some common metamorphic
occurrences of aluminous hydrothermal systems with associated rocks (from Goldsmith and Force, 1978, Figs. 5 and 6; Force, 1980b,
rutile in metavolcanic rocks of low metamorphic grades. Rutile in Fig. 1). Rectangles are metamorphic rutile (some twinned); circles are
aluminous hydrothermal rocks persists to higher-grade metamor- detrital rutile, dashed where retrograding.
phic assemblages (Geijer, 1964; Marsh and Sheridan, 1976).
Some economically interesting rutile occurrences are described
later in this chapter.
Rutile may form in sulfide-bearing rocks by prograde con-
High-grade metamorphic rocks of "normal" P-T ratios version of pyrite to pyrrhotite:
Ramberg (1948, 1952) was thefirstto show that titanium is pyrite + ilmenite = pyrrhotite + rutile. (2-7)
transferred from titanium-bearing silicates at lower metamorphic
grades into titanium oxide minerals at high grades. This process Such reactions involving sulfides may also liberate titanium from
was further documented by Force (1976b) and Goldsmith and titaniferous biotite and hornblende by forming more magnesian
Force (1978). Figure 2 shows the influence of bulk composition silicates plus oxide minerals, including rutile (Robinson and
on resulting mineralogy. Tracy, 1977; Robinson and others, 1982; Mohr and Newton,
In the upper amphibolite facies, rutile becomes common in 1983).
some pelitic rocks, and ilmenite is common over a wide range of In the granulite facies, titanium oxide minerals become the
lithologies. These phases form partly at the expense of sphene, dominant carriers of titanium in most lithologies. The granulite
which becomes less common in these rocks as the calcium con- facies by definition involves formation of pyroxenes at the ex-
tent of plagioclase increases (as in reaction [2-1]). Spear (1981) pense of biotite and hornblende; this leads to titanium liberation,
found that a similar reaction in amphibolites consumes sphene to as the pyroxene contains much less titanium than biotite and
form ilmenite, with some compositional change of plagioclases hornblende. The approximate reactions are:
and amphiboles; this reaction is favored by high oxygen fugacity
biotite + quartz = orthopyroxene + garnet + orthoclase + rutile + vapor
as well as increasing temperature.
(2-8)
Goldsmith and Force (1978) found that rutile occurs in
and:
pelitic rocks at grades as low as the kyanite zone in units in which
the Al/Ca ratio is greater than 50. Wall rocks having lower
hornblende + quartz = orthopyroxene + plagioclase + ilmenite + vapor.
Al/Ca ratios contain sphene instead. Textural relations suggest (2-9)
that rutile forms with garnet at the expense of ilmenite, possibly
by the reaction: With increasing metamorphic grade, biotite and hornblende ac-
commodate increasing amounts of TiC>2, until in the granulite
ilmenite + muscovite + quartz = almandine garnet + rutile + biotite facies, biotite may contain up to 6 percent TiC>2 and hornblende
(2-6) up to 4 percent TiC>2 (reviewed by Force, 1976b; Guidotti and
others, 1977; Spear, 1981; Dymek, 1983). The Ti0 2 contents of
(see also Ghent and Stout, 1984, for related reactions). At higher these minerals are a function of Fe/Mg ratios as well as meta-
metamorphic grades, an increasing variety of rocks contain rutile morphic grade (Guidotti and others, 1977; Robinson and others,
(Fig. 3). 1982). However, the proportion of titanium present in biotite and
14 E. R. Force
is, sphene and other silicates carry most of the titanium, ilmenite
is locally common, and rutile is present in some rocks of unusual
composition. In higher-grade rocks of this type, rutile becomes
common (Blake and Morgan, 1976; Itaya and Banno, 1980); the
high-grade facies was named the epidote-rutile blueschist facies
by Taylor and Coleman (1968). In these high-grade rocks, rutile
typically occurs with garnet, sodic and calcic amphiboles, and
epidote. Ilmenite is not commonly reported, and sphene is present
as a retrograde mineral. Blake and Morgan describe occurrences
of such rutile-bearing rocks from California, New Caledonia,
Japan, the European Alps, and Venezuela.
Rutile is a characteristic phase of eclogitic rocks, the highest
temperature and pressure member of this series. Again, ilmenite is
commonly not reported, and sphene is common only as a retro-
grade mineral. As other titaniferous silicates are absent in eclogite,
Figure 4. Photomicrograph of rutile (dark) in graphite-sillimanite-biotite- virtually all TiC>2 in the rock may be present as rutile. Where
garnet-K-spar gneiss, Franklin, North Carolina. Transmitted plane light,
eclogite has a ferrogabbroic composition, rutile contents may be
6-mm field.
more than 5 percent (Chesnokov, 1960; Cortesogno and others,
1977). Such eclogites are of economic interest, and three districts
are described in the next section.
hornblende in a rock decreases
with increasing grade, because of In many occurrences, these high-grade metamorphic rocks
reactions consuming those minerals (Force, 1976a, Table 6). occur as tectonic inclusions in lower-grade terranes. However, in
Sphene becomes uncommon in the granulite facies except in New Caledonia (Black, 1977), a transition from lawsonitic schists
the most calcic rocks (Ramberg, 1952; Turner, 1968) because of to eclogitic gneisses is exposed, with rutile occurrences limited to
reactions such as: the high-grade epidote zone. Other relatively complete transitions
are known from Ecuador, Venezuela, Norway, Italy, and Japan.
sphene + hornblende + sodic plagioclase =
clinopyroxene + calcic plagioclase + ilmenite. (2-10) M A J O R DEPOSITS
Magnetite also commonly becomes less abundant in this facies, At present, no titanium minerals are produced from meta-
typically shown by lessened gradients on aeromagnetic maps. morphic deposits. The eclogite-type deposit has the greatest eco-
Titanium liberated from biotite, hornblende, sphene, and magne- nomic potential, and such deposits are therefore emphasized as
tite goes into ilmenite and rutile, resulting in the greatest partition- thefirstthree descriptions below. The different types of deposits
ing toward titanium-oxide minerals possible for gneisses, schists, are put in their metamorphic context in Figure 2.
and amphibolites (Table 1). Rutile is the economic mineral in metamorphic titanium-
In these granulite terranes, relatively mafic lithologies typi- mineral deposits. Metamorphism does not enrich rocks in tita-
cally contain ilmenite, and leucocratic lithologies contain rutile nium; thus, available titanium in a rock must be present in a most
(Figures 3, 4). In some situations, especially in the pyroxene valuable form in order for the rock to qualify as a resource.
granulite subfacies, rocks of intermediate composition contain
rutile also. Both rutile and ilmenite in such rocks are typically Piampaludo deposit, Italy
coarse and locally are free of intergrowths (Herz and Force,
1987), but retrogression may produce rims of other minerals. Development work since 1974 has been conducted at Pi-
Eclogites are locally the end members of high-grade meta- ampaludo, mostly by Geomineraria Italiana, on a single large
morphism of "normal" facies series (Coleman and others, 1965). tectonic inclusion of eclogite in serpentine in the Ligurian Alps
These rocks, which contain rutile as a characteristic phase and (Fig. 1) of northern Italy (Mancini and others, 1979; Clerici and
little ilmenite, are described later in this chapter. others, 1981). Similar but smaller bodies are abundant in the
region (Martinis and Pasquare, 1971; Cortesogno and others,
High-grade metamorphic rocks of high-pressure type 1977) and commonly have a ferrogabbroic composition (4.6
percent Ti02, 18.2 percent total Fe oxides).
Metamorphic rocks formed at low ratios of temperature to The Piampaludo body is exposed in several hills and the
pressure form a separate metamorphic facies series. At low intervening gorges of the Orba and Orbarina Rivers (Fig. 5), in
metamorphic grades, these rocks are similar in titanium mountainous country dotted with attractive villages. Its outcrop
mineralogy to normal low-grade metamorphic rocks (Ernst and area is about 500 x 1,800 m.
others, 1970; Seki and others, 1971; Itaya and Otsuki, 1978); that Along the Orba River section through the potential Piampa-
2.7 to 9.3 percent rutile. Rutile that is probably too fine to be
separated (<50 /xm) was not counted. Correcting for density,
rutile averages 6.2 percent by weight, a figure consistent with
chemical analyses of the body (6 percent Ti02) by Mancini and
others (1979). Modal analyses of eclogites of the entire district
average 5.8 volume percent rutile (Cortesogno and others, 1977).
The composition of rutile from Piampaludo has not yet been
reported. Elaine McGee (written communication, 1986) found
by electron microprobe that the chemical impurities Nb20s,
Cr 2 0 3 , MgO, MnO, FeO, CaO, and A1203 in this rutile consti-
tuted less than 1 percent each and that Ti02 content was over 95
percent.
Clerici and others (1981) report that rutile liberation from
its eclogite host during crushing is adequate. A good rutile con-
centrate was made with adequate recovery.
Mancini and others (1979) list proven ore of the Piampa-
ludo deposit as 150 million metric tons, based on surface lithol-
ogy and nine drill holes of 100 m depth. Their estimate of
probable and possible ore, an additional 700 million metric tons,
is optimistic in my opinion. At the northern end of the body,
where no drill holes were sited, the basal contact and foliation in
both host and eclogite appear to dip gently south. Thus, the base
of the body may be shallow there. The amount of proven ore,
however, implies 9.3 million metric tons of rutile, potentially
making Piampaludo a deposit of world importance.
Sunnfjord region, Norway

Eclogite is present as discrete bodies of various sizes forming
a subordinate constituent of gneiss terrane throughout the moun-
tainous Western Gneiss Region of Norway. The origin of these
eclogites has been debated since the pioneering work of Eskola
Figure 5. Annotated photograph of the Piampaludo, Italy, deposit from
the Orba River gorge. E, eclogite.
(1921). Recent reviews are by Griffin and Mork (1981), Griffin
and others (1985), and Griffin (1987).
Formation of this terrane is complex, as one might imagine.
Mafic rocks of several origins apparently were metamorphosed to
ludo orebody, eclogite is dominantly unsheared and unaltered. eclogite during the early Paleozoic. Among the precursors were
Serpentine partings are rare, and the eclogite is otherwise quite mafic portions of a layered ultramafic-anorthositic-gabbroic-
homogeneous. A strong planar fabric dipping south is pre- or mangeritic complex possibly emplaced in supracrustal rocks.
synmetamorphic. Other precursors were mafic dikes and sills. Metamorphism of
Ore in thin section shows porphyroblastic garnet and pyrox- the entire terrane under eclogite-facies conditions preceded pref-
ene megacrysts in a mass of blue to green amphibole, much of it erential retrogression of felsic and supracrustal host rocks to
foliated, along with epidote, micas, and rutile (Fig. 6). Rutile granulite and amphibolite facies. In addition, dismemberment
occurs in aggregates strung out parallel to foliation. The aggre- into isolated bodies occurred by boudinage of competent mafic
gates commonly measure 1 to 2 mm wide and contain 40 to 90 rocks and by injection of granitoid melts, either before or after
percent rutile, mostly as crystals of 100 fim or more minimum eclogite-facies metamorphism. Commonly, eclogite bodies are
diameter (Fig. 6). Other minerals in the aggregates are amphi- present as horizons or even as thin cumulate layers, in bodies that
bole, epidote, ilmenite, and minor garnet. Alteration rims of also contain garnet peridotite, anorthosite and troctolite, manger-
sphene around rutile are minor in most specimens. ite, and/or oxide-rich bands. The margins of bodies of eclogite
The literature reports the rutile contents of Piampaludo ore are commonly retrograded to amphibolite.
only by chemical analysis for Ti02 (Mancini and others, 1979), Eclogites of ferrogabbroic composition are known as large
an ambiguous procedure in the presence of sphene. Based on my bodies only in the Sunnfjord region (Fig. 1). The eclogite bodies
point counts of five typical, well-distributed specimens, rutile av- are limited mostly to one particular unit in the gneiss terrane
erages 5.3 percent by volume. Individual specimens range from (Skjerlie and Pringle, 1978; mapped by Kildal, 1970). In this
16 E. R. Force
region, the eclogite bodies are essentially large xenoliths, as most
of the unsheared eclogite margins are against granitoid rocks
containing abundant small retrograded mafic xenoliths (Naustdal
southern contact and Kvineset lens 2, Krogh, 1980, Fig. 2; Enge-
bafjellet southern contact, my observation).
The less retrograded portions of these ferrogabbroic eclogite
bodies contain as much as 6 percent rutile (by weight). Rutile
resources of the district have been evaluated by Korneliussen and
Foslie (1986). Detailed maps are contained in Korneliussen
(1980). Three of the larger bodies, Naustdal, Engebofjellet, and
Fureviknipa, contain at least 100 million metric tons of eclogite
each and average 2.7 to 3.1 percent rutile. Part of the Naustdal
body is the site of a town.
Rutile occurs as grains averaging 0.1 to 0.2 mm in diameter,
commonly in aggregates, among megacrystic omphacite, euhed-
ral garnet, and minor interstitial amphibole (Fig. 7). Microprobe Figure 7. Photomicrograph of rutile-bearing eclogite from Engeböfjellet,
Norway, showing rutile (dark), omphacite megacrysts, garnet, and ma-
analyses by Korneliussen and Foslie (1986) show that rutile con-
trix amphibole and quartz. Transmitted plane light, 6-mm field.
tains 98.9 to 99.4 percent Ti0 2 , with vanadium the primary
contaminant. Ilmenite is present with rutile as fine intergrowths
throughout the bodies and as rims on rutile aggregates in retro-
graded zones irregularly distributed through the bodies.
average rutile content of the bodies is not given, but Ti0 2 con-
Shubino Village deposit, U.S.S.R. tents range from 3.3 to 5.35 percent. Some of this Ti0 2 , however,
is present as small rutile inclusions in garnet and as sphene.
Eclogites and related rutile-bearing rocks (Chesnokov, Chemical analysis of rutile from this eclogite shows 94.2
1960) occur in schist terrane of the southern Ural Mountains near percent Ti0 2 , with Fe 2 0 3 and A12C>3 the principal impurities
Shubino Village (Fig. 1). The eclogites are of ferrogabbroic com- (Chesnokov, 1960, Table 7). No analyses are listed for several
position and form bodies up to 200 x 1,000 m, elongated con- elemental contaminants of interest.
formable to foliation of the schist matrix. Retrograding along the
margins of these bodies is advanced, but in their interiors, rutile is Summary of eclogite-type deposits
said by Chesnokov to be fresh and its distribution homogeneous.
Such zones are up to 60 m thick. They apparently have been Some features shown in common by the three known rutile
mined for rutile (Blake and Morgan, 1976). deposits in eclogite are ferrogabbroic parent with consequent
Rutile is present mostly in the groundmass. Median grain Ti0 2 enrichment relative to normal eclogite; unaltered, un-
size of the rutile is 0.12 mm (Chesnokov, 1960, Fig. 7). The sheared eclogite with relatively homogeneous rutile distribution;
dimensions from 0.1 to 4 km2; and rutile grain size averaging 0.1
mm or greater. Rutile grades are 3 percent or greater, and rutile
tonnages range from about 0.1 to 10 million metric tons per
body. Host terranes of these eclogite bodies vary widely. Eclogites
of ferrogabbroic composition appear to occur in swarms that may
not include eclogites having more normal compositions.
Garnet composition, as well as rutile abundance, shows the
effect of ferrogabbroic composition in all three rutile deposits.
The garnet is almandine (Chesnokov, 1960; Binns, 1967; Ernst,
1976), whereas more pyropic garnet is normally characteristic of
eclogite. The omphacitic composition of pyroxene is unaffected.
A luminosilicate-rutile deposits
Probably the second most important type of metamorphic
titanium-mineral deposit is the aluminosilicate type. These depos-
Figure 6. Photomicrograph of Piampaludo rutile deposit showing fo-
its also contain rutile as the economic titanium mineral. The
liated aggregate of rutile crystals (dark), euhedral garnet, and omphacite deposit type has been reviewed by Geijer (1964), Wise (1975),
megacryst (upper right). Transmitted plane light, 2-mm field. Marsh and Sheridan (1976), and Schmidt (1985).
These deposits apparently form mostly from volcanogenic
parent rocks by premetamorphic to synmetamorphic hydrother-
mal stripping of some chemical constituents, especially alkalies.
Removal of calcium and iron is most pertinent to the formation
of rutile. Titanium and aluminum are left behind, residually en-
riched by the removal of the other constituents. Fluorine and
perhaps phosphorus may be added. Rutile crystallizes largely
because of the paucity of chemical constituents that stabilize
competing titanium minerals.
The mineralogy of these deposits is quite distinctive, varying
largely with metamorphic grade. Aluminosilicate minerals are
abundant and may dominate over quartz; these may vary, how-
ever, from sillimanite at higher metamorphic grades through kya-
nite and andalusite to pyrophyllite at lower grades. Topaz
(aluminum fluorine silicate) and aluminous phosphates such as
lazulite are common. Pyrite is locally abundant.
Figure 8. Annotated photograph of the White Mountain (Champion)
Chemical stripping at constant volume has resulted in open deposit, California, after Marsh (1979). Sch, schist; qmp, quartz monzo-
space in some deposits of this type. Rutile and other minerals may nite porphyry; ba, bleached argillic rock; qtar, quartz-topaz-andalusite-
thus be euhedral. Some of the more famous rutile mineral locali- rutile rock.
ties, such as Graves Mountain, Georgia, occur in deposits of this
type.
Rutile grade in these deposits is commonly about 1 percent
or less, and tonnages are commonly small. Thus they are promis- for andalusite, is near Bishop, east of the Sierra Nevada (Gross
ing as resources only where the other constituents of the rock can and Parwell, 1969; Wise, 1977; mapped by Crowder and Sheri-
be mined as primary products or where the rock is locally en- dan, 1972). It is resistant andalusite-topaz-quartz rock (Fig. 8)
riched in Ti0 2 . Both cases do occur. that apparently formed by hydrothermal leaching of greenschist-
The kyanite, pyrophyllite, and other aluminosilicate deposits facies volcanogenic rocks. Rutile is most commonly present in
of the southeastern United States are examples of deposits in concentrations of less than 1 percent by weight, but it is present in
which rutile is a possible byproduct. Figures in Espenshade and concentrations averaging about 2 percent in an area 60 x 1,300 m
Potter (1960) imply rutile resources of about 300,000 metric tons (reconnaissance observations in 1981 of Force and Marsh). On
in active and marginal kyanite deposits. Rutile grade averages this basis, Force and Lynd (1984) list a rutile resource of 300,000
about 0.4 percent, and average grain size appears to be less than metric tons. Rutile grain size is highly variable, from 0.01 to 0.2
0.1 mm. mm. The most economically attractive lithology is an atypical
The pyrophyllite-topaz-andalusite rocks of the same region, saccharoidal breccia with about 4 percent rutile confined to the
mined for pyrophyllite, also contain rutile as a characteristic ac- pink matrix (Fig. 9).
cessory mineral (Schmidt, 1985), mostly as aggregates of rutile
crystals only rarely coarser than 20 * 50 nm (Schmidt, oral
communication, 1986). No rutile resource calculations for this
type of occurrence have been attempted. Two deposits of this
type having greater economic potential conferred by greater Ti0 2
contents are the Evergreen deposit of Colorado and the White
Mountain (or Champion) deposit of California.
The Evergreen deposit, in the Colorado Front Range, is a
sillimanite-quartz-topaz gneiss forming a deformed stratiform unit
up to 30 m thick in high-grade gneiss terrane (Sheridan and
others, 1968; Marsh and Sheridan, 1976). Some sections through
the unit contain 2 to 5 percent rutile. Rutile is equant to prismatic
and commonly is 0.5 mm in lesser diameter. Rutile is fairly free of
contaminants other than iron and calcium, and its Ti0 2 content is
everywhere greater than 98 percent (Marsh and Sheridan, 1976,
Table 2). Resources listed by Force and Lynd (1984) are 200,000
Figure 9. Photomicrograph of White Mountain (Champion) breccia
metric tons of rutile. Residential land use prevents the mining of with topaz-rich fragments. The matrix contains more than 10 percent
this deposit. rutile, in grain sizes from 10 to 100 microns. Transmitted plane light,
The White Mountain deposit of California, formerly mined 6-mm field.
18 E. R. Force
Metamorphic-metasomatic blackwalls Eclogite-type deposits are apparently the only metamorphic
deposit type of potential world importance. The smaller alumino-
The Dinning prospect of Harford County, Maryland, is the silicate and blackwall deposits probably warrant exploration only
only blackwall-type deposit thought to have economic potential. for another commodity as primary product, and warrant evalua-
A chlorite rock there separates a serpentinite body from tion as rutile deposits only where they are already located. The
greenschist-facies schist and contains coarse rutile, apatite, and following paragraphs propose an exploration method for rutile in
magnetite (Southwick, 1968). Average rutile content is about 1 eclogite-type deposits; probably no such exploration has as yet
percent but ranges as high as 4 percent (Herz and Valentine, been attempted.
1970). Force and Lynd (1984) list a rutile resource of 700,000 Eclogite-type deposits tend to occur in different lithotectonic
metric tons. The rutile contains appreciable Fe, Si, Al, and Mg. terranes. The explorationist is thus forced to begin with a terrane
The high Fe, Ti, and P contents of this chlorite rock suggest that in which eclogite bodies are known to occur. The first goal in
the parent rock was a ferrogabbro. exploring for a rutile deposit is to find a sizable body of un-
sheared, unaltered eclogite of ferrogabbroic composition. In the
ECONOMIC PROGNOSIS case of districts described in this chapter, locating such eclogites
could have been done to a great extent with existing geologic
The eclogite type of rutile deposit can yield hundreds of maps and existing lithologic and chemical descriptions. This is
millions of tons of ore containing up to about 6 percent rutile. probably true in many other terranes also, because intellectual
The rutile reportedly has a composition that would allow chlo- curiosity about eclogites far predated any economic interest in
ride processing, and it is not too difficult to separate. In the their rutile contents. It appears that eclogite bodies of ferrogab-
current state of change in the world supply of titanium minerals, broic compositions occur in swarms, and these certainly offer the
eclogite-type rutile deposits could emerge as a major source of best hunting.
titanium minerals. Where existing information about eclogite distribution is
The other types of rutile deposits in metamorphic rocks tend scanty, the simplest exploration procedure probably is alluvial
to be either low grade or small. Not even the described examples prospecting for rutile in nonglaciated terranes known to contain
contain as much as a million tons of rutile. Probably such deposits eclogite bodies, followed by physical inspection of eclogite bodies
could not compete in current world markets. for outcrop area, degree of alteration and shearing, and rutile
content and grain size via thin section. Determination of garnet
METHODS OF EXPLORATION composition may be a guide, either in alluvial or bedrock sam-
ples, but in an unexpected way; the ferrogabbroic eclogites that
Titanium oxide minerals in metamorphic rocks are the contain economically interesting rutile contents contain alman-
primary source of those minerals in placer deposits (cf., Force, dine rather than pyropic garnet. Clearly, exploration with pyrope
1980b). Thus, knowledge of metamorphic terranes can be used in will at best find lean rutile in eclogite. In terranes where garnet is
placer exploration. Placer deposits and their exploration are dis- all contained in eclogite bodies, the presence of almandine in
cussed in Chapters 7 through 10. alluvial samples may point to valuable rutile deposits.
Chapter 10.
Titanium oxide minerals in anorthosite-ferrodiorite massifs
Only two suites of igneous rocks contain titanium-mineral The massif anorthosites vary in composition from andesine-
deposits that are currently considered valuable: the anorthosite- type, locally antiperthitic, to labradorite-type, and some have
ferrodiorite massifs discussed in this chapter and alkalic ring marginal facies containing mafic minerals. Plagioclase megacrysts
complexes discussed in Chapter 4. The reason that igneous depos- are commonly separated by afinermatrix representing megacryst
its are restricted to these two suites lies in two factors: (1) the granulation and/or igneous groundmass. Dikes of anorthosite in
titanium in these rocks is predominantly in titanium oxide miner- country rock are common only in the more sodium- and
als, and (2) these rocks include facies that have high Ti02 con- potassium-rich bodies. True anorthosites have very low Ti02
tents (Table 1). The behavior of titanium in other igneous rock contents (Table 1) and normally contain few titanium oxide
clans is discussed in Chapter 5. minerals.
The role of metamorphism discussed in the previous chapter In the anorthosite-ferrodiorite massifs, it is the ferrodiorites
extends, in my opinion, to the anorthosite-ferrodiorite igneous and related rocks that are enriched in Ti02 and contain titanium
massifs. These rocks are characteristically found in high-grade oxide minerals in abundance. These bodies of ferrodiorite to fer-
metamorphic terranes (Anderson and Morin, 1969; Ashwal, rogabbro contain significantly more Ti02 and P2O5 than do
1982b).1 In many districts, such as Sanford Lake, New York, and other diorites and gabbros. Their total iron oxide content is also
Roseland, Virginia, there is adequate evidence that emplacement unusually high, especially relative to MgO (Emslie, 1978). These
was during and/or before metamorphism, that is, some high- bodies form igneous sheets structurally overlying the older
grade metamorphism and attendant deformation affected the ig- anorthosites, and they commonly send dikes into anorthosite and
neous rocks. Thus the equilibria governing the partitioning of other country rocks.
titanium between silicates and oxides in metamorphic rocks also The petrogenesis of this igneous suite is not discussed here,
apply to this clan of igneous rocks. This chapter explores a transi- except for those aspects bearing directly on titanium oxide miner-
tion to similar igneous assemblages in more stratiform bodies als. This topic is complex and is the subject of a large and
emplaced in lower-grade metamorphic terranes. contentious literature.
The anorthosite massifs occurring in metamorphic terranes Titanium-bearing oxide minerals present in this igneous
of high grade were emplaced partly by igneous processes and suite include ilmenite, magnetite, hematite, ulvospinel, and minor
partly by tectonic or diapiric processes and are deformed, com- rutile. Intergrowths among these minerals are common. The pres-
monly into domes. The massifs are accompanied by ferrodiorites, ence or absence of ilmenite as separable single crystals relatively
ferrogabbros, charnockites, and rapakivi granites (Emslie, 1978). free of intergrowths is a key feature determining the economic
The ferrodiorites and gabbros are typically younger than anortho- value of a deposit. In general, Ti02 present in magnetite solid
site, and they show distinct geochemical relations (Ashwal, solution or asfineintergrowths in magnetite is valueless.
1982b), but their association with anorthosite is so strong that Deposits related to anorthosite-ferrodiorite massifs are dis-
there must be a cogenetic relation. The reported range in age of cussed in this one chapter, though they are of two distinct types.
anorthosite massifs is rather narrow, from 1700 to 900 m.y. old In one type, the deposits are true igneous rocks formed from
(cf., Anderson and Morin, 1969). titanium-rich liquids. In the other, high-temperature metasoma-
tism between igneous rocks and titanium-bearing wall rocks
formed the deposits. The geologic setting of the two types of
deposit is similar; indeed, at Roseland the two types are present in
'For a contrary opinion emphasizing more calcic anorthosites, see Morse
(1982).
the same district but did not form at the same time.
19
20 E. R. Force
MAGMATIC ILMENITE DEPOSITS TABLE 7. A CLASSIFICATION FOR THE DESCRIBED DEPOSITS OF
THE ANORTHOSITE-FERRODIORITE TYPE*
Ilmenite deposits of magmatic origin are currently the most
important type of titanium-mineral deposits in igneous rocks. Magmatic ilmenite deposits Contact-metasomatic rutile deposits
They constitute approximately 30 percent of both world Massif deposits Hybrid massif- Anorthositic Albititic
stratiform deposits
titanium-mineral reserves and of current production (from data in
Lynd, 1988).
Sanford Lake Duluth Complex Roseland (part) Krageriä
Fades
district (I, fg, m+n) Pluma Hidalgo Beaver Creek
(a+l, an+fg,
m+n) San Gabriel Flange
In detail, magmatic ilmenite deposits differ considerably (a, an+fg, n)
from one another and show great variation internally. The varia- Allard Lake
tion can perhaps be best understood if the deposits are considered district
to occupy positions in a polygon or matrix, with end members (a, an+fg, m)
reflecting three types of facies variation, each of which varies Tellnes
somewhat independently (Table 7): (1) andesine versus labrador- (a, an, m)
ite anorthosite association, (2) ferrodioritic versus anorthositic Roseland (part)
(aa, an+fg, n)
deposit host, and (3) nelsonitic versus massive oxide facies. There
Laramie Range
is also a transition to stratiform hosts, which is discussed sepa-
(I, an, m+n)
rately; the description here is restricted to anorthosite-ferrodiorite
massifs containing massive to layered rocks rich in oxides, pre-
"Facies shown in parentheses. Abbreviations of facies: Anorthosite
dominantly ilmenite. Within the framework of facies variations, association, I = labradorite, a = andesine, aa = alkalic andesine. Deposit
these deposits show remarkable similarities. host, an = anorthosite, fg = ferrodiorite-gabbro. Mineralogie fades, n =
Andesine versus labradorite anorthosite association. nelsonitic, m = massive-oxide
Massif anorthosites show a spectrum from bodies of labradorite
composition to bodies of andesine composition (Anderson and
Morin, 1969). Andesine in the latter commonly is antiperthitic rock, and locally the dikes may cut other units, such as gabbro or
and in some bodies contains as much as 4 percent K2O. preanorthosite gneiss.
The transition in anorthosite composition corresponds to In the ferrodiorite-gabbro facies, ilmenite-rich lithologies
sympathetic transitions in the compositions of associated mafic characteristically form layers as thin as fractions of a centimeter.
rocks and to the nature of associated iron-titanium oxides. Labra- These may define the layering in ferrodiorite or may be parallel to
dorite anorthosite massifs are associated with gabbros and with other layering and are concordant to the base of the body. Iron-
oxide-rich rocks containing magnetite, ulvospinel, and ilmenite. titanium oxide minerals and apatite together form a net-vein
These phases are commonly too finely intergrown for effective system or igneous cement, interstitial to cumulate pyroxene, oli-
separation. Andesine anorthosite massifs are associated with fer- vine, or even plagioclase, in particular cumulate layers. For brev-
rodiorites and with oxide-rich rocks containing magnetite, hema- ity, these deposits are referred to as concordant deposits in
tite, and ilmenite, with only hematite and ilmenite finely subsequent discussions.
intergrown ("hemoilmenite"). Andesine anorthosites rich in K2O Nelsonitic versus massive oxide facies.
Nelsonite consists
(alkalic andesine anorthosites of Herz, 1969) are accompanied by of ilmenite and apatite, commonly in proportions of about 2:1,
ferrodiorites rich in K2O and SÌO2, which in turn are associated and commonly is an equigranular medium-grained rock. Kolker
with oxide-rich rocks containing nearly stoichiometric inter- (1982) found green spinel and zircon to be characteristic minor
growth-free ilmenite or ilmenite intergrown with hematite but phases. Various varieties of nelsonite have been named on the
largely free of magnetite. basis of major impurities (Watson and Taber, 1913). As origi-
Ferrodioritic versus anorthositic deposit host. In each nally described by Watson and Taber, nelsonite occurred only as
district described in this chapter, the ilmenite-rich rocks cut discordant bodies. However, concordant equivalents in ferrodio-
anorthosite; in younger, more mafic rocks, however, ilmenite-rich ritic rocks have subsequently been discovered; several are de-
rocks parallel igneous layering. In most districts, both types of scribed herein.
ilmenite deposit occur in adjacent rock units. It is here proposed Nelsonitic facies deposits are the predominant type in some
that these deposit types are correlative facies, both postanortho- titanium-mineral districts, such as Roseland and San Gabriel
site but coeval with ferrodiorite or gabbro. Range. In other districts such as Allard Lake, Quebec, and San-
In the anorthositic facies, ilmenite-rich rocks form unambig- ford Lake, they form subordinate facies where the major re-
uous dike-like bodies of variable size, commonly with apophyses. sources are massive-oxide facies.
Contacts are sharp, though they may be modified by metamor- Massive-oxide rock consists of coarse ilmenite with or with-
phic recrystallization. Anorthosite functions merely as country out magnetite, both intergrown with other oxide-mineral phases.
placed in an extensional environment in country rocks of low
metamorphic grade. A nest of layered troctolitic intrusions in an
older anorthosite series makes up the complex. Layering is well
developed but lacks obvious order in terms of mineral composi-
tion, and no ultramafics are present. Late ferrodioritic intrusions
are numerous. Some granulation of megacrysts in anorthosite is
apparent. Mineral compositions are intermediate between those
of massif and stratiform bodies (Fig. 10); plagioclase is about
An55.60, orthopyroxene about En55^5, olivine about F055. In
some subunits of the complex, ilmenite deposits much like those
of massif complexes are present.
Evidence of origin
If the magmatic ilmenite deposits had the compositions of
ordinary rocks, thefieldevidence of intrusion (and their textures)
would undoubtedly have been deemed sufficient to call them
An content of plagioclase
igneous rocks. For a long time, experimental evidence against
Figure 10. An-En diagram for stratiform versus massif anorthosites (from
liquids of this composition existing at reasonable temperatures
Anderson and Morin, 1969; see also Raedeke, 1982), showing composi-
tions of Duluth Complex, Minnesota. Shaded field from South Kawi-
prevented general acceptance of an igneous origin. However,
shiwi anorthosites (M. P. Foose, written communication, 1986); Buddington and others (1955), Hargraves (1962), and Lister
N, Nathan's series; D , Duluth area, both averaged from Weiblen and (1966) were not persuaded by this evidence and proposed not
Morey (1980, Fig. 5). only igneous origins for these rocks but a parent liquid immiscible
in the main silicate magma (another taboo for early experimental
penologists). More recent experimental evidence (Philpotts,
1967; Wiedner, 1982; Bolsover and Lindsley, 1983; Epler and
Other phases such as silicates and apatite may be minor; in fact, others, 1986) shows that liquids of this composition may indeed
sulfides are locally the most abundant impurity. In some deposits, be immiscible in ferrodioritic magma as it drops below 1,000°C,
graphite is reported. fluxed by either phosphate or elemental carbon. Thus, an igneous
origin involving immiscible liquids for these rocks is respectable
A transition between massifs and layered intrusions once more.
In more detail, thefieldrelations and texture of the deposits
support the immiscibility hypothesis and provide a glimpse of
Magmatic ilmenite deposits typically are associated with
three stages in deposit formation. Figure 11 shows the relation
massif anorthosites and related rocks. In contrast, stratiform
layered mafic intrusions do not have comparably valuable among these stages in a cross section of the base of a crystallizing
ferrodiorite sheet.
titanium-mineral deposits; much of the Ti02 in these bodies is
tied up in magnetite, and separable ilmenite is subordinate (seeThe first stage is the unmixing from cooling ferrodioritic
Chapter 5). Both magnetite and ilmenite are commonly cumulate
magma of small spherical droplets of a titanium-rich liquid im-
in origin in layered intrusions. miscible in the main parent magma. These droplets sink because
of their great density. In most districts thisfirststage is texturally
Some igneous complexes containing anorthosites and ferro-
represented only by spherical inclusions in silicate phases (Fig.
diorites show characteristics that make them difficult to classify as
12), probably as a result of the difficulties of preserving sinking
either massif or stratiform type (cf., Romey, 1968). Two such
droplets in mid-fall and of retaining droplet shape during crystal-
bodies that contain titanium oxide deposits are described in the
lization of both oxide minerals and surrounding silicates.
section on major deposits. Of these two, the rocks of San Gabriel
The second stage is the arrival of these droplets at the
Range are more closely related to the massif type, and the Duluth
Complex is more closely related to the layered intrusions.magma floor, that is, at the top of the cumulate pile. This stage is
recorded by the ilmenite deposits concordant to cumulate layer-
The anorthosite-syenite-ferrodiorite suite of the San Gabriel
ing in ferrodiorite and related rocks. The deposits in San Gabriel
Complex is described by Carter (1982a; 1982b, p. 208) as both a
Range of California provide an example.
layered complex and a massif. It is typical of an andesine anorth-
osite massif except for the orderly layering in the body, which These concordant ilmenite-rich bodies are present along the
floors of ferrodiorite sheets and as cumulate-like layers within
consists of nine superposed layers. The body is also unusual in its
them (Fig. 11). Oxides (with or without apatite) are interstitial to
lack of postintrusion metamorphism and deformation; this may
have enhanced the preservation of layering. cumulate phases and poikilitically enclose them (Fig. 13). This
The Duluth Complex (Weiblen and Morey, 1980) was em- texture is present even in the thinnest cumulate layers (Fig. 14), as
22 E. R. Force
Magma
&
Figure 13. Photomicrograph of ilmenite, magnetite, and apatite enclosing

cumulate olivine in layered ferrodiorite, Rattlesnake body, San Gabriel
Range, California. Transmitted plane light, 6-mm field.
Cumulate floor < .••' / first noted by Bateman (1951). Thus the concordant deposits of
titanium oxides, though exceedingly cumulate-like in outcrop
form, cannot be normal cumulates because texturally they are
interstitial to known cumulate solids. This interstitial position in
some described deposits is clearly not the result of metamorphism
Country rock and deformation. Nor, given the geometry of deposition, can it be
due to filter pressing or residual liquids. The evidence points
rather strongly toward accumulation of a heavy liquid enriched in
Figure 11. Diagram of a cumulate pile and overlying ferrodioritic titanium, in the manner shown in Figure 11. Droplets, if supplied
magma chamber, showing the relation of three stages in the formation of in sufficient quantity, coalesce to form an interstitial matrix in the
magmatic ilmenite deposits. Solid pattern is dense immiscible liquid;
uppermost part of the cumulate pile and poikilitically enclose
crystals represent any cumulate solid, dotted where separated by other
silicates. cumulate phases. The titanium-rich interstitial fluid remains in the
uppermost part of the pile, however, probably because poikilitic
silicate crystallization has already filled interstitial positions
farther down in the pile (Fig. 15). Therefore, this interstitial fluid
remains concordant to cumulate layering in form. It is able to
permeate even the narrowest selvages in uncemented cumulates,
however, because of its extremely low viscosity (Kolker, 1982).
Silicate liquid may be trapped (Fig. 16) in this interstitial net-
work, just as cumulate solids may contain trapped droplets (Figs.
11,12). Where droplet abundance is insufficient for titanium-rich
liquids to coalesce in the interstitial spaces of the cumulate pile,
silicate liquids may fill the remaining spaces and encase the
droplets.
The presence in some districts of many discrete cumulate
layers having interstitial oxide minerals, separated by normal
cumulates, probably tells us that unmixing occurs only sporadi-
cally in the magma. Unmixing may be in response to additions of
fresh magma to the chamber, a hypothesis that fits with the
observed local coincidence of interstitial oxide minerals with
cumulate olivine. In other layers in the same localities, however,
Figure 12. Photomicrograph of nearly spherical opaque inclusions in
the transitions from ordinary ferrodiorite to ilmenite-rich layers
olivine, Rattlesnake body, San Gabriel Range, California. Transmitted involve change only in the interstitial spaces; the cumulate phases
plane light, 2-mm field. remain the same (Fig. 15).
The accumulation process, as outlined here, and the field
evidence supporting it, are analogous to sulfide immiscibility in
stratiform mafic bodies, such as that outlined by Scholtz (1936)
and Naldrett (1979). Some deposits show textural evidence of
both oxide and sulfide immiscibility.
CONTACT-METASOMATIC RUTILE DEPOSITS
Although magmatic ilmenite deposits commercially are the

most important type of anorthosite-ferrodiorite titanium-mineral
deposit, another type of deposit does exist. The two types are
commonly assumed to be the same or closely related, but the
second type is better regarded separately, as contact-metasomatic
rutile deposits; they are analogous to skarn deposits. Magmatic
ilmenite deposits originate with ferrodiorite and related rocks,
Figure 14. Weathered cumulate ferrodiorite with thin ilmenite-apatite- whereas contact-metasomatic rutile deposits actually originate
cemented layers toward the top, Santa Clara Divide highway, San with anorthosite intrusion. The relations of anorthosite and ferro-
Gabriel Range.
diorite described in the preceding section imply that the contact-
metasomatic deposits are everywhere older than magmatic
ilmenite deposits found in the same complex.
The third and last stage in the formation of magmatic ilmen- The rutile albitite (kragerite) type of deposit is also described
ite deposits is the escape of this coalesced heavy liquid via frac- here, even though true anorthosite need not be present, because
tures into lower structural units (most commonly anorthosite); the mechanism of deposit formation is similar to such deposits in
here this liquid becomes a discordant magmatic ilmenite deposit anorthosite. Kragerites have historically been likened to anortho-
(Fig. 11), such as those of the Allard Lake district of Quebec. sitic rutile deposits (Watson, 1912; Green, 1956).
This large deposit consists of virtually pure ilmenite-hematite
rock in anorthosite. Locally, features indicative of mineral crystal- Rutile deposits on anorthosite margins
lization followed by resupply of fresh liquid are preserved (Fig.
17). Minor silicate impurities represent xenoliths and/or trapped Alkalic andesine anorthosites contain rutile along their in-
liquids. These bodies are further described in a following section. trusive margins with older country rocks. Rutile is present in a
The sites of accumulation of titanium-rich liquids are prob- marginal lithology that is coarse-grained, like anorthosite in the
ably structural depressions that grew on magma floors. Asso- core of the body, but loaded with impurities such as pyroxenes
ciated fractures aided liquid escape. In some districts, discordant and quartz. Rutile crystals are large in this zone; ilmenite may
magmatic ilmenite deposits form irregular intrusions that die out also be present. Rutile also is present in country rock within a few
with depth (Bateman, 1951; Herz and Force, 1987). meters of the intrusive contacts; here rutile has approximately the
Figure 15. Photomicrograph of transition from layer having oxide- Figure 16. Photomicrograph of spherical inclusion of disordered silicate
mineral oikocrysts to layer having silicate oikocrysts in ferrodiorite of in apatite, representing trapped liquid, Rattlesnake body, San Gabriel
Rattlesnake body, San Gabriel Range. Transmitted plane light, 6-mm Range. Transmitted plane light, 2-mm field.
field.
24 E. R. Force
localities. At one locality, near KragerO in southern Norway, this
rock was named kragerite (kragertfite in modern Norwegian
usage) by Watson (1912). One kragerite body was mined for
rutile and is described separately. The other locality, in the Beaver
Creek area of the northwestern Adirondack lowlands of New
York, was noted by Brown (1983) and isfirstdescribed here. It is
of minor economic importance.
The two localities are strikingly similar in geologic setting,
albitite relations and petrography, and controls of rutile forma-
tion. In both areas, early supracrustal rocks are isoclinally folded
and show dome-and-basin configuration. Metamorphism is
mostly of the upper amphibolite facies, with some metamorphic
clinopyroxene. Both areas contain a variety of aplitic and pegma-
titic intrusives, but rutile is limited primarily to albititic aplites.
Host rocks of the albitites are biotitic amphibolites containing
Figure 17. Massive ilmenite-hematite ore of the Cliff ore body. Allard sphene and scapolite. Rutile albitites form mostly concordant but
Lake district, Quebec. Exposure about 60 m high.
locally discordant intrusives up to 60 m thick; they commonly
show gradational contacts with country rock by decrease of mafic
minerals over widths of up to several meters. Xenoliths of country
same distribution and grain size as ilmenite in unaltered country rock are extensively altered. Albitite is equigranular, consisting of
rock. Where swarms of anorthosite dikes and sills permeate coun- 1- to 3-mm albite grains with minor quartz and variable micro-
try rock, deposits containing appreciable tonnages of rutile may cline and tourmaline. Biotite is present near contacts with country
form. Bodies of alkalic andesine anorthosite are presently known rock. Typical rutile content is 1 to 2 percent; it occurs as equant
in only four places—Roseland and Montpelier, Virginia; Pluma grains and stubby prisms averaging 0.5 to 1 mm in diameter. The
Hidalgo, Oaxaca, Mexico; and St. Urbain, Quebec, Canada; ru- distribution of rutile around contacts of albitite with country rock
tile is present in all four deposits. Rutile deposits at Roseland and reveals an important influence of metasomatism in these contact
Pluma Hidalgo have the greatest economic potential. zones (Figs. 18,19). This large number of features common to the
Rutile mineralization is observed in these bodies along the two localities suggests the existence of a class of deposit, and
upper margins of anorthosite bodies; hence the rutile is probably implies quite an intricate control of the deposits by their geologic
not cumulate in origin. The silicate megacrysts of the marginal environment.
facies locally lie athwart thin dikelets, making significant meg- The New York occurrence may represent an ambient condi-
acryst (and rutile) transport unlikely. The distribution of rutile in tion for this class of deposit, one well short of producing econom-
anorthosite and country rock seems to be most consistent with its ically significant rutile deposits. Albitite bodies are as thick as 10
formation in response to strong chemical gradients present during m, in zones of country rock 20 to 60 m thick (Fig. 19). Rutile
the intrusion of anorthosite at high temperatures (about 850°C or content in albitite ranges from 1.0 to 2.3 weight percent in 15
more for alkalic anorthosites) into country rocks that probably
were recrystallizing under granulite-facies conditions. Many ele-
ments had sharp concentration gradients across the intrusive con-
tact, but the most pertinent for rutile formation are iron and
titanium. Anorthosite contains negligible amounts of both con-
stituents. Titanium remained immobile, as it does in most geo-
logic processes, but iron diffused into anorthosite along with other
elements. In country rock, some ilmenite recrystallized to rutile.
Adjacent to anorthosite, partial melting and recrystallization of
country rock resulted in a coarse-grained rock containing the
same mineral constituents as country rock, except coarse rutile
that reflects diffusion of iron and only local redistribution of
titanium. High metamorphic grade prevented this titanium from
forming sphene.
Rutile-bearing albitites
Figure 18. Amphibolites and crosscutting albitites, Kragera, Norway.
Nearly monomineralic albite rocks, with rutile as the most The shape, size, and orientation of sphene clots in amphibolite are pre-
important accessory, form small intrusives in two well-studied served as sphene-ilmenite-rutile clots in albitite.
analyzed specimens and averages 1.7 percent. Rutile grain size
averages 0.5 mm, with 75 percent coarser than 0.05 mm. Coun-
try rock within 1 m of albitite contacts averages 0.6 percent rutile,
formed at the expense of sphene. In Figure 19, for example, 67
EXPLANATION percent of the section consists of rutile-bearing rock, and average
rutile grade for the entire section is 1.0 weight percent.
Glacial cover
The distribution of rutile in deposits of this type shows that
Aplite, rutile-bearing rutile forms at least partly by recrystallization of the titanium
minerals in country rocks. Formation of ore-grade material, as at
Aplite, not rutile-bearing KragereS, apparently requires that some horizons in country rock
be extraordinarily high in Ti02-
^v Pegmatite
The formation of rutile in albitites shows both similarities to
and differences from its formation in anorthosite-margin rutile
Gneiss, amphibole-
pyroxene-scapolite deposits. In the case of albitite, there is an intrusive low in iron
1.7% and calcium that destabilizes both ilmenite and sphene, under
Gneiss, biotite-amphibole conditions where both are otherwise stable. In anorthosite-margin
deposits, a low-iron intrusive destabilizes ilmenite, and granulite-
Sample locality and rutile facies conditions destabilize sphene.
content (wt. %)
50
Dip of contact MAJOR DEPOSITS
Plunge of fold The following descriptions of anorthosite-ferrodiorite-

related deposits include at least one important deposit of each
subtype (Table 7). In addition, these descriptions constitute doc-
umentation of the geologic relations described previously.
Sanford Lake (Tahawus, Maclntyre) district, New York

The Sanford Lake district, now nearly inactive, is among the
world's most important in terms of cumulative production. The
district is in the heart of the Adirondack Mountains and on the
23% southwestern edge of the Marcy massif of anorthosites and related
rocks. The upstream end of the Hudson River per se is on one of
the Upper Works, or Calamity Brook, oxide bodies.
Ilmenite has been mined in this district since 1942 by N L
Industries and its predecessors. More than 10 million metric tons
of ilmenite concentrate containing 46 percent Ti0 2 have been
produced. The remaining marginally economic resources are at
least 20 million metric tons of ilmenite (Force and Lynd, 1984),
divided among several deposits. The area mined most recently
(until 1983) is the southern end of the Sanford ore body. The
body having the largest remaining resources is near Cheney Pond,
2 km west of the Sanford body. Substantial resources may also be
present in the Upper Works-Calamity Brook area 5 km to the
north.
Anorthosite is the predominant rock of the district, ranging
in composition from An43 to Anss and averaging Anso (Ashwal,
1982a). Antiperthitic lamellae are common. Two facies are pres-
5 Meters
ent: the Marcy Anorthosite containing andesine and labradorite
megacrysts, and the Whiteface Anorthosite, which is finer
grained, more sodic, more mafic, and foliated. The Whiteface is a
Figure 19. Strip map of one section through rutile albitite and its host
marginal facies along the outside of the domical massif and struc-
rocks in the Beaver Creek area, New York. Bigelow Farm, near locality
50 of Brown (1983).
turally overlies the Marcy facies.
26 E. R. Force
site in this type of ore are common; indeed, I have seen intrusion
breccias of anorthosite blocks veined by ore. At the scale of
geologic maps, however, most of the larger lenses of ore in
anorthosite are parallel to ore lenses in adjacent gabbro (Gross,
1968). Garnet selvages are present between ore and anorthosite
host. Ilmenite grain size is typically 2 to 3 mm. Ilmenite is less
abundant than magnetite in anorthosite-hosted deposits (Stephen-
son, 1945; Gross, 1968). Ore of this type is typically almost free
of silicate gangue. Ore bodies may be up to 60 m thick (Stephen-
son, 1945). To my knowledge, economic bodies are nowhere
more than 100 m structurally beneath gabbro in this belt.
Ilmenite-rich bodies of considerable size are also present
in the Upper Works-Calamity Brook area. They form tabular
bodies concordant to gabbro and discordant to anorthosite, sim-
ilar to those of the Sanford deposits (Stephenson, 1945). Their
Figure 20. Photomicrograph of oxide minerals interstitial to cumulate resources were not evaluated by Gross (1968) and have not been
orthopyroxene in gabbro-hosted layers of an Upper Works body, San- reported. Gabbro-hosted ilmenite enrichments show textures as in
ford Lake district, New York. Transmitted plane light, 6-mm field.
the Sanford deposits; oxide minerals are interstitial to cumulate
phases (Fig. 20).
Ilmenite deposits in the Cheney Pond area are hosted en-
tirely by gabbro that is intrusive into anorthosite (Stephenson,
Gabbro2 is subordinate to anorthosite in areal extent in the 1945). Gross (1968) reported that gently dipping gabbro is both
Sanford Lake district, but gabbro is spatially associated with il- underlain and overlain by anorthosite. The thickest portion of the
menite deposits. There is a continuous gradation from gabbro to gabbro that isrichenough in ilmenite to constitute ore occurs in a
ilmenite ore by increase in the number and thickness of oxide-rich minor syncline. The gabbro is finely banded; layers dominated by
layers concordant to fine layering in gabbro. Gabbro intrudes oxide minerals constitute about 20 percent of the rock sequence
anorthosite and contains xenoliths of it (Stephenson, 1945; Gross, and may be more than 1 m thick. The nelsonite collected by
1968). Kolker (1982) near Cheney Pond is now shown by extensive
In the Sanford Lake-Sanford Hill belt of deposits, which is test-pitting to be a minor lithology.
about 1.5 km long, gabbro structurally overlies anorthosite In all these deposits, the oxide minerals consist of ilmenite
(Gross, 1968). Ilmenite deposits are associated with both rock with variable finely exsolved hematite and of magnetite with
types in such a way that gabbro-hosted deposits are called considerable amounts of finely exsolved ilmenite and spinel
hanging-wall deposits at the mine. The economic gabbro-hosted (Ashwal, 1982a). The magnetite contains up to 3 percent V2O5
deposits are concordant to layering in gabbro and are localized (Balsley, 1943; Ashwal, 1982a).
along the base of gabbro bodies. Thicknesses of such deposits are
as great as tens of meters (Gross, 1968), but some ilmenite-
enriched layers are only a few grains thick. Ilmenite in these
deposits is typically 1 to 2 mm in diameter and is more abundant
than magnetite (Stephenson, 1945; Gross, 1968). The host gab-
bros contain from less than 1 to 10 percent apatite and much
disseminated metamorphic garnet. Ashwal (1982a; see also
Bateman, 1951) noted that the typical texture of gabbro rich
in oxide minerals has cumulate pyroxene with interstitial oxides
(cf., Fig. 20); individual cumulate layers grade upward from
oxide rich to oxide poor. Within oxide-mineral domains, magne-
tite and ilmenite grains show an interlocking texture (Fig. 21).
Pyroxene, garnet, apatite, and other gangue minerals in gabbro-
hosted ilmenite ore together average about 30 wt %.
Ilmenite ore in anorthosite of this belt of deposits is highly
discordant, forming dikes and sills of all thicknesses, with apoph-
yses (Stephenson, 1945; Gross, 1968). Xenoliths of anortho-
Figure 21. Photomicrograph of ilmenite-magnetite (pitted) relation in
2
As described in the literature; actually ferrodiorite (P. Ollila, written com- gabbro-hosted ore of the south end of the Sanford ore body. Reflected
munication, 1986). light, 3.5-mm field.
A llard Lake (Lac Tio) district, Quebec richments along their bases. Ferrodiorite consists mainly of plagi-
oclase (An4|_44, a composition related by Hargraves and
The Lac Tio deposit of the Allard Lake district, discovered Bergeron to contamination from anorthosite), orthopyroxene, 8
in 1946, has produced more than 40 million metric tons of ore, to 10 percent apatite, and ilmenite and lesser magnetite together
and larger quantities remain in the ground. At present, this single totaling at least 20 percent by weight. Coarsely intergrown ilmen-
deposit supplies 19 percent of the world's titanium needs (Table ite and hematite, along with magnetite, apatite, and sulfides, oc-
4). An innovative ore-smelting technique has given the ilmenite- cupy interstices (Fig. 22) between cumulate phases and
hematite deposits of this district a competitive edge. The Lac Tio poikilitically enclose them.
deposit is inland 40 km from Havre St. Pierre, on the northern Hargraves (1962) and Bergeron (1986) relate massive
shore of the St. Lawrence seaway (Fig. 1). The mine site itself is ilmenite-hematite ore to ferrodiorite parents. All observers seem
an uninhabited, heavily glaciated, lightly forested area of moder- agreed that the massive oxide rock is igneous in origin; indeed,
ate relief, reached via company railroad. Hargraves (1962) and Lister (1966) founded their pioneering
The Allard Lake district, described by Hargraves (1962) and hypothesis of liquid immiscibility for magmatic ilmenite deposits
Bergeron (1972, 1986) and mapped by Hocq (1982), contains on evidence from the Allard Lake district. This is historically
about six important titanium-mineral prospects. The district is remarkable because the evidence is so much better elsewhere.
toward the eastern edge of a large andesine anorthosite massif, The Lac Tio deposit has been described by Hammond
with several inliers of later ferrodiorite rich in apatite and oxide (1952). It is a massive, coarse-grained, subhorizontal igneous
minerals. These ferrodiorite bodies contain massive ilmenite en- sheet more than 60 m thick, in anorthosite. It consists essentially
of platy crystals measuring 10 x 10 * 2 mm composed of coarse
intergrowths of ilmenite (75 percent) and hematite (25 percent).
< \J5M
This ore contains 32 to 36 percent Ti0 2 . Minor constituents of
ore include magnetite, sulfides, apatite, hercynitic spinel, and zir-
con. Magnetite is less abundant than sulfides, which here include
71
several Ni-Co minerals. Rose (1969) presented a K-Ar age of
1025 Ma for a pegmatite that cuts anorthosite but is cut by ore of
the Lac Tio deposit.
A part of the orebody known as the Cliff dips gently east
and contains silicate-bearing bands (Hammond, 1952). The form
of these bands (Fig. 17) is analogous to a fluvial channel deposit
with festoon cross beds. In this case, movement and precipitation
V S^BH
P L N N " + r m T r J i of oxide magma in a passageway through anorthosite produced
the structures.
^IWjjjMLf V^BH Geometric features of the deposit are consistent with deriva-
tion of oxide magma from nelsonitic components immiscible in
ferrodioritic magma. The presence of apatite and magnetite in
concordant enrichments in ferrodiorite (Fig. 22) but not in dis-
cordant massive ore is unexplained.
Tellnes district, Norway

The Tellnes and Storgangen deposits of the southern coast of
Norway have been important titanium-mineral producers since
1902. The Storgangen deposit closed in 1964 and has been sup-
planted by the Tellnes deposit, discovered in 1954 about 2 km to
the south. At present, this deposit supplies 12 percent of the
world's titanium. Total production from the two deposits has
been more than 19.5 million metric tons of ilmenite, and remain-
ing reserves are more than 140 million tons (Krause and others,
1986; Korneliussen and others, 1986). Ilmenite concentrates con-
Figure 22. Photomicrographs of ilmenite enrichments in ferrodiorites of tain 44 to 45 percent Ti0 2 .
the Allard Lake district, Quebec. A . Interstitial texture of oxide minerals The district is in jumbled coastal mountains of the Ana-Sira
and apatite. Cumulate phases in the field of view are orthopyroxene and
anorthosite massif (Fig. 23). It consists of three main deposits:
plagioclase. Transmitted plane light, 8 m m field. B. Same texture in
reflected light, showing exsolved hemoilmenite poikilitically enclosing
Tellnes, Storgangen, and Blafjell; the geology of all three has been
cumulate minerals, 3.5-mm field. summarized by Krause and others (1986). Geologic maps of the
28 E. R. Force
lesser clinopyroxene, olivine, and biotite (Table 8). Magnetite,
apatite, sulfide minerals (including Ni-Co sulfides), and baddeley-
ite are minor constituents; magnetite is localized in one zone
(Fig. 24). The baddeleyite is included within ilmenite (Gierth and
Krause, 1974).
Tellnes ore appears quite homogeneous and typically has a
subophitic texture formed by orthopyroxene and plagioclase
grains up to about 2 mm long. Clinopyroxene is interstitial, and
brown amphibole forms small oikocrysts apparently nucleated by
olivine. In drill core, alternating zones of oxide-rich and oxide-
poor rock are visible (Ragnar Hagen, written communication,
1987).
Enclosed by this subophitic texture of silicate minerals is
another texture that contains most of the ilmenite in the rock
(Fig. 25). In this texture, globules about 0.2 to 0.5 mm in diame-
ter consist mostly of single ilmenite crystals and their hematite
Figure 23. View from the western end of Tellnes mine, Norway, with
mountains of Ana-Sira anorthosite massif in background. Anorthosite exsolution lamellae, with subordinate magnetite, sulfides, apatite,
(A) overlies dark-colored ore (O) forming a J-shaped cylindrical intru- green spinel, and baddeleyite (Table 8). Some globule shapes are
sion in this down-plunge view (opposite to view in Fig. 24). slightly modified by ilmenite crystal faces. Two or more globules
are commonly in contact, thus forming a texture analogous to
4
district are by Krause and others and by Falkum (1982). The bunches of grapes (Fig. 25A); adjoining globules have planar
Ana-Sira massif consists of andesine anorthosites and interlayered mutual walls. The "grape-bunch" texture may be enclosed either
leuconorites that are dated at about 900 Ma (Pasteels and others, in orthopyroxene or by the outer portions of plagioclase crystals,
1979), are intrusive into high-grade metamorphic rocks (Wilmart and triaxial interglobule spaces are also filled by orthopyroxene
and Duchesne, 1987), and are structurally overlain by the or plagioclase (Fig. 25). The cores of plagioclase crystals are free
Bjerkreim-Sokndal lopolith (Duchesne, 1972). This layered in- of globules, except in marginal zones of the orebody. In some
trusion consists of lower leuconorites and of upper monzonorites specimens, magnetite is present both in globules and as coarser
with unusually abundant ilmenite and magnetite (in interstitial separate anhedral grains. In specimens particularly rich in oxides,
positions), fayalite, and apatite (Duchesne and others, 1987). The both ilmenite and magnetite are coarse and anhedral, and the
monzonoritic rocks should be called ferrodiorite, based on their "grape-bunch" texture is not apparent. Minor late ilmenite shows
mineralogy, the composition (antiperthitic, An <40) and abun- other morphologies (Gierth and Krause, 1973).
dance (>50 percent) of feldspar, and their high Fe/Mg ratios The marginal zone of ilmenite norite, in contact with
(from Duchesne, 1972, Fig. 2; Krause and others, 1986, Table 2). anorthosite, shows a variation of globule morphology that proba-
Tellnes deposit. This deposit is a steeply dipping discordant bly represents morphology at intrusion, encased in the earliest
tabular intrusion of ilmenite norite3 in anorthosite. Contacts with crystallized silicates (Fig. 25B). Antiperthitic plagioclase nu-
anorthosite are sharp, and intrusion breccias and anorthosite xe- cleated on the contact encloses a variety of spherical, ellipsoidal,
noliths are locally present along it. The orebody is 2,700 m long and budding globules in a wide range of sizes.
and as much as 400 m wide. In cross section it is a J-shaped body The oxide-rich globular domains may represent former drop-
(Figs. 23, 24) in anorthosite, generally dipping steeply to the lets of immiscible oxide-rich liquid. If so, high surface tensions
south (Ragnar Hagen, written communication, 1987); the true and high silicate magma viscosity must have retarded droplet
thickness of ore is 200 to 250 m. This shape is apparently an merger until silicate crystallization preserved them.
original feature of intrusion; it plunges gently east-southeast under Storgangen deposit. The Storgangen deposit is also an
anorthosite. Ore is associated and apparently coeval with a set of elongate discordant intrusion in anorthosite, subparallel in strike
mangeritic dikes that physically link the Tellnes body with the to Tellnes but dipping to the north. It has been described in some
lopolith. Mangeritic rock also forms a marginal zone of the ore detail by Krause and Pape (1975, 1977). At the western end of
intrusion. the body, it is in contact with the base of the Bjerkreim-Sokndal
The ore averages 18.4 percent Ti0 2 and contains (by vol- lopolith and is concordant to it, but eastward the body cuts
ume) 53 percent homogeneous andesine averaging A n ^ O ^ g in down-structure into anorthosite, which thus forms both the hang-
composition, 29 percent (39 percent by weight) ilmenite with ing wall and footwall of the body. The Storgangen body is offset
hematite and spinel lamellae, and 10 percent orthopyroxene, with by mangerite dikes that are related to the Tellnes body.
4
'Again, I would call this rock ilmenite ferrodiorite based on its mineralogy This texture is not, however, the same as botryoidal texture in authigenic
and chemistry as reported by Krause and others (1986). Henceforth I will simply minerals.
call it ore.
SE NW
/ " - M a g n e t i t e zone
\ Tellnes /
Ana-Sira
anorthosite
i
2 km (approximately)
Figure 24. Schematic relation of Tellnes and Storgangen deposits and base of Bjerkreim-Sokndal
lopolith in NW-SE cross section. Restoration of the latter two to nearly horizontal rotates the Tellnes
body to nearly vertical, with a nearly horizontal mineralogic boundary. Internal layering dashed.
In contrast to the Tellnes body, the Storgangen deposit is

TABLE 8. BULK CHEMISTRY, MODAL ANALYSES, AND MINERAL
strongly layered (Fig. 26) throughout its 50-m thickness. Oxide
CHEMISTRY (WHERE KNOWN) OF
minerals are more abundant in dark layers near the base of the
TELLNES AND STORGANGEN BODIES, NORWAY*
body. Bulk composition is similar to the Tellnes body (Table 8),
Modes Storgangen Tellnes but layers vary from leuco-"norite" to ilmenite "norite."
(vol. %) Oxide minerals are interstitial to cumulate crystals of ortho-
pyroxene and plagioclase throughout the Storgangen deposit
Quartz 0 0 (Fig. 27; see also Krause and Pape, 1977, Fig. 15). Ilmenite only
Plagioclase 52.3t (An 43-55) 53.2 (An 35-55) locally encloses pyroxene poikilitically here, as interstitial do-
Orthopyroxene ? (Fs 25-30) 10.2 (Fs 20) mains alternately consist of ilmenite, magnetite, and minor green
Clinopyroxene P 0.8 spinel. Minor apatite and sulfides are also present (Table 8). The
Biotite P 3.9
texture of this body suggests that dense immiscible liquid accumu-
Olivine 0.9
lated in the cumulate pile.
Apatite
Ilmenite
P
22.2t
0.5
28.6(6.1% MgO)
Relation of the Tellnes and Storgangen deposits. Table
Magnetite P 0.7 (0.7% V 2 0 3 , 0.7% C r 2 0 3 ) 9 compares the compositions of the interstitial domains at Stor-
Sulfide P 0.5 gangen with the globular domains at Tellnes. These compositions
Baddeleyite P P are sufficiently close that the Tellnes globules could plausibly
represent an early stage of evolution of an immiscible liquid
Chemistry Storgangen§ Tellnes similar to that represented at Storgangen.
(wt. %) Cross-cutting relations suggest that Storgangen is slightly
older than Tellnes, but both are closely linked to the Bjerkreim-
Si02 27.9 30.4 Sokndal lopolith, and their compositions are similar. The plane
TI02 19.6 18.4
defined by the Storgangen body is nearly parallel to the base of
AI2O3 8.3 11.7
the gravity-layered lopolith (Fig. 24), so Storgangen should also
Fe203 11.7 7.3
have differentiated by gravity. If the plane of the Storgangen body
FeO 19.2 18.1
7.0 6.1
is restored to an approximately horizontal attitude, most of the
MgO
CaO 3.3 4.4 Tellnes body becomes nearly vertical (Fig. 24). Thus the Tellnes
Na^ n.a. 2.4 body might be expected to show little cumulate layering. Proba-
K2O n.a. 0.6 bly Tellnes cooled and crystallized mostly inward from nearly
S 0.3 0.2 vertical walls. If immiscible heavy liquid droplets were present,
P2O5 0.04 0.09 they would be intercepted in their fall and enclosed in
plagioclase-seeded ophitic intergrowths crystallizing inward from
*From Krause and others (1986), Krause and Pape (1975, 1977), and these walls.
Gierth and Krause (1973, 1974). P = present; n.a. = not analyzed,
Bjafjell deposits. The ilmenite ores at Blafjell were worked
tAverage of 40 specimens from throughout the section, calculated
for iron before either the Storgangen or the Tellnes deposit. The
from Krause and Pape (1977, Fig. 2).
Blaljell deposits, described by Krause and Zeino-Mahmalat
§lntegrated composition from 1963 average mill feed; Ragnar Hägen,
(1970) and Krause and others (1986), are associated with a
written communication, 1988.
coarse ferrodiorite ("norite-pegmatite") intruded into anorthosite
30 E. R. Force
Figure 25. Photomicrographs of globular hemoilmenite crystals (and other minor phases) enclosed by
subophitic orthopyroxene and plagioclase, Tellnes ilmenite norite. A. "Grape-bunch" texture of
hemoilmenite in orthopyroxene. Transmitted plane light, 2-mm field. In reflected light, globules have
planar mutual walls that bound different crystal directions in hemoilmenite. B. Globules in rapidly
cooled zone within 1 cm of anorthosite contact. Reflected light, 6-mm field. Darker host is altered
pyroxene.
Figure 26. Layering in Storgangen ore. Plagioclase and orthopyroxene in

different proportions are the main cumulate phases.
Figure 27. Photomicrograph of interstitial texture of oxide minerals rela-

tive to cumulate orthopyroxene, Storgangen. Transmitted plane light,
6-mm field.
TABLE 9. PERCENTAGES OF MINERALS IN INTERSTITIAL dikes and sills, rutile is almost regional in extent. Rutile concen-
DOMAINS AT STORGANGEN AND IN GLOBULAR DOMAINS AT trations typically are only about 2 percent by weight, which
TELLNES, BASED ON I.OOO-POINT COUNTS. approximately matches the Ti0 2 content of metavolcanic coun-
try rock. In some small deposits, however, rutile exceeds 2 per-
Base of Contact zone of cent; at one of these deposits, the Roseland Rutile mine, rutile
Storgangen body Tellnes body
was mined until 1949.
% rock % domain % rock % domain
Both rutile and ilmenite are present in these deposits; ilmen-
(vol.) (vol) (vol.) (vol.)
ite increases at the expense of rutile passing into country rock.
69.0 Ilmenite is present as single crystals without appreciable
Silicate minerals 72.2
Domain 27.8 31.0 intergrowths.
Ilmenite-hematite 84.4 86.3 The temperature of intrusion is thought to have been about
Magnetite 14.0 6.1 850 °C based on two-pyroxene geothermometry. Country rock
Green spinel 0.9 0.3 must have been at granulite-facies temperatures, here also of
Sulfide minerals 0.6 1.3 about 850 °C, either during intrusion or shortly afterward.
Apatite 0.1 3.2
The two ferrodiorite units are younger and form igneous
sheets concordant to the regional domal structure. However, nu-
merous ferrodiorite dikes are present in the older, structurally
underlying units. The ferrodiorites contain xenoliths of anortho-
along the crest of the Ana-Sira dome. The coarse-grained site and granulite gneiss. The ferrodiorites contain one fewer
ilmenite-hematite ores are, in my opinion, of two types: metamorphic fabric than the older anorthosite and granulite
(1) layered ores along the ferrodiorite-anorthosite contact gneiss and show only amphibolite-facies recrystallization. The
(originally the base of ferrodiorite), with coarse ilmenite intersti- ferrodiorites characteristically have high apatite and ilmenite
tial to cumulate coarse euhedral plagioclase (Fig. 28); and contents.
(2) discordant, nearly monomineralic, ilmenite bodies in Concordant impure nelsonites form discontinuous bodies
anorthosite, apparently with strongly linear shape in the Under- along the bases of the ferrodiorite sheets. These bodies consist of
gruve mine workings. The relation of the Blafjell deposit to the tabular cumulate layers with ilmenite-apatite net veins poikiliti-
Tellnes and Storgangen deposits is not clear. cally enclosing cumulate pyroxenes (Fig. 29). Also present in the
bodies are elliptical leucocratic quartz-feldspar-pyroxene domains
Roseland district, Virginia flattened in a direction parallel to the base of ferrodiorite sheets.
Large ilmenite-rich bodies of this type along the bases of ferro-
The Roseland district contains examples of both contact- diorite sheets were mined until 1971. The texture of the mined
metasomatic rutile deposits and magmatic ilmenite deposits. The bodies is not known, however, because they are intensely
ilmenite deposits are of the nelsonitic type and contain both weathered.
concordant (ferrodioritic host) and discordant (anorthositic host)
deposits. Deposits of all of these types have been mined, but no
deposits are being mined currently. Remaining resources are ap-
proximately 1.5 million metric tons of rutile and 12.5 million
metric tons of ilmenite (Herz and Force, 1987). The ilmenite of
the district is nearly stoichiometric in composition and commonly
is free of intergrowths. The following descriptions are after Herz
and Force; earlier descriptions of the district are by Watson and
Taber (1913) and Ross (1941).
The district consists of a domical alkalic andesine anortho-
site intruded along its upper and outer margins into metasedimen-
tary and metavolcanic gneisses. Anorthosite and gneiss share a
fabric showing granulite-facies metamorphic assemblages. These
units are structurally overlain by younger ferrodiorites that are
unusuallyrichin silica and potash. The age of anorthosite is about
1,050 m.y.; the ferrodiorites are about 980 m.y. old.
Coarse-grained rutile is developed along the contacts of
anorthosite with older country rock, especially the ilmenite-
bearing metavolcanic gneiss. Rutile is present in a marginal im-
pure facies of coarse anorthosite and in country rock, in a zone Figure 28. Hand specimen of Blafjell ore (Norway) showing coarse oxide
that straddles the contact. Where there are swarms of anorthosite minerals interstitial to euhedral cumulate plagioclase.
E. R. Force
Figure 29. Elliptical domains of felsic minerals among cumulate ortho-

pyroxene and interstitial ilmenite-apatite, Roseland district, Virginia.
A. Hand specimen; field of view is 50 mm. B. Photomicrograph of mafic
domain showing net-vein texture of interstitial oxides and apatite among
cumulate orthopyroxene grains. Transmitted plane light, 1.8-mm
field. C. Same texture in reflected light, showing poikilitic enclosure of
cumulate pyroxene by hemoilmenite, 1.8-mm field.
The nelsonite bodies for which the rock type was named (1958) and Carter (1982a, b), contains numerous ilmenite pros-
occur as discordant intrusions. Most of these are too irregular in pects of two types: Discordant ilmenite pyroxenites are present in
shape to properly be called dikes. Most of them are small; I have anorthosite, and concordant bodies enriched in ilmenite, magne-
seen discordant nelsonites thinner than 1 cm. Many bodies consist tite, and apatite are present in ferrodiorite. Economic interest in
of only ilmenite and apatite, as equigranular medium-grained the district has recently been rekindled after a virtual lapse of
rock. There are, however, all transitions from pure nelsonite to more than forty years (Industrial Minerals, 1986).
ilmenite- and apatite-rich ferrodiorite intrusions. Some of the The district forms much of the crest of the western portion
latter have thin cumulate nelsonite accumulations along their of the San Gabriel Range, immediately north of the Los Angeles
bases. suburbs (Fig. 30), and is largely in national forest and shooting
All of the discordant nelsonite bodies occur in country rock preserve. Slopes are steep and the climate semiarid. The range is
just below the base of ferrodiorite intrusive sheets. Several are caught between the San Andreas and San Gabriel faults, so frac-
known from drilling or from stream-gorge outcrops to die out at turing is locally severe. In coherent domains between fractures,
depth. Only the weathered overburden over the thickest nelsonite however, original textural relations are unusually well preserved,
body has been extensively mined, near the town of Piney River. because metamorphism and plastic deformation of the complex
This body has abundant chlorite and minor sulfide impurities. are minimal. It is for this reason that textures from the San
Significant resources remain in fresh rock of this body. Gabriel Range were used as an example earlier in this chapter.
The igneous complex dates from about 1200 Ma and covers
San Gabriel Range, California about 250 km2. It consists of a basal anorthosite unit, a syenite
unit, and an upper ferrodiorite unit (jotunite of Carter, 1982a).
The anorthosite-syenite-ferrodiorite suite of the San Gabriel The complex seems intermediate in character between massif and
Complex, described and mapped by Higgs (1954), Oakeshott stratiform types. Country rock of the complex is the Mendenhall
The cumulate phases are commonly rounded euhedra and
may show textural grading within an ilmenite-rich body. Olivine
is the cumulate phase in many of the ilmenite-rich layers; olivine
is not common in the ferrodiorite elsewhere. The cumulate phases
contain spherical ilmenite inclusions; locally these inclusions are
arranged in rings that apparently outline growth stages of the
cumulate grain. Conversely, apatite shows spherical inclusions of
disordered silicate (Fig. 16).
Laramie Range, Wyoming

An anorthosite massif about 1,400 m.y. old (Smithson and
Hodge, 1972) forms much of the Laramie Range. Plagioclase
composition ranges from An4o to A n ^ and averages about Anss.
The body has been described and mapped by Newhouse and
Figure 30. The San Gabriel Range from Santa Clara Divide, looking Hagner (1957) and Klugman (1966). Little ferrodiorite or gabbro
south into the San Fernando Valley, California. Ferrodiorite occupies the has been reported, but numerous rocks enriched in iron-titanium
foreground and Pacoima Creek (first valley on left side). Mendenhall
oxide minerals form sharply discordant to concordant lenses in
Gneiss occupies the next ridge and San Gabriel Fault, the next valley.
labradorite anorthosite (Diemer, 1941; Goldberg, 1984). En-
riched rocks vary from oxide-rich anorthosite, leuconorite, and
troctolite to younger olivine-bearing massive magnetite-ilmenite
Gneiss, a finely banded gneiss of granulite metamorphic fades. rock. Apatite is present in some enriched bodies (Bolsover and
The igneous complex is arranged in a rude dome, and the upper Lindsley, 1983).
ferrodiorite unit is in exposed intrusive contact with the gneiss. The largest of the enriched bodies, at Iron Mountain, was
The anorthosite unit, the oldest of the complex, consists of mined for heavy aggregate in the 1960s. This deposit has been
coarse calcic andesine. Leucogabbro is locally present in the anor- described by Pinnell and Marsh (1954), Hagner (1968), and
thositic unit and retains a subophitic texture. Overlying syenite is Eberle and Atkinson (1983) and mapped by Newhouse and
similar to anorthosite but has mesoperthitic feldspars. The ferro- Hagner (1951) and Dow (1961). Massive magnetite-ilmenite
diorite unit, the youngest, characteristically contains plagioclase, rock crosscuts an east-dipping contact zone between structurally
commonly antiperthitic; two pyroxenes; ilmenite; magnetite; and overlying troctolite and underlying leucogabbro. The troctolite is
apatite. Several of its five subunits (of Carter, 1982a) show cumu- locally rich in interstitial magnetite and ilmenite and contains
late layering, defined by grain size and/or mineralogic transitions, apatite. Massive magnetite-ilmenite rock contains as much as 50
commonly graded. percent olivine. Ilmenite is present both as intergrowths in magne-
Ilmenite deposits in anorthosite and Mendenhall Gneiss are tite and as coarse separate grains containing exsolved spinel
present as small discordant ilmenite pyroxenites. Apatite contents (G. Turner in Hagner, 1968). The olivine-poor rock toward the
of these deposits vary from 2 to 20 percent. core of these bodies has a high ratio of magnetite to ilmenite, and
The concordant ilmenite deposits in ferrodiorite are in cu- its V2O5 content averages 0.64 percent (Dow, 1961).
mulate layered rocks. My experience with these deposits is largely Massive magnetite-ilmenite rock from Iron Mountain con-
with the so-called Rattlesnake and Saturday Night claims in Pa- tains about 20 percent Ti02, but acceptable ilmenite concentrates
coima Creek and the bodies exposed along the Santa Clara Di- have not yet been made from it (Pinnell and Marsh, 1954; Dow,
vide highway. High-grade layers range in thickness from one cm 1961). Thus, it is questionable whether the deposit should be
(Fig. 14) to more than 10 m. Ilmenite forms more than 20 listed as a resource using the Force and Lynd (1984) definition.
percent by weight of some layers, and can be concentrated into
fractions containing 45 percent Ti02- The bodies share one im- Duluth Complex, Minnesota
portant textural characteristic (Force and Carter, 1986): ilmenite,
magnetite, and apatite form net vein systems interstitial to cumu- The Duluth Complex consists of a series of anorthositic,
late phases in particular layers (Fig. 13). Green spinel may also gabbroic, and troctolitic intrusions dating from about 1100 Ma
form part of the interstitial assemblage and commonly separates (Weiblen and Morey, 1980). Titanium-mineral resources of the
large single crystals of magnetite and ilmenite. Ilmenite poikiliti- Duluth Complex are of two types. The first is ilmenite recovera-
cally encloses cumulate phases and is typically free of intergrown ble as a byproduct of mining for base-metal sulfides in troctolitic
oxides. The edges of the bodies are defined by the substitution of intrusions (Iwasaki and others, 1982). The ratio of ilmenite to
an interstitial silicate phase, poikilitically enclosing cumulate magnetite varies among the sulfide deposits (Pasteris, 1985). Sul-
phases, for ilmenite, magnetite, and apatite in the same positions fide and some oxide minerals are present in interstitial positions in
(Fig. 15). these rocks (Fig. 31; see also Pasteris, 1985, Fig. 6).
34 E. R. Force
Figure 31. Photomicrographs of textural relations of oxide and sulfide

minerals in rocks of the Duluth Complex, Minnesota. A. Interstitial
sulfide and oxide minerals (both opaque) at the base of troctolitic layered
intrusion, Dunka Pit. Cumulate crystals are mostly clinopyroxene (bot-
tom in this view) and olivine (top). Transmitted plane light, 3.5-mm
field. B. Same view in reflected light. Ilmenite gray, sulfides bright.
C. Interstitial pyroxene with cumulate plagioclase, olivine, and ilmenite
in unit G of Nathan (1969). Transmitted plane light, 3.5-mm field.
The other type of resource, of economic interest for titanium

alone, is mostly confined to one small part of the Duluth Com-
plex in the layered series of Nathan (1969), adjacent to the Cana-
dian border. A separate intrusion, the Water Hen intrusion, is
mentioned by Nafziger and Elger (1987) as containing additional
resources of this type.
The layered series of Nathan (1969), probably the oldest of lized in interstices. My specimens suggest that all parties are partly
the complex (Weiblen and Morey, 1980), is systematically en- correct; some rocks are plagioclase-olivine-ilmenite cumulates
riched in iron-titanium oxides. It is in a heavily glaciated, remote with interstitial pyroxene and lesser oxide and sulfide minerals
area of low relief with a maze of lakes. The series consists of (Fig. 31).
successive intrusions of troctolitic and gabbroic compositions, The oxide minerals of unit G consist of ilmenite and magne-
with minor quartz and K-feldspar in a few lithologies. The earlier tite in coarse grains; ilmenite is largely free of included lamellae
intrusions are concordant sheets dipping gently south, and a later but may contain minor magnetite, ulvospinel, and hercynite
set of intrusions is discordant. Nathan's unit G of coarse gabbro is (Nathan, 1969). Magnetite contains abundant lamellae of
the most extensive unit rich in oxide minerals and is among the ilmenite and ulvospinel. Lister (1966) found that this magnetite
youngest concordant bodies. Nathan's discordant units T and V contains 0.3 to 0.4 percent vanadium. The ratio of magnetite to
are closely associated with unit G and contain even higher con- ilmenite averages about 1:1 but varies with stratigraphic position
tents of oxide minerals. within the unit. Lister (1966, p. 293) found magnetite pre-
Unit G, which is as much as 400 m thick, showsfinemodal- dominant over ilmenite in three out of the four oxide-rich layers
graded layering. Major cumulate phases are plagioclase, olivine, of unit G.
and augite. Plagioclase composition in unit G trends from An54 at Units T and V are small discordant bodies containing
the base to Anso at the top. No other simple trends in mineral abundant olivine and augite, respectively. Coarse oxide minerals
composition were found by Nathan (1969). vary from 20 to 70 percent by volume. Apatite contents are
The oxide minerals of unit G vary from about 10 to 60 minor. Separable ilmenite is only up to 10 percent; that is, the
percent by volume; apatite reaches 5 percent in the uppermost of ratio of magnetite to ilmenite is high in the rocks richest in oxide
three subunits. Nathan (1969) considered the oxide minerals minerals. Intergrowths in the oxide minerals are similar to those
cumulate, whereas Grout (1949-1950) considered them intersti- described for unit G.
tial and therefore postcumulate. Lister (1966) proposed that the Grout (1949-1950) appraised the economic potential of
oxide minerals represent a heavy immiscible liquid that crystal- titanium minerals in this area. He focused on oxide mineral en-
large areas. A strong subhorizontal lineation trends northwest-
southeast.
Passing structurally downsection, these gneisses are intruded
by concordant and discordant bodies of impure anorthosite con-
taining antiperthite and pyroxene megacrysts and quartz. These
rocks host the Pluma Hidalgo deposits. Still farther downsection,
toward the bed of the Toltepec River, anorthosite is massive but
deformed and shows the same northwest-southeast lineation. It
seems appropriate to call this body the Toltepec Anorthosite.
Farther down toward the Pacific coast the anorthosite is highly
altered with the addition of much silica and tourmaline.
In the immediate region of Pluma Hidalgo, small intrusions
of the impure anorthosite are abundant and generally contain 1 to
2 percent coarse rutile. The rutile deposits of greatest economic
interest, however, are impure anorthosites that contain from
Figure 32. Rutile-rich rock of the Agua Titania deposit, Piuma Hidalgo
about 2 to more than 50 percent rutile (Fig. 32), probably averag-
district, Mexico, in hand specimen view with rutile reflectant. Scale in
ing 20 percent by weight in the deposits known as Agua Titania
millimeters.
or Las Minas de Tisur. These may be the highest-grade rutile
deposits of any origin in the world. Rutile-bearing intervals aver-
age approximately 20 to 40 m in width over a strike length of at
richments subsequently mapped as part of unit G and possibly as least 600 m, exposed in four perilous adits and in the steep bed of
part of unit V by Nathan (1969). Other enrichments on Little Agua Titania. Wall rocks of these zones are mostly gneiss but
Iron Lake, mostly in unit T, were regarded by Nathan and subse- locally are impure anorthosite with lower rutile contents.
quent students as having greater potential but were not investi- Rutile occurs as coarse single crystals; in high-grade samples,
gated by Grout. Thus resources listed by Grout can only be it has the appearance of a matrix between altered feldspar or
suggestive of total resources of the area. pyroxene megacrysts. Ilmenite is also present in most specimens
Grout (1949-1950) showed that an ilmenite fraction con- as coarse single crystals without intergrowths. In some high-grade
taining more than 40 percent Ti02 could be made from some material, low rutile content is compensated by high ilmenite
oxide-mineral enrichments of the area and that this fraction con- content.
stituted as much as 12 percent of some rock bodies. I now regard Some form of late- or postmagmatic mass transport of tita-
only the 2.1 million metric tons of ilmenite listed by Grout as a nium must have occurred to form the high-grade rutile deposits of
demonstrated resource of separable ilmenite; the mineralogy and Pluma Hidalgo, as deposits of this type normally have lower
size of other enriched bodies included by others, and repeated by rutile contents. The contact-metasomatic hypothesis fits the geo-
Force and Lynd (1984), are insufficiently known. logic relations but can be only one ingredient in the origin of these
rutile deposits.
Pluma Hidalgo district, Mexico
Kragert> district, Norway
The state of Oaxaca contains a large terrane of high-grade
metamorphic rocks and structurally underlying anorthosites Rutile was mined until 1927 from rutile-rich albitites near
(Ortega-Gutierez, 1981) approximately 1,000 m.y. old. These Kragere) on the southern coast of Norway. Probably several tens
rocks host rutile deposits near the town of Pluma Hidalgo, access- of thousands of tons of rutile were produced. The deposits are
ible only by a narrow road hung along a steep seaward-facing now nearly mined out, and today the region is a popular resort
escarpment among rain forest and coffee plantations. Weathering area that is part of a lovely coastal archipelago of glaciated islands
is deep, except in road cuts and streambeds, so exposure of fresh (Fig. 33). Rock exposure is exceptionally good.
rock is poor. The following description of the deposits is from my The deposits and their host rocks have been described by
own observations, supplemented by Paulson (1964) and an un- Breigger (1934-1935) and Green (1956). Regional geology has
published (1957) description by T. P. Thayer. No adequate geo- been reviewed by Starmer (1985a) and mapped by Starmer
logic maps of the district have been published. (1985b). The immediate area of the deposits has been mapped by
The structural sequence is displayed in the face of the es- Green (1956).
carpment and shows gneissic country rocks toward the top and Some aspects of regional geology in the district were de-
anorthosite at the bottom. The gneissic country rocks are mostly scribed in a previous section of this chapter. Rutile albitites are
dark, finely banded quartz-antiperthite-two pyroxene-garnet- present in two types of amphibolitic host rocks. Probably the
graphite-ilmenite gneiss, with numerous thin sills of quartz- more important type is foliated metagabbro containing scapolite
feldspar rock. Banding is folded but remains subhorizontal over and sphene. This type, though relatively homogeneous, locally
36 E. R. Force
Figure 33. The Kragera archipelago, photographed from the former rutile mine at Lindvikskollen,
Norway, looking east to islands of amphibolite, quartzite, and gneiss. Langay is in extreme upper left.
shows well-developed cumulate layering. Where gabbro bodies many of equant angular shape, measuring up to 20 cm perpen-
are little altered, these cumulates include magnetite bands rich in dicular to elongation of the zone. For this reason the term xeno-
Ti and V. The other type of amphibolite is deformed and recrys- lithis more appropriate than schlieren, the term used by Brogger
tallized pillow lava. Locally, as on western LangOy island, albi- (1934-1935).
tite preferentially replaces interpillow hyaloclastite (Fig. 34). The original lithology represented by these xenoliths is un-
Rutile is most abundant in albitite where amphibolite coun- certain, as the train could not be traced into amphibolite country
try rocks are rich in sphene. Locally, the shape, size, and orienta- rock. Two possibilities are discrete pillows in amphibolite and
tion of clots of sphene crystals in amphibolite are reflected by the xenolith trains in metagabbro. Since the metasomatic replace-
distribution of clots of rutile, ilmenite, and sphene in crosscutting ment of xenoliths by albitite is simplest if the precursor contained
albitite bodies (Figs. 18, 35). Thus, metasomatism of country 10 to 30 percent Ti0 2 ,1 suggest that the precursor was a cumu-
rock without Ti0 2 transport is required in at least the marginal late band of titaniferous magnetite in metagabbro, disrupted by
zones of albitite bodies. Green (1956) also reported this the formation of albitite.
phenomenon.
The main rutile ore body at Kragertf was a zone about 2 m ECONOMIC PROGNOSIS
wide, parallel to the walls of a large concordant body of albitite
containing 1 to 2 percent rutile. This zone averages 6 to 10 Magmatic ilmenite deposits in anorthosite-ferrodiorite mas-
percent rutile but is quite inhomogeneous. The zone apparently sifs currently supply about 30 percent of the world's titanium
represents a former train of xenoliths; the sizes and shapes of minerals, and large high-grade resources remain. Objections to
individual xenoliths are outlined by rutile-rich albitite (Fig. 36). pollution from sulfate-process refining could make retention of
The xenoliths differ from one another in texture and configura- this status difficult, but smelting and synthetic-rutile technologies,
tion. Some contain more than 25 percent rutile; others contain which can be used with these ores either singly or in tandem,
tourmaline or corundum and less rutile. The xenoliths include greatly decrease pollution. Thus, magmatic ilmenite deposits
Figure 34. Metamorphosed pillow lavas of western Langay, Kragera Figure 35. Photomicrograph of albitite from locality of Figure 18,
district, with interpillow area replaced by albitite. Lens cap 50 mm Kragera, showing skeletal ilmenite + quartz pseudomorphic after sphene
across. (upper left). Transmitted plane light, 2-mm field.
supply a significant share of world titanium markets; however,
small-scale mining of some high-grade deposits may be profitable.
METHODS OF EXPLORATION
The most valuable magmatic ilmenite deposits occur in an-

desine anorthosite massifs associated with ferrodiorites; contact-
metasomatic rutile deposits occur on the margins of alkalic
andesine anorthosite massifs. Thus, exploration should focus on
andesine anorthosite-ferrodiorite massifs. Labradorite anorthosite
massifs and stratiform mafic complexes are not promising as
sources of economic titanium minerals.
Magmatic ilmenite deposits tend to occur toward the bases
of ferrodiorite sheets and/or in structurally underlying country
rock within about 100 to 200 m of the reconstructed base of the
overlying ferrodiorite sheet. Such deposits seem to be preferen-
tially located in or under synclines in the ferrodiorite. Contact-
metasomatic rutile deposits occur along contacts of alkalic
andesine anorthosites with older country rock, especially along
contacts where numerous dikes and sills of the anorthosite are
intruded. Enrichments in Ti02 to make such deposits economic
may be present only at structurally and/or lithologically favora-
ble sites.
Most of the deposits described here were discovered by
physical exploration, but hidden extensions of magmatic ilmenite
deposits have commonly been found by using aeromagnetic sur-
veys. It would be far more efficient to start with the aeromagnetic
surveys. In doing so, however, it must be borne in mind that the
Figure 36. Rutile ore exposed in pillar of Lindvikskollen mine, Kragero.
most valuable deposits contain ilmenite that is of a stoichiometric
Host rock is light-colored albitite. Rutile-rich albitite is present as darker composition or is intergrown with hematite. Magnetite may be
square-cornered ghosts of former xenoliths. The upper more shadowy present only in the less attractive bodies, as in the ferrodiorite
xenoliths contain biotite, tourmaline, and locally corundum in addition deposits of the Allard Lake district. Therefore, once within an
to rutile. area of magmatic ilmenite occurrence, all anomalies should be
checked regardless of amplitude. Contact-metasomatic rutile
deposits, of course, will show no magnetic anomalies.
probably will be able to compete with other deposit types for Induced polarization methods were found by Elliot and
future world markets. Guilbert (1975) to respond to hemoilmenite ± magnetite concen-
In the United States, these advanced technologies have not trations. Magmatic ilmenite deposits consist in large part of elec-
as yet been used to refine ilmenite from igneous deposits, whereas trically conductive grains in physical contact at least through the
smelting is used in Canada and Norway. This situation is hard to intercumulus spaces, so some conductivity method should be
explain, as some major ores in the United States are amenable to useful in their exploration.
smelting (Elger and others, 1986; Nafziger and Elger, 1987). The Since mineralogy in both types of anorthosite-related
economic future of magmatic ilmenite deposits in the United titanium-mineral deposit determines the economic significance of
States depends less on its geology than on its technology. a deposit, fluvial prospecting for ilmenite and rutile may be effi-
Contact-metasomatic rutile deposits are normally low in cient in nonglaciated areas. For example, detrital ilmenite too
grade; portions sufficiently high in grade to be economic are coarse to be derived from gneisses could be sought and examined
typically small. Therefore, these deposits will probably never for favorable intergrowth type.
Chapter 10.
Titanium oxide minerals in alkalic igneous rocks
A second suite of igneous rocks that contains valuable perovskite relative to sphene (Carmichael and others, 1970); (2)
titanium-mineral deposits is alkalic ring complexes of "miaskitic" high contents of niobium, thorium, and rare earths, which can
type. The term miaskitic, introduced by Ussing (1911), refers to readily be accommodated in the lattices of perovskite, brookite,
nepheline syenites having atomic alkali/alumina ratios of less and rutile but are not reported in comparable concentrations in
than one; those having higher ratios are called agpaitic. Ussing sphene; (3) high C0 2 content, which forms rutile plus calcite at
pointed out that miaskitic nepheline syenites tend to occur in the expense of sphene (Hunt and Kerrick, 1977); and (4) the low
alkalic complexes in which high titanium and iron contents are oxygen fugacity prevalent in alkalic rocks, which favors
present as oxide minerals. Thus the term miaskitic is applied to perovskite over sphene (Haggerty, 1976b).
entire complexes containing miaskitic nepheline syenites. As we
shall see, the two meanings of the term are not everywhere con- TITANIUM-OXIDE MINERAL ENRICHMENTS
sistent. I will use the term rather loosely in the latter sense. A
major class of miaskitic alkalic rocks occurs as ring complexes, The most valuable enrichments of titanium oxide minerals
with carbonatites typically forming the core unit and with acces- occur in miaskitic complexes that contain pyroxenites and alkali
sory carbonate disseminated through the complex. pyroxenites (jacupirangites, etc.). Most such deposits are actually
in the pyroxenites. In the Kola Peninsula of the U.S.S.R., enrich-
MINERALOGY ments of titanium oxide minerals occur in pyroxenites of both
miaskitic (Africanda) and agpaitic (Khibiny, Lovozero) affinity
The important titanium-bearing oxide-mineral phases in based on nepheline syenite composition (Yudin and Zak, 1971).
miaskitic alkalic complexes, in order of probable abundance, are The agpaitic pyroxenites contain sphene in addition to magnetite
perovskite, magnetite, and the Ti0 2 polymorphs rutile and and perovskite.
brookite. These minerals, except for magnetite, commonly In many of the important deposits, the titanium-mineral
contain large amounts of niobium in alkalic rocks (Fleischer and enrichments occur as magnetite-perovskite rocks. These rocks are
others, 1952). Ilmenite is normally minor but may be abundant; not well described in any deposit, but in my experience occur as
indeed, the type locality of ilmenite is in a miaskitic intrusive of small dike-like bodies permeating pyroxenite. Their origin has
the Ilmen Range of the Ural Mountains, near Miask, U.S.S.R. been attributed by Herz (1976) to an oxide liquid immiscible in
(Sarensen, 1974). alkalic magmas of intermediate composition. Relations described
In contrast, the agpaitic alkalic complexes typically contain herein for the Kodal deposit (Norway) may represent an inter-
their titanium mostly in silicates such as sphene, titanaugite, mel- mediate stage in the formation of such liquids.
anitic (or schorlomitic) garnet, kaersutitic amphibole, and several An important titanium resource at Magnet Cove, Arkansas,
unusual Ti-Zr silicate minerals. Factors that suppress the forma- is a contact-metamorphic deposit, formed where alkalic magma
tion of these silicates are responsible for the economic titanium came in contact with silicic sedimentary country rock. Several
deposits in miaskitic alkalic complexes. Possibly sphene is sup- deposits in alkalic rocks have a weathering overprint, and in
pressed by some combination of the following bulk chemical Brazilian deposits the weathered overburden is the resource. Var-
variables in alkaline rocks: (1) low silica activity, which favors iations among deposit types are summarized in Table 10.
39
40 E. R. Force
TABLE 10. RELATION OF THE DESCRIBED TITANIUM-MINERAL DEPOSITS
IN ALKAUC IGNEOUS ROCKS
Deposit Deposits In igneous rocks Contact- Weathered

Pyroxenitic Syenite Other metamorphic deposits
host host host deposits
Powderhorn
district X
Brazilian
deposits X X
Magnet Cove
district
Magnet Cove
Rutile deposit X
Mo-n
prospect X
Christy deposit X X
Hardy-Walsh
deposit X X
Kodal deposit X
M A J O R DEPOSITS Pyroxenite consists of diopsidic augite, locally titaniferous,

with lesser and varying amounts of magnetite, perovskite, biotite,
Powderhorn (Iron Hill, Cebolla Creek) district, Colorado and phlogopite. Minor calcite and apatite are ubiquitous. Impor-
tant varieties of pyroxenite are defined by local abundance of
Claims by the Buttes Oil and Gas Company (Wall Street apatite, olivine, nepheline, feldspar, and melanite garnet contain-
Journal, 1976; Thompson, 1987) that the Powderhorn district ing 5 percent Ti02- Garnet is, like sphene, a late mineral and
contains about 500 million (short?) tons of perovskite ore imply formed apparently at the expense of perovskite. Variations in
that the district is the largest single titanium-mineral resource in both composition and texture make the pyroxenite an exceed-
the United States. However, beneficiation of perovskite to extract ingly inhomogeneous unit. Larsen (1942) considered the varieties
Ti02 has not yet been demonstrated to be commercially feasible. mutually intrusive. The range in Ti02 contents of pyroxenite is
The district has been described by Larsen (1942), Heinrich correspondingly great, 0.3 to 11.9 percent, averaging 6.5 percent
(1966), and Temple and Grogan (1965) and mapped by Hedlund (Rose and Shannon, 1960).
and Olson (1975). The titanium-mineral resources are in an al- Magnetite-perovskite rock forms discontinuous dikes and
kalic ring complex with a carbonatite core in Precambrian crys- lenses ranging in thickness from less than 0.5 m to almost 50 m
talline rocks. It is part of a Cambrian swarm of alkalic intrusives (Hedlund and Olson, 1975). Abundant apatite defines one vari-
in southern Colorado. The oxide-rich nature of the complex is not ety of this rock but normally is a minor constituent. Minor biotite
reflected in nepheline syenite composition (analysis in Larsen, is also present. Larsen (1942) reported perovskite contents as high
1942). as 50 percent and Ti02 contents as high as 40 percent in
The Powderhorn area, at more than 8,000 ft (2,440 m) magnetite-perovskite rock.
elevation, is on the northern flank of the San Juan Mountains and The petrology of this rock has not been adequately de-
slopes toward the Gunnison River. The area is sparsely inhabited scribed; the following observations are my own. Perovskite forms
and forested. The pyroxenite unit, which composes about 70 crystals commonly 1 to 4 mm in diameter, some euhedral
percent of the 30-km2 area of the complex, occupies a long (Fig. 38). Magnetite contains coarse ilmenite lamellae. Ilmenite is
intermontane depression (Fig. 37). In three dimensions the also present as separate grains that grade in morphology into the
pyroxenite is cone-shaped (Temple and Grogan, 1965). lamellae in magnetite (Fig. 38B). Ilmenite locallyrimsperovskite.
Late microcrystalline sphene-anatase intergrowths (apparently
the leucoxene reported in the literature) both rim perovskite and
fill cracks in it. Some lithologies also contain apatite and
clinopyroxene.
The resourcefiguresby Rose and Shannon (1960) and by
the Buttes Oil and Gas Company are based on estimates of the
volume of pyroxenite that is unusually permeated by magnetite-
perovskite veinlets; perovskite is thus treated as if it were dissemi-
nated. About 500 million or more tons of such pyroxenite
average about 12 percent Ti02 but contain only about 8 percent
separable perovskite with 45 to 50 percent Ti02, 1.5 percent
rare earths, 0.5 percent Nb (Thompson, 1987, and oral commun-
ication, 1981), and appreciable thorium (Hedlund and Olson,
1975).
Figure 37. View of the Powderhorn district, Colorado, looking W S W .
Tapira and Salitre (Minas Gerais State) and Pyroxenite occupies the grassy slopes beyond the wooded foreground of
Catalao I (Goias State) deposits, Brazil older Powderhorn Granite. Carbonatite occupies wooded Iron Hill in
center. Photo courtesy of Theodore Armbrustmacher.
The anatase deposits in weathered overburden of alkalic

bodies in Brazil are among the world's largest titanium-mineral
resources; there is some industry speculation that these deposits
could replace more traditional sources of titanium minerals (In-
dustrial Minerals, 1978). Estimates of total resources are 300
million metric tons or more of anatase ore containing more than
20 percent Ti02 (Mineragao Metalurgia, 1977; Beurlen and
Cassedanne, 1981; Turner, 1986) for the three bodies known as
Tapira, Salitre, and Catalao I. Trade-journal articles furnish most
of the available non-Portuguese information on these deposits
(e.g., Harben, 1984; Turner, 1986), although some information
on their parent alkalic bodies is available in older technical litera-
ture (Troger, 1928, 1935) and in more recent brief descriptions
(Heinrich, 1966; Herz, 1976).
Ulbrich and Gomes (1981) list 44 alkalic stocks dating from
40 to 90 Ma, disposed around the Parana basin in Brazil. In
addition to the three bodies containing titanium-mineral re-
sources, other bodies such as Araxa contain important resources
of rare earths, phosphate, baddeleyite, supergene manganese
oxides, and garnierite. Some of the bodies are agpaitic, and
numerous others are poorly known.
The three stocks of interest for titanium have been most
thoroughly described by Alves (I960), Carvalho (1974), Filho
(1974), and Geisel-Sobrinho (1974). All have been mapped by
Barbosa and others (1970), who show them as circular stocks
intruded into Precambrian rocks. Their surface areas range from
about 25 to 40 km2. Annotated air photos of each body, shown
by Barbosa and others, indicate that the stocks are physiographi-
cally expressed as subtle domes largely mantled by residual
debris. Country rock crystalline terranes show moderate relief
and largely grassland vegetation. Annual rainfall averages 1.7 m
(Harben, 1984), permitting deep weathering.
The alkalic stocks are not well known, in part because of
Figure 38. Photomicrographs of magnetite-perovskite rock, Powderhorn
their mantles of residuum. They seem to be sufficiently similar district. A. Transmitted plane light, 2-mm field. Perovskite, along the
that they can be described in aggregate, based on references listed top, transmits light. B. Same field in reflected light, showing ilmenite in
above. They all show marginal emplacement breccias and have magnetite (pitted), both as separate grains and intergrowths.
42 E. R. Force
basis for calculating reserve figures for these bodies from pub-
lished data on tonnage and TiC>2 content.
Magnet Cove district, Arkansas

A small alkalic ring complex of Mesozoic age, intruded into
folded Paleozoic sedimentary rocks, forms the Magnet Cove dis-
trict near Hot Springs, Arkansas (Fig. 40). The alkalic igneous
rocks form a basin, or cove, which interrupts the Zigzag Moun-
tains of sedimentary rocks. Titanium-mineral production from
the Magnet Cove area has been minor.
The geology and geochemistry of the district have been
described by Erickson and Blade (1963), and the district has been
a «ai U S.G.S WT/^MmrmmMa mapped by Erickson and Blade and by Danilchik and Haley
^ ^ /J ^ wan. (1964). Compositions of nepheline syenite (Tables 3 and 7 of
Figure 39. Anatase pseudomorphic after perovskite and/or magnetite, Erickson and Blade) show both miaskitic and agpaitic values.
Tapira, Brazil. Scale in millimeters. Erickson and Blade relate their 28 alkalic lithologies to three
rings; the innermost contains carbonatite. The igneous hosts for
titanium-mineral deposits are jacupirangite of the outer ring and
altered phonolite of the intermediate ring. In addition, garnet
small carbonatite cores. Bostonite (aplitic syenite) is present at nepheline syenite and jacupirangite of the outer ring may be
Tapira. Jacupirangite, bebedourite, and other alkalic pyroxenites related to a contact-metamorphic deposit of brookite in altered
rich in oxide minerals are common to predominant lithologies in sedimentary country rocks.
all three stocks and contain up to 14 percent perovskite and/or Jacupirangite, or magnetite-perovskite pyroxenite, contains
30 percent sphene (Troger, 1935). Some less common rocks are 4.0 to 4.3 percent Ti0 2 , present in the following titaniferous
further enriched in magnetite, perovskite, sphene, and apatite; in phases: titaniferous clinopyroxene (78 percent average mode)
some, apatite is the most abundant mineral. containing 3.0 percent TiC>2, perovskite (4 modal percent), minor
The petrography of magnetite-perovskite-sphene rocks at titaniferous melanite garnet, and an undetermined amount of
Tapira is illustrated by Alves (1960). All three minerals form magnetite intergrown with ilmenite (Erickson and Blade, 1963).
coarse independent grains, commonly with apatite. Magnetite Apatite and calcite are also present. Near contacts with siliceous
contains ilmenite lamellae and is partially replaced by hematite. Arkansas novaculite wallrock, late sphene takes the place of
The potential ores of these deposits, however, are not the perovskite and magnetite.
titaniferous alkalic rocks, but parts of the weathered residuum Altered phonolite, locally brecciated, averages 2.5 percent
over them, which can be as thick as 200 m. Anatase (octahedrite) TiC>2 and contains the Ti-bearing phases magnetite-ilmenite and
in this residuum is microcrystalline and replaces perovskite, mag- sphene. Abundant late calcite and minor apatite are present.
netite, and sphene. Anatase may be pseudomorphic after oc- Garnet-nepheline syenite contains 0.8 to 1.1 percent Ti0 2 .
tahedral perovskite and/or magnetite (Fig. 39); that is, the Titaniferous phases are zoned melanitic garnet (as much as 6
octahedral form of anatase masses is from other octahedral modal percent), which in this district contains up to 15 percent
minerals. Anatase masses vary in size from more than 1 cm to TiC>2, titaniferous clinopyroxene, and minor magnetite and
grains less than 0.1 mm in diameter. Some masses in my speci- sphene.
mens have cores of primary minerals such as perovskite. The The Magnet Cove alkalic intrusives are unusual in showing
microcrystalline anatase is porous and cemented by iron little fenitization of country rocks (Heinrich, 1966). Thus, alkalic
hydroxides (Turner, 1986). igneous rocks impinge directly on sedimentary rocks. Flohr and
Weathered residuum of the three stocks contains phosphate Ross (1989) note fenitized xenoliths at one locality.
resources in addition to titanium-mineral resources. The weather- Titanium-mineral deposits of the district have been exhaus-
ing profile shows a basal zone rich in apatite, an intermediate tively described. In all, their treatment by Kinney (1949), Reed
anatase- and magnetite-rich zone, and more barren surficial mate- (1949a and 1949b), Fryklund and Holbrook (1950), Toe we and
rial (Cruz and others, 1977; Harben, 1984). At Tapira, the ana- others (1971), and many others comes to more than 450 pages,
tase ore of the intermediate zone is 40 to 55 m thick and contains plus maps. The identified resources, however, appear to be of
more than 20 percent TÍO2. Titanium is present in magnetite, modest economic importance, as tonnages are small and mineral
perovskite, and Ti-rich schorlomitic garnet as well as anatase compositions unfavorable. The following descriptions are from
(Harben, 1984). Herz (1976) believes that more than 60 percent those sources.
of the TÍO2 is present as anatase, but no substantiation for this The deposits occur in three main settings: rutile in feldspar-
figure appears in the literature. Thus there is currently no good carbonate veins in altered phonolite, brookite in feldspar-pyrite
Figure 40. Geologic map and cross section of selected units of the Magnet Cove area, Arkansas, after
Erickson and Blade (1963) and Danilchik and Haley (1964). Deposits are 1, Magnet Cove Rutile;
2, Mo-Ti; 3, Hardy-Walsh; and 4, Christy. Other brookite prospects labeled B. See text for locality X.
Unit n, novaculite; i, alkalic igneous rock, including j, jacupirangite. See text for locality X and resources
of dark-shaded unit in cross section.
44 E. R. Force
brookite-limonite veinlets and disseminations in quartzite recrys-
tallized from novaculite. Open spaces have permitted the
formation of euhedral crystals of several minerals, including
brookite. Taeniolite (KLiMg2Si40ioF2) mica occurs as dissemi-
nations and irregular masses, and increases westward toward the
intrusives. Fine rutile is a minor constituent. Brookite averages
0.5 to 1 mm in diameter and reaches 6 to 7 mm. It is slightly
enriched in clayey weathered zones about 5 m thick. Where truly
unoxidized rock was encountered by Fryklund and Holbrook
(1950), pyrite was present in place of limonite.
Apparently, titanium in these deposits has been introduced
from Ti02-rich intrusive rocks, such as jacupirangite, into
novaculite that contains almost no Ti02 (Cressman, 1962,
Table 6). Other elements apparently introduced from the
intrusive were Fe, V, Li, F, Nb, and perhaps S. Elements moving
Figure 41. Network of quartz-brookite-limonite veinlets (dark) in quartz- into the intrusive included Si (to form jacupirangite with second-
ite, Christy deposit, Magnet Cove. All units weathered; exposure about
ary sphene). Thus these contact-metamorphic deposits suggest
4 m high.
some unusual elemental fluxes. In normal skarn deposits,
titanium is not mobile; indeed, titanium distribution is sometimes
used to delimit the original extent of igneous rock (cf., Large,
1972).
veins in jacupirangite, and brookite in skarn-like (or fenite-like?) Six brookite prospects have the same geologic settings as the
contact metamorphic deposits. Christy and Hardy-Walsh deposits (Fig. 40). Their distribution
Deposits in igneous rocks. The Magnet Cove Rutile suggests to me that the total resources of this deposit type at
deposit (Fig. 40) produced minor amounts of rutile from 1932 to Magnet Cove should be calculated as all coarse brookite present
1944. Rutile occurs in swarms of feldspar-carbonate veins that in the lower member of novaculite (100 m thick) between 100
cut porphyritic aegirine phonolite and a variety of alkalic dike and 300 m from the alkalic complex, through the Chamberlain
rocks, all hydrothermally altered in places. The rutile-bearing Creek syncline and adjacent structures. The cross section (Fig.
veins contain albite or microcline; dolomite, calcite, or ankerite; 40) presents this view of resource potential, using a depth to
and pyrite. Rutile occurs as prisms averaging 0.1 mm in length, novaculite at the syncline axis (890 m) calculated from accurate
locally aggregated into irregular masses. The rutile content of data on the plunge in the adjacent barite mine (Scull, 1958). The
intervals as long as 0.6 m is up to 5 percent and averages 2.7 implied brookite resource in the Chamberlain Creek syncline
percent. The rutile contains 1 to 2 percent niobium and 0 to 1 alone is on the order of 5 x 106 metric tons. This is a hypothetical
percent V2O5, and these impurities make it difficult to use. Re- resource of respectable magnitude, worthy of further investigation
sources are not thought to be large anyway. from the points of view of industrial utility of this brookite,
The Mo-Ti prospect (Fig. 40) contains brookite, and pyrite corollary vanadium resources, and further exploration. For ex-
coated by molybdenite, in microcline veins through a large ample, drilling to the lower member of novaculite at location X
jacupirangite mass. The total mineralized area measures at most in Figure 40 could be warranted.
10 m x 130 m (Holbrook, 1948), so resources are thought to be
minor. Perovskite containing 0.6 percent niobium is reported Kodal deposit, Norway
from the host rock.
Contact-metamorphic deposits. The Christy and Hardy- The Oslo alkalic igneous province contains syenitic compos-
Walsh (Kilpatrick) brookite deposits formed at the eastern mar- ite ring intrusions of miaskitic affinity (from analyses of nepheline
gin of the igneous complex in the lower member of siliceous syenite in Oftedahl, 1960). These syenites are called larvikite
Arkansas novaculite, where it is in contact with alkalic igneous (augite syenite) and lardalite (nepheline syenite). The Kodal
rocks (Fig. 40). The lower member is about 100 m thick, and the deposit is associated with a nepheline larvikite ring showing
deposits are within about 300 m of intrusive rocks. Igneous rock rhythmic igneous banding (Petersen, 1978; Lindberg, 1986). It
nearest the Christy deposit is garnet nepheline phonolite, whereas can be described either as a jacupirangite or as an impure nelson-
that nearest the Hardy-Walsh deposit is jacupirangite. Both ite (Bergstetl, 1972); I refer to it here as jacupirangite.
deposits are weathered. The average abundance of recoverable The main jacupirangite body dips steeply south, is 1,900 m
brookite is about 5 percent. The brookite contains 2 percent Nb long, and varies in thickness from 2 to 32 m (Nielsen, 1972). In
and 0.5 percent V2O5. The Christy deposit is presently being addition to pyroxene, it contains 17 percent apatite, 40 percent
mined for vanadium. magnetite, and from 5 to 15 percent ilmenite (averaging 8 to
Both deposits consist of a network (Fig. 41) of quartz- 9 percent). Ilmenite and magnetite form separate grains 0.1 to
ECONOMIC PROGNOSIS
Titanium-oxide mineral deposits in alkalic rocks suffer from

a great economic disadvantage compared with deposits of other
types: their oxide minerals contain niobium and, commonly, rare
earth elements and thorium. This makes refinement difficult.
Some deposits contain perovskite, for which Ti02-recovery tech-
nology is still uncertain. Deposits like Kodal are dominated by
magnetite, which has no value as a titanium mineral.
Some deposits of titanium oxide minerals in alkalic rocks
also are too small, for example, the Magnet Cove Rutile deposit.
One type of deposit that suffers from an unfavorable mineral
composition but not from small size is the contact-metamorphic
type, of which the Christy deposit is an example.
Titanium-oxide mineral deposits in weathered overburden
Figure 42. Photomicrograph of interstitial opaque oxides and apatite over titanium-rich alkalic rocks have neither disadvantage, and
among cumulate augite crystals, ore band at Kodal, Norway. Transmit- these deposits are thus placed most favorably for commercial
ted plane light, 6-mm field.
competition. The titanium-oxide mineral assemblage of these
deposits is reconstituted in the weathering zone and thereby is
mostly purged of detrimental trace elements and calcium,
enabling treatment by the chloride process. The volume of some
1.0 mm in diameter. Ilmenite grains contain rutile and magnetite weathered overburdens of this type, such as those in Brazil, rivals
intergrowths, and magnetite grains contain ingrown ilmenite. those of large magmatic ilmenite deposits. Thus, these deposits
Apatite is finer grained. Lindberg (1986) describes sharp footwall are in a good position to capture a significant share of world
contacts below which is a 100-m zone of larvikite impregnated titanium markets.
by oxide minerals and apatite. The hanging wall is gradational,
and the overlying larvikite contains numerous stratiform lenses METHODS OF EXPLORATION
enriched in oxide minerals and apatite.
As noted by Petersen (1978) and Lindberg (1986), the The formation of valuable anatase deposits in weathered
jacupirangite body itself shows cumulate bands outlined by alkalic rocks is clearly a two-stage process. The second stage,
modal and grain-size variations. Oxide minerals and apatite are weathering, is not discussed in this chapter but in chapter 6. Here
interstitial to pyroxene in jacupirangite (Fig. 42; cf., Bergstol, we need to discuss exploration for alkalic stocks rich in titanium
1972) and in the underlying impregnated zone (my observation). oxide minerals. Location of promising stocks may be aided by
Locally, pyroxene cumulates contain macroscopic blebs of oxide their distinctive shapes in air photos such as those shown by
minerals and apatite in a range of sizes. This evidence suggests Barbosa and others (1970), by soils maps if available, and by
immiscibility of a nelsonitic liquid in larvikitic magma, by the aeromagnetic surveying, as magnetite invariably seems to accom-
same reasoning developed in Chapter 3. Similar conclusions were pany perovskite or ilmenite in large Ti-rich bodies. If fresh rock is
reached by Bergstetl (1972) and Lindberg (oral communication, exposed, selection of the more favorable stocks can be on the
1987). basis of miaskitic composition of nepheline syenites; that is, the
The Kodal deposit is less alkalic than others described in this ratio Na20+K20/Al203 should be less than one (on an atomic
chapter. Its oxide mineralogy (magnetite, ilmenite with inter- basis); however, this has proved an unreliable indication in sev-
grown rutile) and mode of formation (cumulates of dense eral districts. Heavy mineral concentrates could delineate sub-
immiscible liquids) suggest a transition to ilmenite deposits of areas containing a titanium mineralogy suitable for upgrading by
anorthosite-ferrodiorite massifs. weathering in the regolith.
Chapter 10.
Titanium oxide minerals in other igneous suites
Titanium oxide minerals show interesting distribution panied by biotite. Haggerty (1976a) points out that the high
patterns, but form no economic deposits, in igneous suites other oxygen fugacity represented by the I-type or magnetite granites
than anorthosite-ferrodiorite massifs and alkalic complexes. The favors oxides over silicate minerals, but that the oxide minerals
hydrothermal alteration of hypabyssal granitoid systems forms represent the Fe-rich rather than the Ti-rich members of their
fine rutile that is of some economic interest. respective solid-solution series.
Haggerty (1976a, Table Hgl2) shows that typical primary
PRIMARY TITANIUM OXIDE MINERALS oxide-mineral assemblages in granitoid rocks are coexisting
OF SOME IGNEOUS SUITES ulvospinel-magnetite and ilmenite-hematite, each showing exso-
lution textures. Exsolution oxidation of ulvospinel to ilmenite is
Titanium-mineral assemblages of primary and high-temper- common, and later pseudomorphic oxidation to intergrown
ature oxidation origin are summarized here for four igneous hematite and rutile is known. The MnO contents of ilmenite in
suites: two igneous suites, the granitic and basaltic rocks, are granitoid rocks are high, and especially in peralkaline granites this
selected for discussion because of their great volume and to cor- manganoan ilmenite can properly be called pyrophanite.
rect misleading statements about their rutile contents (reviewed by Reports of titanium-mineral concentrations in granitoid
Force, 1980b). Two less voluminous suites, layered mafic intru- rocks are few. Nelsonite-like concentrations of titaniferous mag-
sives and kimberlites, are included because their oxide minerals netite and apatite in breccia matrix and dikelike bodies occur in
are the subject of a voluminous literature. These discussions rely syenitic hypabyssal and other granitoid stocks of the Canadian
heavily on two review monographs by Haggerty (1976a, b), re- Cordillera (Badham and Morton, 1976; W. J. McMillan, written
markable for their 107 pages of tables and 185 photomicro- communication, 1981). Some of these have been explained as
graphs. An additional review is by Elsdon (1975). Partitioning immiscible liquids.
analyses between oxides and silicates are mostly from Force The occurrence of rutile as discrete grains in granitoid rocks
(1976a). and related pegmatites has been reviewed by Force (1980b).
Granular accessory rutile is reported from very few granitoid
Granitoid rocks rocks (Lee and Dodge, 1964), although secondary rutile included
in other oxide-mineral grains is fairly common. In some granites
The Ti02 contents of granitoid rocks range from about 0.3 from which rutile has been reported, such as those of Dartmoor
to 0.7 percent; the high values are in lithologies of more interme- and Cornwall, Great Britain (Bramall, 1928; Groves, 1931),
diate composition. Silicate minerals typically contain 60 to 95 rutile is present in late veins that accompany hydrothermal
percent of this T1O2. However, variations in partitioning of Ti02 alteration, commonly as pseudomorphs of titaniferous biotite
among oxide and silicate minerals are among the criteria used to (D. R. Wones, oral communication, 1980).
classify granites. The I-type of Chappell and White (1974) or Niobian rutile (ilmenorutile) has been reported from
magnetite series of Ishihara (1977) contains 0.1 to 2 modal per- hypabyssal leucocratic stocks bearing molybdenum and tin mineraliza-
cent magnetite that contains ilmenite as lamellae, accompanied tion (Sainsbury, 1968; Desborough and Sharp, 1978; Desbo-
by up to 1 percent sphene, and Ti-poor hornblende. The S-type rough and Mihalik, 1980). This rutile is apparently of primary
or ilmenite series contains only 0 to 0.2 percent ilmenite, accom- origin, as it forms discrete grains in unaltered rock.
47
48 E. R. Force
TABLE 11. TITANIUM OXIDE MINERALS OF S O M E IGNEOUS SUITES OTHER T H A N THOSE
DISCUSSED IN TEXT*
Rock Suite TÌO2 Range Primary Mineralogy^ Exsolution Pseudomorphic

(%) Mag-Usp Hem-llm Oxidation^ Oxidationt
Rhyolitic rocks 0.2-0.6 P P ilm (R) hem (R)

Andesites 0.8-2.6 P P ilm (C) hem, mgh (C)
Trachybasalts 2.6-2.8 P P ilm (C) R
Syenites 0.7-1.1 P P R R
Alpine mafic 0.02-1.0 R R
*From Haggerty, 1976a, and Force, 1976a.

tAbbreviations: Mag = magnetite; Usp = ulvospinel; Hem = hematite; Ilm = ilmenite; Mgh =
maghemite; P = predominant; C = common; R = rare.
Basaltic rocks In general, titanium-oxide mineral accumulation in layered

mafic intrusions seems to begin at a later stage than in ferro-
The TiC>2 contents of typical basaltic rocks range from 1.5 diorites of anorthosite massifs, probably because of liquid
to 2.7 percent. Partitioning is poorly documented, but calcula- immiscibility in the latter. Locally, however, massive nelsonitic
tions from partial analyses show that oxide minerals commonly rocks consisting of magnetite (with intergrown ilmenite) and
contain more than 50 percent of this T1O2. For example, figures apatite may form discordant to concordant bodies attributed to
in Gottfried and others (1968) show that oxide minerals contain immiscible liquids in layered complexes (Grobler and Whitfield,
65 percent of the TiC>2 in early diabases, 86 percent in Fe-Ti-rich 1970).
pegmatitic facies, and 47 percent in granophyres of the Dillsburg
diabase sill of Pennsylvania. The variation is due in part to crys- Kimberlites
tallization of ilmenite, which typically begins midway through the
crystallization history of basaltic rocks (Wright and Peck, 1978). The titanium oxide minerals of kimberlites have been inten-
Primary oxide minerals normally include members of both sively investigated as clues to the composition of the lower crust
the ulvospinel-magnetite and ilmenite-hematite series, both in and mantle and as prospecting tools in diamond exploration. The
tholeiitic and alkaline olivine basalt types (Haggerty, 1976b). oxide mineralogy, which is quite complex because of lack of
Exsolution oxidation to ilmenite is common, though possibly less equilibrium, has been reviewed by Haggerty (1976a), Pasteris
common in alkaline olivine basalt. Further pseudomorphic oxida- (1980), and Mitchell (1986). Ilmenite (as much as 10 percent),
tion to rutile, hematite, and/or pseudobrookite is widespread rutile, and perovskite are common in kimberlites, and armalcolite
except in deep-sea basalts and is more advanced in flow centers [(Fe,Mg)Ti2C>5] has been reported. Both ilmenite and rutile occur
than on flow margins. Granular accessory rutile in basalt is as megacrysts, in the groundmass, and in intergrowths with other
unknown. minerals. Ilmenite is characteristically rich in Mg and Cr and is
called picro-ilmenite (ilmenite-geikielitess). Magnesium contents
Layered mafic intrusions of ilmenite tend to be higher in the groundmass than in
megacrysts.
Stratiform, or layered, mafic intrusion complexes may con-
tain layers enriched in iron-titanium oxide minerals, typically of Other suites
cumulate origin. These layers contain cumulate plagioclase,
clinopyroxene, and olivine and occur in the upper parts of Haggerty (1976a) also gives information on titanium oxide
complexes with ultramafic basal lithologies. Magnetite com- minerals of other igneous suites, which is summarized in
monly dominates over ilmenite and contains exsolved ulvospinel. Table 11. Correlative partitioning analyses are not currently
Locally, oxidation of ulvospinel forms ilmenite. Primary ilmenite available in the literature; a rigorous study of titanium partition-
contains exsolved hematite. Stratigraphically lower layers typi- ing between oxide and silicate minerals is needed.
cally contain oxide minerals that are richer in iron in each solid-
solution series. These lower layers may also have higher ratios of RUTILE IN H Y D R O T H E R M A L
the magnetite series to the ilmenite series (reviewed by Haggerty, ALTERATION ASSEMBLAGES
1976a, from Skaergaard, Kiglapait, Muskox, Kap Edward Holm,
Somerset Dam, and Sudbury). Chromite-bearing layers locally An extensive literature now exists on rutile associated with
contain rutile of probable cumulate origin (Cameron, 1979). the hypabyssal granitoid rocks of porphyry alteration systems
TABLE 12. RUTILE-PRODUCING REACTIONS
IN PORPHYRY SYSTEMS
Biotite + sulfur •phlogopitic biotite + pyrite + rutile

K(Fe,Mg) 2 (Fe,AI,Ti)Si 3 AI0IQ(OH,F)2 + S 2 ---•
K(Mg,Fe)3 Si 3 AI0 10 (OH,F)2 + FeS 2 + T 0 2
Hornblende + sulfur - - ••actinolitic hornblende + pyrite + rutile + alumina

Ca2(Fe,Mg1"niAI)5Si7AI022(OH,F)2 + S2 •
Ca 2 (Mg,Fe) 5 Si 8 0 2 2(OH,F)2 + FeS 2 + T I 0 2 +AI2O3
Titaniferous magnetite + sulfur • l e s s e r magnetite + pyrite + rutile

2(Fe,"n)304 + S 2 • F e 3 0 4 + FeS 2 + Ti0 2
llmenite + sulfur • p y r i t e + rutile

FeTO 3 + S 2 FeS 2 + T I 0 2
Sphene + carbon dioxide • rutile + calcite + quartz Figure 43. Photomicrograph of rutile (dark, high relief) + carbonate +
CaTiSiOs + C 0 2 •TO 2 + CaC03 + S 0 2 quartz pseudomorphic after sphene, from propylitically altered rocks at
Ajo, Arizona. Transmitted plane light, 6-mm field.
(Lawrence and Savage, 1975; Williams and Cesbron, 1977; "highest grade" sulfide mineralization. There is some evidence
Force, 1980a; Czamanske and others, 1981; Udubasa, 1982; that phyllic (quartz-sericite) overprinting on higher grade zones
Force and others, 1984). Rutile is a secondary mineral, coeval does not destroy previously formed rutile. In propylitic (chlorite-
with hydrothermal alteration, and its formation is virtually inevi- epidote-carbonate) zones, rutile is very fine grained or absent.
table if this alteration is sufficiently severe. The following The amount of rutile formed varies with the titanium con-
discussion of rutile formation mostly follows Czamanske and tent of the parent rocks, as titanium abundance is apparently
others (1981). conserved during alteration. Deposits developed on two igneous
parents having differing TiC>2 contents and mineralogy have con-
Process of formation sequent differences in rutile content and morphology.
Rutile in porphyry alteration systems records the amount Bingham, Utah, example
and position of titanium present in fresh parent rocks, which are
commonly calc-alkaline granitoids containing 0.3 to 1.0 percent Rutile has been recorded in ten porphyry deposits in the
TiC>2. Titaniferous minerals in these parent igneous rocks are United States. Rutile occurrence at the Bingham Canyon deposit
magmatic biotite, magmatic hornblende, titaniferous magnetite, of Utah is described here as an example, because it is best known
sphene, and ilmenite. Any one of these minerals may be the most and possibly is the most valuable rutile resource. Bingham, until
important carrier of titanium in a given intrusion. None of these recently the largest U.S. producer of copper, is a copper porphyry
minerals is stable in the alteration environment, however; biotite,with subordinate molybdenite. The deposit is in Tertiary intru-
hornblende, and magnetite alter incongruently to new minerals sions of mesocratic equigranular quartz monzonite (MEQM) and
containing less TiC>2, and ilmenite and sphene disappear. later leucocratic porphyritic quartz monzonite (LPQM) in
Table 12 shows some of the apparent mineral reactions, largely Paleozoic sedimentary rocks (Bray and Wilson, 1975, and maps
driven by the introduction of sulfur and CO2. Figures 43 and 44 therein). The ore is approximately coincident with potassic (bio-
show some pseudomorphic relations that suggest these reactions. tite, K-feldspar) alteration. Propylitically altered and unaltered
Table 13 contrasts the initial magmatic and subsequent post- MEQM is present outside the potassic zone, and sericitic altera-
alteration TiC>2 contents of biotite and amphibole. tion is overprinted on potassic alteration in both intrusions.
Studies to date indicate that liberation of titanium is most MEQM averages 0.9 percent Ti0 2 and 15 to 25 percent biotite,
complete, and resultant rutile of coarsest grain size, in the most whereas LPQM contains 0.6 percent Ti0 2 and less than 10 per-
altered zones of the alteration system—the potassic (biotite, It- cent biotite. The Ti02-bearing minerals in unaltered MEQM
feldspar) zone at San Manuel, Arizona, and Bingham, Utah, and include minor ilmenite and Ti-poor magnetite in addition to bio-
the advanced argillic (andalusite) zone at Tangse, Sumatra (Force tite (G. K. Czamanske, written communication, 1979). The dif-
and others, 1984). Commonly these zones also have therichestor ference in Ti0 2 content between magmatic biotite and phlogopitic
50 E. R. Force
TABLE 13. CONTRAST OF TIO Z CONTENT OF
MAGMATIC AND HYDROTHERMAL BIOTITES AND
AMPHIBOLES IN PORPHYRY SYSTEMS
Ti0 2 Content
Magmatic Hydrothermal
Biotite
Bingham, Utah* 4.925 1.65
Copper Canyon, Nevadat 4.7 1.5
Santa Rita, New Mexicot 4.1-4.7 1.5-3.1
Babine Lake, British Columbia§ 4.3 2.8
Butte, Montanat 4.58 1.99
Amphibole
Babine Lake, British Columbia§ 2.0 0.4
•Moore and Czamanske, 1973. Figure 44 photomicrograph of leucocratic porphyritic quartz monzonite
TCzamanske and others, 1981. f r o m Bingham Canyon, Utah, showing rutile of two grain sizes in hy-
SCarson and Jambor, 1974. drothermal biotite after euhedral magmatic biotite—reticulate rutile
needles and stubby prism (high relief, left center). Transmitted plane
light, 2-mm field.
hydrothermal biotite (Table 13) averages about 3.3 percent Ti0 2 . on rutile in soil samples. The presence of rutile in the regolith, and
Thus in the MEQM unit, 0.81 percent Ti0 2 out of 0.9 percent its grain size, were found to be potent indicators of alteration
TiC>2 in the rock is available to form rutile. zones and sulfide distribution in unweathered underlying rock.
Czamanske and others (1981) found that rutile in MEQM
and LPQM averages 0.34 and 0.24 weight percent, respectively, Economic prognosis
in the potassic alteration zone. The ubiquitous sagenitic rutile
needles in altered magmatic biotite were not included. The aver- Rutile forms very large but very low-grade deposits in
age grain size of the coarser rutile was 60 x 30 um in MEQM, porphyry systems. It can be recovered only as a byproduct of
60 x 40 /um in LPQM. Trace-element impurities in this rutile mining for sulfides from these deposits. This rutile recovery has
total less than 1.5 percent (Czamanske and others, 1981). been evaluated most recently by Rampacek (1982) and Sillitoe
(1983). The potential magnitude of rutile production is great.
Use of rutile in exploration Rutile isfinegrained in porphyry deposits; less than half the
rutile is as coarse as fine sand. Preliminary attempts using flota-
Lawrence and Savage (1975) and Williams and Cesbron tion and gravimetric methods to separate rutile from porphyry
(1977) suggested the use of rutile in exploration for porphyry copper ores and tailings in the United States have been
deposits. Rutile is stable in the weathering environment moderately successful (Llewellyn and Sullivan, 1980; Sullivan
(Chapter 6); of the entire hydrothermal mineral assemblage in a and Llewellyn, 1981; Davis and others, 1988). Such attempts
porphyry deposit, rutile is probably least susceptible to weather- have also been made with Chilean porphyry copper ores (Centro
ing. Force and others (1984) tested an exploration method based de Investigación Minera y Metalurgia, 1986).
Chapter 10.
The weathering of titanium oxide minerals and the role of climate change
The weathering and erosion of titanium oxide minerals is Weathering processes have upgraded a large number of
the bridge between "primary" (igneous and metamorphic) and titanium oxide mineral deposits and have made some of them
"secondary" (sedimentary) titanium-mineral deposits. This chap- economic. In a number of deposits discussed in previous chapters,
ter treats the weathering not only of such crystalline rocks but the most valuable material is a saprolite developed on the pri-
also of sedimentary deposits and rocks. For this purpose, these mary deposit (Table 14). All detrital deposits can be regarded as
sediments are treated solely as weathered rock; the following upgraded by predepositional weathering. Superimposed in situ
chapters discuss the sedimentary processes of titanium-mineral (post depositional) weathering of detrital deposits is also of great
accumulation. importance in some deposits.
ECONOMIC IMPORTANCE THE BEHAVIOR OF TITANIUM IN THE

WEATHERING ENVIRONMENT
Weathering exerts a powerful influence on the relative com-
positions of titanium oxide-bearing rock, its weathered mantle,
and its erosional debris. This influence is critical to the economic Titanium as a chemical entity is residually enriched by
value of deposits, both in weathered primary rock and in sedi- weathering, as the more soluble elements are leached out of
mentary concentrations. saprolites. Commonly this enrichment is less than 50 percent
Titanium oxide minerals are mostly resistant to weathering, (Wells, 1960; Short, 1961; Dennen and Anderson, 1962; Lough-
but they may be residually enriched in Ti02- Weathering thus man, 1969). However, enrichments in titanium by more than 100
beneficiates titanium-mineral deposits in the following ways percent have been reported even in noncalcareous rocks, where
(Fig. 45): (1) Reduction in mass of the parent rock by leaching, at weathering is or was once intense (Goldich, 1938; Wahlstrom,
approximately constant volume. Normally this reduction is less 1948; Valeton, 1972; Patterson and others, 1986). In extreme
than 50 percent, but greater reductions from calcareous and/or weathering, Ti02 may be mobilized within the rock to form
feldspathic rocks, locally to lesser volumes, are possible. Titanium concretions (Sherman, 1952).
oxide minerals are thus residually enriched. (2) The corollary
destruction of such minerals as garnet, amphiboles, and Silicate minerals
pyroxenes, which have no economic value but are difficult to
separate from titanium oxide minerals. (3) Disaggregation of the Biotite, hornblende, titanaugite, sphene, and other titanium-
rock into monomineralic grains. This is economically important bearing silicate minerals are among the less stable minerals in the
even in unconsolidated sediments, because detrital rock fragments weathering environment (Goldich, 1938; Dryden and Dryden,
may include both valuable and valueless minerals. (4) The chem- 1946). When they break down, their contained Ti0 2 generally
ical enrichment of Ti02 in several titanium oxide minerals. This remains and commonly forms exceedingly fine-grained Ti02
is particularly important for ilmenite, from which iron is leached minerals such as anatase. Aggregates of these Ti02 minerals may
in weathering. Other titanium oxide minerals may be leached of form pseudomorphs of the precursor silicates (Hartman, 1959;
other ions; for example, perovskite may be leached of calcium to Valeton, 1972).
form microcrystalline anatase. The rate of weathering varies with climatic conditions at the
51
52 E. R. Force
TABLE 14. TITANIUM OXIDE-MINERAL DEPOSITS KNOWN TO BE
GREATLY BENEFICIATED BY WEATHERING
Reference
Igneous and metamorphic deposits

Roseland, Virginia Herz and Force, 1987
Magnet Cove, Arkansas Fryklund and Holbrook, 1950
Tapira, Salitre, and Catalao I, Brazil Harben, 1984
Many rutile-bearing porphyry copper
deposits Czamanske and others, 1981
Postdepositional weathering of sedimentary placer deposits

Trail Ridge, Florida Pirkle and Yoho, 1970
Yoganup, Australia Welch and others, 1975
Gbangbama, Sierra Leone Raufuss, 1973
Lakehurst, New Jersey Puffer and Cousminer, 1982
Predepositional weathering of detrital grains of placer deposits

All placer deposits Gillson, 1959
Residual enrichment of titanium oxide minerals

Rutile is among the most stable minerals in the weathering
environment (Fish, 1962; Hubert, 1962; Overstreet and others,
1963; Force, 1976b). Microcrystalline rutile commonly forms
therein from the weathering of ilmenite. Anatase, a polymorph of
rutile, also forms in the weathering environment, mostly from
silicate minerals in the parent rocks. The polymorph brookite is at
least moderately stable, as it is found in the saprolites at Magnet
Cove, Arkansas (see Chapter 4). However, all three TiC>2 poly-
morphs are apparently rendered less stable if they contain apprec-
iable amounts of niobium and rare earths, as in alkalic rocks.
The stability of perovskite in the weathering zone is poorly
known, in part because of its sporadic occurrence. Perovskite
persists in the regolith over the alkalic stocks of Brazil as nucleii of
grains heavily rimmed by anatase. In the Powderhorn district of
Colorado, perovskite is locally altered to anatase even though the
weathered zone is thin.
Ilmenite, considered as a fixed mineralogic entity, is residu-
ally enriched by weathering processes (Goldich, 1938; Hartman,
1959; Overstreet and others, 1963; Loughman, 1969) because it
is moderately stable in the weathering environment. The ratio of
Figure 45. Diagram contrasting fresh and weathered rock (sand in this
case) containing titanium oxide minerals. F, feldspar; shaded, pyroxenes ilmenite to magnetite commonly increases with progressive
and amphiboles; black, titanium oxide minerals; remainder quartz. weathering. Over basalts, ilmenite may increase sufficiently for
the saprolite to become a marginal resource (e.g., Hosterman and
others, 1960).
Alteration of ilmenite
weathering site (McLaughlin, 1955) and from one silicate min-
eral to the other. Sphene and melanitic or schorlomitic garnet are Ilmenite undergoes a remarkable transformation in the
two titanium silicate minerals that may weather slowly and lo- weathering environment; it is moderately stable there, but iron is
cally may survive even in saprolite and other thick soil profiles progressively leached from it. The result of extreme leaching can
(Overstreet and others, 1963; Harben, 1984). be tan-colored grains having a specific gravity of about 3.0, a
rutile x-ray diffraction pattern, and a Ti0 2 content of over 80 Ti0 2 values (Welch, 1964; Baxter, 1977, 1986; Mertie, 1979).
percent. The economic consequences of this alteration are dis- An increase in AI2O3 may occur by infiltration into porous al-
cussed in Chapter 1. tered ilmenite (Frost and others, 1983).
The progressive alteration of ilmenite in weathering is the Oxidation and humic acid leaching were shown by Kark-
subject of a large body of literature (Creitz and McVay, 1948; hanavala and Momin (1959) and Lynd (1960) respectively, to
Lynd and others, 1954; Bailey and others, 1956; Karkhanavala facilitate the weathering alteration of ilmenite. Reportedly, initial
and others, 1959; Flinter, 1959; La Roche and others, 1962; alteration oxidizes iron to Fe203, and later alteration leaches iron
Temple, 1966; Dimanche, 1972; Grey and Reid, 1975; Dimanche (Temple, 1966; Gamar, 1972; Baxter, 1977).
and Bartholome, 1976; Subrahmanyam and others, 1982; Morad At Ti02 contents of about 58 to 60 percent, the ilmenite
and Aldahan, 1986). Variations in microtexture, chemistry, lattice is no longer detectable to x-ray diffraction, in my
crystallography, and magnetism of the ilmenite alteration series experience. The anisotropy and pleochroism of ilmenite in pol-
are well documented. The weakness of this literature is that each ished section also disappear (Bailey and others, 1956; Dimanche
study is of detrital mixtures of variously altered grains in a given and Bartholome, 1976). Textural relics of intergrown iron oxide
sand; thus the sites of weathering are not known. phases can be recognized in altered ilmenites, but at higher Ti0 2
Ilmenite from unweathered parent rock as supplied to the contents, these are represented only by pits. Even recognizable
weathering environment generally contains less TiC>2 than hematite is etched away (Temple, 1966; Puffer and Cousminer,
stoichiometric ilmenite (52.6 percent TiC>2), because of inter- 1982; Darby, 1984). Single grains may be concentrically zoned,
growths with hematite, magnetite, other oxides, and silicate inclu- from ilmenite at the core to leucoxene on the rim (Welch, 1964;
sions. Industrial ilmenite concentrates from such rocks contain as Frost and others, 1983), so that the external appearance of a grain
much as 32 to 46 percent Ti0 2 . does not always correlate with composition and magnetic
Altered ilmenite, on the other hand, contains from about 54 properties.
to more than 80 percent TiC>2. Altered ilmenite that contains less Recent studies appear to agree that pseudorutile (about
than about 65 percent TiC>2 is typically black, paramagnetic, and Fe2Ti309, or 60.0 percent Ti0 2 ; Teufer and Temple, 1966) is
greater than 3.3 in specific gravity. Brown to tan, nonmagnetic, commonly an intermediate product of ilmenite alteration. This
less dense, porcelanous opaque material that contains more than compound may be the same as that debated in older literature
70 percent TiC>2 is known as leucoxene. This leucoxene contains under the name arizonite. Fine pseudorutile is a pseudomorph of
microcrystalline rutile; leucoxene has no real mineralogic signifi- ilmenite that retains some optical properties of ilmenite grains
cance, however, as low-grade metamorphic leucoxene has a sim- (Mathis and Sclar, 1980).
ilar appearance but consists of sphene (± anatase).
The trace-element content of ilmenite also varies with the SITES OF TITANIUM-MINERAL WEATHERING
intensity of weathering alteration. Based on rather fragmentary
evidence, it appears that MgO contents decrease as alteration The weathering alteration of titanium oxide minerals takes
proceeds and that MnO contents increase in the first stages of place in four broad environments (Fig. 46). Based on the rather
alteration (to low-Ti02 altered ilmenite) but decrease at higher sketchy information currently available, it is possible to tenta-
Shoreline
Bedrock ^ " ^ e T ^ ^
deposit
Sea
level
Figure 46. Diagram showing sites of weathering beneficiation of a titanium-oxide mineral grain.
54 E. R. Force
TABLE 15. COMPARISON OF ILMENITES FROM PAIRS OF FRESH ROCK AND SAPROLITE,
IN GRANITOID ROCKS OF SOUTHEASTERN UNITED STATES'
Ilmenite (+ hematite) concentrate

Ilmenite llm/mag TI02 MgO MnO Fe20g/ Fe oxides/ Intergri
State/Sample content ratio FeO TÌO2 magne'
(%)
North Carolina
Saprolite 1 0.71 4.4 23.08 0.01 0.97 3.89 3.29 P
Granitoid 1 0.085 5.0 19.27 0.03 0.96 4.60 4.14 P
Gain (%)§ -16 -12 20 -67 1 -15 -21
South Carolina
Saprolite 2 0.060 0.79 48.27 0.01 8.30 0.64 0.90 A
Granitoid 2 0.011 0.16 50.05 0.06 8.27 0.22 0.83 P
Gain (%)§ 445 394 -4 -83 <1 190 8
South Carolina
Saprolite 3 0.053 1.32 27.34 0.01 1.41 3.95 2.61 A
Granitoid 3 0.046 0.86 26.32 0.04 1.44 3.66 2.74 P
Gain (%)§ 15 53 4 -75 -2 8 -5
Georgia
Saprolite 4 0.28 >28 51.85 0.05 4.48 0.33 0.84 A
Granitoid 4 0.22 7.6 47.97 0.11 4.17 0.37 1.00 A
Gain (%)§ 27 >268 8 -55 7 -11 -16
Georgia
Saprolite 5 0.037 1.76 46.18 0.01 1.75 1.06 1.13 A
Granitoid 5 0.021 1.75 42.66 0.02 1.64 0.79 1.31 A
Gain (%)§ 76 0 8 -50 7 34 -14
Average Gain (%) 109 >140 7 -66 3 41 -10
'From Mertie (1979) and my own observations.

tllmenite remained constant in all pairs; hematite remained constant except in pair 4, where it
decreased in saprolite. P = present; A = absent.
§Gain expressed as [saprolite value/fresh rock value -1] x100%.
tively partition the weathering of titanium oxide minerals among rocks of the southeastern United States. The data show some
these environments. scatter, but average values show clear trends. Ilmenite is residu-
ally enriched in saprolite, commonly by 15 to 75 weight percent,
Weathered rock and the ilmenite/magnetite ratio increases by an average of 140
percent. The ilmenite concentrate from saprolite contains slightly
The behavior of titanium minerals in the weathered mantle more Ti0 2 but less MgO than that from rock. The Fe 2 0 3 /Fe0
over igneous and metamorphic rocks has been outlined above. ratio of this concentrate is higher in saprolite, but total iron in
Ilmenite may alter significantly in weathered rock (Jackson and saprolite decreases relative to TiC^. X-ray diffraction patterns of
Sherman, 1953; Carroll and others, 1957; Puffer and Cousminer, the concentrates from fresh rock and from saprolite are similar
1982, p. 386), but modest alteration is more common (Rumble, and show sharp ilmenite peaks, but ilmenite from saprolite shows
1973; Dimanche and Bartholome, 1976). Ilmenite alteration in subdued magnetite peaks. Hematite intergrowths with ilmenite
the weathering environment is locally to anatase rather than to remain unaffected. Thus, the alteration of ilmenite in these sapro-
pseudorutile and microcrystalline rutile. lites is modest. In comparison to feldspar and ferromagnesian
Table 15 shows the ilmenite contents a n d compositions o f minerals, which are almost completely decomposed in these
five pairs o f fresh bedrock a n d overlying saprolite from granitoid saprolites, ilmenite alters slowly in this environment.
Fluvial transport 1. Gradients in mineralogic composition that are related to
depth within weathering or soil profiles of the deposit are attribu-
The second potential site for weathering of titanium oxide table to postdepositional weathering.
minerals is the fluvial transport system. Little is known about its 2. Differences of mineralogic compositions that are a func-
role. Cannon (1950) asserted that significant alteration of tion of age in series of similar deposits must be due to postdeposi-
ilmenite occurs in the fluvial transport system; however, Austin tional weathering, assuming that the nature of minerals supplied
(1960) suggested that this alteration occurs during occasional to all the deposits was constant. This topic is discussed further in
burial. Riezebos (1979) found that ilmenite abundance and alter- the next section.
ation state were little changed by transport over hundreds of 3. Detrital grains of the same original character that now
kilometers in the humid tropics. Darby and Tsang (1987) and show heterogeneous alteration states in one deposit must repre-
Basu and Molinari (1989) found that fluvial ilmenite clearly re- sent a detrital mixture of grains having varied alteration histories,
flects the trace-element composition and intergrowth structure of acquired before deposition. Conversely, uniform alteration of
its crystalline sources. Rivers draining crystalline rocks commonly grains of the same original character is evidence of postdeposi-
deliver labile mineral assemblages long distances to the seacoast tional alteration.
with little modification (Russell, 1937; van Andel, 1950; 4. Altered grains that show a polish acquired by abrasion in
Neiheisel, 1976); ilmenite at the river mouth may remain fresh the depositional environment must have been altered before dep-
(LaRoche and others, 1962; Neiheisel, 1976; Force and others, osition, as postdepositional alteration would destroy this polish
1982). Chapter 8 treats deposition in the fluvial environment. by authigenic mineral growth.
5. Chemical environments such as peat beds within a
Intermediate sedimentary hosts deposit may prevent postdepositional alteration. Grains preserved
in these environments show only predepositional alteration.
The role of intermediate sedimentary hosts as sites of weath- 6. The average grain size of leucoxene compared with other
ering of titanium oxide-mineral grains is also poorly known but minerals may reflect its present density or the greater density of
is probably great. A clastic grain typically comes to rest in a an ilmenite precursor (see "Depositional equivalence," Chapter
succession of sedimentary deposits for varying lengths of time. At 7). Where leucoxene is coarser than ilmenite, the leucoxene must
each site, the grain is subject to early diagenesis and intrastratal have been altered before deposition (Wilcox, 1971).
solution, then exposure to chemical weathering, followed by The literature on placer deposits of titanium oxide minerals
further fluvial or marine transport to the next host. presents several examples of the first guideline for postdeposi-
The evidence of weathering in intermediate hosts is mostly tional alteration (e.g., Pirkle and Yoho, 1970; Welch and others,
indirect; for example, the mineral assemblages in unconsolidated 1975). The Ti0 2 contents of ilmenite concentrates, as well as
Upper Cretaceous through Cenozoic "coastal plain" deposits in leucoxene/ilmenite ratios, increase upward within soil profiles
the eastern United States vary in units of different ages (Owens, (Figs. 47, 48); these profiles may be superimposed at an angle to
1985), apparently as a function of paleoclimatic variation (dis- primary sedimentary bedding, as at Trail Ridge, Florida (Force
cussed below). Compared with juvenile assemblages being and Garnar, 1985). Changes in the entire heavy mineral
supplied via fluvial systems today, mineral assemblages of such assemblage may parallel those of ilmenite alteration; the mineral
sediments are generally more mature, partly because they have assemblages are commonly most mature in the uppermost zones
been subjected to intrastratal solution (discussed in Chapter 10) if of weathering profiles (Force and others, 1982; McCartan and
not to greater weathering. Where detritus derived from coastal others, 1990; for dissenting views, see Hails and Hoyt, 1972;
plain deposits is admixed with juvenile detritus, in streams drain- Beck, 1973).
ing crystalline terranes, the fluvial mineral assemblage becomes Iron leached from ilmenite in the surficial zone may be
more mature. reprecipitated deeper in the same deposit as local iron hydroxide
The mature mineral assemblages of some of these interme- cement (Welch, 1964; Puffer and Cousminer, 1982). This cement
diate hosts include altered ilmenite, with elevated Ti0 2 contents encases ilmenite grains and retards further leaching (Lissiman and
(e.g., Wilcox, 1971; Puffer and Cousminer, 1982). Detritus de- Oxenford, 1975; Baxter, 1986). Heavy-mineral-rich laminae may
rived primarily from such intermediate hosts inherits their be preferentially cemented by iron-rich humate (Fig. 49), or
ilmenite alteration (e.g., the Horse Creek, South Carolina, placer humic acid salts.
of Williams, 1967). The positions of present and former water tables seem to
function as a base level for some types of postdepositional altera-
Postdepositional weathering tion. These levels are also depositional loci of humate and/or iron
In a given deposit, it is difficult to separate the changes due hydroxide cements (Fig. 47). Grey and Reid (1975) proposed
to weathering to which detrital particles were subjected sep- that some ilmenite alteration at Trail Ridge proceeds below the
arately, before deposition, from weathering changes they endured water table, but in Chapter 9, I explain their data in another
together, after deposition. The guidelines that seem most useful in manner. Weathering below the water table passes gradationally
separating these cases are: into intrastratal solution, a topic discussed in Chapter 10.
56 E. R. Force
DEPTH Percentage of Ti0 2 content of altered
(m) (ft) leucoxenized ilmenite ilmenite (in weight percent)
Land surface
Bleached zone
Húmate cement
Húmate cement
Figure 47. Vertical profiles of ilmenite alteration in surficial sands on Trail Ridge, Florida, from data in
Pirkle and Yoho (1970) and Temple (1966). Geologic profile and sedimentary structures (left) after
Force and Garnar (1985).
WEATHERING OF QUATERNARY SHORELINE others, 1982) show that within a given area, ilmenite is more
DEPOSITS AS A FUNCTION OF enriched in Ti0 2 on the higher terraces and that within a terrace
AGE AND LATITUDE the higher values are at lower latitudes (Fig. 51). Both the
mineral-assemblage and ilmenite-composition arrays show un-
A considerable literature exists on variations of heavy- usual simplicity and thus have predictive power. The great long-
mineral assemblages with the elevations of Quaternary marine strike homogeneity of the Appalachians may have provided a
terraces in the eastern United States (Neiheisel, 1962; Hails and relatively constant source area and thus permitted the simplicity.
Hoyt, 1972; Thorn and others, 1972; Beck, 1973; Cazeau, 1974; However, data compiled from Baxter (1977) suggest that altered
Oaks and others, 1974; Force and Geraci, 1975; Force and oth- ilmenite compositions on the west coast of Australia behave in
ers, 1982; McCartan and others, 1990). The higher terraces are the same manner (Fig. 52). Possibly, detrital mineral assemblages
known to be older, and each represents an interglacial high stand and ilmenite compositions are everywhere sensitive functions of
of sea level. The higher terraces contain the more mature, or age and latitude and everywhere correlate with each other. For
weathered, heavy-mineral assemblages and less feldspar. example, correlation of Figures 50 through 52 suggests that am-
Each of these studies was done in a restricted area, and phibole and pyroxene become negligible when the Ti0 2 content
compilation of such information makes possible a display (Fig. of altered ilmenite exceeds about 57 percent.
50) of mineralogy as a function of latitude and terrace height (the Compositions of altered ilmenite compiled from Quaternary
latter proxying for age). The percentage of the labile amphiboles shoreline deposits worldwide are shown in Figure 53. The rela-
and pyroxenes in the heavy-mineral assemblage was chosen as a tive ages of most of these deposits are unknown, and the nature of
measure of weathering, because these minerals show great varia- the ilmenite supplied by various source areas varies. Nevertheless,
tion in degree of postdepositional weathering from terrace to the data, including those of Figures 51 and 52, define a composi-
terrace. The array in Figure 50 is quite orderly; at any given tional envelope for altered ilmenite that is a function of latitude.
latitude, weathering is greater in the older, higher terraces, where- At high latitude, the composition of detrital ilmenite is the com-
as at any given terrace elevation, there is an equally striking position supplied by source rocks. At lower latitude, Ti0 2 con-
increase in weathering toward lower latitudes. The small reversals tent of ilmenite is extremely variable as a function of weathering
in trend are due to the influx of juvenile unweathered debris at duration and deposit type but shows a maximum at low latitude
the mouths of rivers draining crystalline rock. beyond which leucoxene loses coherence and density. Steep side
Similarly, the Ti0 2 contents of altered ilmenites from the gradients of the envelope are such that ilmenites having Ti0 2
same marine terrace deposits (Force and Geraci, 1975; Force and contents greater than about 50 percent are mostly between 35°N
Ti0 2 content of altered ilmenite (in weight percent)
54 56 58 60
E 4 -
Figure 49. Photomicrograph of impregnated Trail Ridge eolian ilmenite
ore sand, showing contrasts in mineralogy, grain size, and humate ce-
Q.
0> mentation between slip-face laminae rich (top) and poor (bottom) in
Q heavy minerals. Heavy minerals are altered ilmenite (black), zircon, and
sillimanite. Cloudy dark matrix is humate. Transmitted plane light, 2-
m m field.
acids. Last, we know that humate can also precipitate in heavy-

mineral-rich laminae (Fig. 49).
Accordingly, it is possible to specify a model for ilmenite
weathering. Shallow burial above the water table permits inter-
mittent saturation with surface-derived humic acid solutions.
These humic acids donate oxygen forfirst-stagealteration but are
sufficiently reduced to put iron into solution. Humic material can
precipitate as humate either by consumption-neutralization on
grains of altering ilmenite or by pH-neutralization at the water
table. Iron put into solution by humic acid leaching can precipi-
Figure 48. Vertical profile of ilmenite alteration in surficial sands of the tate as hydroxides and/or complex with humate in the same two
Yoganup shoreline, western Australia. Values in Welch and others
places.
(1975) have been averaged for each depth interval.
A typical detrital concentration of titanium oxide minerals
contains grains that have survived such alteration in varying
numbers of intermediate hosts and to varying degrees. Each de-
and 35°S. All of the economic Quaternary shoreline deposits of posit records two types of alteration, a predepositional alteration
ilmenite occur in this zone. Thus, in deposits of this type, the of mixed degree and an overprinted postdepositional alteration
distribution of altered ilmenite defines a latitude-parallel zone limited to certain surficial zones.
permissive of exploitation.
CLIMATE AND PALEOCLIMATE
A PROPOSED MECHANISM OF ILMENITE
ALTERATION Preceding discussions have treated weathering as an un-
changing phenomenon, but weathering is a function of climate,
Some clues to the nature of ilmenite alteration are implied and climate is known to have changed through geologic time
by relations described above. The unique potency of humic acids (reviewed by Frakes, 1979). Even during the short (Quaternary)
in the leaching of ilmenite implies that its oxygen-buifering capa- period of time considered in Figures 50 through 53, important
bility, as well as some acidity, is required. We know from the variations of climate have occurred. The strong latitude-related
chemical path of ilmenite alteration that oxidation is followed by control implied in those figures for the alteration of ilmenite
some combination of acidification and/or reduction to put triva- actually represents a collage of interglacial climatic conditions
lent iron into solution. We know from postdepositional through Quaternary time. Ilmenite compositional gradients
weathering profiles that ilmenite is effectively leached in a zone of shown at 35° latitude may not correspond precisely with today's
shallow burial above the water table (Fig. 47). We know that climate but rather with a weighted average of Quaternary inter-
humate accumulates at the water table by neutralization of humic glacial climates.
58 E. R. Force
S.E. U.S. - pyribole in heavy mineral fraction S.E. U.S. - ilmenite composition
'25m terrace
to 8m terraces
1- to 3m terraces
10
~M c
o
O o
Modem Beaches
H—
20
5
'Z £
> —
Q . CD
•S ® 3 0
®
C» c
C
2 >40
c >
Q) CD
O ®
50
28' 29' 30' 31" 32" 33" 34 35" 36' 37' 38'
North latitude
Figure 50. Plot of percent pyriboles (pyroxene plus amphibole) in Qua-
32" 33' 34" 35' 36" 37' 38'
ternary^) shoreline deposits of the southeastern United States as a func-
North latitude
tion of terrace elevation and of latitude. Data from Martens, 1935;
Figure 51. Plot of TÌO2 content of ilmenite concentrates from the south-
McCaulay, 1960; Neiheisel, 1962; Thorn and others, 1972; Beck, 1973;
eastern United States, as in Figure 50. Sources include my unpublished
Oaks and others, 1974; and Force and others, 1982.
data; unpublished data of J. B. Mertie; McCauley, 1960; Force and
Geraci, 1975; Force and others, 1982; and Pirkle and others, 1984. The
anomalous trend of the 25-m terrace is unexplained; the data come from
different sources.
Climate change and weathering

Climatic zones on continents are basically latitude parallel, others, 1985), perhaps to 80°, and carbon dioxide and oxygen in
with an overprinted difference between east and west coasts. the atmosphere were higher (Fischer, 1981; Berner and others,
Climate change can be visualized as relative motion between a 1983; Berner and Landis, 1988). Deep weathering, locally to
given continent and the climate belts crossing it. This motion may produce economic mineral deposits, occurred (Goldich, 1938;
be of two general types. (1) Latitude-parallel belts or isotherms Tourtelot, 1983; Purucker, 1983). Ilmenite-bearing placer depos-
may move away from or toward the equator by climate change; its that formed during Cretaceous time (Wilcox, 1971; Houston
this is potentially the more rapid motion. (2) Tectonic motion of and Murphy, 1977) or were reworked from Cretaceous deposits
a continent may carry it through climatic zones with fixed posi- (Horse Creek placer of Williams, 1967) commonly have mature
tions. The second factor was important in the Cenozoic for mineral suites, including ilmenite having high Ti0 2 contents.
Australia and the Indian subcontinent. These deposits are also characterized by high ratios of monazite
Two examples of resultant climatic change that affects to total heavy minerals. The stability of monazite, a soft, dense
weathering rates are: (1) a cool moist area becomes a warm moist phosphate mineral, may be sensitive to atmospheric composition.
area, accelerating the weathering rate; and (2) a hot dry area Perhaps high atmospheric CO2, oxygen, and/or some other at-
becomes a hot moist area, also accelerating weathering. mospheric gas stabilized detrital monazite during the Cretaceous.
Each geologic time period is characterized by certain cli- Puffer and Cousminer (1982) have analyzed paleoclimatic
matic regimes and consequently by different areal distributions of controls of ilmenite weathering in a deposit of Miocene-Pliocene
intense weathering. These temporal variations in weathering are age in some detail. This time period was one of relatively intense
recorded in the detrital mineral assemblage of sedimentary rocks weathering also. Paleoclimatic aspects of pre-Quaternary placer
deposited through these time periods. The mineralogy of some of deposits are further discussed in Chapter 10.
these rocks records weathering more severe than at present.
Where these rocks function as intermediate sedimentary hosts for Variation with Quaternary sea levels
younger deposits of titanium oxide minerals, they release grains to
the modern environment that are weathered beyond the power of During the Quaternary, climatic variations have been tied
present climatic conditions. Herein lies much of the importance of more or less mechanically to variations in ice volume and thus to
intermediate hosts in the weathering history of titanium minerals. sea levels. During peak glacial periods, sea level was about 85 m
The Cretaceous is a good example of a time period when lower than today's, sea-surface temperatures averaged about 4°C
climate and weathering were much different than today. Temper- cooler, and the CO2 content of the atmosphere was about 30
ate climates extended to high latitudes (Frakes, 1979; Arthur and percent less (Ruddiman, 1985; Oeschger, 1985; reviewed in
Western Australia - ilmenite composition history of shoreline deposits: (1) the duration of subaerial expo-
sure, that is, the age of the deposit minus duration of submer-
Q.
gence; and (2) the weighted average of weathering rate while
03
O exposed, which is a function of paleoclimate. Total weathering
65 1 1 1 1 ! should be a function of the product of these two factors, both of
which in turn are functions of the present depth of submergence:
o W = Rt =/, (D)/2(D)
o 60 30- to 115- m terraces /
(6-1)
cu
where W is total postdepositional weathering, R is weathering
rate, a function of climate, t is time intervals of subaerial expo-
55 -
sure, and D is present depth of submergence.
0- to 10-m y Variations in weathering rate [R =/i(D)] are addressed first,
terraces /
and the east coast of the United States is used as an example.
0
<u 50 Coastal temperatures for each season during glacial periods have
01
been quantified by Mclntyre and others (1976). Figure 54 con-
verts their coastal paleotemperatures to isotherms on the conti-
45 1 1 1 1 nental shelf; these isotherms represent summer temperatures
35" 34" 33' 32" 31" 30" 29"
when each point was at sea level, assuming that sea-level fall was
S o u t h latitude
proportional to temperature drop. Isotherm shifts of about 3° of
Figure 52. Plot of T i 0 2 content of ilmenite concentrates in Quater-
n a r y ^ ) shoreline deposits of western Australia. Data from Lissiman and
latitude toward the south are common on the outer shelf. If these
Oxenford (1973; Eneabba only) and Baxter (1977). temperature variations were representative of all climate varia-
tions, the favorable zone for the formation of ilmenite placer
deposits would shrink during a glacial period to latitudes lower
than about 32°.
Sundquist and Broecker, 1985). The smoothly cyclical and repeti- The other factor in weathering, duration of subaerial expo-
tive nature of both the sea-level and paleoclimatic Quaternary sure [t = /2(D)], acts as a permissive condition. When a
cycles suggests that certain Quaternary sea levels represent certain previously formed shoreline deposit is submerged below sea level,
climates, independent of time. If so, we can address the subaerial it cannot weather in the conventional sense; there is no evidence
weathering history of Quaternary shoreline deposits formed at of submarine leaching of titanium oxide minerals. Thus, subaerial
various sea levels, including those now offshore. This approach is exposure is necessary for weathering of mineral suites of interest.
invalid for tectonically active coasts, including those rebounding A shoreline deposit that formed at sea level but is now at 60-m
from ice load. depth on the ocean floor would seldom have been exposed as sea
Two factors interact to determine the subaerial weathering level cycled through the range of glacially controlled stands.
South latitudes Equator North latitudes

Figure 53. Plot of T i 0 2 content of ilmenite concentrates as a function of latitude, in Quaternary
shoreline deposits of the world, compiled from numerous sources, published and unpublished. Data
from Figures 51 and 52 are summarized as lines.
60 E. R. Force
Present
Sea
Level
5 >3) 40-
o®
0) CO
_Q CD
« ">
60- ß
0) c
0> S
5 <U
Present 40 80 120 160 200

T h o u s a n d years before present
Figure 55. Quaternary sea-level curve with shading proportional to

weathering intensity at each sea-level position. Weathering history of
deposits A and B contrasted in text.
The two factors act in concert in such a way that Quater-

nary shoreline deposits found at progressively greater depths off-
shore are progressively less weathered. A shoreline deposit now at
great depth has seldom been exposed, and when it was, the
temperature was such that it did not weather rapidly. Observed
variations in the mineralogy of offshore sands (Reich and others,
1982) suggest that immature assemblages are found at greater
depths. This principle could apply even to shoreline deposits that
are smeared out by later current reworking.
Figure 55 is a diagrammatic solution of the interaction of the
weathering rate and duration variables (Equation 6-1). It is a
sea-level curve on which are superimposed weathering-intensity
variations that correspond to each sea-level position. Points A
and B represent two shoreline deposits of about the same age that
were formed during different phases of the same sea-level cycle.
Deposit A received weathering proportional to all shading along
the sea-level curve below line A-A'. The deposit at B receives
much less weathering, proportional to shading below line B-B',
probably about 25 percent as much as A.
EXPLORATION IMPLICATIONS
Weathering variations have exploration implications for

weathering-beneficiated deposits of titanium oxide minerals in
igneous and metamorphic deposits and for all placer deposits. In
deposits of Cretaceous age, weathering beneficiation probably
occurred on all stable, moderate-relief land surfaces at latitudes
up to at least 50°, except where the landmass more recently
moved into such latitudes from higher latitudes. For younger
Figure 54. M a p of the continental shelf off the eastern United States
showing summer isotherms in degrees centigrade (solid lines) applicable
deposits this beneficiation situation is more constricted. For Qua-
when each point was at sea level. Bathymétrie contours (dashed) from ternary deposits formed during interglacial periods, the favorable
National Oceanographic and Atmospheric Administration. Quaternary latitudes of syn- and postdepositional weathering are lower than
paleotemperatures from Mclntyre and others (1976). 35°. For deposits formed at the lower sea levels of glacial peri-
ods, favorable latitudes may not exist.
Chapter 10.
Titanium oxide minerals in sedimentary rocks and principles ofplacer concentration
DISTRIBUTION AMONG SEDIMENTARY ROCKS mostly as altered ilmenite and rutile, as by definition all labile
minerals, including titanium silicates and magnetite, are absent.
The chemical immobility of titanium has produced a predic- The restricted mineralogy of these sands is acquired by multiple
table distribution pattern of titanium in sedimentary rocks. Chem- cycles of chemical and mechanical beneficiation, commonly in
ical sediments are systematically impoverished. Detrital sedi- passive-margin or cratonic tectonic settings. Because only a few
ments have titanium contents controlled by the mechanical minerals can contain titanium in these sands, titanium enrich-
behavior of titanium-bearing grains in various hydraulic regimes. ments in mature sands are unusual. However, mature sands in
These grains may be either titanium-bearing silicates or oxides. which placer concentration has occurred have moderate to high
Limestones average 0.1 to 0.2 percent Ti0 2 (Table 1). Even Ti0 2 contents as rutile and ilmenite and are economically valua-
in these chemical sediments, Ti0 2 is commonly present as detrital ble deposits. The remainder of this book is devoted to discussion
admixtures, as in eolian sand fractions containing ilmenite. of these deposits and the processes forming them.
Shales average 0.6 to 0.7 percent Ti0 2 (Table 1), and values
over 1 percent are common; shale is the sedimentary rock with PLACER CONCENTRATION
the highest average Ti0 2 content. Because offinegrain sizes, the
titanium mineralogy of shale has never been adequately studied. I A placer-enriched deposit is formed by the concentration of
suspect that much of the Ti0 2 in shales represents the fine- denser minerals. This concentration occurs by a complex interac-
grained anatase present in regoliths as a result of the weathering tion among transport type, particle size and shape, and particle
of titanium-bearing silicates. Siltstones have slightly lower aver- density. I shall attempt to explain this interaction in nonmathe-
age Ti0 2 contents than do true shales. matical terms. For any type of transport or erosion, a given flow
Sandstones and coarser clastics vary greatly in Ti0 2 con- removes some particles from the bed and leaves others. The
tents because of variations in concentration of titanium-bearing division is primarily a function of particle size and density. For
grains; the average value must be 0.25 to 0.5 percent Ti0 2 . The the given flow, a particle of quartz or feldspar of a certain grain
Ti0 2 also shows great variation in partitioning between oxide size will behave like a denser mineral of a different grain size. The
and silicate minerals. In the mineralogically immature sandstones, laws relating particle size and density vary among different trans-
titanium mineralogy closely follows that of source rocks. This port modes, and this variation is the key to placer concentration;
typically means that much titanium is present as silicate grains, adjustment by the bed to successive types of transport and erosion
particularly in the active continental-margin settings where most can enrich dense minerals.
of these sands are formed. For example, in voluminous terranes of Transport laws vary at a number of scales; for example, the
volcanogenic graywackes, representing many depositional envi- laws for a beach deposit and for the adjacent eolian dune differ.
ronments, sandstones contain titanium as magnetite, ilmenite, At progressively smaller scales, within the beach deposit there are
titanaugite, hornblende, and/or sphene. This suite of titanium different laws for breaking-wave and backwash transport and
minerals may occur in placer concentrations of heavy minerals, even for different portions of a single migrating ripple.
resulting in deposits having high Ti0 2 contents but no economic The laws governing equilibrium transport as a function of
value (e.g., Thomas and Berryhill, 1962). density and grain size have been studied for many modes of
Mineralogically mature sandstones typically contain Ti0 2 transport and erosion. The law studied first (Rubey, 1933) is
61
62 E. R. Force
probably the simplest—deposition from suspension, such that
grains of the same settling velocity accumulate together. Settling
velocity is, of course, a function of density and grain size. The
term hydraulic equivalence was preempted for this type of equi- Suspension equivalence only
9o°om°
librium (Rittenhouse, 1943), but clearly a whole family of
hydraulic equivalences exists for different transport types. These
can be loosely divided into depositional equivalences (which in-
clude suspension equivalence) and entrainment equivalences,
with transport equivalence a complex mixture of the two (Slin-
gerland, 1984; Slingerland and Smith, 1986). In some deposits,
dispersive equivalence (natural heavy-media separation) is also a
factor. Each equivalence type is described below. Entrainment equivalence only
(supply random)
ooo«oooo
The enrichment of dense minerals (relative to light minerals)
is not directly explained by any one law of hydraulic equivalence,
because in deposits formed according to that law, heavy minerals
will be deposited with hydraulically equivalent light minerals, O • O O O • O
and no concentration is achieved (Force, 1976b; Komar and
Wang, 1984; Slingerland and Smith, 1986). In only one circum-
stance can an enrichment result from deposition obeying one Sequential (suspension equivalence
law—the case in which the dense mineral of interest is supplied in followed by entrainment equivalence)
a range of grain sizes equivalent to a size range for light minerals
that is in short supply. Except perhaps for minerals having spe-
cific gravities more than 8, that is, minerals hydraulically equiva-
lent to much coarser light minerals (Tourtelot, 1968), such
enrichments are rare.
The general case for placer concentration is the sequential Figure 56. Diagram of grain populations of light (open) and heavy
operation of two slightly different laws in such a way that the (dark) minerals resulting from various hydraulic equivalence laws.
sediment can obey both laws only by becoming enriched in small
dense minerals. Most commonly, one law is of the deposition
type and the other is of the entrainment type (cf., Komar and
Wang, 1984). For example, deposition from a turbulent breaking comes in the separation of depositional and entrainment compo-
wave on a beach face is essentially by a law of suspension equiva- nents of transport. In our wave example, thorough sampling be-
lence, with any given spot in the swash zone representing a tween breaking wave and backwash is difficult (however, see
certain settling rate. Large light grains are deposited with small Slingerland, 1984).
dense grains (Fig. 56, top). The more laminar backwash of the Rubey (1933) formulated the concept that grains found to-
same wave, however, has an erosive effect since its suspended gether forming a single bed represent grains having equal settling
load has been dropped. It erodes according to an entrainment velocity, that is, coarser light minerals and heavy mineralsfinerin
equivalence law that is almost entirely a function of grain size; the proportion to their densities. Rittenhouse (1943) called this con-
grains that project out of the bed are eroded, and these are mostly dition hydraulic equivalence and cast his similar results with flu-
coarse (Slingerland and Smith, 1986). Thus the large light grains vial sands in the form of hydraulic ratios relating the diameters of
selected by the first process are removed by the second, and an light and various heavy grains deposited together. Ironically,
enrichment in small dense grains results (Fig. 56, bottom). The though the more sophisticated treatments of heavy mineral depo-
enriched deposit remains in equilibrium with the next wave. Note sition have moved significantly beyond these landmark papers,
that if thefirstprocess had been inefficient, large dense grains and most studies of heavy mineral assemblages are not up to the
fine light grains would also have been deposited. The latter would conceptual level of Rubey and Rittenhouse (reviewed by Force
not have been eroded by the second process, and little or no and Stone, 1990).
enrichment would have occurred (Fig. 56, center). This simplified Figure 57 illustrates the working of depositional equivalence
example is expanded in Chapter 9. in a single bed. This example is from foreset beds of a glaciolacus-
trine delta, and deposition was from suspension in continuous
Depositional equivalence turbidity flows (Force and Stone, 1990). The size distributions of
various minerals supplied to the depositional environment are
Because of difficulties of direct observation, laws of deposi- shown in the upper part of the figure. The size distribution of light
tional equivalence have mostly been studied experimentally and mineral grains deposited from a given flow, to form a given bed,
theoretically. The difficulty of observation in a natural situation are shown below in solid lines. Cumulative curves for some
heavy mineral species are shown superimposed on this size distri- White and Williams, 1967; Grigg and Rathbun, 1969; Lowright
bution; the position of the curve for each mineral is a function of and others, 1972; Slingerland, 1977, 1984; Komar and Wang,
density and shape of the mineral. The relation of its size distribu- 1984; Slingerland and Smith, 1986). In general, these studies
tion to the size distribution of the whole sample isfixed;that is, have shown that entrainment is a function more of grain size than
the family of curves in the bottom part of the figure slides in its of grain density. Large grains may project out of the bed into a
entirety to the right or left as modal grain size of the bed changes zone of turbulent flow and thus be preferentially subjected to
as a function of flow parameters. An arbitrary distribution repre- removal. In addition, larger grains can be preferentially dislodged
senting an adjacent bed is shown with dashed lines. The abun- by rotation through smaller angles. In a beach example studied by
dance of a given mineral in each bed depends on the relative Komar and Wang (1984), larger grains of all mineral species are
position of the size of that mineral required for deposition from removed from the upper swash zone and transported progres-
the flow (bottom of figure) and the size of the mineral in abun- sively offshore. As the entire beach is fed coarser lights and finer
dant supply (top of figure). Where the two coincide, the mineral heavies, entrainment results in a placer enrichment in the upper
will be relatively abundant. In the example shown, ilmenite and swash zone, which gives way seaward across the beach face to
biotite will be abundant in the bed represented by the solid line, coarser-grained, less enriched deposits showing a spectrum of
whereas garnet will predominate in the bed represented by dominant heavies having progressively lesser density.
dashed lines. Such glaciolacustrine deposits are further discussed
in Chapter 8; in the present discussion, these and analogous Transport equivalence
depositional-equivalent deposits are protores for further enrich-
ment by entrainment and transport, which have other equiva- Fine grains commonly lag behind coarser grains in transport
lence laws. over a coarse bed as a consequence of entrainment differences.
Where heavy minerals are finer than depositionally equivalent
Entrainment equivalence light minerals, this results in dynamic lag enrichments of heavy
minerals (Slingerland, 1984). Where heavy and light minerals are
Most recent studies of heavy mineral deposition have em- supplied to the transport system at the same average size, heavy
phasized entrainment equivalence (Mclntyre, 1959; Hand, 1967; minerals lag because their greater settling velocities cause them to
strike the bed more frequently. Where the depositional system is
Grain size of supply, in millimeters arrested, the lesser transport rates of denser minerals result in
local concentrations.
Dispersive equivalence
In dense grain suspensions, grain collisions and fluid pres-
sures may permit natural heavy-media separations of light and
heavy minerals. Such separation depends not on flow parameters
but on factors that permit the bed to remain liquified. Dense
minerals sink to the base of the liquified layer. On beaches, for
example, this may occur by wave-induced lateral gradients in the
local water table and/or by pounding of the surf. Sallenger
(1979) has shown that differential settling of denser and/or finer
grains through the bed may result in both basal placer enrichment
and inverse grading of beach deposits.
ECONOMIC PLACER DEPOSITS OF TITANIUM

MINERALS—A SYNTHESIS THUS FAR
It is now time to gather some threads leading from previous

parts of this book. These threads are some necessary precondi-
tions for the formation of economic placer deposits of titanium
Figure 57. Grain size distributions showing supply and depositional in- minerals.
teractions in depositional equivalence. Supply curves shown above; size Source rocks must supply an appropriate mineral suite. The
distributions of two beds, and minerals thereof, shown below (one bed most favorable source terranes for detrital ilmenite and rutile are
solid, other bed dashed). All curves are frequency diagrams except for high-grade metamorphic rocks, locally with related igneous rocks
the nested cumulative curves for individual minerals in the beds shown
of the charnockite-ferrodiorite-anorthosite suite (Chapters 2 and
below. The interaction between supply and depositional curves is dis-
cussed in the text. 3). These terranes are commonly exposed over areas of thousands
64 E. R. Force
of square kilometers, and ilmenite and rutile dominate their
titanium-mineral assemblages. Rutile is most common in pelitic
lithologies (Figs. 2,3). Erosion of these terranes releases about 0.1
to 1.0 percent ilmenite and rutile to the transport system (re-
viewed for rutile by Force, 1980b).
Other large source terranes supply the wrong mineral suites;
for example, basalt terranes supply ilmenite intergrown with
magnetite; some granitic terranes supply a little ilmenite, inter-
grown with hematite and magnetite; and neither terrane supplies
rutile. The only other source terranes that contribute valuable
assemblages of titanium minerals are small—for example, alkalic
stocks and hydrothermal systems (Chapters 4 and 5). Locally,
intermediate sedimentary hosts are the most important sources,
but these in turn have primary source rocks. If the source terranes
do not supply the right minerals, an economic titanium-mineral
placer cannot form.
Weathering, as predepositional weathering of constituent
grains and/or as postdepositional weathering of a placer deposit,
is necessary also (Chapter 6). It restricts the mineral suite to one
in which titanium minerals and other economic heavies are
among the few heavy species, and it chemically upgrades the
ilmenite. Currently, no titanium-mineral placers are operating in
which this weathering upgrading has not occurred.
Discussion in upcoming chapters treats further controls on
placer location and formation. Conduits, in both fluvial and
economic placer deposits of titanium minerals. A. General case. Each
shoreline environments, are necessary to bring favorable detritus box has the potential of containing economic deposits. B. Adaptation of
to the depositional site in undiluted form. Once placer concentra- general case for Quaternary shoreline placer deposits.
tion has occurred, it must be preserved; locally this has required
removal of the concentrated sediment to another depositional
environment.
Thus, consideration of placer concentration processes, as in Even concentration itself is a two-stage process, and thefirststage
this chapter, requires consideration also of the numerous other may occur in part outside the depositional environment. Inherited
influences on placer formation (Fig. 58A). Most of these influ- characteristics of the placer extend from the depositional site all
ences are not apparent at the site of concentration. A great deal of the way back through the geology of titanium oxides, and these
the geologic evolution of an economic placer deposit is hidden inherited characteristics control the economics of exploitation just
from view at the depositional site and must be studied elsewhere. as surely as does final placer concentration.
Chapter 10.
Placer deposits of titanium oxide minerals in nonmarine sediments
Althouth numerous nonmarine placer deposits of titanium vorable sources. The relative importance of nonsource factors
minerals have been identified, only one, a fluvial deposit in Sierra differs in the two major types of nonmarine deposits; fluvial
Leone, is currently of economic importance. Nature operates deposits are more beneficiated by weathering, whereas glacio-
under too many disadvantages in nonmarine environments for lacustrine deposits are better winnowed but completely
there to be many economic nonmarine placer deposits of titanium unweathered.
minerals. Deposition in these environments is of interest partly as
a step in the formation of shoreline placer deposits. FLUVIAL DEPOSITS
This chapter concentrates on nonmarine placers in young
hosts, because virtually all identified nonmarine titanium-mineral Heavy minerals form concentrations in a large number of
deposits formed during the present geomorphic cycle (Table 16). fluvial environments, and such concentration of different heavy-
The significance of youth in these deposits is apparently that they mineral species in different environments is the subject of a vo-
are unindurated, and they were easy to find using present geo- luminous literature (abstracted by Smith and Minter, 1980;
morphology. However, older fluvial deposits, some of them indu- Slingerland and Smith, 1986). Scales of placer enrichment vary
rated, were important intermediate hosts of titanium oxide from individual bed forms to facies of large depositional systems.
minerals in economic placers of younger shoreline deposits. Much of the information available to us concerns modern en-
richments (some from experiments) at the intermediate scale of
RELATIVE IMPORTANCE OF point bars and channel junctions (Adams and others, 1978;
GEOLOGIC CONTROLS Schumm and others, 1987). Larger-scale enrichments apparently
form by selective preservation in particular environments of
All nonmarine placer deposits of titanium oxide minerals smaller-scale enrichments. The largest scales of fluvial placer con-
appear to be intimately related to favorable source rocks (Table centrations, in wet alluvial fan environments, are mostly of min-
16). The source factor is more important than the other factors erals having specific gravities more than 6.
that normally enter into the formation of a titanium oxide placer
deposit (Fig. 58A), because the other factors are feebly devel- Concentration and weathering limitations
oped. Weathering beneficiation varies greatly but is far less potent
than in some shoreline deposits. True placer concentration of Placer concentration in fluvial environments has formed
titanium minerals in nonmarine deposits is apparently effective important economic deposits of a number of heavy mineral
only at scales much smaller than that of a mineable deposit. No commodities, such as gold, diamonds, monazite, and cassiterite.
large deposits formed in nonmarine environments seem to have Thus it is strange that titanium oxide minerals are extensively
had titanium minerals greatly enriched by placer concentration produced from placer deposits but rarely from fluvial ones. There
compared with the debris fed from the source. are two reasons. First, the scale of fluvial placer concentration is
Because of the importance of the source terranes in the too small. Economic placer deposits of titanium oxide minerals
formation of nonmarine placer deposits of titanium minerals, normally contain at least a million tons of mineral concentrate.
valuable deposits represent unusual geomorphologies and clastic Few fluvial bodies of this size contain concentrations of minerals
distribution systems that minimize dilution by debris from unfa- having specific gravities of 4 to 5, such as rutile and ilmenite.
65
66 E. Force
TABLE 16. NONMARINE PLACER DEPOSITS OF TITANIUM OXIDE MINERALS*
Deposit or Location Age Titanium Source Distribution References

district Mineral Terrane System^
Fluvial
Gbangbama Sierra Leone Quaternary Rutile Kasila granulite-facies Raufuss, 1973
(Sherbro) gneisses
Sand Canyon,
Pacoima Canyon California Quaternary llmenite San Gabriel Mountain 2,3 Oakeschott, 1958;
ferrodiorites§ Industrial Minerals,
1986
Otter Creek Oklahoma Quaternary llmenite Wichita Mountain gabbros, 2 Hahn and Fine, 1960
etc.
Shooting Creek North Carolina Quaternary Rutile Garnet-mica schist 2 Hartley, 1971
Horse Creek South Carolina Quaternary llmenite, rutile Cretaceous sands 2b Williams, 1967
Roseland Virginia Quaternary llmenite, rutile Roseland crystalline rocks§ 2 Minard and others,
1976; Herz and
Force, 1987
Glaciolacustrine
Port Leyden New York Pleistocene llmenite Adirondack Mountain Stone and Force,
charnockites, gabbros, and 1980
granulite-facies gneisses§
*Not including the U.S.S.R., China, or operations recovering titanium minerals as byproducts.
t S e e text and Figure 59.
§Chapter 3.
Second, fluvial deposits contain ilmenite of roughly the same situation in the Gbangbama district has produced the most valua-
composition as that supplied by source rocks. Shoreline placer ble single deposit in nonmarine deposits. (2) Drainage basins
deposits have the advantage of greater ilmenite alteration. In the entirely within favorable source rocks. A subtype is along-strike
only major economic fluvial deposit, the Gbangbama district of drainages with few tributaries in elongate favorable terrane (type
Sierra Leone, the detrital mineral of value is rutile, which needs 2b, Fig. 59). (3) Drainage basins having headwaters in favorable
no chemical beneficiation by weathering. source rocks and surrounded by other sediments derived from
Table 16 lists the better-documented economic and near- those favorable sources. The third type of system can be coupled
economicfluvialdeposits of titanium minerals. Titanium-mineral with either of thefirsttwo, thereby extending them downstream.
contents in these deposits are only slightly greater than in source Most known fluvial titanium-mineral deposits are of one or more
rocks. Most beds in these deposits are poorly sorted, and conse- of these types (Table 16).
quently there is not much concentration. Within favorable drainage basins, titanium oxide minerals
Ilmenite in these deposits is relatively unaltered, but weath- are distributed in predictable ways. In the fluvial deposits that
ering beneficiation of fluvial deposits does have a considerable have been studied sufficiently, titanium oxide minerals decrease
effect in restricting heavy mineral suites toward an economic away from the source (Hahn and Fine, 1960; Raufuss, 1973;
assemblage. All thefluvialdeposits listed in Table 16 are in areas Minard and others, 1976). This may be in spite of an increase in
of deep weathering. Even the San Gabriel, California, deposits, the proportion of titanium oxides in the heavy fraction by attri-
which formed adjacent to an area of high relief, show some tion of less stable heavies (Raufuss, 1973). The trend is due to
modification of mineral assemblages by weathering. The eco- some combination of lag effects and dilution by other debris.
nomic role of disaggregation and weathering probably exceeds Lithologically, the titanium-oxide mineral deposits listed in
that of placer concentration in most fluvial deposits of titanium Table 16 occur mostly in gravelly silty sands. For example, in
oxide minerals. alluvium of the Roseland district of Virginia, the 37 layers with
greater than 3 percent heavy minerals, out of 122 layers reported
Geomorphic and lithologic controls by Minard and others (1976), average 11.1 percent gravel (>2
mm) and 25.5 percent silt and clay (<0.06 mm). Trask sorting
Three distribution systems (Fig. 59) that permit little dilu- values (So) range from about 2 to 6, that is, the deposits are
tion favor the formation of fluvial deposits of titanium minerals: generally poorly sorted. Concentrations greater than 10 percent
(1) Radial drainages from positive favorable source areas. This of heavy minerals occur in deposits that are better sorted, includ-

ing some that are muddy or gravelly but not both. Heavy mineral GLACIOLACUSTRINE DEPOSITS
mode was 0 to 1 </>finerthan the mode of the entire sample (Fig.
60A) and varied from about 1 to 3 $ (0.125 to 0.5 mm). That is, During the Pleistocene, much glacial debris was sorted and
heavy minerals are the transport equivalents of coarse to fine deposited in Gilbert-type tripartite deltas in numerous large gla-
quartz sand in these deposits. cial lakes. In deltas derived from crystalline rocks, Force and
Commonly, the heavy-mineral content increases with depth others (1976), Stone and Force (1980, 1983), and Force and
within the alluvial sequence (Hahn and Fine, 1960; Raufuss, Stone (1990) have found that entire foreset sequences, containing
1973; Minard and others, 1976). Probably the basal deposits are up to 108 metric tons of sand, are heavy-mineral deposits contain-
less muddy and represent axial-channel environments rather than ing 4 to more than 10 percent heavy minerals. In the sense of
point-bar or flood-plain deposits. Chapter 7, these show only thefirststage of placer enrichment.
The foreset sequences, as thick as 80 m or more, are largely
sandy and consist of beds in unconformity-bounded sets, dipping
10° to 40° toward the lake basin. Unconformably overlying the
foreset sequence are cobbly fluvial topset beds (Fig. 61 A).
Individual foreset beds range from coarse to veryfinesand.
Gravel beds and pebbly sand beds are also common. Sedimentary
structures are planar lamination and climbing-ripple cross-
lamination, the latter more common infinerbeds. Sorting is quite
good (So 1.3 to 1.6) in most foreset beds, including virtually all
sand beds and some gravel beds, but is poor in bimodal gravelly
sands. Sorting values for monomineralic populations are com-
monly as good as So 1.2. Titanium oxides are enriched in size
fractions about 1 4> gradefinerthan that of the sand-sized mode
for the host bed (Fig. 60B). Heavy-mineral contents are lower in
gravelly beds, because of dilution of the sand fraction. Variations
with stratigraphic positions or sedimentary structure were not
detected in the Connecticut deltas studied by Force and Stone
(1990).
The deposits are derived from specific local source terranes
via glacial action, and are transported in meltwater and deposited
on delta fronts too rapidly for any significant weathering to occur.
Thus, in these deposits, source rocks are of great importance;
Force and Stone (1990) show that adjacent coeval deltas with
different sources have correspondingly different detrital mineral-
ogies. Apparently little mixing of debris occurs in the subglacial
environment, and the meltwater dispersal path is short.
Deposition of the foreset beds is by bottom-hugging turbid
currents that transport the entire fluvial sediment load down the
steep foreset slope (Gustavson and others, 1975). These currents
are continuous but variable and shift laterally. Deposition is from
suspension but with a variable traction component. Since the
currents are continuous and since deposition is directly from the
current, a given current deposits a bed with a narrow range of
settling velocities; measured SQ values imply settling velocities
differing by ±45 percent for the central half of the grain size
distribution. Sorting this good is unusual for deposition from
suspension, because most suspension deposition is from rapidly
waning currents or from sediment rain through still water (as in
marine deltas); under these conditions, grains of different settling
velocities are deposited together.
In the prodelta environment, erosion probably occurs as a
Figure 59. Diagram of fluvial distribution systems (numbered) that per-
consequence of lateral shifting of currents, thus producing uncon-
mit transport with little contamination by mineral suites other than those
from favorable sources (discussed in text). Drainage type 1, radial;
formities between sets. Within sets, the sedimentary structures,
2, within source; 3, within debris of favorable source. especially steeply climbing ripples, show that reentrainment of
K R. Force
0.015 .0075
0.5 0.25 0.125 0.06 0.03 0.015 .0075
Grain size, in millimeters

Figure 60. Examples of size distributions of light and heavy (shaded) minerals in some placer-forming
environments. A. Fluvial deposit, Roseland district, Virginia; sample 328 of Minard and others
(1976). B. Glaciolacustrine deposit, Connecticut; sample K12 of Force and Stone (1990). C. Eolian
heavy mineral enrichment, Trail Ridge, Florida; Force and Rich (1989).
B
Figure 61. Structure of typical glaciolacustrine deltas. A. Topset and upper foreset beds, Hopeville
delta, Quinebaug valley, Connecticut. Exposure about 5.5 m high. B. Cross section of the Port
Leyden delta, New York (7.2 km across).
previously deposited grains is minor. Thus placer concentration is Rocks of the Gbangbama Hills are deeply weathered garnet
arrested at thefirststage, the establishment of depositional (sus- amphibolites and leucocratic garnet granulites (Raufuss, 1973).
pension) equivalence. The latter characteristically contain rutile in concentrations of
In glaciolacustrine foreset beds, the grain size of a bed exerts about 0.2 to more than 1 percent. Besides garnet and rutile, the
strong control over the mineral assemblage in the bed. Figure 57 granulites contain two pyroxenes, ilmenite, apatite, zircon, and
illustrates the relation of heavy-mineral deposition to grain size of locally sphene. The garnet amphibolites also contain local rutile,
host and to the nature of bulk feed from source in these deposits but as these rocks are more resistant to weathering, this lithology
(see text of Chapter 7 for full explanation). Given the nature of is less well represented in derived alluvial sands.
this bulk feed and the modal grain size of each bed, the mineral- Rutile-bearing deposits are radially disposed around the
ogy of the bed can be predicted rather precisely. For example, Gbangbama Hills. The present drainage does not reflect this ra-
coarser beds may be dominated by garnet and finer beds by dial pattern. At least some of the deposits are late Pleistocene in
titanium oxide minerals. Actual enrichment in titanium oxide age, based on 14C of lignitic layers. Characteristically, they are
minerals over that in source rocks is uncommon in this environ- poorly sorted sands, commonly gravelly and averaging 35 to 45
ment; it occurs only where the grain sizes of titanium minerals percent muddy matrix, locally with clay layers. The deposits are
and of hydraulically equivalent light minerals are fortuitously mostly about 10 m and up to 20 m thick, lying directly on
those in which the ratio of oxide minerals to light minerals is high weathered bedrock. Many of them form discrete inliers in crystal-
in the bulk feed from source areas. line terrane.
The deposits themselves contain as many as two laterization
M A J O R DEPOSITS surfaces. Differentiation of laterized sediment and saprolite is
locally difficult (Raufuss, 1973), and this difficulty is said to have
Gbangbama district, Sierra Leone hampered earlier mining. Multiple weathering cycles clearly have
Rutile has been mined from the Gbangbama district in the affected this deposit, and the extent of mineralogic beneficiation is
coastal region of Sierra Leone since 1967. Economically, these significant.
ventures have had a spotted history, but currently, under Sierra The detrital mineral assemblage of the deposit is actually
Rutile, the operations are a major rutile supplier. Sierra Leone is dominated by rutile. Ilmenite contents are roughly inverse to
surely the only country for which rutile is the most important rutile contents (probably because this ratio varies in the source
earner of export revenue. terrane). Corroded pyropic garnet is a locally major constituent; it
The district consists of the Gbangbama Hills, about 200 m is intergrown with rutile. Other locally common minerals are
high and trending northwest to southeast, surrounded by low amphiboles, both pyroxenes, kyanite, sillimanite, zircon, and
coastal plain. The vegetation is rain forest, and the area was graphite. Minor constituents include magnetite, monazite, sphene,
remote until the coming of mining operations (Spencer, 1964). corundum, and tourmaline. Iron oxides record the former pres-
My description of the deposits is largely from Raufuss ence of mineral grains unstable in the weathering environment.
(1973). Additional information is from Spencer and Williams Total heavy-mineral contents commonly range from 1 to 5 per-
(1967) and Lang (1970). cent, with rutile contents ranging from 0.5 to 2 percent.
70 E. R. Force
Raufuss (1973) ascribes deposition of the most poorly sorted Camp, 1974). Areas to the north and south apparently contain
deposits (up to 60 percent silt and clay) to sheet floods and substantial additional resources. Ilmenite-bearing grains average
remarks on the ineffectiveness of true placer concentration here. 35 percent of the heavy mineral fraction; these contain locked
Rutile content shows a high at an optimum transport distance of a silicate impurities and are completely unweathered. The ilmenite
few kilometers, apparently as a result of the interaction of favora- is approximately stoichiometric in composition, but ilmenite-
ble and unfavorable transport factors. Immediately adjacent to bearing grains as a concentrate contain only about 25 percent
the source is a zone of poorly sorted gravel and slumped laterite; TiC>2. Other minerals of value in these sands include zircon,
outside this is a zone of some hydraulic sorting. In this second sillimanite, and minor rutile. Pyroxene, amphibole, and garnet,
zone, heavy-mineral concentration decreases away from the however, are the most abundant constituents of the heavy-
source because of decreasing grain size and possibly lag effects. mineral fractions. Feldspar is abundant in the light fractions. All
The rutile portion of the heavy-mineral fraction increases in this the grains are poorly rounded.
zone, however, because of progressive elimination by weathering
of garnet, amphibole, pyroxene, and magnetite. Raufuss notes Titanium oxide minerals as byproducts from
that deeper deposits having coarser grain sizes commonly have fluvial deposits
higher rutile contents.
Mining thus far has been in the area northeast of the Gbang- Fluvial deposits of value for other commodities locally con-
bama Hills near Mogbwemo. Deposits on the southwestern side tain titanium oxide minerals recoverable as a byproduct; in only
are similar and perhaps equally extensive, but they are slightly two examples is recovery of titanium minerals occurring at pres-
lower in grade and show some marine influence. ent. The first is ilmenite from alluvial tin deposits of Malaya,
The unusual mineral assemblage of the Gbangbama district Indonesia, and Thailand (Flinter, 1959; Macdonald, 1971a, b;
is related to an unusual source, little dilution, and weathering Industrial Minerals, 1972; Achalabhuti and others, 1975). Malay-
beneficiation. Placer concentration played a minor role in the sia is a significant world producer of ilmenite based on this by-
formation of economic deposits here. product recovery (Table 3). The second example is in the San
Similar occurrences of rutile-dominated detrital deposits are Gabriel Mountains of California, where mining is primarily for
known to extend toward the northwest, in similar relation to sand and gravel. Ilmenite has been recovered recently from these
Kasila granulites (Raufuss, 1973). To the southeast these rocks deposits (Industrial Minerals, 1986). Recovery of titanium oxides
extend across the Mano River into Liberia, where I have studied has been evaluated for a number of other deposits offluvialsand
them near Lake Piso. Here they consist of equigranular leuco- and gravel or silica sand (Davis and Sullivan, 1971; Gomes and
cratic banded quartz-feldspar-garnet gneisses with minor pyrox- others, 1979, 1980; Force, 1980a). Ilmenite was once recovered
ene and 0.25 to 1 percent rutile. Interlayered mafic granulite from fluvial placer deposits of monazite and other minerals in
gneisses (gnm of Thorman, 1977) containing ilmenite locally Idaho (Storch and Holt, 1963).
predominate. The rocks are deeply weathered, and some of the
fluvial drainages allow little dilution. Thus, the Liberian end of ECONOMIC PROGNOSIS
the province has some potential for additional fluvial rutile
deposits. Sierra Rutile has demonstrated the potential profitability of
fluvial rutile deposits. Far too little is known about the distribu-
Port Leyden delta, New York tion of rutile-rich crystalline rocks and their erosional debris to
claim that this deposit is unique. Where shoreline placer deposits
Along the western margin of the Black River valley, forming of rutile become depleted, exploration in adjacent valleys may
the southwestern margin of the Adirondack Mountains, plateaus lead to fluvial rutile deposits.
whose flat tops are at 1,200 ft (370 m) rise as much as 80 m Forfluvialand glaciolacustrine ilmenite deposits, the lack of
above the valley floor (Fig. 6IB). The plateaus represent coa- ilmenite alteration is a problem shared with magmatic ilmenite
lesced Pleistocene deltas deposited in glacial Lake Port Leyden deposits. Nonmarine ilmenite placers might compete where un-
(Fairchild, 1912), deposited from meltwater of the receding con- weathered ilmenite can be utilized, and if the economic recovery
tinental glacier and derived from the Adirondacks. The deltas balance between high-grade hard rocks and low-grade disaggre-
consist largely of sandy foreset and proximal bottomset beds gated sediments shifts in favor of the sediments.
averaging 3.5 percent heavy minerals (Force and others, 1976;
Stone and Force, 1980). The sands are well to moderately sorted METHODS OF EXPLORATION
and contain sedimentary structures indicative of deposition from
bottom currents; among these areripplecross-laminae with vari- Exploration for nonmarine placer deposits of titanium min-
able angles of climb, in which heavy minerals are enriched. erals must center around the lithology of source rocks, as all such
Ilmenite resources of about 26 million metric tons are con- deposits known are proximal to and closely reflect their sources.
tained in these deltaic beds in the Port Leyden Quadrangle (Force A rutile-bearing source is far more promising than an ilmenite-
and others, 1976; mapped by Howard University Geology Field bearing source, as nonmarine placers seem not to contain altered
ilmenite. A sedimentary source containing already-weathered il- complication of subglacial transport. However, known deposits
menite may be favorable, however. are near their source. The presence of thick deltas in partially
For fluvial deposits, deep weathering of the source terrane is filled basins of former glacial lakes may be obvious from topog-
an additional requirement implied by our present knowledge of raphy or known from geologic maps showing Quaternary units.
such deposits. In some deposits, this weathering is as important as The tendency of fluvial deposits to become richer at depth
fluvial placer concentration in upgrading debris from the source. makes exploration with motorized equipment necessary. Favora-
Unusual geomorphic situations are required in fluvial sys- ble lithologies are coarse and medium sands with little gravel or
tems to prevent the dilution of favorable by unfavorable debris mud matrix; axial-channel deposits may be more favorable than
(Fig. 59). Exploration should be concentrated where these situa- overlying point-bar deposits.
tins permit the accumulation of large-volume deposits derived Geophysical exploration techniques for these deposits are
from a favorable source terrane. not well known. Techniques that are successful in shoreline
Not as much is known about favorable paleodispersal pat- placer deposits, discussed in the next chapter, may be useful with
terns in Pleistocene glaciolacustrine systems, because of the nonmarine deposits.
Chapter 10.
Placer deposits in shoreline-related sands of Quaternary age
Titanium-mineral deposits in sand bodies on Quaternary GENERAL CHARACTER

shorelines are currently of greater importance than deposits of
any other type. Even magmatic ilmenite deposits rank a poor Shoreline deposits of titanium oxide minerals show a great
second in importance behind shoreline placers. In the United deal of variation, as one might expect from the large number of
States, virtually all current production and 36 percent of all iden- factors involved. Yet the most valuable deposits, that is, those in
tified resources are contributed by this deposit type (Table 5). The which all the factors are optimal, show a great number of features
equivalent figures for the entire world are about 55 percent (1987 in common.
production) and 45 percent (Table 4), respectively. Individual The deposits (of Quaternary age) are found on trailing mar-
shoreline placer deposits can contain tens of millions of tons of gins of continents, at latitudes lower than 35°, and are fed by
titanium oxide minerals. detritus mostly from high-grade metamorphic source terranes.
Shoreline placer deposits acquired considerable economic Host sands form coast-parallel surficial bodies, with basal eleva-
stature in the early days of the titanium industry, during the late tions that generally correspond to local Quaternary sea-level
1940s, and their importance increased with development of the high stands. Many of the sand bodies form topographic highs that
chloride process. Only the shoreline deposits have been able to represent former barrier islands and/or eolian dunes. Sand bodies
supply the high-Ti02 altered ilmenite favored for that process; in containing economic titanium-mineral deposits are typically 10
addition, shoreline deposits commonly supply rutile, which is m thick, 1 km wide, and more than 5 km long. Amalgamated
even more valuable. The advantage conferred by a loose, well- smaller bodies of similar origin may be economic, as may super-
sorted sediment with uniquely attractive mineralogy is difficult to imposed composite bodies of different facies. Sets of sand bodies
counter, even with a rock of much higher Ti0 2 grade. The cur- that are parallel but not contiguous commonly represent the
rently economic rutile "placer" deposits of eastern Australia, for strandlines on flights of marine terraces.
example, have cutoff grades well below the average crustal abun- The sands of such bodies are medium- tofine-grained,well
dance of Ti0 2 (about 1.4 percent). sorted, and generally well rounded. Typically the sands are unin-
The complexity of shoreline placer deposits is far greater durated. Depositional facies vary as discussed below. Well-
than meets the eye (Fig. 58B). Other chapters treat hidden factors developed weathering profiles are superimposed on most sand
such as source terranes (Chapters 2 and 3), the weathering history bodies. Local cementation by iron oxides, clay, and/or humate
(6), fluvial conduits (8), and hidden components of the concen- commonly marks postdepositional geochemical interfaces (such
tration system (7) necessary for the formation of these deposits. In as former water tables) within deposits.
addition, the shoreline sands embody a history of sea-level stands The heavy-mineral assemblage is restricted to species resist-
and sediment budgets and represent a number of individual depo- ant to weathering and consists largely of the economic suite al-
sitional and preservational environments. Thus, the study of tered ilmenite, rutile, zircon, aluminosilicate minerals, and
shoreline placer deposits represents a culmination of several dis- monazite. The proportions of these minerals vary considerably,
ciplines and requires all the skill that any investigator can muster. and with these proportions varies the total heavy-mineral content
73
74 E. R. Force
Figure 62. Diagram of various stages of heavy-mineral enrichment in the swash zone of a beach. A
breaking wave is decelerating from left to right, whereas the accelerating backwash of the previous wave
is moving from right to left. Insets show two stages of deposition from the turbulent breaking wave by
suspension equivalence (toward the left), and two stages of placer enrichment by superimposed entrain-
ment and transport equivalence (toward the right). Heavy minerals shown dark.
necessary to make a deposit economic. The range of total heavy results in a continuous but narrow spectrum of grain sizes for
minerals in today's economic deposits is from less than 1 percent each mineral, with finer grains deposited toward the top of the
to more than 25 percent; thus, the color of ore sand varies from swash zone (Miller and Zeigler, 1958). At any given place,
white to black. coarser light minerals are deposited withfinerheavy minerals. At
the upper limit of wave swash, water motion either ceases or is
PROCESSES OF ENRICHMENT slow and along shore, and most of the sediment load is dropped.
When backwash begins and accelerates downslope, the
General principles of placer enrichment, discussed in Chap- water has little suspended load and therefore has an enhanced
ter 7, are applied here to the particular environment of the beach erosive capacity. Water motion in the backwash is a sheet flow,
face. Indeed, this is the environment in which such principles are unlike the turbulent breaking wave. In the viscous boundary layer
most clearly illustrated. Except for attributed statements, the fol- of this flow, velocity and thus erosive capacity are functions of
lowing observations are my own. distance above the bed. The larger grains, which project into the
Detritus supplied to a beach face is commonly somewhat current, are plucked from the bed, and smaller grains are left
sorted by settling velocity (Komar and Wang, 1984). However, behind. Because of prior differential settling from suspension, the
the main separation of particles on the basis of their settling removed coarser grains are light minerals. Thus, in the upper
velocities takes place on the upper part of the beach face, the portion of the swash zone, heavy minerals are concentrated on
swash zone. A breaking wave carries a charge of turbulent the surface of the beach by the law of entrainment equivalence (of
sediment-laden water from the lower, submerged beach face onto Slingerland and Smith, 1986).
the swash zone (Fig. 62). The wave decelerates as it advances. Farther down on the swash face, an entire carpet of grains
Grains are deposited from turbulent suspension and spread across has been entrained in the backwash and is transported down the
the surface of the swash zone, as functions of their settling veloci- swash zone as traction load in the flow. Finer-grained dense
ties. Grains having the same settling velocities are deposited to- minerals ride low in this carpet and travel more slowly than the
gether. The grains that are deposited first, at the bottom of the coarser light minerals above, in transport and dispersive equiva-
swash zone, are those having the highest settling velocities. This lence (of Slingerland and Smith, 1986). Thus the lower part of
the swash face commonly contains a thin, shallowly buried heavy Heavy-mineral concentration does occur in the eolian en-
mineral concentration after the return of backwash to the surf vironment, mostly on the scale of individual bed forms. Heavy-
zone (Clifton, 1969). This concentration can be traced in "subsur- mineral-rich laminae form in response to daily variations in wind
face" to the surficial concentration on the upper swash face. The speed that change the shapes of eolian bedforms (Hunter and
heavy minerals of the lower subzone may be less dense than those Richmond, 1988).
of the upper swash face, however (Komar and Wang, 1984),
because of slightly coarser equilibrium grain size. FORMATION AND PRESERVATION OF
In the lowermost swash face, the return flow may eventually WHOLE DEPOSITS
become more turbulent, ending separation by grain size and
heavy-mineral enrichment. This subzone is commonly marked by Heavy-mineral concentration as discussed in the preceding
a line of shell lags and pebbles that begins the coarser deposits of section is a process involving lag enrichment on the swash face.
the lower beach face (cf., Clifton and others, 1971). The buried The dominance of erosion over deposition on the swash face
heavy-mineral lamination terminates at coarse sediment of this ensures that backwash efficiently sorts the available material, to
subzone. produce a layer enriched in fine dense minerals. Thus the process
Particles left by the backwash of one wave are roughly in of erosion is essential to heavy-mineral concentration. We shall
equilibrium with the next breaking wave, as the grain population see that many individual enriched layers lie on unconformities of
left by each is the same except for the plucking of coarser consti- minor to major significance.
tuents by backwash. The overburden of the buried heavy-mineral Yet a typical economic deposit as described above contains
lamination is preferentially exposed to any erosion by succeeding millions of cubic meters of sand and is clearly the result of preser-
waves. vation of countless superimposed concentrated layers. The clue to
Numerous authors have noted that rich beach concentra- the apparent paradox in the roles of erosion and preservation
tions contain disproportionate percentages of the densest miner- comes from the geometry of enriched layers. In this context, there
als. This is due to sorting by density and grain size within the are two types of shoreline heavy-mineral deposits: those in which
heavy-mineral population on beach segments where most light heavy-mineral concentration occurred at the depositional site,
grains have already been removed. and those in which enrichment occurred elsewhere. These can be
Enrichment processes on the beach face operate most effec- referred to as in-place and transported enrichments.
tively during storms or other periods of high wave energy. During
these storms, high onshore winds transport sand from the beach In-place enrichments
and deposit it in eolian landforms above high tide. Heavy miner-
als are transported preferentially, because the upper swash zone, Many in-place enrichments show progradation of swash
where heavy minerals are exposed, is driest and least cohesive. face environments at constant sea level. Heavy-mineral-rich lay-
Thus, storm periods are optimal not only for heavy-mineral con- ers in these deposits show an imbricate arrangement of former
centration on the beach face but also for storage of the concen- shore faces that dip seaward in a progradational package. Each
trate in the eolian environment. When fair weather returns, the layer enriched in heavy minerals represents an erosional change
dune deposits can be richer in heavy minerals than newly ac- in beach profile to a storm-influenced configuration, and the
creted fair-weather beach deposits, which bury the enriched overlying low-grade layer represents subsequent burial by fair-
beach deposits. weather deposits. Where progradation of fair-weather deposits is
Thus, dune deposits are enriched in heavy minerals not only sufficiently great, they will protect the deposits of one storm from
by concentration that occurs there but also by a highly selective the next storm, and the younger storm profile will be seaward of
supply system. Bulk enrichment of heavy minerals in dunes rela- the older (Fig. 63). Examples of strandline enrichments formed
tive to adjacent beach deposits probably does not occur; studies by progradation at nearly constant sea level are those of eastern
by Shideler and Smith (1984) and Bradley (1957) that claim Australia, formed predominantly by progradational burial of
eolian enrichment were based solely on surficial samples of storm concentrations (Fig. 64A). Some of the heavy minerals in
summer beaches. That is, they did not sample the higher-energy these deposits were supplied by erosional reworking of older
system that supplied the dunes. Authors who have studied whole shoreline deposits.
beach-dune systems (Neiheisel,1958; Gillson, 1959; Lissiman and Deposits on the western coast of Australia show prograda-
Oxenford, 1975; Welch and others, 1975; Collins and Hamilton, tion amplified by local tectonic uplift. Each concentrated layer is
1986) agree that the heavy-mineral contents of coastal eolian related to a slightly lower base level than the preceding layer
deposits are less than those of adjacent beach deposits and that (Figs. 64B, 64C). This is a second type of preservation of in-place
eolian heavy-mineral assemblages contain smaller proportions of enrichments; falling sea level protects previously enriched depos-
the densest minerals. Others have noted that lower dunes have its from marine erosion.
greater heavy-mineral contents than higher dunes (Neiheisel, A third type of preservation of in-place enrichments on the
1958; Fockema, 1986). swash face forms at extreme high tides of meteorological origin.
76 E. R. Force
The more valuable beach deposits include both Pleistocene
Profile (with berm) of and Holocene examples. Even the Pleistocene deposits com-
beach established
between storms
monly preserve strong elements of their original physiography
and apparently represent high stands of Pleistocene sea levels.
They are found on marine terraces roughly parallel to the present
shore. Economic beach placer deposits represent several types of
beaches, such as barrier islands, spits, and cliffed shorelines. Pro-
gradation plays an important part in preservation of mineable
deposits regardless of beach type.
Beach-facies deposits in cross section consist essentially of
/ Heavy-mineral- alternations of sands poor in heavy minerals and sands that are
poor layers
_ X r "
enriched (Fig. 64). Enriched intervals may be as much as several
meters thick, in which case, mining ventures can focus on indi-
I
vidual concentrations. More commonly, individual concentra-
tions are about a centimeter thick, and progradational sequences
Heavy-mineral-
Profile of Profile of enriched layers containing many thin concentrations are mined as if the deposit
beach in beach in were disseminated.
storm #2 storm #1
Individual seaward-dipping heavy-mineral laminae typically
represent swash-zone profiles under storm-wave conditions (Fig.
Figure 63. Diagram of preservation by progradation of heavy-
63). On the tectonically stable coast of eastern Australia, the
mineral-enriched layers produced by storm erosion.
bases of Holocene placer enrichments are found to be at modern
mean sea level (Fig. 64A). Thus, enrichment appears limited to
Lag enrichments formed on swash faces at these high levels are former swash zones. In depositional packages representing pro-
protected from subsequent marine erosion and may be buried by gradation of the entire beach face, which may be 10 m thick or
later eolian deposits. more, the portion showing greatest concentration should be an
All three presented examples of the preservation of in-place upper interval representing the swash zone.
enrichments involve beach deposits. In-place enrichments proba- In map view, the locales of heavy mineral concentration
bly also occur in other types of shoreline environments, such as may be highly localized within a shoreline complex having low
tidal inlet deposits. These have been studied little. heavy-mineral content (McKellar, 1975; Force and others, 1982).
Most segments of most beaches are not efficient "concentrators."
Transported enrichments Particular portions of shoreline compartments may systematically
show the greatest enrichment. This topic is further discussed
Enrichments may be reworked from the environment where under exploration methods.
enrichment occurred and fed in bulk to a new environment where
they can better be preserved. The most important of these new Eolian deposits
environments is eolian. In eolian deposits, an enrichment is re-
moved grain by grain from the swash zone where enrichment Coastal eolian dunes contain significant portions of the
occurred to a new environment with greater stability and storage titanium-mineral resources in shoreline deposits of the United
capacity. Eolian deposits may acquire great volume. Transported States, South Africa, and Australia, that is, those districts where
enrichments may also occur in other environments, such as mechanized mining can be of low-grade material. Their great
washover fans. volume and homogeneity make some eolian sand bodies eco-
nomic even at low grades. Composite eolian dunes, such as
FACIES Fraser Island, Australia, and Trail Ridge, Florida-Georgia, can
contain several billion tons of sand, though only a fraction is ore
Beach deposits grade. Other eolian sands, such as those of Richards Bay, South
Africa, are high-grade deposits.
Many of the world's more important placer deposits of tita- The coastal dunes containing heavy-mineral resources are
nium minerals are true beach deposits, and every major district of both Pleistocene and Holocene in age. The dunes can be subdi-
shoreline titanium-mineral deposits includes some beach deposits. vided into three main types: foredunes, transgressive dunes, and
Many deposits (such as that shown in Fig. 65) are composites of stationary dunes. Table 17 gives examples of deposits of each
beach sands and overlying eolian sands, in which the beach-facies type.
deposits are typically higher in grade. Thus it would be difficult to Foredunes are those immediately adjacent to beach deposits.
calculate the relative magnitude of beach and eolian titanium- They form the backbone of many barrier islands and the ridges in
mineral resources. many accretionary beach-ridge complexes. Foredune deposits
E
Meters
Sgaward Ground surface
6
Sea Level
30 Meters Shoreface sands
Figure 64. Profiles of some progradational Australian shoreline placers. Heavy mineral concentrations
shaded. The scales differ but vertical exaggeration has been eliminated. A. Cudgen, after McKellar
(1975). B. Yoganup, after Baxter (1982). C. Eneabba, after Lissiman and Oxenford (1973).
may be interbedded with true beach deposits formed during ex- are characteristically parabolic, composed of sand transported up
treme high tides. Mining of some foredune deposits probably eolian ramps (commonly later eroded) on cliffed or other steep
occurs in virtually all beach-facies deposits. In some Holocene shorelines. The Jennings Eneabba heavy mineral deposit of
deposits of eastern Australia, separate mining of beach and fore- Western Australia apparently formed as a cliff-top dune.
dune facies was practiced when operations were smaller in scale. Stationary dunes are tied to bedrock features. They probably
Transgressive dunes are those that have become detached evolved from other dunes but in later stages of development
from beach deposits and have migrated inland. Deposits in trans- accumulated successive additions of transgressive dunes without
gressive dunes commonly overlie lagoonal or swamp deposits movement. Those of the southern Queensland coast in Australia
(Thorn and others, 1981; Force and Garnar, 1985; Force and are collages of successively accumulated large parabolic trans-
Rich, 1989). Transgressive dunes can be further subdivided into gressive dunes, separated in cross section by buried soil horizons
long wall and cliff-top dunes (Thorn and others, 1981; Short, (Thompson and Ward, 1975; Ward, 1977, 1978).
1987,1988). Longwall dunes remain parallel to the shore as they Migration of transgressive dunes preserves slip-face laminae,
migrate inland over topography of low relief. Many economic which thus characterize most of the dune interior (Bigarella and
heavy mineral deposits are of this type (Table 17). Cliff-top dunes others, 1969). Heavy-mineral-rich fine laminae outline the
78 E. R. Force
assemblages (at low concentrations) constitute the headlands,
high sea level activates the concentration of this valuable mineral
assemblage on beaches.
TIME SERIES OF DEPOSITION, AS RECORDED

ON FLIGHTS OF MARINE TERRACES
In some coastal compartments, a series of local high stands

of Quaternary sea level is represented by contemporaneous shore-
line deposits and related facies. These deposits are preserved on
marine terraces at elevations ranging from sea level to tens of
meters and inland as much as tens of kilometers from the present
shore. Flights of such terraces are well illustrated in the southeast-
ern United States, where as many as six shoreline deposits can be
recognized in South Carolina (Colquhoun, 1965; Thom and oth-
ers, 1972; McCartan and others, 1990) and Virginia (Oaks and
Figure 65. Annotated photograph showing eolian dune (d) overlying
regressional beach deposits (b), both of Pleistocene age, together formingothers, 1974; Force and Geraci, 1975). In some places, former
ore being dredged in Tomago area near Newcastle, New South Wales. barrier segments, capes, and tidal inlets can be recognized from
Nearby, 12,000-year-old freshwater peats intervene between the two the shapes of shoreline sand facies, and these features may closely
facies of deposits (Thorn and others, 1981). mimic present features of the same coastal compartment (as in
Fig. 66).
Each shoreline deposit of a set may contain heavy-mineral
enrichments (Baxter, 1977; Force and others, 1982). These de-
former slip faces in some economic deposits (Macdonald, 1983; posits typically differ in the extent of postdepositional weathering
Force and Garnar, 1985; Force and Rich, 1989). In other eolian (Chapter 6). Weathering differences make the older, more ele-
deposits, heavy-mineral distribution is more nearly disseminated. vated deposits of each coastal compartment potentially the more
valuable. Mineral assemblages may also vary, because of sea-level
SEDIMENT DISTRIBUTION controls of distribution patterns.
The supply and distribution of sediment within a coastal DEPOSITS NOW OFFSHORE
compartment can be an important control on the formation of
titanium-mineral deposits. Commonly, sediment of more than During most of the Quaternary, sea level has been lower
one type is supplied, and only one has the mineralogy and grain than it is now (Fig. 55). Any sea-level stand, especially a still-
size to form a valuable deposit. In economic deposits, sediment
distribution is such that most of the promising material is fed to a TABLE17. TITANIUM-MINERAL DEPOSITS IN DIFFERENT TYPES OF
"concentrator" such as the swash zone, whereas most of the rest QUATERNARY SHOREUNE SAND BODIES
of the material is deposited elsewhere. The mechanisms for ac-
complishing this are as varied as are source areas, conduits, and Example
depositional environments and defy orderly description. Some Type Holocene Pleistocene(?)
examples are given in descriptions of individual districts.
Changes in the distribution and type of sediment with Beach deposits Cudgen, East Australia Green Cove Springs,
Minninup, West Australia Florida; Jerusalem
changing sea level are particularly common. For example, drown-
Travancore coast, India Creek, East Australia;
ing of river mouths or individual headlands may activate different
Yoganup, Cape), and
distribution systems. Such changes are recorded as mineralogic
Eneabba, West Australia
differences in strandlines at different elevations within the same
coastal compartment. Cyclic sea-level changes may cause depos- Eolian deposits
its formed at different times but at the same sea level to have Foredune Cudgen, East Australia ?
comparable mineral assemblages. Transgressive Williamtown, East Australia Trail Ridge, Florida;
Headland erosion of friable sandstones is apparently an im- Richards Bay, South Africa Bridge Hill, East Australia;
portant factor in changes of sediment distribution in several dis- Jennings Eneabba, West
tricts described in this chapter. At high sea levels, the drowning of Australia
river mouths cuts off fluvial supply of immature mineral suites, Stationary Stradbroke, Moreton, and
and headland erosion becomes a larger component of mineral Fraser Islands, East
supply. Where friable sandstones containing valuable mineral Australia
stand, can leave behind a shoreline sand deposit, and most of
these should now be present on the continental shelf. By this
reasoning, the continental shelf should be favorable for shoreline
placer deposits. Several factors complicate this hypothesis (Atta-
nasi and others, 1987). For example, transgression commonly
destroys the upper portions of shoreline sand bodies (Swift,
1968); these are the portions containing heavy-mineral enrich-
ments. Constituent heavy-mineral grains may be transported
landward (possibly as transported enrichments) or may be incor-
porated in reconstituted sand bodies on the continental shelf. All
shoreline sands formed at sea levels lower than today's have
subsequently suffered transgression.
Another limitation of offshore sands is their weathering his-
tories; shoreline sand bodies now offshore formed at times of slow
weathering and have been submerged through most of their his-
tory. Thus they should be less weathered than their onshore coun-
terparts (Chapter 6).
Despite theoretical limitations, sand bodies on continental
shelves locally contain substantial quantities of valuable heavy
minerals (Grosz and others, 1986; Grosz, 1987). On the continen-
tal shelf of the eastern United States, surficial heavy-mineral en-
richments are numerous. Vibracore drilling shows that appre-
ciable volumes of heavy-mineral-bearing sands are present in
some areas.
Off a few coasts, shoreline sand bodies have survived trans-
gression and are partly buried on continental shelves, with origi-
nal morphology partially preserved (Flemming, 1981; Schluter,
1982). In these bodies, titanium oxide minerals should be present
approximately as in bodies above sea level. Off other coasts, sand
bodies exposed on continental shelves are Holocene in age and
formed in place (Duane and others, 1972). These bodies may be
derived from older shoreline sands that were redistributed by
transgression. darker than Holocene ones.
CEMENTATION AND WEATHERING
Weathering of titanium minerals in shoreline deposits is

extensively discussed in Chapter 6. Much of that discussion is contain large amounts of iron complexed with organic matter.
aimed at discrimination of weathering that occurred before and Aluminum is similarly complexed in some deposits (Ward and
after deposition of heavy-mineral grains. Cementation of shore- others, 1979).
line sands is in part a result of postdepositional weathering and is The cements generally form thin, tabular layers that repre-
of great economic importance. sent the positions of present and former water tables (Cannon,
The degree of cementation of permeable surficial sands of 1950; Welch and others, 1975; Ward and others, 1979; Baxter,
shoreline sand bodies varies with the age of a deposit and depth 1982; Thompson and Bowman, 1984; Force and Garnar, 1985).
within it. Holocene deposits are little cemented (except locally by Massive cements can apparently form by amalgamation of these
calcareous material); younger Pleistocene deposits, such as that thin layers (Force and Rich, 1989). Humate cement forms at the
shown in Figure 65, contain cements but are still friable. Older water table where humic acids, derived from vegetation at the
Pleistocene deposits, such as those of Eneabba and Trail Ridge, ground surface, are neutralized by groundwater. The iron com-
contain cements that can make a hammer ring. ponent of the cements is derived from acid leaching of sediments
The noncalcareous cements are apparently of two intergrad- above the water table, and iron also precipitates at the water table
ing types. At one extreme are pure organic cements, consisting of by neutralization (and perhaps oxidation).
amorphous black humin or humate. At the other extreme are The iron component of cements is related to ilmenite altera-
pure iron hydroxides. Variable amounts of clay matrix can be tion. In many deposits, ilmenite above the water table has been
incorporated in either type. Most cements are intermediate and intensely leached (Chapter 6). Iron in cements precipitated at the
80 E. R. Force
water table probably was leached in large part from ilmenite geologic relations of this dune indicate that it is of the transgres-
(Welch, 1964; Puffer and Cousminer, 1982). This hypothesis is sive type. Preserved bedding is that of slip faces dipping steeply
supported by the preservation of less altered phases such as mag- southwest.
netite within iron-cemented layers (New Jersey: Jim Stern, per- The ore sand at Trail Ridge has a convex exposed upper
sonal communication, 1974; Western Australia: Lissiman and surface but a subhorizontal base and overlies a 1.5-m lignitic peat
Oxenford, 1975; Baxter, 1986). Early-formed iron cements pre- unit in the area of current mining. This peat is of freshwater
vent further alteration of encased material. swamp origin (Rich, 1985). Probably the swamp was impounded
by the dune. Sand impurities of the peat record the approach of
M A J O R DEPOSITS the transgressive dune, which subsequently overrode the swamp
(Force and Rich, 1989).
Jacksonville district, Florida and Georgia The exposed upper part of the ore sand consists of several
subhorizontal humate-cemented layers, overlain by a bleached
The only titanium-mineral deposits currently mined in the zone (Fig. 47). The humate layers intersect slip-face bedding at a
United States are in northeastern Florida, near Jacksonville (Fig. high angle and are present in a zone up to 3 m thick. The
67A). The two active deposits, known as Trail Ridge and Green bleached zone, 2 to 3 m thick, is rootlet mottled and contains
Cove Springs, are part of the Jacksonville district, which includes heavily altered ilmenite. The humate layers apparently represent
many deposits in northeastern Florida and southeastern Georgia. precipitation by humic acid neutralization at the water table and
Mining was formerly conducted in the district near Folkston, former water tables.
Georgia, and Boulogne and Jacksonville, Florida. Unmined de- The bulk ore sand has a median grain size of about 0.3 mm,
posits of the district include the Cumberland Island and is well sorted (So 1.16 to 1.32 for the sand fraction alone, So 1.23
Altamaha Plantations deposits of Georgia and the Amelia Island to 1.35 if clay matrix is included), and is positively skewed.
and Yulee deposits of Florida. All these deposits are Pleistocene, Constituent grains are well rounded, and even prismatic minerals
related to former shorelines on at least three different marine show appreciable sphericity (Pirkle, 1975). Grain surfaces are
terraces (Fig. 67B). Modern beach deposits of the district were frosted, analogous to eolian sands elsewhere (Force and Rich,
also mined from 1916 to 1929. Altogether, the district has pro- 1989). Heavy minerals are slightly finer than light minerals (Fig.
duced about 5 x 106 metric tons of Ti0 2 and still contains 60C). The heavy-mineral assemblage includes altered ilmenite
resources of about 14 x 106 metric tons of Ti0 2 (Force and and leucoxene (about 50 percent [%]), zircon (15%), staurolite
Lynd, 1984). (15%), sillimanite (5 to 6%), tourmaline (5%), rutile (2 to 3%), and
The district is the southern end of the Atlantic coastal plain, kyanite (3%), with minor spinel, corundum, and monazite.
which to the north contains additional subeconomic deposits In detail, two distinct sand lithologies are present (Force and
(Force and Lynd, 1984). The district has little vertical relief and is Garnar, 1985; Force and Rich, 1989; Fig. 49). Slip faces are
mostly wooded. Surficial units are all late Cenozoic in age, but outlined by alternations of predominant light-colored sand and
exposure is extremely poor. Mining in the district is entirely by subordinate dark-colored sand. The dark-colored sand is far
dredge, so that geologic relations in the mines are obscured. As a richer in heavy minerals and is finer grained in both the light-
result, geologic information on the deposits is limited to drill mineral and heavy-mineral fractions. The dark laminae preferen-
cuttings and a few cores, with the exception of Trail Ridge. tially contain the denser heavy minerals of the deposit, as fine
Trail Ridge deposit. Sands of Trail Ridge are mined by grains, whereas staurolite, tourmaline, and sillimanite are parti-
DuPont at two dredging operations. Altered ilmenite and zircon tioned toward the light-colored layers as coarser grains (Fig. 68).
are the main products and dominate the heavy-mineral assem- No coeval beach or marine facies have yet been found asso-
blage. Heavy minerals average about 4 percent of fine- to ciated with Trail Ridge, although marine facies overlie Trail
medium-grained sands that form a surficial sand body up to 20 m Ridge sands toward the north (Pirkle and Czel, 1983). Analogy
thick. This sand body forms a ridge about 2 km wide and more with Australian transgressive dunes suggests that such coeval ma-
than 160 km long, extending well into Georgia. Only the south- rine facies were probably present within several kilometers to the
ern 29 km contains known ore-grade sands (Fig. 67A). The east. Figure 69 depicts the probable relation of shoreline, wind
elevation of the ridge crest is about 75 m toward the southern end direction, and dune orientations.
and about 45 m toward the north; the base of the sand shows a The age of the Trail Ridge deposit is still unknown. Micro-
similar slope (Fig. 67C). flora of the peat show it to be post-Miocene, while 14C dates
The Trail Ridge deposit has been studied by Spencer (1948), show that it is older than 45,000 years. Thus, Trail Ridge sands
Pirkle and others (1970, 1971, 1977), Pirkle and Yoho (1970), could be either Pliocene or Pleistocene.
Force and Garner (1985), and Force and Rich (1989). Force and Trail Ridge sands have been a focus of much of the work on
Rich include an extensive review of other previous work. weathering of ilmenite. Pirkle and Yoho (1970) and Temple
Temporary exposures (caused by lowering of the dredge- (1966) documented a great increase in leucoxene at the expense
pond surface) of ore in vertical profile have shown Trail Ridge to of ilmenite in the surficial leached zone, above the water table
be a great eolian dune (Force and Garnar, 1985). The form and (Fig. 47). This weathering is clearly postdepositional. Grey and
Figure 67. Titanium-mineral deposits of the Jacksonville district, Florida

and Georgia. A . M a p showing distribution of deposits (black) and line
of cross section. B. Cross section of the Trail Ridge and Green Cove
Springs deposits (shaded). C. Projection of deposit altitudes onto a
north-south vertical plane, showing location of marine fossils overlying
the flanks of the Trail Ridge dune. Data from Pirkle and others (1974)
for Green Cove Springs and from Pirkle and Y o h o (1970) and Pirkle
and others (1971, 1977) for Trail Ridge.
A'
METERS
75
/ \Trail Ridge
Green Cove - 50
Springs
25
-
0
10 20 30 40 Kilometers
METERS
T r
- 75
50
c 150 ' \ c
o o
Fossil
ra
> locality >D
<
o
w 100
25
50 30 Kilometers
Sea _1_ MSL

level 29°50' 30°00' 30°10' 30°20' 30°30' 30"40'
N o r t h latitude
82 E. R. Force
LIGHT BAND DARK BAND

I I I I
0.6% heavy 12.3% heavy

minerals minerals
(dark)
1- •
I I I 1
37% i l m e n i t e
+ zircon 78% i l m e n i t e +
50 -
(shaded) zircon
(shaded)
40 - -
30 - -
20 - -
10 - -
0 I 1 1
EXPLANATION
Sta uro I ite 13%
- - Sillimanite 4%
Ilmenite 55%
Figure 68. Size-frequency distributions comparing adjacent laminae rich and poor in heavy minerals at
Trail Ridge. The upper graphs show histograms of the whole sample, the middle graphs show histo-
grams of the heavy-mineral fraction, and the bottom graph shows cumulative curves for individual
heavy-mineral species.

Reid (1975) noted that ilmenite beneath the water table is also
altered and proposed that afirststage of postdepositional ilmenite
alteration occurred there. Sand morphology suggests another so-
lution: grains below the water table were partly weathered at the
time of deposition. Detrital oxides there show mixed alteration
states but uniform high polish, acquired during deposition. Thus,
this zone is a detrital mixture of variably altered minerals. Pre-
depositional weathering is also suggested by the mineralogy of the
eolian sand in the sandy peat facies underlying the ore, which
contains variably altered ilmenite much like that of ore (Force
and Rich, 1989). In the reducing peat environment, below the
water table,first-stagealteration by oxidation was effectively pre-
vented, so the mixed alteration assemblage represents the original
detrital assemblage. Weathering of Trail Ridge sands thus took
place in at least two stages, one before deposition and the other
afterward.
The intermediate sources of Trail Ridge sands have been
investigated by Pirkle (1975), who concluded that local upper
Cenozoic formations could have supplied the material. The
coarse grain size, mineralogy, shape, and surface features of Trail
Ridge sand grains greatly limit the possible intermediate sedimen-
tary sources, as well as immediate beach-facies sources.
Green Cove Springs Deposits. East of Trail Ridge and at
slightly lower elevations (Fig. 67A) are the deposits at Green
Cove Springs, now operated by Associated Minerals. These de-
posits formed in a progradational beach system. The most
complete published descriptions, by Pirkle and others (1971,
1974), are rather limited in scope and do not include descriptions
of subsequently discovered satellitic orebodies.
The main orebody is a progradational beach ridge complex
oriented SSE-NNW, and is about 15 km long and 1 km wide. At
the northern end, the elevation of the base of the ore is 98 ft
(30 m); apparent southward tilting is discussed below. The ore is
a surficial unit, normally about 6 m thick; it is a gray sand with
humate-cemented layers, overlying brown sand in which heavy-
mineral content decreases with depth. The thickness of the gray
and brown sands as a whole is 13 to 17 m. If the sand is a single
depositional sequence, it probably represents shoreface deposits
overlain by swash-zone deposits.
The ore sands are fine, with mean grain size of about 0.17
mm. Trask sorting (So) values calculated from Pirkle and others
(1971, 1974) are 1.3 to 1.4, almost as well sorted as Trail Ridge
sands. Rounding of all minerals is moderate, but markedly less so
than at Trail Ridge.
Heavy minerals constitute more than 3 percent of the sand
and include altered ilmenite and leucoxene (about 50 percent
[%]), zircon (about 15%), rutile (about 10%), staurolite and other
aluminosilicates (about 15%), tourmaline, monazite (each about
0.3%), and, locally, epidote and garnet. Ilmenite concentrates
contain 64 percent Ti02- Mean grain sizes of these minerals are
0.10 mm for ilmenite and zircon, 0.13 mm for staurolite, and
Ridge. A. Time of dune formation. B. Later time, with sea level taken 0.08 mm for monazite (from data in Garnar, 1980). Rutile cur-
from Fig. 67C. rently earns the most revenue (Fantel and others, 1986).
84 E. R. Force
Other deposits of the district. Titanium-mineral deposits analogous to present high-dune islands off the southern Queens-
on shorelines roughly correlative with those of Green Cove land coast of Australia.
Springs, at about 30 m elevation, formed at Boulogne (Pirkle and
others, 1971, 1974) and Folkston (Fig. 67A). Shoreline sands at Eastern Australia district
about 6 to 10 m elevation contain titanium-mineral deposits at
Yulee (Pirkle and others, 1984), Altamaha Planatations, and Rutile is present in coastal placer deposits of eastern Austra-
Jacksonville. Younger Pleistocene deposits formed on Amelia lia from south of Sydney (New South Wales) to north of Brisbane
and Cumberland Islands when sea level was only 2 m or less (Queensland), forming a single district more than 1,200 km long
above present sea level. (Fig. 70). This district has been the world's most valuable single
Source and evolution. The ultimate sources of the heavy titanium-mineral resource, because of its great aggregate size and
minerals concentrated in the deposits of the Jacksonville district a unique mineral assemblage dominated by rutile, zircon, and
are predominantly the high-grade metamorphic rocks of the ilmenite.1 Production of rutile began in the 1940s near the New
southern Blue Ridge and inner Piedmont (Force, 1976b). Sedi- South Wales-Queensland border wherefirstgold and then zircon
mentary hosts were intermediate sources for Trail Ridge (Pirkle, were already mined from heavy-mineral concentrations of very
1975) and probably for the other deposits of the district. Since the high grade (Morley, 1981). Rutile production gradually ex-
Piedmont-draining streams in the region contain a mineral suite panded southward to the Hunter River area and northward to the
far too immature (Neiheisel, 1976; Force and others, 1982) to high-dune islands of southern Queensland.
have supplied the placers of the district without extensive predep- The district has been well described in terms of constituent
ositional weathering beneficiation, it is probable that this mineral deposits, geologic relations among host sand bodies, and
beneficiation occurred in the intermediate sedimentary hosts. processes of deposit formation. Individual deposits have been
The immediate source of eolian sand supplied to Trail Ridge described and mapped by Gardner (1955), Connah (1961), and
is still a problem. This source was both heavy-mineral rich and Winward and Nicholson (1974). A summary by McKellar
coarse grained. It seems likely that the dune was supplied by a (1975) is the most recent for the whole district. Geology of Qua-
beach to the east, but the Green Cove Springs beach deposit, the ternary coastal sand bodies has been discussed and mapped by
location of which seems appropriate, is heavy-mineral rich but Roy (1982) and Thorn (1983). Development of the bodies, in-
toofinegrained. cluding offshore extensions, in terms of sea-level history and sed-
A suggestion by Pirkle and others (1974) that the base of the iment supply has been summarized by Roy and Thom (1981).
main Green Cove Springs orebody is tilted to the south is provoc- In detail, the district is diverse, as it includes both Pleistocene
ative, because the base of a progradational beach-ridge sequence and Holocene deposits, each including both beach and eolian
records a horizontal surface at the time of deposition. A total of components. Commonly, more than one of the four resulting
six boreholes show uniform southward tilt of about 0.5 m/km in deposit types are mined in a single face (Fig. 65). Additional
the northern three-quarters of the main orebody (Fig. 67C). Mi- diversity is imposed on the district by being plastered against a
chael Shepherd of Associated Minerals (written communication, geologically complex continental margin that includes three large
1987) warns that the base of the southern end of the orebody sedimentary basins and three Paleozoic foldbelts (Figs. 70, 71).
drops southward in discrete steps. Considering that this is an embayed shoreline with hundreds of
If one assumes that the Green Cove Springs body can be bedrock headlands, it is remarkable that a single general descrip-
used as a tiltmeter, and that the Trail Ridge body is older but on tion can apply to the entire district.
the same crustal block, then the Trail Ridge deposit underwent General description and petrology. The Quaternary
the same southward tilting. At present the base of the Trail Ridge coastal deposits form discontinuous narrow coastal plains in
sand plunges northward at about 0.3 m/km, but before tilting, discrete bays outlined by headlands, between the Tasman Sea and
this land surface would have sloped about 0.8 m/km to the north bedrock hills rising behind the beaches. Sands exploited in the
(Fig. 67C). Eolian sand bodies like Trail Ridge commonly are district are most commonly medium- andfine-grainedsands,
deposited by climbing up sloping surfaces. Opdike and others quite well sorted and rounded. Median grain sizes are about 0.11
(1984) based their uplift history of northern Florida on an as- to 0.13 mm (Beasley, 1950). Quartz is the predominant light
sumption that Trail Ridge formed as a shoreline sand with a mineral; carbonate grains are commonly absent, and feldspar is
horizontal base. Instead it appears that the Green Cove Springs exceedingly minor. Most of the shoreline sands of the district
body is of shoreline origin and formed horizontal, but that Trail have very low heavy-mineral contents, less than similar sands
Ridge is eolian and never was horizontal. Both have subsequently from the east coast of the United States. Whitworth (1959) and
been tilted down to the south. This may explain marine fossils McKellar (1975) estimate the ambient heavy-mineral content to
overlying Trail Ridge in Georgia (Pirkle and Czel, 1983); the
northern end of the dune ridge at the time the Green Cove
Springs beaches were deposited could have had a drowned base 1
Ilmenite had not been recovered in this district until recently because of the
(Fig. 67C, 69). Trail Ridge at this time could thus have been difficulty in separating it from minor chromite.
be well under 0.1 percent. However, the mineral assemblage is so
valuable that total heavy-mineral grades of less than 1 percent
currently constitute ore.
In high-grade beds, rutile, zircon, and ilmenite make up
greater than 90 percent of the heavy-mineral fraction throughout
much of the district. Minor but characteristic members of the
suite are tourmaline, monazite, chromite, and garnet. Cassiterite,
epidote, magnetite, spinel, pyriboles, and metamorphic alumino-
silicates are locally present. Col well (1982a) noted that where
heavy minerals are present at low concentration, the lighter heavy
minerals form a larger portion of the suite. Beasley (1950) noted
local increases of coarse noneconomic heavy minerals such as
garnet around specific headlands, showing a contribution of head-
land erosion.
Figure 71 shows trends in rutile contents of heavy-mineral
concentrates as a function of latitude through the entire district.
Rutile distribution is fairly homogeneous but shows appreciable
influence of differing adjacent geologic provinces; rutile contents
are lower adjacent to the two southern foldbelts. Deposit age
apparently is a less important control on rutile content; Hails
(1969) reported little difference in mineral assemblage between
Pleistocene and Holocene deposits. The limits of the district coin-
cide not with lesser heavy-mineral concentrations but with
changed assemblages; magnetite, ilmenite, pyriboles, epidote, and
tourmaline suddenly become predominant to the south (Hails,
1969), whereas ilmenite gradually becomes predominant to the
north (Connah, 1961). Hails (1969) reported abnormally high
concentrations of andalusite, pyriboles, staurolite, and epidote in
heavy-mineral assemblages in the central region of the district,
adjacent to the New England foldbelt. My own coastal collec-
tions (included in Fig. 71) support his observation.
Electron microscope images of grain surfaces (Fig. 72) show
that a given deposit is typically mixed in character. Quartz grains
vary from angular to well rounded and polished. The economic
heavy-mineral assemblage matches the more mature of these
quartz populations in morphology.
Deposit types. Four types of deposits are exploited in the
district; these are Holocene beach deposits, Holocene eolian de-
posits, Pleistocene beach deposits, and Pleistocene eolian
deposits. Holocene beach deposits were recognizedfirstand were
the focus of most early mining; in fact, mining of the rich deposits
of the immediately previous storm was an important component
of this activity. Enrichment on beach faces is intensified by ero-
sion because of seasonal changes in beach profile and reorienta-
tion of beaches (Beasley, 1948; McKellar, 1975). Thus,
individual enriched beds lie on unconformities. Preservation of
the enrichments occurs if they are protected from further erosion
on the beach face (1) in those portions above normal high tide,
(2) by rapid burial (progradation), or (3) by removal into the
intruded foldbelts, narrow shelf, and type and location of deposits. Un- eolian environment. Individual layers of concentrated heavy min-
derlined deposits in production. erals are (or were) as thick as 2 m, tapering and dipping seaward.
In aggregate, these seams are en echelon, parts of progradational
sequences with base level at about mean sea level (Fig. 64A).
Beasley (1948) showed that little heavy-mineral enrichment in
86 E. R. Force
S o u t h latitude
Figure 71. Variation in rutile component of heavy-mineral concentrates as a function of latitude through
the eastern Australia district. Boundaries of geologic provinces at the coast are also shown. Values from
Beasley (1950), Gardner (1955), Hails (1969, Fig. 6), McKellar (1975), and my own work. Duplicate
samples at same location are averaged for curve. Data of Colwell (1982a) are not comparable and were
not used.
Figure 72. Scanning electron microscope photographs of grain morphologies of quartz and rutile from
deposits in eastern Australia. A. Holocene beach sand deposit, Tuncurry. Two rutile grains on left, two
quartz on right. B. Pleistocene eolian deposit, Tomago. One rutile grain above, three quartz below.
Photographs by John Evans, U.S. Geological Survey.
Figure 73. Holocene eolian dunes in eastern Australia. A. Foredunes in the Jerusalem Creek area,
looking north. Note the iron- and humate-cemented beachrock of Pleistocene age exposed in the
beachface. B. Transgressive "longwall" dunes in the Williamtown area, looking east. The foreground is
occupied by 4,300-year-old Holocene estuarine deposits (Thorn and others, 1981).
these sequences extends below sea level. Supply of heavy miner- the eolian deposits contain disseminated heavy minerals at low
als to the beaches is thought to be cannibalistic in part, as mining grades.
has depleted the total heavy-mineral stock of some coastal com- Pleistocene eolian dunes represent a large share of the re-
partments, except where older heavy-mineral deposits are ex- maining resources in the district and are well represented in the
posed in the beach face (Michael Shepherd, oral communication, Newcastle area (Fig. 65) and as the great high-dune islands
1986). Storm activity thus produces little placer concentration at (Stradbroke, Moreton, and Fraser) of southern Queensland. Like
present. Pleistocene beach deposits, these dunes are cemented by humate
Pleistocene beach deposits differ little from their Holocene and iron-aluminum hydroxides, down to the levels of present and
counterparts. They are also preserved as regressional beach plains former water tables (Thompson and Bowman, 1984). The
containing seaward-dipping enrichments but are topographically Queensland islands are collages of dunes of various ages. Several
more subdued (Thom and others, 1981). The sea levels repre- stages of Pleistocene accretion of parabolic transgressive dunes
sented by Pleistocene deposits are 0 to 5 m higher than those are represented (Thompson and Ward, 1975; Stephens, 1982a).
represented by Holocene deposits (Thom and others, 1981; Mel- Individual dunes may have complex histories, shown by buried
ville, 1984, Fig. 3). The mineral assemblages of Pleistocene and soil horizons. As the eolian sand extends well below sea level
Holocene deposits are similar, probably because the mineral as- offshore from these islands (Kudrass, 1982), and as the dune
semblages were already mature before deposition in most parts of formation in part occurred at times of low sea level, the islands
the district and were little affected by further weathering. The are thought to represent transgressive dune complexes that were
main lithologic difference between Pleistocene and Holocene later drowned (McKellar, 1975; Ward, 1977, 1978).
deposits is the cementation of the former by iron-aluminum hy- Differentiation among these four types of deposits is locally
droxides and humate (Fig. 73). The age of known Pleistocene possible based on petrography. In addition to the presence of iron
deposits is about 140,000 years (last interglacial; Roy, 1982). oxide and humate cements, Pleistocene deposits have ilmenite
Exploitation of Pleistocene beach deposits at Jerusalem Creek fractions that are more uniformly altered than those of Holocene
(McAuleys Lead) was among the earliest in the district; this de- deposits. Ilmenite fractions of Holocene deposits are mixtures of
posit is along an unconformity between two sets of Pleistocene slightly altered material and subordinate fresh ilmenite locally
regressional beach plains. Pleistocene beach deposits are currently intergrown with other iron-titanium oxides. The eolian deposits
receiving a large share of the mining and exploration attention in contain some grains having eolian surface features such as frost-
the district. ing, but the variations among these grains are great.
Holocene eolian dunes, both as foredunes contiguous with Depositional setting. The embayed eastern coastline of
beach deposits and as detached transgressive dunes moving in- Australia is wave dominated and faces a continental shelf less
land (Fig. 73), commonly rest on Holocene beach deposits and than 60 km wide. Predominant wave direction sets up a strong
are mined with them. A sizable resource in Holocene dunes is longshore drift to the north, but some headlands project offshore
present in the Bridge Hill dune of the Myall Lakes area. The into the south-flowing Eastern Australia Current. Sand may be
transgressive dunes may form shore-parallel ("longwall") features stored offshore south of such headlands (Kudrass, 1982; Field
overlying estuarine deposits. In contrast to beach concentrations, and Roy, 1985), or it may bypass a headland as eolian transgres-
88 E. R. Force
sive dunes. Present sea level has drowned therivermouths, and England foldbelt coincides with low rutile values, punctuated by
little immature fluvial sediment currently reaches the continental high values only near headlands of Mesozoic sandstone. This
shelf (Roy, 1977; Roy and Crawford, 1977). mineralogic tie to an intermediate source is a powerful tool in
Quaternary geologic setting. The Holocene sands mined unraveling the developmental history of the deposits.
along the coast of eastern Australia are the shoreward portion of a The ultimate source of the heavy minerals has long been
sand sheet occupying the inner shelf. Throughout this sheet are debated. The most conveniently located possible sources are the
found similar mineral assemblages, which are less sensitive to New England and related Paleozoic foldbelts and their associated
positions of (drowned) river mouths (Kudrass, 1982) than to granites. Beasley (1950) recorded two rutile occurrences in the
adjacent geologic provinces (Fig. 71). The sand bodies contain a New England foldbelt and found rutile averaging less than 1
less mature assemblage in proportion to present depth either of percent of heavy-mineral assemblages in streams draining the
water or burial (Jones and Davies, 1979; Col well, 1982b; Kud- foldbelt. He and Gardner (1955) concluded that the New Eng-
rass, 1982; Reich and others, 1982; Stephens, 1982b). Thus, land foldbelt was the source of the rutile supplied to the adjacent
sands deposited at lower sea levels were less mature. Holocene younger basins. Gardner's conclusion, however, invoked an alter-
sea levels have remained virtually constant for 6,500 years along ation of the abundant ilmenite to coarse single-crystal rutile, a
this coast, permitting extensive reworking and progradation of process not yet observed in nature.
shoreline deposits (Thom and Roy, 1985). Whitworth (1959) showed that New England zircon and
Pleistocene sediments on the continental shelf underlie the monazite were unsuitable source materials of the Quaternary
Holocene sands and form near-surface units on the midshelf coastal deposits, because the morphologies and compositions of
(Jones and Davies, 1979; Schluter, 1982). These include several these minerals differ in the two areas. Layton (1966) pointed out
shoreline sand bodies. Petrography of the Pleistocene sands is a that cassiterite, a stable mineral supplied in the New England
sensitive function of sea level, even more than for Holocene foldbelt, is lacking in most of the Quaternary coastal deposits.
sands; the submerged Pleistocene bodies have immature heavy- Whitworth was able to find very little rutile in stream sediments
mineral assemblages (Reich and others, 1982), comparable to of the New England foldbelt; my own conclusions, based on
those of relatedfluvialsands, whereas Pleistocene shoreline sands stream sampling there, are emphatically in agreement with those
exposed above sea level have mature mineral assemblages com- of Whitworth.
parable to those of exposed Holocene sands (Hails, 1969). If the New England rocks were not the ultimate source,
Interglacial periods (including the Holocene) and their high what rocks were? Whitworth (1959) and Layton (1966) pointed
sea levels are represented by shoreline sands having mature min- out the suitability of some metamorphic rocks in the interior
eral assemblages in present coastal regions; glacial periods and Australian craton. I would like to add that in Mesozoic time,
their lower sea levels are represented there by immature fluvial Australia was still adjacent to other portions of the former
deposits. Near some present estuaries, immature Pleistocene flu- Gondwanaland. Paleocurrent measurements from the Hawkes-
vial deposits can be found stratigraphically between mature bury Sandstone (Rust and Jones, 1987) show transport to the
Pleistocene and Holocene shoreline sands (Roy, 1982). northeast and make continents since removed just as suitable a
Source rocks. Sources of minerals can be categorized as source as the Australian craton (cf., Galloway, 1972).
immediate, intermediate, and ultimate. The immediate derivation History of development.
The present sediment budget for
of beach and eolian deposits of eastern Australia is clearly from rutile- and zircon-dominated shoreline deposits apparently con-
the cannibalism of older beach deposits, from offshore sands, and tains a key to their development. These deposits are fed mainly by
from sands being carried northward by longshore drift. cannibalism of older but similar coastal deposits and by longshore
Authors also seem agreed on the intermediate sources—the drift from the south. The juvenile component in the mineralogy of
Mesozoic sandstones of the Sydney, Clarence, and Moretón ba- this material must be small, as the supply by rivers is presently
sins (Fig. 71), which have heavy-mineral assemblages dominated minor. The main source of the mere trickle of juvenile material
by rutile, zircon, ilmenite, and tourmaline, with minor chromite must be sea-cliff and headland erosion, and this type of supply is
and tourmaline (Beasley, 1950; Gardner, 1955; McElroy, 1962; far more productive of sand in the Sydney and Clarence-Moreton
Galloway, 1972; Win ward and Nicholson, 1974; McKellar, basins than in the New England and other foldbelts. Sea-cliff
1975; Davidson, 1982). Thus, the unique mineral assemblage of erosion of friable sandstones is characteristic of Mesozoic rocks of
the shoreline deposits finds a match in an extremely unusual the basins. Thus the supply at present is small but is largely from
assemblage in Mesozoic sandstones.2 The Hawkesbury Sandstone sedimentary basins that supply almost exclusively rutile, zircon,
of the Sydney basin was probably the most important single ilmenite, tourmaline, and quartz to the system. Locally, headland
source. The abrupt southern boundary of the district coincides erosion of other rocks dilutes the assemblage (Fig. 71).
with the southern end of the Sydney basin (Fig. 71). The New Two lines of evidence suggest that the present sediment
budget is not applicable during times of lower sea level.
2
Ilmenite in shoreline deposits is likely to have a separate source in part, as
(1) Morphology and stratigraphy of river mouths suggest that at
some of it is too little weathered to have been cycled through Mesozoic lower sea level, rivers become active suppliers of the system.
sandstones.
(2) The high epidote and pyribole content of offshore sands of
both fluvial and shoreline origins is incompatible with derivation
primarily from sandstones of the Sydney and Clarence-Moreton
basins, suggesting that immaturefluvialdebris has swamped those
sources. Sands offshore from the New England foldbelt are the
debris of low-grade metamorphic and granitoid terranes, whereas
those off the Sydney basin show the influence of basaltic intru-
sions3 in the interior of the basin and of the adjacent Lachlan
foldbelt (Hudson, 1986).
Lowered sea levels apparently activate a new distribution
system, in which fluvial debris swamps debris from sea-cliff ero-
sion (cf., Schluter, 1982). This sea-level "trigger" is undoubtedly
modified, indeed intensified, by the coupled weathering-sea level
factor described in Chapter 6. Sands now offshore that were
formed in subaerial environments were subject to less maturation
by weathering because temperatures were colder and atmo-
spheres less reactive. In addition, the duration of in-place
weathering was less because the sands were seldom exposed.
The changes in mineral distribution with sea level appear to
explain (1) the great maturity and unique assemblage of all the
interglacial-age shoreline sands, (2) the modified heavy-mineral
assemblage of these sands adjacent to the New England foldbelt,
and (3) the great contrast in both provenance and maturity of the
sands offshore. Note that the mechanism proposed by Colwell
(1982b), shoreward winnowing of offshore heavy minerals, is
unnecessary and does not explain all the mineralogic data. The
minerals onshore and offshore are related to each other not in
terms of mechanical stability and density but in terms of differing
weathering stability and provenance.
Geographe Bay (Bunbury-Capel) district,

Western Australia
This district has been an important producer of ilmenite
since 1956. Constituent deposits have been well described
(Welch and others, 1975; Baxter, 1977), but the geologic context
needs further study. Geologic maps of the area are by Lowrie and
others (1967, 1983). The deposits lie as far north as Waroona
(Fig. 74), but mining is centered on the Bunbury-Capel area. The
district forms the southern end of the Swan coastal plain, facing
the Indian Ocean. The area is farmed except for extensive forests
of jarrah and other native trees. Relief is low, and rock exposure
is poor, in part because of extensive eolian sand cover.
The heavy-mineral deposits are found in three former shore-
line complexes ranging in age from early Pleistocene or Pliocene
(Collins and others, 1986) to Holocene (Collins and Hamilton,
1986); each complex represents a progradational swarm of shore-
lines at progressively lower elevations. From oldest to youngest,
these are the Yoganup complex, representing six individual shore-
lines at 66 to 26 m elevation; the Capel complex, with up to 10
shoreline deposits at 4 to 6 m elevation; and the Minninup com-
Figure 74. M a p of western Australia coast showing the Geographe Bay
and Eneabba districts, the Swan coastal plain, and various "scarps" and
3
My fluvial samples from the Sydney basin show a large contribution from crustal blocks. The Minninup shorelines, not shown, coincide at this
basaltic rocks, probably the numerous diatremes of Crawford and others (1980). scale with the modern shoreline.
90 E. R. Force
with sedimentary and minor volcanic rocks between the Leeuwin
and southwestern Yilgarn crustal blocks (Fig. 74). Both blocks
are composed of high-grade metamorphic and granitoid rocks.
The two older shorelines were cut into sandstone of the Leeder-
ville Formation, which itself has a high heavy-mineral content in
some places (Baxter, 1977, p. 47; Collins and Baxter, 1984). Its
mineral assemblage is not recorded.
The modern shore of Geographe Bay is microtidal with
sporadically high wave energy (Semeniuk and Johnson, 1982;
Collins and Hamilton, 1986). The direction of longshore trans-
port reverses seasonally. A typical succession of near-shore and
beach depositional environments is present (Semeniuk and John-
son, 1982), with heavy-mineral concentration most efficient in
the swash zone of the beach face (Hocking and others, 1982;
Collins and Hamilton, 1986). Heavy minerals are transported by
Figure 75. Waroona deposits at the foot of Darling scarp, western Aus- eolian action to transverse foredunes, where their concentration is
tralia. Here a former sea cliff coincided with an important structural less but their preservation potential greater (Baxter, 1977; Hock-
feature bounding the Swan coastal plain. Portions of the deposit are on ing and others, 1982; Collins and Hamilton, 1986). The stratig-
granitoid bedrock. The plain in the foreground is locally punctuated by
raphy of the younger Holocene deposits shows beach facies
former sea stacks.
preserved by progradation (Semeniuk and Johnson, 1982).
Holocene sea-level history shows a general stillstand for
about the past 7,000 years, locally modified by tectonic move-
plex, at or slightly above modern sea level. These shorelines ments (Semeniuk and Searle, 1986; see also Welch and others,
outline successive paleo-Geographe Bays. 1975, p. 1074). Progradation during this stillstand was preceded
The Yoganup and Capel complexes are cut into Lower by transgressional landward displacement of shoreline sand
Cretaceous clastic sedimentary rocks of the Leederville Forma- bodies (Hocking and others, 1982).
tion. A former sea cliff of the Yoganup system is exposed at the Deposit facies and stratigraphy.
Progradational stratig-
surface as the Whicher scarp and in the Waroona area coincides raphy of economic deposits is shown by Collins and Baxter
with the Darling structural scarp (Figs. 74, 75). (1984) and Collins and Hamilton (1986). Lithologies such as
General description. All the heavy-mineral deposits of the planar-laminated,fine-grained,well-sorted sand rich in heavy
district are dominated by ilmenite. As described by Baxter minerals are interpreted as high-energy beach deposits.
(1977), the deposits in the three shoreline complexes have other Progradation of beach deposits was apparently amplified by
features in common also. They consist of moderately to poorly local fall of sea level, as progressively younger deposits formed on
rounded grains in poorly sorted to bimodal progradational depos- progressively lower wave-cut benches (Fig. 64B). Eolian sands
its. These are locally conglomeratic toward the base, with a ma- having lesser heavy-mineral contents commonly overlie the beach
trix of heavy-mineral-rich clayey sand. Overlying the conglom- sands and constitute the majority of some deposits (Welch and
erate is heavy-mineral-rich clayey sand, overlain in turn by others, 1975).
fine-grained sediment. Interbedded local fluvial deposits are poor Alteration and cementation.
Alteration of the deposits of
in heavy minerals. The surface deposits are transgressive eolian the Geographe Bay district has been documented as a function of
dune sands related in age to the next younger shoreline. elevation and thus of age (Fig. 52). The younger sands contain
Heavy-mineral contents of greater than 10 percent are quite calcareous shell fragments, fresh feldspar, an immature heavy-
common in the deposits of this district. Ilmenite ranges from 56 to mineral suite, and fresh ilmenite. The older sands are noncalcare-
95 percent of the heavy minerals, zircon 2 to 18 percent, and ous and contain a clay matrix partly from feldspar alteration and
rutile 0.5 to 2 percent. Other heavy minerals include leucoxene, partly from infiltration by younger clays (Baxter, 1982). Heavy-
monazite, and kyanite; garnet, magnetite, pyriboles, and epidote mineral suites of the older sands are restricted to the more stable
are common in the younger deposits (Carroll, 1939), which con- species, and ilmenite is altered, commonly concentrically, to con-
tain calcareous shell fragments and fresh feldspar in the light- tain about 60 percent Ti0 2 (Welch, 1964; Baxter, 1977, 1986;
mineral fraction. Frost and others, 1983).
Differences among the shorelines in degree of weathering In the older deposits, degree of weathering is a function of
beneficiation are pronounced. The older shorelines have weath- position in the vertical weathering profile; the most altered as-
ered mineral assemblages, including ilmenites having high Ti02 semblages are above less altered assemblages (Fig. 48). This
contents (Fig. 52). weathering is clearly postdepositional (Baxter, 1982). So-called
Geologic and depositional setting. Geographe Bay and its coffee rock, that is, sand cemented by iron hydroxides, is found
forerunners are developed in the Bunbury Trough, a graben filled along present and former water tables (Welch and others, 1975;
Baxter, 1982). The iron in this cement has been leached from
overlying ilmenite grains, according to Welch (1964), who also
suggested that iron cementation is triggered by the leaching of all
pH-buffering calcareous material. Baxter (1986) showed that iron
cementation arrests the alteration of encased ilmenite at an early
stage.
Source. All authors seem agreed that the immediate source
of ilmenite was the Leederville Formation; sea-cliff erosion of
that heavy-mineral-rich unit must have provided abundant heavy
minerals to beaches below. Southwestern Australia provides sev-
eral possibilities for ultimate sources of this ilmenite. Baxter
(1982) is inclined toward the southwestern Yilgarn block (Fig.
74) as the major source, whereas Carroll (1939) and Welch and
others (1975) believe the Leeuwin block to be more important.
Both areas contain high-grade metamorphic rocks (Wilson, 1964,
1969; Peers, 1975a). Figure 76. Undercut portion of former sea cliff cut in the Yarragadee
Formation, exhumed in mining of the zircon-rich 115-m shoreline,
Eneabba district, Western Australia Eneabba, Western Australia.
North of Perth is the Eneabba district (Fig. 74), the most

important remaining identified resource of titanium oxide miner-
als in Australia except for the newly discovered Waroona depos- Ambient winds are from the southwest, but monsoonal winds are
its in Victoria. Related deposits south of the Eneabba district are from the northwest, producing variable longshore drift directions.
near Jurien, Coojarloo, and Gingin. The Eneabba district was not Discussions of paleogeography follows.
discovered until 1970. It is valuable because of its great size, high The mineral assemblage of all the deposits is dominated by
heavy-mineral grades, and a rutile-rich mineral assemblage. Total ilmenite, rutile, and zircon, with minor monazite, tourmaline,
heavy mineral contents of greater than 10 percent are common, kyanite, staurolite, and pyribole. Average heavy-mineral grain
and heavy-mineral resources are greater than 30 million metric size ranges from 0.15 to 0.18 mm.
tons. The district has been described by Lissiman and Oxenford Fades. The upper terrace deposits, from 128 m down to
(1973, 1975), Baxter (1977), and Shepherd (1990); except for 100 m, represent beach deposition on benches cut in the Yarra-
attributed statements, these works are the basis for the following gadee Formation. Locally, the undercut surface of a sea cliff is
description. Geologic maps of the district are by Baxter (1972) exhumed by mining (Fig. 76). These deposits are high grade, and
and Lowrie and others (1973). Further study, especially sedimen- concentrates are rich in zircon (36 to 61 percent) and locally
tological, is needed. monazite, at the expense of ilmenite (28 to 46 percent) and rutile
The Eneabba district is toward the northern end of the Swan (5 to 8 percent). Minerals are well rounded and moderately
coastal plain (Fig. 74). The greatest local relief is on the Gingin sorted. Mean quartz grain size is 0.3 mm, and that for the bulk
scarp, an aggregate of former sea cliffs cut into the Yarragadee heavy-mineral fraction is 0.8 4>finer.Skewness of both light and
Formation of Jurassic age. An eolian sand cover over thick later- heavy fractions is low. Deposition is apparently under conditions
ite has produced poor exposures and a smoothed topography. of high wave energy.
Intensive farming to the south along the coast gives way within Deposits on lower terraces show several important differ-
the district to sparsely settled dry scrubland to the north and east. ences. The heavy-mineral assemblage is dominated by ilmenite
The major resource of the district consists of more than (53 to 68 percent) and rutile (8 to 11 percent), at the expense of
seven high-grade layers, each 1 or more meters thick. These zircon (15 to 23 percent). Lighter heavy minerals (staurolite,
layers were deposited with other clastics on wave-cut platforms kyanite, tourmaline) are more common. Mean grain sizes of
forming subparallel steps from 130 m down to 29 m (Fig. 64C). heavy minerals are similar to those on higher shorelines, and
These sediments, thought to be early or pre-Quaternary in age, heavy minerals are well rounded and sorted. Average sizes of
are called the Yoganup Formation by Baxter (1982); that is, he heavy minerals are related to each other as would be predicted
correlates them with the Yoganup shoreline sands to the south. from their densities. Light minerals, however, are reported to be
The sands are locally cemented by iron hydroxides and a clay finer grained than on the higher shorelines. Heavy minerals show
matrix. Overlying and locally interbedded with these deposits are appreciable positive skewness, whereas light minerals show
lower-grade eolian deposits. Shapes of eolian landforms suggest strongly negative values. These data could be explained by the
paleowinds from the southwest. presence of two grain populations.
The modern Indian Ocean coastline of the district is carbon- I was able to examine the 91- and 94-m shoreline deposits in
ate dominated and normally of low to moderate wave energy. a freshly excavated trench. Individual high-grade beds about 5 cm
92 E R.
thick consist of rounded heavy minerals and quartz, supported
partly by interstitial clay (Fig. 77). These beds show herringbone
cross-lamination (Fig. 78), indicative of a low-energy intertidal
environment. Separating the high-grade beds are interbeds of
clayey sand virtually barren of heavy minerals. There is a sugges-
tion of distortion of sedimentary structures by dissolution of car-
bonate shell fragments, but fresh feldspar is present. Such deposits
probably formed on beaches of a coast with high tides but low
wave energy. The two grain populations are a coarser rounded
population generated by wave energy, probably higher on the
beach and derived in part by sea-cliff erosion, and afinerpopula-
tion representing calmer intertidal conditions.
Overlying the shoreline deposits of the Eneabba district are
(1) clayey estuarine sands, (2) deposits of alluvial fans off the
Gingin scarp, and (3) several generations of eolian sands plastered
across the Gingin scarp. The eolian sands are lower in heavy-
mineral grade but are commonly mined with underlying shore-
line deposits (Fig. 79); locally they form an independent resource.
Lighter heavy minerals such as kyanite are relatively more
abundant in the eolian sands. Under the eolian sands, a pisolitic
laterite surface is commonly developed.
Alteration and cementation. The Eneabba shoreline
deposits were intensely weathered, as shown by the lateritic
surface developed on them. They now contain no carbonate, and
some original feldspar has gone to granular clay lumps. Heavy
mineral assemblages are mature, and ilmenite is altered to contain
62 percent TÍO2. Frost and others (1983) show that pseudorutile
is the major constituent of this altered ilmenite.
The shoreline deposits are locally cemented by iron hydrox- Figure 78. Herringbone cross-lamination in deposits of the 91-m shore-
ide, to form hard rock known locally as "coffee rock" that shows line, Eneabba. Lens cap 48 m m in diameter.
structures suggestive of nucleation around tree rootlets and/or
marine bioturbation (Fig. 80). It probably represents precipita-
tion, at former water tables, of iron derived in part from leaching
of ilmenite in overlying sands. Ilmenite encased in coffee rock is 1975). Cementation apparently started early, as detrital fragments
less altered than that in permeable sand (Lissiman and Oxenford, of coffee rock are incorporated in the youngest Eneabba shoreline
deposits.
Source. Jurassic rocks of the Cockleshell Gully, Yarra-
gadee, and South Perth formations are important intermediate
sedimentary sources of the titanium minerals of the Eneabba
district. Local high heavy-mineral contents are reported from the
Yarragadee and South Perth formations; friable heavy-mineral-
rich sandstones of these formations may themselves constitute
low-grade resources. My collection of Yarragadee sandstones
from throughout the area suggests that the ratio of zircon to
ilmenite and rutile varies considerably, from about 0.06 to 0.15.
The low value occurs in a micaceous sandstone with relatively
high contents of the lighter heavies staurolite, kyanite, tourmaline,
epidote, and pyroxene.
The ultimate sources of the minerals of the Eneabba district
are the Northhampton block to the north and the southwestern
Yilgarn block to the south. Both contain abundant high-grade
Figure 77. Photomicrograph of impregnated ore from the 91-m shore-
line, Eneabba. Note dark clay matrix. Detrital minerals are ilmenite, metamorphic rocks (Wilson, 1969; Peers, 1975b). A clue to ulti-
rutile (both dark), zircon, and quartz. Transmitted plane light, 2-mm mate sources is provided by rutilated blue-gray quartz in both the
field. Eneabba deposits and in the Yarragadee Formation; this type of
similar deposits of the Indian subcontinent are those in Orissa and
northeastern Sri Lanka.
Two main deposits form the district, the larger deposit north
of Quilon, on the Kayankulam-Needakara (K-N) bar of the Ker-
ala coast, and the other at Manavalakurichi (MK) in Tamil Nadu
(formerly Madras), about 100 km to the south, virtually at the
southern tip of India. The borders of the district are exceedingly
diffuse; many smaller coastal deposits are known, but in those
more distant, ilmenite concentrates have low Ti0 2 contents
(Jacob, 1956; Sinha, 1967).
The MK deposit is in an embayment bounded by headlands
of deeply weathered crystalline rock of the region (Gillson,
1959). The coast has been uplifted so that former sea cliffs are
now elevated and inland. The K-N bar of the Quilon deposit is a
barrier island on a drowned coast. At the southern end are head-
Figure 79. Stratigraphy of Eneabba deposits in a mine face. Light- lands 2 to 10 m high of weathered Cenozoic Warkalli sedimen-
colored basal deposits are the upper surface of Mesozoic sandstones, into tary rocks (Poulose, 1972; Prakash and Verghese, 1987). No
which heavy minerals have been introduced in joints and potholes. The
fluvial sediment currently reaches the sea near the Quilon deposit
dark overlying material is a heavy-mineral-rich shoreline deposit. Not
visible is a laterized upper surface on this material. The uppermost light- (Tipper, 1914; Gillson, 1959), and net erosion occurs along this
colored material is eolian sand. shoreline (Prakash and Verghese, 1987).
Both deposits are apparently Holocene in age and consist of
high-grade enrichments on modern beaches, buried prograda-
quartz is characteristic of granulite terranes (see Herz and Force, tional beaches, and lower-grade eolian dunes (Gillson, 1959).
1987, for details). Welch and others (1975) showed that the Enrichment is only in sands at or above present sea level. Ilmenite
trace-element contents of Eneabba ilmenites are quite different is the predominant heavy mineral in both deposits (70 to 80
from Geographe Bay ilmenites, implying separate sources for the percent [%], with lesser zircon (about 5%), rutile (4%), sillimanite
two districts. (4%), and monazite (0.5 to >1%). Garnet is common at MK but
History of development. Lissiman and Oxenford (1973, not at Quilon. Ilmenite concentrates from Quilon contain 60 to
1975) and Baxter (1977) discuss the paleogeography of the dis- 62 percent Ti0 2 , whereas those from MK contain 54 percent
trict in terms of a north-facing former bay, the western side of (Sinha, 1967). The alteration of MK ilmenite has been described
which is a peninsula ("Rocky Springs cape") formed of Cockle- by Subrahmanyam and others (1982), and of Quilon ilmenite by
shell Gully Formation, stratigraphically below the Yarragadee. Karkhanavala and others (1959). Rutile concentrates are re-
This cape, with a maximum present elevation of 95 m, would ported to show high niobium and tantalum contents (Subrahmanyam
have sheltered the depositional environments of the lower shore- and Rao, 1980).
lines and would have added the Cockleshell Gully as a source for Mining on modern beaches of the district has shown the
lower shorelines (Fig. 81). formation of natural concentrates to be highly seasonal. Black
Shepherd (1990), reporting on more recent exploration,
finds no evidence for the Rocky Springs cape or the north-facing
bay and no subcrops of Cockleshell Gully Formation in the mine
area. In Shepherd's view, the different character of the higher
shoreline deposits is due to selective preservation; only high-
energy swash zone deposits were preserved on the narrow higher
terraces.
Travancore Coast district, Kerala

and Tamil Nadu States, India
This district, mined for monazite since 1911, became the
world's largest ilmenite producer in the 1940s. Extensive use of
hand labor and shallow harbors have limited production and
caused the district to lose its position.
The geology of the Travancore deposits has been studied
most comprehensively by Tipper (1914). Other studies are by Figure 80. "Coffee rock" from Eneabba, containing iron hydroxide ce-
Gillson (1959) and Mallik and others (1987). Other, somewhat ment probably pseudomorphic after organic structures.
94 E. R. Force
ently thefirstto link source rocks and changes in fluvial conduits
and sea level with the locations of any shoreline placer deposits.
Richards Bay district, South Africa

Heavy-mineral deposits north of Richards Bay, mined since
1967, have put South Africa in the forefront of titanium-
mineral-producing countries. An innovative smelting process for
the relatively low-TiC^ ilmenite is partly responsible for the
commercial success of the venture. Richards Bay is on the Natal
0 1 2 3 Kilometers coast, facing the Indian Ocean. The deposits have been described
1 I I I
only by Hammarbeck (1976) and Fockema (1986). Quaternary
Figure 81. Cross-sectional relation implied by Lissiman and Oxenford development of coastal eolian landforms has been described by
(1973, 1975) and Baxter (1977) for shorelines, Rocky Springs cape, and Maud (1968).
exposed Mesozoic sandstones in the Eneabba district. See text for
evaluation.
The economic heavy-mineral deposits consist of eolian
sands in a belt along the modern coast, up to about 1 km inland
and 5 to 25 km north of Richards Bay. They are apparently
mostly Holocene, but some of the older sands that form part of
sands carpet the swash zones during strong onshore winds of the the deposit may be Pleistocene, based on their weathering and
monsoon season, whereas otherwise the beaches are white and no cementation and the dating of buried stone implements (Davies,
concentrates are visible (Canadian Mining Journal, 1956). Con- 1970). The eolian dunes are locally more than 100 m high and
centration efficiency is so great that glass sand deposits are present rest on a platform of 20 to 40 m elevation, underlain by the
also in modern shoreline sands of the same district (Poulose, Pleistocene Port Durnford Formation. Some laterite is developed
1972). In the Quilon sector, mining activity is gradually depleting on this platform. The eolian beds rest on a beach sand that
the heavy-mineral stocks available for concentration (Gillson, Fockema (1986) includes in the Port Durnford, but which Maud
1959). (1968) and Hammarbeck (1976) regard as post-Port Durnford.
The ultimate sources of the heavy minerals of the district are The upper part of the Port Durnford Formation is exposed in
granulite-facies gneisses of the interior. The most abundant wave-cut benches of the district and, according to Maud and
lithologies of the interior headwaters of the district are Hammarbeck, consists of yet more eolian sands.
khondalites (Balasundaram, 1970; Mallik and others, 1987); that The dune deposits are commonly about 20 m thick and
is, sillimanite-garnet-graphite gneisses that here contain ilmenite average 10 to 14 percent heavy minerals. Heavy-mineral content
and monazite (Soman, 1985) and commonly elsewhere contain is greater in the lower dunes and in the lower parts of the high
rutile (Force, 1976b; Fig. 4). dunes. The economic heavy minerals ilmenite, zircon, and rutile
Three intermediate sources are of importance in the district. average 5.9 percent of bulk ore. Ilmenite is the dominant heavy
First, thick laterites are developed on the crystalline rocks, and mineral and contains 46 to 50 percent Ti02- Some leucoxene and
these release the right mineral suite (Gillson, 1959). Second, the monazite are present and are also of value. Other minerals present
Cenozoic Warkalli sandstones contain heavy mineral concentra- include hornblende and augite, magnetite, garnet, epidote, and
tions and are themselves weathered (Tipper, 1914). Poulose tourmaline. Feldspar is present in the younger sands. Average
(1972) observed that the area of abundant sea-cliff erosion of diameter is about 0.3 mm for the light minerals and about 0.1 to
Warkalli sandstones is precisely the area of rich Holocene 0.15 mm for the heavy minerals.
heavy-mineral concentrates. Third, Pleistocene cemented coastal The chain of probable intermediate sources is long. The Port
eolian dunes are being eroded (Tipper, 1914; Narayanaswami Durnford beds are clearly important sources of the heavy miner-
and Mahadevan, 1964) and their contents resupplied to beaches. als in the dunes. Upper Cretaceous sandstones also crop out along
A significant role of weathering in the formation of sands of this coast. Hammarbeck (1976) believes that the Ecca Series
the district is implied by the beneficiated mineral assemblages and (Permian) of the Karoo System is an intermediate host of some
the numerous intermediate hosts (Poulose, 1972). Garnet, for importance, based on similar ilmenite composition. Karoo rocks
example, is abundant in the source rocks and is present in the are the most abundant within a 200-km radius of the district, and
Warkalli beds but is virtually absent in the Quilon deposit elsewhere in the Karoo basin contain Permian shoreline-facies
(Tipper, 1914). Scanning electron microscope photographs by concentrations of the same mineral suite (Behr, 1965,1986). The
Mallik (1986) show a great deal of chemical etching that predates presence of little-altered ilmenite with magnetite, pyroxene, and
final rounding (e.g., his Figs. 2A, 3B). epidote in a mineral suite including leucoxene (Dimanche, 1972)
Gillson (1959) proposed a Cenozoic history of stream cap- suggests multiple intermedate sources for heavy-mineral sands of
ture in the fluvial drainages of the district. His model was appar- the district. The ultimate sources are probably granulite-facies
gneisses, some of which are present in the immediate area (Sag- (Force, 1976b; Goldsmith and Force, 1978). These rocks contain
gerson and Turner, 1978). the proper mineral suites (Chapter 2) and form large source
terranes (Force, 1980b). Anorthosite-ferrodiorite massifs com-
ECONOMIC PROGNOSIS monly found with these terranes also supply the proper mineral
suite.
Quaternary shoreline deposits are currently the most impor- Intermediate sedimentary sources are of great importance in
tant source of titanium minerals, and for most uses they are the every district described. In the eastern Australia district, such
preferred source. The only problem with the continued commer- intermediate sources are the only ones we see, possibly because of
cial success of shoreline deposits relates to their possible exhaus- continental fragmentation. The exploration logic relative to
tion. Deposits of this class will probably continue to capture a source should be to consider coasts currently fed by favorable
respectable share of the market as long as sizable deposits remain ultimate source terranes and/or by clastic sediments and sedi-
accessible. mentary rocks known to have favorable mineralogy.
Though many currently known shoreline placers have lim- Weathering is the next permissive condition to be consid-
ited lives, this does not mean that the future of the entire deposit ered. Postdepositional weathering is known to be important, but
class is sealed. As this is written, two large, newly discovered every district described in this chapter shows strong evidence of
shoreline placers, in Victoria, Australia, and Madagascar, are predepositional weathering at other sites. Among these sites, the
being developed, reflecting the fact that much remains to be intermediate sedimentary hosts are probably most important.
learned and done. Exploration efforts to date have used a small Simple zonation by latitudes apparently controls the distribution
fraction of the pertinent information. of currently economic Quaternary shoreline placer deposits, with
In some areas of the world, environmental constraints have auxiliary influence by rainfall distribution. However, changes in
prevented or halted exploitation of shoreline placer deposits that climate through time clearly influence intermediate hosts and
are still forming or are of very recent age. This is legitimate, and pre-Quaternary placers. The working of these factors is detailed
we should expect environmental protection of modern coastlines in Chapter 6. The exploration logic relative to weathering should
to become more common worldwide. In most districts, protec- be to look for Quaternary shoreline deposits on coasts of moder-
tion of modern coastlines takes away only a small fraction of the ate relief at latitudes lower than 35°, unless climate change is
resources there. In the same area where the modern shoreline is thought to have produced weathering-beneficiated assemblages in
subject to high land values, competing land uses, and strict en- intermediate hosts. For Quaternary shoreline deposits now off-
vironmental regulations, multiple deposits of former shorelines a shore, the favorable latitudes are restricted as a function of depth
few kilometers inland are typically free of these constraints. The (Chapter 6).
only advantage of the modern over the former shoreline is in ease Conduits that supply the proper detrital feed to the shoreline
of discovery; the potential for deposits is the same, and the in little-diluted form are the next consideration. Fluvial systems
amount of exploration required to confirm a deposit is the same. that characteristically supply little-diluted material are described
The older deposits commonly have as little overburden as the in Chapter 8. In this chapter, however, we have seen that conduits
modern deposits and are more beneficiated by weathering. integral to the coast itself are equally important. These commonly
involve sea-cliff erosion of favorable intermediate hosts and/or
METHODS OF EXPLORATION the suppression of fluvial or other unfavorable debris by sea-level
change. The most valuable deposits probably result where fluvial
Conceptual exploration drainages bring favorable weathered material to the coast in a
little-diluted state and/or where coastal processes at particular sea
Every portion of this book, and probably much more levels bring material from favorable intermediate hosts.
information besides, bears on the exploration for shoreline placer Guild (1971) stated that shoreline placer deposits of
deposits of titanium minerals. These deposits can form where (1) titanium oxide minerals occur on passive continental margins.
source rocks have the proper mineralogy, (2) weathering restricts Statistically he is right. Such margins permit favorable source
the heavy-mineral assemblage, and (3) conduits can supply the rocks to be adjacent to the coast, permit deep weathering because
right material in little-diluted form. Only then can concentration of moderate relief, and permit delivery to the coast of fluvial
on the beach face form an economic deposit. Given these pre- mineral suites little contaminated by volcanics, glaucophane, ser-
requisites, additional factors may aid the concentration process pentine, and other troublesome components of active continental
and the preservation of the concentrate. All these factors can be margins. However, Guild's concept neglects the possibility of
considered in sequence in exploration programs to delineate microplate accretion of favorable terranes onto an active margin.
promising areas. The coastal anorthosite-granulite terrane of Oaxaca, Mexico, is
Source areas are logically consideredfirst,as they are neces- an example. Somewhere else on active margins a sedimentary
sary to supply the proper mineral assemblage. The great majority intermediate host bearing weathered titanium oxide minerals
of economic shoreline placer deposits of titanium minerals clearly could have been rafted in to form coastal outcrops; fluvial debris
have ultimate source areas of high-grade metamorphic rocks from the remainder of the margin could be suppressed by sea-
96 E. R. Force
level rise, and the valuable titanium minerals supplied only to Economic shoreline deposits of titanium oxide show only
local coastal compartments. very subtle magnetic response (Robson and Sampath, 1977) be-
cause (1) these bodies are shallow, (2) favorable source terranes
Favorable depositional environments contribute little magnetite, and (3) the weathering required for
beneficiation of the mineral assemblages removes what magnetite
The next step is the search for physical environments that is present.
produced and preserved placer concentrations of titanium miner- Aeroradioactivity has been the most successful geophysical
als. Progradational high-energy beach systems are a promising exploration method thus far, and large portions of some coastal
target; storm and seasonal events produce concentrations on plains have been thus surveyed. The Altamaha Plantations and
swash faces, and burial by progradation preserves them. Enrich- Green Cove Springs deposits were delineated in part from total-
ment in such deposits commonly occurs in the upper portions of count gamma aeroradioactivity maps. Detailed studies (Force
beach-ridge sand bodies and represents the portion of the beach and others, 1982; Grosz, 1983) have shown that a great deal of
complex that accumulated at and above sea level. geologic knowledge is necessary for correct interpretation of
Transported enrichments may occur in eolian sand masses. aeroradioactivity maps and intelligent exploration with them.
These are derived from the coast but may or may not be physically Surficial heavy-mineral concentrations can form aeroradioactiv-
contiguous with true beach deposits; some coastal eolian sands ity highs where monazite and zircon are present. However, these
are found atop sea cliffs. Assuming no overburden, these sand highs may be hidden by thin overburden, including water, or they
masses will be topographically positive elements, commonly but may be swamped by other common surficial lithologies that pro-
not always elongate parallel to the paleoshore. Heavy-mineral duce high radioactivity values. In addition, human activities and
concentration in eolian deposits is likely to be fairly homogene- additions to the surficial environment—for example, phosphate
ous, except at the scale of individual laminae. However, the tops fertilizer—may greatly modify the aeroradioactivity pattern
of the highest dunes are reported to be impoverished in heavy (Force and others, 1982; Grosz, 1983). Spectral radioactivity sur-
minerals in some districts. veys, whether air- or ground-based, can help sharpen exploration
The positive geomorphic expressions of former barrier efforts with aeroradioactivity, as heavy-mineral concentrations
islands and associated foredunes have been widely used in ex- are normally dominated by the contribution from thorium. Aero-
ploration for shoreline placer deposits. Geomorphology may radioactivity prospecting coupled with sampling has shown that
indeed give some strong clues to the locations of different deposi- most shoreline sands have radioactivity responses lower than am-
tional environments and even conduit systems. Lithologic and/or bient values because surficial exposures of most shoreline sands
soils maps, if available, can indicate the locations of appropriate are poor in heavy minerals (Force and others, 1982). Thus, sur-
sand bodies, and used in conjunction with geomorphology, can face exposures of shoreline sands rich in heavy minerals com-
efficiently direct exploration. Air-photo interpretation may help monly produce local positive aeroradioactivity anomalies in
delineate progradational beach-ridge systems and transgressive larger areas of low values.
dune morphology. Ground-based induced polarization (IP) methods have
In many districts, certain parts of individual shoreline com- proved to respond to shoreline placer deposits dominated by
partments or barrier segments are consistently those showing altered ilmenite (Wynn and others, 1985) but not to those domi-
greater heavy-mineral concentration (cf., Peterson and others, nated by rutile (Robson and Sampath, 1977). Spectral IP surveys
1985). These are apparently controlled by the locations of head- appear to differentiate between the response of ilmenite and other
lands and/or inlets. Once the pattern has been established in a minerals.
district, geomorphic analogs can be systematically prospected. Geophysical methods of exploring for deposits now offshore
require modification. Radioactivity exploration is hampered be-
Geophysical exploration cause gamma radioactivity is shielded by water. Induced polariza-
tion, on the other hand, is simpler than on land, because the
Exploration based on the physical properties of shoreline conductivity of sea water permits use of a surface-towed streamer
placer concentrations has been instrumental in the discovery and (Wynn and Grosz, 1986).
delineation of several deposits, and there is great potential for
further improvement of these methods. Methods investigated thus
far include magnetism, radioactivity, and induced polarization.
Chapter 10.
Placer deposits in pre-Quaternary sedimentary rocks
Placer deposits described in preceding chapters are those in Eolian deposits have not been preserved in any of the dis-
which original depositional landforms are still preserved. Some tricts described in this chapter. If eolian deposits were originally
possibly pre-Quaternary deposits that met this criterion were in- present, they probably were planed off in the nonmarine envi-
cluded. The pre-Quaternary deposits described in this chapter are ronments represented by the beds overlying the beach concentra-
buried by younger sediments, thus masking original depositional tions, as the dunes were mobile positive elements of the land
landforms. Some such deposits even have steep dips. In two surface. The prevalence of titanium-mineral resources in Quater-
districts described, host rocks are indurated, whereas in two oth- nary coastal eolian deposits probably reflects lack of reworking
ers they are not. The inherent economic disadvantage of an during burial.
indurated deposit can be overcome only with very high grades. Three of the four districts described in this chapter formed
Ironically, some details of sedimentary deposition are best seen in during geologic time intervals characterized by deep weathering
cross section in the indurated deposits. (Late Cretaceous, Miocene-Pliocene boundary). Their mineral
Only one pre-Quaternary placer district of titanium oxide assemblages are thus more beneficiated by weathering. The im-
minerals, in the Lakehurst area of New Jersey, has been mined portance of such weathering, and temporal variations thereof, are
extensively. This deposit is closely analogous to Quaternary discussed in Chapter 6.
placer deposits in many respects. The economic status of the other
pre-Quaternary deposits is uncertain. INTRASTRATAL SOLUTION
DEPOSITIONAL ENVIRONMENTS AND The leaching of heavy-mineral assemblages, discussed in

TEMPORAL VARIATION Chapter 6 in terms of surficial weathering processes, acquires a
subsurface dimension in buried older deposits. The entire heavy-
All pre-Quaternary placer concentrations of titanium min- mineral assemblage of a sandstone can vary greatly with its his-
erals described to date as economic or near economic are shore- tory of intrastratal solution. Equilibration with a variety of
line deposits and thus formed under conditions described in ground waters may have occurred over relatively long time peri-
Chapter 9. Stratigraphic relations detailed in this chapter show ods. An extensive summary is given in Pettijohn (1957). Iron
that the deposits formed on prograding shorelines, that is, they are silicate minerals are particularly susceptible to alteration, both in
underlain by marine rocks and overlain by nonmarine rocks. oxidizing (Walker, 1967) and reducing (Siever and Woodford,
Individual concentrations probably represent minor erosional 1979) ground waters. Skeletal remnants of these former detrital
transgression, like concentrations observed in modern prograding minerals are visible in some sedimentary rocks (Houston and
beach systems. In two districts described here, stratigraphic de- Murphy, 1962; Friis, 1974; Scholle, 1979). The titanium silicate
scriptions are sufficiently precise that the environment of concen- sphene is also somewhat unstable and may be replaced by
tration can be specified as the upper swash zone. In one district, "leucoxene" and quartz and/or carbonate (Morad and Aldahan,
the distribution of heavy-mineral concentrations relative to other 1985).
lithologies has been used to hypothesize details of the paleo- Detrital ilmenite may alter to authigenic anatase, rutile,
environment that would be difficult to reconstruct in any other and/or brookite, forming rims, crack fillings, and/or cement
way (Houston and Murphy, 1962, 1977). (Houston and Murphy, 1962; Mader, 1980; Morad and Aldahan,
97
98 E. R. Force
1986). Authigenic anatase is locally euhedral and too coarse to be is, toward the south-southeast. Overlying this facies in an area of
considered microcrystalline, but it isfinerthan associated detrital about 1 km2 is the Legler Lignite (Rachele, 1976), a freshwater
sand grains (Houston and Murphy, 1962; Morad, 1986). Rutile swamp deposit 1 to 5 m thick with thin associated clays that
as skeletal aggregates may form pseudomorphs of detrital contain marine microfossils in the upper part of the unit.
ilmenite where alteration of ilmenite is complete. Iron sulfide These lithologies, forming Carter's lower sequence, clearly show
minerals, where present with titanium dioxide alteration prod- beach progradation. The upper sequence of the Cohansey
ucts, suggest that leaching of iron was under reducing conditions contains cross-bedded subtidal and intertidal sands, and in my
(Dimanche and Bartholome, 1976; Reynolds and Goldhaber, opinion represents a separate depositional sequence following
1978). transgression.
Environment of concentration. Puffer and Cousminer
M A J O R DEPOSITS (1982) found that sands enriched in heavy minerals (as much as
Lakehurst district, New Jersey 63 percent) arefine(0.2 mm mean) and well sorted, but not well
rounded. They attributed high positive skewness to an eolian
From 1962 to 1982, the Lakehurst district was an important origin but neglected to consider that heavy-mineral enrichment in
source of altered ilmenite in the United States. Mining, by two shoreline environments occurs by preferential removal of coarse
companies in separate pits, was largely from the upper Tertiary lights, which necessarily produces positive skewness. The strati-
Cohansey Sand. graphic position of greatest enrichment recorded by Carter
The district is in the "pine-barren" region of the coastal plain (1978) was at the top of the swash-zone facies, and this suggests
of southern New Jersey. Population growth in the area is rapid, upper swash-zone concentration similar to modern beach
and part of the district lies beneath a naval air station used for deposits.
lighter-than-air craft. Topographic relief is low; average elevation Ilmenite alteration. Mathis and Sclar (1980) and Puffer
of the land surface at the deposits is about 30 m. The district was and Cousminer (1982) found that precursor ilmenite-hematite
probably continuous originally, but fluvial dissection during the intergrowths were pseudomorphically replaced by microcrystal-
Pleistocene and Holocene has produced separate deposits. line pseudorutile (topotactic after ilmenite) and voids (after
Ilmenite deposits of the Lakehurst district were discovered hematite lamellae). This alteration is less complete in the Kirk-
in 1956 by Frank Markewicz and his colleagues at the New wood than in the Cohansey, where an appreciable number of
Jersey Geological Survey, using a genetic model involving source grains also show microcrystalline rutile. In the Cohansey, altered
rock and fluvial conduit (Markewicz, 1969). The most substantial ilmenite averages 65 percent TÍO2. Unable to relate weathering
of the more recent studies are by Carter (1978), who documented state to depth below land surface, Puffer and Cousminer sug-
stratigraphic trends in depositional environments, and Puffer and gested that an appreciable component of ilmenite alteration was
Cousminer (1982), who presented an integrated sedimentologic- predepositional. Some postdepositional alteration, however, is
paleoclimatic-mineralogic analysis of the deposit. suggested by local iron hydroxide cement, which encases minor
General description and stratigraphy. The ilmenite magnetite as well as ilmenite (James Stern, personal communica-
deposits are overlain by up to 7 m of Pleistocene(?) fluvial gravels tion, 1974). The intensity of alteration, coupled with the subtrop-
(Quirk and Eilertson, 1963). The upper Tertiary formations host- ical nature of Legler Lignite palynomorphs, prompted Puffer and
ing the deposits are unindurated. The highest-grade portions, Cousminer to hypothesize a warmer climate during deposition
which contain greater than 5 percent heavy minerals, form about than at present.
5 m of laminated fine to medium sand near the middle of the Geologic evolution. Puffer and Cousminer (1982) envision
Cohansey Sand, which here is about 20 m thick (Quirk and the Lakehurst deposits derived from ilmenite-rich gneissic source
Eilertson, 1963; Puffer and Cousminer, 1982). Altered ilmenite is rocks in the Hudson Highlands and possibly the Adirondack
the predominant heavy mineral (85 percent); also present are Mountains and from older sedimentary hosts, during a period of
zircon (7 percent), sillimanite (3 percent), and staurolite and deep weathering at the boundary between the Miocene and Plio-
tourmaline (each 1 percent). Underlying the Cohansey are finer cene epochs. Deposition was during a sea-level regression of that
sands of the Kirkwood Formation that contain lesser concentra- period, followed by marine transgression. The ancestral Delaware
tions of the same heavy-mineral assemblage. Recent workers are River was the fluvial conduit, as suggested by Markewicz (1969).
inclined to regard the base of the Cohansey as conformable with
the underlying Kirkwood. McNairy Sand, Tennessee
Carter (1978) has divided the Cohansey Sand into two "se- The Upper Cretaceous McNairy Sand contains substantial
quences," with the heavy-mineral-enriched zone in the lower resources of heavy minerals in three deposits that define a district
sequence. Shallow-marine trace fossils are found throughout the in western Tennessee (Wilcox, 1971). The McNairy is an unindu-
formation. At the base of the Cohansey are interbedded granule rated, very fine-grained sand representing shoreline deposition in
and sand layers that represent deposition in the surf zone. An the Mississippi embayment from Mississippi to Illinois. The
overlying laminated facies is enriched in heavy minerals and rep- heavy-mineral assemblage is economically attractive, and ilmenite
resents swash-zone deposition. The lamination dips seaward, that is altered to Ti0 2 contents of about 60 percent, butfinegrain size
Figure 82. Heavy minerals in facies of McNairy Sand exposed in a silica sand pit
operated by the Jessie Morie Co. in the Seventeen Creek quadrangle, Tennessee.
A. Flat-laminated facies below, with heavy-mineral concentration above shovel,
overlain by lowest cross-bedded facies. B. Large cross beds outlined by heavy-
mineral laminae in the upper unit.
makes mineral separation difficult. The district is not currently heavy minerals, with an average heavy-mineral grain size of
being developed. about 0.1 to 0.06 mm throughout the district. The average rela-
Physiographically, the McNairy Sand forms a dissected low tive abundance of heavy minerals is altered ilmenite, 55 percent
plateau bordering the presently dammed Tennessee River. The (%); leucoxene, 8%; rutile, 2%; zircon, 10%; monazite, 1%;
forested Natchez Trace State Park occupies the southern end of aluminosilicate minerals, about 20%; and tourmaline, 2%. Ilmen-
the district and contains substantial resources (Hershey, 1966, ite, zircon, and monazite are concentrated in thefinestfractions
1968). The McNairy is at the surface over much of the district but (<0.06 mm), whereas less-dense leucoxene, aluminosilicates, and
is poorly exposed except in steep bluffs along digitate margins of tourmaline are concentrated in the coarser (>0.06 mm) fraction.
the plateau. The best exposures are in several silica sand mines in Wilcox (1971) pointed out that the relative grain size of
the McNairy. leucoxene implies that it was already less dense at deposition and
The regional geologic context of the district has been de- hence was altered before deposition. In two of the deposits,
scribed by Russell (1975). Quadrangles containing deposits have leucoxene is more abundant toward the base of heavy-mineral
been mapped by Hershey (1966,1968), Russell (1967), and Fer- enrichments (Ferguson and Garman, 1970; "Manleyville" of
guson and Garman (1970). The McNairy Sand, about 50 to Wilcox, 1971), also suggesting predepositional alteration.
100 m thick, is the shoreline facies of the upper regressive In several silica sand pits southeast of Bruceton, the strati-
sequence in a transgressive-regressive wedge. The McNairy is graphic sequence in and above the basal member of the McNairy
underlain by a shallow-marine shale. Toward the eastern (land- can be seen (Fig. 82A). Heavy mineral enrichments occur in fine-
ward) edge of the outcrop belt, at the closure of the wedge, the to medium-grained, well-sorted but angular sand with planar
McNairy locally lies directly on Paleozoic bedrock. Northward laminations dipping very gently southwest. Ophiomorpha bur-
into Kentucky, I have found that the McNairy becomes a rows are common in this interval. These beds are apparently
tidal-flat facies with tidal channel sands, probably reflecting the beach deposits. Overlying sands show increasingly high-angle
stronger influence of delta growth to the north, as reported by cross beds with variable mica and clay contents and clay beds
Pryor (1960). Heavy-mineral contents are low in the McNairy of draped over dune and channel bed forms (Fig. 82B). These beds
Kentucky, as they are in southern Illinois (Hunter, 1968). are apparently fluvial in origin and suggest sedimentary transport
The basal member of the McNairy Sand, up to 15 m thick, to the southwest. They contain fewer heavy minerals than the
contains most concentrations of heavy minerals. The sands of this beach deposits, and their mineral assemblage contains far greater
member show evidence of beach deposition. The veryfineangu- proportions of the lighter heavies leucoxene, kyanite, and
lar sands of the basal member locally contain more than 5 percent goethite. The upper beds are preferentially mined for silica.
100 E. R. Force
Upper Cretaceous deposits of the the depositional environment based on variations in deposit
western interior, United States geometry.
Heavy-mineral concentrations in sandstones of Late Cre- Karoo Basin, South Africa

taceous age are known from Montana, Wyoming, Utah,
Colorado, New Mexico, and Arizona. Several of these sandstones Heavy mineral concentrations of economic interest occur in
are black from heavy-mineral concentrations; most were found as shoreline deposits of the Ecca Group, of Permian age, in the
a result of their radioactivity. The concentrations are preferen- northern Karoo Basin. The deposits, though rich, are not eco-
tially cemented and therefore are resistant to erosion; they form nomic because of difficulties with mineral separation. The
hogbacks and mesas in several individual sedimentary basins. thorough description of the Bothaville deposit by Behr (1965) is
Sedimentologic and tonnage aspects of the deposits have been the basis of the following discussion. Behr (1986) has described
described by Houston and Murphy (1977) and Dow and Batty other deposits of the district. The petrography of heavy-mineral
(1961), respectively. Houston and Murphy (1962) and Roehler alteration has been treated by Mukeijee (1964).
(1989) have described the Wyoming deposits in some detail. Enrichments occur in the Vryheid Formation (Middle Ecca
These references are the sources of the following descriptions. Allsandstone), which rests on upper Paleozoic glacial deposits and
the deposits from six states are described here as a single district,
Precambrian lavas. A swarm of individual deposits, each oriented
for want of space for more detailed treatment. northwest-southeast and dipping very gently southwest, occurs in
The deposits represent regressive shorelines from several progradational relation. Individual deposits average 3 m thick
transgressive-regressive cycles in the Upper Cretaceous of the and 350 m wide; one is about 5 km long. The general stratigraphy
western interior seaway. Heavy-mineral contents are locally as of the Ecca Group suggests deltaic progradation, and heavy-
high as 90 percent, and intervals as thick as 3 to 5 m may contain mineral enrichments are thought to represent shoreline reworking
10 percent or more heavy minerals. Opaque oxide minerals dom- of inactive delta lobes. Eolian facies are apparently absent.
inate the heavy-mineral assemblage throughout the district but Sands enriched in heavy minerals are well sorted and
vary from mostly magnetite in Montana to mostly altered positively skewed, with average grain size 0.1 to 0.2 mm, in
ilmenite (with up to. 60 percent Ti0 2 ) in New Mexico (see also contrast to less enriched adjacent sandstones that are more poorly
Chenowith, 1957). In the Wyoming deposits, altered hemoil- sorted and have average grain sizes of 0.25 to 0.30 mm. Detrital
menite predominates over altered ilmenite. Zircon is common, grains are well rounded.
and minor garnet, tourmaline, rutile, and monazite are present. In Heavy-mineral contents of more than 50 percent are com-
sandstones containing leaner heavy-mineral concentrations, mon in enriched beds. Ilmenite averages 60 percent, zircon
garnet, tourmaline, sphene, amphibole, epidote, and staurolite 7 percent, and garnet, monazite, and rutile a few percent each, of
form an appreciable part of the assemblage. The deposits are the heavy-mineral suite. Some chromite is present. Ilmenite,
cemented by goethite, carbonate, and authigenic anatase (locally which commonly contains hematite intergrowths, is rimmed by
brookite). The goethite component of this cement apparently leucoxene consisting of microcrystalline anatase. Quartz, feldspar,
results from alteration of opaque minerals in the modern weather- and mica are the light minerals. Iron sulfides, siderite, chlorite,
ing environment. and anatase are present as cement; near the ground surface, goe-
Most of the deposits are in sequences of fine sandstone, but thite is present in place of sulfides and siderite (Mukerjee, 1964).
the rich concentrations are generally very fine sands. Constituent Vos and Hobday (1977) concluded from stratigraphic trends
grains are mostly well rounded, although opaque oxide minerals within the beach deposits that heavy-mineral enrichment oc-
are commonlyrimmedby authigenic anatase, which gives them a curred in the upper swash zone. Sands of the swash zone contain
variety of shapes. gently southwest-dipping laminae, fine upward, and show local
Horizons showing the greatest concentrations are toward the inverse grading. Vos and Hobday proposed a storm-wave micro-
top of stratigraphic intervals that represent beach deposition. tidal environment prograding within a deltaic depositional regime.
These are underlain by shallow-marine deposits and are overlain Behr (1986) reports the economic potential of the Bothaville
in turn by lagoonal, swamp, and/or fluvial deposits. Enriched and related deposits to be low, particularly because of problems
horizons and underlying beach sands contain planar laminae dip- with recovery of altered ilmenite. Strong cementation of the host
ping at very low angles down Cretaceous beach faces (see cover rock requires vigorous crushing that destroys friable altered
illustration). Thus the concentrated horizons apparently represent ilmenite.
prograding upper swash zone sands. Eolian sands have only been
recognized in one locality (Roehler, 1989). ECONOMIC PROGNOSIS
The elongation of these beach deposits in map view has
been used to establish Cretaceous shoreline trends. Locally, Unconsolidated shoreline placer deposits are equally attrac-
Houston and Murphy (1962, 1977) have been able to hypothe- tive regardless of age, as long as their mineral assemblages and
size wind directions, tidal ranges, and other subtle features of grades are appropriate and overburden is minor. The older depos-

its may be even more beneficiated by postdepositional weathering be useful for unindurated near-surface pre-Quaternary deposits
or may represent deposition in a paleoclimate more conducive to also. The important tool lacking in pre-Quaternary deposits is
intensive weathering (Chapter 6). Since the proportion of pre- geomorphic expression, which has been important in exploration
Quaternary deposits that has been discovered is probably small, for Quaternary deposits. Depositional landforms reveal favorable
they may eventually assume a larger share of world titanium- depositional environments and former fluvial conduits, which are
mineral supply if Quaternary deposits approach exhaustion. useful in predicting deposit locations. In exploration for pre-
Indurated deposits, on the other hand, must be very high Quaternary deposits, this step is blind. For deposits that have
grade to be economically competitive. The requisite grades are been buried, the exploration geologist must use stratigraphie evo-
documented only in beach deposits, and in these, individual high-lution as a substitute for depositional landforms.
grade portions are typically ony 1 to 3 m thick, 50 to 150 m wide, Pre-Quaternary beach placers apparently form in prograd-
ing shoreline sequences. Thus, exploration should focus on sandy
and less than 5 km long. Thus they typically contain less than one
million tons of heavy minerals. The cementation of the deposits, transitions from underlying marine sections to overlying nonma-
coupled with the alteration of contained ilmenite, also creates arine sections, in sequences that fill the source-rock, weathering,
separation problem; many of the ilmenite grains would not sur- and conduit requirements of placer formation. In many areas,
vive crushing. existing stratigraphie information is sufficient to pinpoint the
An intermediate situation in Cretaceous black sandstones ofmore favorable shoreline depositional environments in such
the western United States has been described by Houston and transitions.
Murphy (1962). Here cementation is in part a surface phenom- Certain geologic periods are characterized by deep weather-
ing, both regionally and worldwide (Frakes, 1979). Sections re-
enon, caused by alteration of detrital iron-titanium oxides at the
present land surface. Below this surface, the deposit is more presenting these periods are everywhere more favorable for
friable, and the cement is partly carbonate and easily removed formation of economic placer deposits of titanium oxide miner-
chemically. als. Chapter 6 discusses the prediction of favorable depositional
regions within favorable time envelopes.
METHODS OF EXPLORATION All geophysical methods useful in exploration for Quater-
nary deposits should be useful for analogous pre-Quaternary
Exploration for pre-Quaternary placer deposits of titanium ones. Radioactivity surveys deserve special mention because most
oxide minerals should be aimed at near-surface unconsolidated pre-Quaternary placer deposits, including those described in this
shoreline placers, as these present the favorable combination of chapter, were found by radiometric methods. One reason is his-
low mining costs and potentially high grades throughout mineable torical; many sandstones were examined for uranium in the
tonnages. Probably a great many of these deposits remain to be 1950s. The high monazite contents of the pre-Cenozoic placers
discovered. enhances their radioactivity.
The conceptual and physical exploration tools that are use-
ful for Quaternary shoreline deposits (Chapter 9) would mostly
102 E. R. Force
REFERENCES CITED
Abdullah, M. I., and Atherton, M. P., 1964, The thermometric significance of , 1950, Heavy-mineral beach sands of southern Queensland; Part 2, Physi-
magnetite in low-grade metamorphic rocks: American Journal of Science, cal and mineralogical composition, mineral description, and origin of the
v. 262, p. 904-917. minerals: Proceedings of the Royal Society of Queensland, v. 61, p. 59-104.
Achalabhuti, C., Isarangkoon, P., Ratanawong, S., Aeo-Phanthong, V., Kulvanich, Beck, W . W., Jr., 1973, Correlation of Pleistocene barrier islands in the lower
S., and Suwanasing, A., 1975, Heavy minerals in southern Thailand: Thai- Coastal Plain of South Carolina as inferred by heavy minerals: South Caro-
land Department of Mineral Resources Economic Geology Bulletin 9,106 p. lina Division of Geology Geologic Notes, v. 17, no. 3, p. 55-67.
Adams, J., Zimpfer, G. L., and McLane, C. F., 1978, Basin dynamics, channel Behr, S. H., 1965, Heavy-mineral beach deposits in the Karoo System: South
processes and placers, and placer formation; A model study: Economic African Geological Survey Memoir 56, 116 p.
Geology, v. 73, p. 416-426. , 1986, Heavy mineral deposits in the Karoo Sequence, South Africa, in
Alves, B. P., 1960, Distrito niobio-titanifero de Tapira: Brazil Divisao de Fomento Anhaesser, C. R., ed., Mineral deposits of southern Africa: Johannesburg,
da P r o d u j o Mineral Boletim 108, 48 p. Geological Society of South Africa, p. 2105-2118.
Anderson, A. T., Jr., and Morin, M., 1969, Two types of massif anorthosites and Bergeron, M., 1972, Quebec Iron and Titanium Corporation ore deposit at Lac
their implications regarding the thermal history of the crust, in Isachsen, Tio, Quebec: International Geological Congress, 1972, Excursion guidebook
Y. W., ed., Origin of anorthosite and related rocks: New York State Museum B 09, 8 p.
and Science Service Memoir 18, p. 57-69. , 1986, Minéralogie et géochimie de la suite anorthositique de la région du
Arthur, M. A., Dean W . E., and Schlanger, S. O., 1985, Variations in the global Lac Allard, Québec: Evolution des membres mafiques et origine des gites
carbon cycle during the Cretaceous related to climate, volcanism, and massifs d'ilménite [Ph.D. thesis]: University of Montreal, 480 p.
changes in atmospheric C 0 2 , in Sundquist, E. T., and Broecker, W . S., eds., Bergstal, S., 1972, The jacupirangite at Kodal, Vestfold, Norway: Mineralium
The carbon cycle and atmospheric C 0 2 ; Natural variations, Archean to Deposita, v. 7, p. 233-246.
present: American Geophysical Union Geophysical Monograph 32, Berner, R. A., and Landis, G. P., 1988, Gas bubbles in fossil amber as possible
p. 504-546. indicators of the major gas composition of ancient air: Science, v. 239,
Ashwal, L. D., 1982a, Mineralogy of mafic and Fe-Ti oxide-rich differentiates of p. 1406-1409.
the Marcy anorthosite massif, Adirondacks, New York: American Mineral- Berner, R. A., Lasaga, A. C., and Garrels, R. M., 1983, The carbonate-silicate
ogist, v. 67, p. 14-27. geochemical cycle and its effect on atmospheric carbon dioxide over the last
, 1982b, Proterozoic anorthosite massifs; A review: Lunar and Planetary 100 million years: American Journal of Science, v. 283, p. 641-683.
Institute Report 82-01, p. 40-45. Beurlen, H., and Cassedanne, J. P., 1981, The Brazilian mineral resources: Earth
Attanasi, E. D., de Young, J. H., Force, E. R., and Grosz, A. E., 1987, Resource Science Reviews, v. 17, p. 177-206.
assessments, geologic deposit models, and offshore minerals, with an exam- Bigarella, J. J., Becker, R. D., and Duarte, G. M., 1969, Coastal dune structures
ple of heavy-mineral sands, in Teleki, P. G., ed., Marine minerals: London, from Parana (Brazil): Marine Geology, v. 7, p. 5-55.
Reidel, p. 485-513. Binns, R. A., 1967, Barroisite-bearing eclogite from Naustdal, Sogn og Fjordane,
Austin, S. R., 1960, Ilmenite, magnetite, and feldspar alteration under reducing Norway: Journal of Petrology, v. 8, p. 349-371.
conditions: Economic Geology, v. 55, p. 1758-1759. Black, P. M., 1977, Regional high-pressure metamorphism in New Caledonia:
Badham, J.P.N., and Morton, R. D., 1976, Magnetite-apatite intrusions and calk- Phase equilibria in the Ouegoa district: Tectonophysics, v. 43, p. 89-107.
alkaline magmatism, Cansell River, N.W.T.: Canadian Journal of Earth Blake, M. C., Jr., and Morgan, B. A., 1976, Rutile and sphene in blueschist and
Science, v. 13, p. 348-354. related high-pressure facies rocks: U.S. Geological Survey Professional Paper
Bailey, S. W., Cameron, E. N., Spedden, H. R., and Weege, R. J., 1956, The 959C, p. C1-C6.
alteration of ilmenite in beach sands: Economic Geology, v. 51, p. 263-279. Bolsover, L. R., and Lindsley, D. H., 1983, Sybille oxide deposit; Massive Fe-Ti
Balasundaram, S.M.S., 1970, Geological and mineral map of Kerala: Geological oxides intrusive into the Laramie Anorthosite Complex, Wyoming [abs.]:
Survey of India, scale 1:1,000,000. EOS Transactions of the American Geophysical Union, v. 64, no. 18, p. 328.
Balsley, J. R., Jr., 1943, Vanadium-bearing magnetite-ilmenite deposits near Lake Bradley, J. S., 1957, Differentiation of marine and subaerial sedimentary en-
Sanford, Essex County, New York: U.S. Geological Survey Bulletin 940-D, vironments by volume percentage of heavy minerals, Mustang Island, Texas:
p. 99-123. Journal of Sedimentary Petrology, v. 27, p. 116-125.
Barbosa, O., Braun, O.P.G., Dyer, R. C., and Cunha, C.A.B.R., 1970, Geología Bramall, A., 1928, Dartmoor detritals; A study of provenance: Proceedings of the
da Regiáo do Triángulo Mineiro: Departamento Nacional Produgao Mineral Geological Association, v. 39, p. 27-48.
(Brazil) Boletim 136, 140 p. Bray, R. E., and Wilson, J. C., 1975, Guidebook to the Bingham mining district:
Basu, A., and Molinari, E., 1989, Provenance characteristics of detrital opaque Society of Economic Geologists, 156 p.
Fe-Ti oxide minerals: Journal of Sedimentary Petrology, v. 59, p. 922-934. Bragger, W . C., 1934-1935, O n the several Archean rocks from the south coast
Bateman, A. M., 1951, The formation of late magmatic oxide ores: Economic of Norway; 2, The south Norwegian hyperites and their metamorphism:
Geology, v. 46, p. 404-426. Oslo, Norsk Videnskaps-Academi Oslo Skrifter, Section I. Matem.-Natur
Baxter, J., 1972, The geology of the Eneabba area, western Australia, in Report of Klasse, 421 p.
the Department of Mines, western Australia, for the year 1971: Perth, West- Brown, C. E., 1983, Mineralization, mining, and mineral resources in the Beaver
ern Australia Department of Mines, p. 105-106. Creek area of the Grenville lowlands in St. Lawrence County, New York:
, 1977, Heavy mineral sand deposits of western Australia: Geological U.S. Geological Survey Professional Paper 1279, 21 p.
Survey of Western Australia Mineral Resources Bulletin 10, 148 p. Buddington, A. F., Fahey, J., and Vlisidis, A., 1955, Thermometric and petro-
, 1982, Mineral sands mining in Western Australia, in Exploitation of genetic significance of titaniferous magnetite: American Journal of Science,
mineral sands: Perth, Western Australia Institute of Technology, p. 61-80. v. 253, p. 497-532.
, 1986, Supergene enrichment of ilmenite; Is it related to lateralization? in Cameron, E. N., 1979, Titanium-bearing oxide mineral of the Critical Zone of the
Australia; A world source of ilmenite, rutile, monazite, and zircon: Perth, eastern Bushveld Complex: American Mineralogist, v. 67, p. 140-150.
Australasian Institute of Mining and Metallurgy, p. 23-27. Canadian Mining Journal, 1956, Beach sand mining from the sands of Travan-
Beasley, W. F., 1948, Heavy-mineral beach sands of southern Queensland; Part 1, core: Canadian Mining Journal, v. 77, p. 67-76.
The nature, distribution and extent, and manner of formation of the deposit: Cannon, H. B., 1950, Economic minerals in the beach sands of the southeastern
Proceedings of the Royal Society of Queensland, v. 59, p. 109-140. United States, in Snyder, F. G., ed., Proceedings of the Symposium on
Mineral Resources of the southeastern United States: Knoxville, University and eclogites—their differences and similarities: Geological Society of Amer-
of Tennessee Press, p. 202-210. ica Bulletin, v. 76, p. 483-508.
Carmichael, I.S.E., Nicholls, J., and Smith, A. L., 1970, Silica activity in igneous Collins, L. B, and Baxter, J. L., 1984, Heavy mineral-bearing strandline deposits
rocks: American Mineralogist, v. 55, p. 246-263. associated with high-energy beach environments, southern Perth Basin,
Carpenter, R. H., 1982, Mineralization and alteration in the Lincolnton, Ga., Western Australia: Australian Journal of Earth Sciences, v. 31, p. 287-292.
McCormick, S.C., area, in Allard, G. O., and Carpenter, R. H., eds., Explo- Collins, L. B., and Hamilton, N.T.M., 1986, Stratigraphic evolution and heavy-
ration for metallic resources in the Southeast: Athens, University of Georgia mineral accumulation in the Minninup shoreline, southwest Australia, in
Department of Geology, p. 120-142. Australia; A world source of ilmenite, rutile, monazite, and zircon: Perth,
Carpenter, R. H., and Allard, G. O., 1982, Aluminosilicate assemblages; An Australasian Institute of Mining and Metallurgy, p. 17-22.
exploration tool for metavolcanic terrains of the southeast, in Allard, G. O., Collins, L. B., Hochwimmer, B., and Baxter, J. L., 1986, Depositional facies and
and Carpenter, R. H., eds., Exploration for metallic resources in the South- mineral deposits of the Yoganup shoreline, southern Perth basin, in Australia;
east: Athens, University of Georgia Department of Geology, p. 19-22. A world source of ilmenite, rutile, monazite, and zircon: Perth, Australasian
Carroll, D., 1939, Beach sands from Bunbury, Western Australia: Journal of Institute of Mining and Metallurgy, p. 9-16.
Sedimentary Petrology, v. 9, p. 95-107. Colquhoun, D. J., 1965, Terrace sediment complexes in central South Carolina:
Carroll, D., Neuman, R. B., and Jaffe, H. W., 1957, Heavy minerals in arenaceous Atlantic Coastal Plain Geological Association Field conference 1965, 62 p.
beds in parts of the Ocoee Series, Great Smoky Mountains, Tennessee: Colwell, J. B., 1982a, Heavy mineral data from east Australian coastal sediments,
American Journal of Science, v. 255, p. 175-193. Newcastle to Fraser Island: Geologische Jahrbuch, v. D 56, p. 49-53.
Carson, D.J.T., and Jambor, J. L., 1974, Mineralogy, zonal relationships, and , 1982b, Sedimentology of surface sediments of the New South Wales shelf:
economic significance of hydrothermal alteration at porphyry copper de- Geologische Jahrbuch, v. D 56, p. 111-124.
posits, Babine Lake area, British Columbia: Canadian Mining and Metal- Connah, T. H., 1961, Beach sand heavy mineral deposits of Queensland: Geologi-
lurgical Bulletin, v. 67, p. 110-133. cal Survey of Queensland Publication 302, 31 p.
Carter, B. A., 1982a, Geology and structural setting of the San Gabriel anortho- Cortesogno, L., Ernst, W . G., Galli, M., Messiga, B., Pedemonte, G. M., and
site-syenite body and adjacent rocks of the western San Gabriel Mountains, Piccardo, G. B., 1977, Chemical petrology of eclogitic lenses in serpentinite,
Los Angeles County, California, in Geologic excursions in the Transverse Gruppo di Voltri, Ligurian Alps: Journal of Geology, v. 85, p. 255-277.
Ranges, southern California; Geological Society of America, Cordilleran Crawford, E. A., Herbert, C., Taylor, G., Helby, R., Morgan, R., and Ferguson, J.,
Section, Guidebook Anaheim meeting, Trips 5, 6, 11: p. 1-56. 1980, Diatremes of the Sydney Basin: Geological Survey of New South
, 1982b, Mineral potential of the San Gabriel anorthosite-syenite body, San Wales Bulletin 26, p. 294-323.
Gabriel Mountains, California, in Fife, D. L., and Minch, J. A., eds., Geol- Creitz, E. E., and McVay, T. N., 1948, A study of opaque minerals in Trail Ridge,
ogy and mineral wealth of the California ranges: Santa Ana, California, Florida, dune sands: American Institute of Mining and Metallurgical Engi-
South Coast Geological Society, p. 208-212. neers Technical Publication 2426, 7 p.
Carter, C. H., 1978, A regressive barrier and barrier-protected deposit; Deposi- Cressman, E. R., 1962, Nondetrital siliceous sediments: U.S. Geological Survey
tional environments and geographic setting of the late Tertiary Cohansey Professional Paper 440-T, 22 p.
Sand: Journal of Sedimentary Petrology, v. 48, p. 933-950. Crowder, D. F., and Sheridan, M. F., 1972, Geologic map of the White Mountain
Carvalho, W . T., 1974, Aspectos geológicos e petrográficos do complexo Peak Quadrangle, Mono County, California: U.S. Geological Survey Geo-
ultramáfico-alcalino de Cataláo I, Goias: Porto Alegre, Sociedad Braziliera logic Quadrangle Map GQ-1012, scale 1:62,500.
de Geología, Annais do 28th Congresso, v. 4-5, p. 107-123. Cruz, F. T., Machado, P. T„ Townsend, R., and Machado, J. E„ 1977, Aspects of
Cazeau, C. J., 1974, Heavy minerals of Quaternary sands in South Carolina, in the superficial phosphorus and titanium deposits of the Tapira Carbonatite
Oaks, R. Q., Jr., and Du Bar, J. R., eds., Post-Miocene stratigraphy, central Complex, Minas Gerais, Brazil: 25th International Geological Congress,
and southern Atlantic Coastal Plain: Logan, Utah State University Press, Sydney, Abstracts, v. 1, p. 209-210.
p. 174-178. Czamanske, G. K., Force, E. R., and Moore, W . J., 1981, Some geologic and
Centro de Investigación Minería y Metalurgia, 1986, Recuperación de rutilo potential resources aspects of rutile in porphyry copper deposits: Economic
desde relaves de cobre porfídico: Boletín, v. 1, 4 p. Geology, v. 76, p. 2240-2245.
Chappell, B. W., and White, A.J.R., 1974, Two contrasting granite types: Pacific Danilchik, W., and Haley, B. R., 1964, Geology of the Paleozoic area in the
Geology, v. 8, p. 173-174. Malvern Quadrangle, Garland and Hot Springs Counties, Arkansas: U.S.
Chenowith, W . L., 1957, Radioactive titaniferous heavy-mineral deposits in the Geological Survey Map 1-405, scale 1:48,000.
San Juan basin, New Mexico and Colorado, in Southwestern San Juan Darby, D . A., 1984, Trace elements in ilmenite; A way to discriminate proven-
Mountains Guidebook: New Mexico Geological Society 8th Field Confer- ance or age in coastal sands: Geological Society of America Bulletin, v. 95,
ence Guidebook, p. 212-217. p. 1208-1218.
Chesnokov, B. V., 1960, Rutile bearing eclogites from the Shubino village deposit Darby, D. A., and Tsang, Y.-W., 1987, Variation in ilmenite element composition
in the southern Urals: International Geology Review, v. 2, p. 936-945. within and among drainage basins; Implications for provenance: Journal of
Chidester, A. H., 1962, Petrology and geochemistry of selected talc-bearing ul- Sedimentary Petrology, v. 57, p. 831-838.
tramafic rocks and adjacent country rocks in north-central Vermont: U.S. Davidson, D., 1982, Exploration methods for mineral sands, in Exploitation of
Geological Survey Professional Paper 345, 207 p. mineral sands: Perth, Western Australia Institute of Technology, p. 107-122.
Clerici, C., Mancini, A., Mancini, R., Morandini, A., Occella, E., and Protlo, C., Davies, O., 1970, Pleistocene beaches of Natal: Annals of the Natal Museum,
1981, Recovery of rutile from an eclogite rock, in Laskowki, J., ed., Mineral v. 20, p. 403-442.
processing; Developments in mineral processing: New York, Elsevier, v. 2, Davis, E. G., and Sullivan, G. V., 1971, Recovery of heavy minerals from sand
pt. B, p. 1803-1827. and gravel operations in the southeastern United States: U.S. Bureau of
Clifton, H. E., 1969, Beach lamination; Nature and origin: Marine Geology, v. 7, Mines Report of Investigation 7517, 25 p.
p. 553-559. Davis, E. G., Sullivan, G. V., and Lamont, W . E., 1988, Potential for recovery of
Clifton, H. E„ Hunter, R. E., and Phillips, R. S., 1971, Depositional structures and rutile and other by-products from western copper tailings: U.S. Bureau of
processes in the non-barred high-energy nearshore: Journal of Sedimentary Mines Report of Investigation 9158, 19 p.
Petrology, v. 41, p. 651-670. Dennen, W . H., and Anderson, P. J., 1962, Chemical changes in incipient rock
Coleman, R. G., Lee, D. E., Beatty, L. B., and Brannock, W . W., 1965, Eclogites weathering: Geological Society of America Bulletin, v. 73, p. 375-384.
104 E. R. Force
Desborough, G. A., and Mihalik, P., 1980, Accessory minerals in the igneous host Fairchild, H. S., 1912, The glacial waters in the Black and Mohawk valleys: New
of molybdenum ore, Henderson Mine, Colorado: U.S. Geological Survey York State Museum Bulletin 160, 47 p.
Open-File Report 80-661, 16 p. Falkum, T., 1982, Geologisk kart over Norge, berggrunnskart "Mandal": Norges
Desborough, G. A., and Sharp, W . N., 1978, Tantalum, uranium, and scandium in geologiske undersokelse (Norway), scale 1:250,000.
heavy accessory oxides, Climax molybdenum mine, Climax, Colorado: Eco- Fantel, R. J., Buckingham, D. A., and Sullivan, D. E., 1986, Titanium minerals
nomic Geology, v. 73, p. 1749-1751. availability; Market economy countries: U.S. Bureau of Mines Information
Diemer, R. A., 1941, Titaniferous magnetite deposits of the Laramie Range, Circular 9061, 48 p.
Wyoming: Geological Survey of Wyoming Bulletin 31, 23 p. Ferguson, C. C., and Garman, R. K., 1970, Mineral resources summary of the
Dimanche, F., 1972, Evolution minéralogique de quelques sables titaniferes Buchanan Quadrangle, Tennessee: Tennessee Division of Geology and Min-
d'Afrique du Sud: Annales Société Géologique de Belgique, t. 95, eral Resources Summary M R S 19-NW (and map), 16 p.
p. 183-190. Field, M. E., and Roy, P. S., 1985, Offshore transport and sand-body formation;
Dimanche, F., and Bartholome, P., 1976, The alteration of ilmenite in sediments: Evidence from a steep, high-energy shoreface, southeastern Australia: Jour-
Minerals Science Engineering, v. 8, p. 187-200. nal of Sedimentary Petrology, v. 54, p. 1292-1302.
Dow, V. T., 1961, Magnetite and ilmenite resources, Iron Mountain area, Albany Filho, J.G.C.S., 1974, Prospec$ao de uranio nas diamines alcalinas de Serra
County, Wyoming: U.S. Bureau of Mines Information Circular 8037,133 p. Negra e Salitre: Comissao Nacional de Energia Nuclear (Brazil) Boletim 9,
Dow, V. T., and Batty, J. V., 1961, Reconnaissance of titaniferous sandstone 41 p.
deposits of Utah, Wyoming, New Mexico, and Colorado: U.S. Bureau of Fischer, A. G., 1981, Climatic oscillations in the biosphere, in Nitecki, M., ed.,
Mines Report of Investigation 5860, 52 p. Biotic crises in ecological and evolutionary time: New York, Academic,
Dryden, L., and Dryden, C., 1946, Comparative rates of weathering of some p. 101-131.
common heavy minerals: Journal of Sedimentary Petrology, v. 16, p. 91-96. Fish, G. E., 1962, Titanium resources of Nelson and Amherst Counties, Virginia;
Duane, D. B., Field, M. E„ Meisburger, E. P., Swift, D.J.P., and Williams, S. J., Part 1, Saprolite ores: U.S. Bureau of Mines Report of Investigation 6094,
1972, Linear shoals on the Atlantic Inner Continental Shelf, Florida, to Long 44 p.
Island, in Swift, D.J.P., Duane, D. B., and Pilkey, O . H., eds., Shelf sediment Fleischer, M., Murata, K. J., Fletcher, J. D., and Narten, P. F., 1952,
transport: Process and pattern: Stroudsburg, Pennsylvania, Dowden, Geochemical association of niobium (columbium) and titanium, and its
Hutchinson and Ross, p. 499-575. geological and economic significance: U.S. Geological Survey Circular 225,
Duchesne, J. C., 1972, Iron-titanium oxide minerals in the Bjerkrem-Sogndal 13 p.
massif, southwestern Norway: Journal of Petrology, v. 13, p. 57-81. Flemming, B. W., 1981, Factors controlling shelf sediment dispersal along the
Duchesne, J. C., Denoiseux, B., and Hertogen, J., 1987, The norite-mangerite southeast African continental margin: Marine Geology, v. 42, p. 259-277.
relations in the Bjerkreim-Sokndal layered lopolith (southwest Norway): Flinter, B. H., 1959, The alteration of Malayan ilmenite grains and the question of
Lithos, v. 20, p. 1-17. "arizonite": Economic Geology, v. 54, p. 720-729.
Dymek, R. F., 1983, Titanium, aluminum, and interlayer cation substitutions in Flohr, M.J.K., and Ross, M., 1989, Alkaline igneous rocks of Magnet Cove,
biotite from high-grade gneisses, West Greenland: American Mineralogist, Arkansas; Metasomatized ijolite xenoliths from Diamond J o Quarry: Amer-
v. 68, p. 880-899. ican Mineralogist, v. 74, p. 113-131.
Eberle, M. C., and Atkinson, W . W., 1983, Results of mapping at Iron Mountain, Fockema, P. D., 1986, The heavy mineral deposits north of Richards Bay, in
Laramie Anorthosite Complex, Wyoming: Geological Society of America Anhaensser, C. R., ed., Mineral deposits of southern Africa: Johannesburg,
Abstracts with Programs, v. 15, p. 565. Geological Society of South Africa, p. 2301-2307.
Elger, G. W., Wright, J. B., Tress, J. E., Bell, H. E., and Jordan, R. R., 1986, Force, E. R., 1976a, Titanium contents and titanium partitioning in rocks: U.S.
Producing chlorination-grade feedstock from domestic ilmenite-laboratory Geological Survey Professional Paper 959-A, 10 p.
and pilot-plant studies: U.S. Bureau of Mines Report of Investigation 9002, , 1976b, Metamorphic source rocks of titanium placer deposits; A geochem-
24 p. ical cycle: U.S. Geological Survey Professional Paper 959 B, 16 p.
Elliot, C. L., and Guilbert, J. M., 1975, Induced polarization response attributed , 1976c, Titanium minerals in deposits of other minerals: U.S. Geological
to magnetite and certain Fe-Ti oxide minerals: Geophysics, v. 40, p. 147. Survey Professional Paper 959-F, 5 p.
Elsdon, R., 1975, Iron-titanium oxide minerals in igneous and metamorphic — — , 1980a, Is the United States of America geologically dependent on im-
rocks: Minerals Science Engineering, v. 1, p. 48-70. ported rutile? Proceedings of the Industrial Minerals, 4th Industrial Minerals
Emslie, R. F., 1978, Anorthosite massifs, rapakivi granites, and late Proterozoic International Congress, p. 43-47.
rifting of North America: Precambrian Research, v. 7, p. 61-98. , 1980b, The provenance of rutile: Journal of Sedimentary Petrology, v. 50,
Epier, N. E., Bolsover, L. R., and Lindsley, D. H., 1986, Nature and origin of the p. 485-488 (reprinted 1985, in Leupke, G., ed., Economic analysis of heavy
Sybille Fe-Ti oxide deposit, Laramie Anorthosite complex, SE Wyoming: minerals in sediments: New York, Van Nostrand Reinhold, p. 230-233).
Geological Society of America Abstracts with Programs, v. 18, p. 595. Force, E. R., and Carter, B. A., 1986, Liquid immiscibility proposed for nelsonitic
Erickson, R. L., and Blade, L. V., 1963, Geochemistry and petrology of the alkalic components of the anorthosite-syenite-gabbro complex, San Gabriel Moun-
igneous complex at Magnet Cove, Arkansas: U.S. Geological Survey Profes- tains, California: Geological Society of America Abstracts with Programs,
sional Paper 425, 95 p. v. 18, p. 604.
Ernst, W . G., 1976, Mineral chemistry of eclogites and related rocks from the Force, E. R., and Garnar, T. E., Jr., 1985, High angle aeolian crossbedding at
Voltri Group, western Liguria, Italy: Schweizerishe Mineralogishe and Pe- Trail Ridge, Florida: Industrial Minerals, August, p. 55-59.
trographishe Mitteilungen, v. 56, p. 293-343. Force, E. R., and Geraci, P. J., 1975, Heavy minerals in Pleistocene(?) shoreline
Ernst, W. G., Seki, Y., Onuki, H., and Gilbert, M. C., 1970, Comparative study of sand bodies of southeastern Virginia: U.S. Geological Survey Map MF-718,
low-grade metamoprhism in the California coast ranges and the outer meta- scale 1:250,000.
morphic belt of Japan: Geological Society of America Memoir 124, 276 p. Force, E. R., and Lynd, L. E., 1984, Titanium-mineral resources of the United
Eskola, P., 1921, O n the eclogites of Norway: Skrifter, Videnskapsselskapet i States; Definitions and documentation: U.S. Geological Survey Bulletin
Christiania, Matematisk-Naturvidenskabelig Klasse I, v. 8, 118 p. 1558-B, 11 p.
Espenshade, G. H„ and Potter, D. B., 1960, Kyanite, sillimanite, and andalusite Force, E. R., and Rich, F. J., 1989, Geologic evolution of Trail Ridge eolian
deposits of the southeastern states: U.S. Geological Survey Professional heavy mineral sand and underlying peat, northern Florida: U.S. Geological
Paper 336, 121 p. Survey Professional Paper 1499, 16 p.
Force, E. R., and Stone, B. D., 1990, Heavy mineral dispersal and deposition in tions in central and southern California: U.S. Bureau of Mines Report of
sandy deltas of glacial Lake Quinebaug, Connecticut: U.S. Geological Sur- Investigation 8471, 20 p.
vey Bulletin 1874, 21 p. Gottfried, D., Greenland, P. L., and Campbell, E. Y., 1968, Variation of Nb-Ta,
Force, E. R., Lipin, B. R., and Smith, R. E., 1976, Heavy mineral resources in Zr-Hf, Th-U, and K-Cs in two diabase-granophyre suites: Geochimica et
Pleistocene sand of the Port Leyden Quadrangle, southwestern Adirondack Cosmochimica Acta, v. 32, p. 925-947.
Mountains, New York: U.S. Geological Survey Map MF-728-B, scale Green, J. C., 1956, Geology of the Storkollen-Blankenburg area, Kragere, Nor-
1:24,000. way: Norsk Geologisk Tidsskrift, v. 36, p. 89-140.
Force, E. R., Grosz, A. E., Loferski, P. J., and Maybin, A. H„ 1982, Aeroradioac- Grey, I. E., and Reid, A. F., 1975, The structure of pseudorutile and its role in the
tivity maps in heavy-mineral exploration; Charleston, South Carolina: U.S. natural alteration of ilmenite: American Mineralogist, v. 60, p. 898-906.
Geological Survey Professional Paper 1218, 19 p. Griffin, W . L., 1987, On the eclogites of Norway—65 years later: Mineralogical
Force, E. R., Sukirno, D., and van Leeuwen, T., 1984, Exploration for porphyry Magazine, v. 51, p. 333-343.
metal deposits based on rutile distribution; A test in Sumatera: U.S. Geologi- Griffin, W . L., and Mork, M.B.E., 1981, Eclogites and basal gneisses in western
cal Survey Bulletin 1558-A, 9 p. (also Indonesia Directorate of Mineral Norway, Excursion Bl: Oslo, Geologisk Museum, 88 p.
Resources Special Publication, v. 1, p. 1-13). Griffin, W . L., and 8 others, 1985, High-pressure metamorphism in the Scandina-
Frakes, L. A., 1979, Climates through geologic time: Amsterdam, Elsevier, 310 p. vian Caledonides, in The Caledonide Orogen; Scandinavia and related areas:
Friis, H., 1974, Weathered heavy-mineral associations from the young Tertiary New York, Wiley, p. 783-801.
deposits of Jutland, Denmark: Sedimentary Geology, v. 12, p. 199-213. Grigg, N. S., and Rathbun, R. E., 1969, Hydraulic equivalence of minerals with a
Frost, M. T., Grey, I. E., Harrowfield, I. R., and Mason, K., 1983, The de- consideration of the reentrainment process: U.S. Geological Survey Profes-
pendence of alumina and silica contents on the extent of alteration of sional Paper 650-B, p. B77-B80.
weathered ilmenites from Western Australia: Mineralogical Magazine, v. 47, Grobler, N. J., and Whitfield, G. G., 1970, The olivine-apatite magnetites and
p. 201-208. selected rocks in the Villa Nora occurrence of the Bushveld Igneous Com-
Fryklund, V. C., and Holbrook, D. F., 1950, Titanium ore deposits of the Magnet plex: Geological Society of South Africa Special Publication 1, p. 208-227.
Cove area, Hot Springs County, Arkansas: Arkansas Division of Geology Gross, E. B., and Parwell, A., 1969, Rutile mineralization at the White Mountain
Bulletin 16, 173 p. andalusite deposits, California: Arkiv Mineralogi och Geologi, v. 4, paper 29,
Galloway, M. C., 1972, Statistical analyses of regional heavy mineral variation, p. 493-497.
Hawkesbury sandstone and Narrabeen Group (Triassic), Sydney Basin: Geo- Gross, S. O., 1968, Titaniferous ores of the Sanford Lake district, New York, in
logical Society of Australia Journal, v. 19, p. 65-76. Ridge, J. D., ed., Ore deposits of the United States, 1933-1967, v. 1: New
Gardner, D. E., 1955, Beach-sand heavy-mineral deposits of eastern Australia: York, American Institute of Mining, Metallurgical, and Petroleum Engi-
Bureau of Mineral Resources Bulletin 28, 103 p. neers, p. 140-153.
Garnar, T. E., 1972, Economic geology of Florida heavy-mineral deposits: Florida Grosz, A. E., 1983, Application of total-count aeroradiometric maps to the explo-
Bureau of Geology Special Publication 17, p. 17-21. ration for heavy mineral deposits in the coastal plain of Virginia: U.S. Geo-
, 1978, Heavy mineral mining in Florida [abs.]: American Institute of Min- logical Survey Professional Paper 1263, 20 p.
ing, Metallurgical, and Petroleum Engineers Technical Program, p. 19. , 1987, Nature and distribution of potential heavy-mineral resources off-
, 1980, Heavy minerals industry of North America, in Coope, B. M., ed., shore of the Atlantic coast of the United States: Marine Mining, v. 6,
Proceedings of the 4th Industrial Minerals International Congress: London, p. 339-357.
Industrial Minerals, p. 29-42. Grosz, A. E., Hathaway, J. C., and Escowitz, E. C., 1986, Placer deposits of heavy
Giejer, P., 1964, Genetic relationships of the paragenesis Al2Si05-lazulite-rutile: minerals in Atlantic continental shelf sediments: Offshore technology confer-
Arkiv Mineralogi och Geologi, v. 3, p. 423-466. ence, 18th, Houston, Preprint OTC 5196, p. 387-392.
Geisel-Sobrinho, E., 1974, Prospecgao de uranio na chamine alcalina de Tapira, Grout, F. F., 1949-1950, The titaniferous magnetites of Minnesota: St. Paul,
Minas Gerais: Comissao Nacional de Energia Nuclear (Brazil) Boletim 10, Office of the Commissioner of the Iron Range Resources and Rehabilitation,
16 p. 117 p.
Ghent, E. D., and Stout, M. Z., 1984, T i O j activity in metamorphosed pelitic and Groves, A. W., 1931, The unroofing of the Dartmoor Granite and the distribution
basic rocks; Principles and applications to metamorphism in southeastern of its detritus in the sediments of southern England: Quarterly Journal of the
British Columbia: Contributions to Mineralogy and Petrology, v. 86, Geological Society, v. 87, p. 62-%.
p. 248-255. Guidotti, C. V., Cheney, J. T., and Guggenheim, S., 1977, Distribution of
Gierth, E., and Krause, H., 1973, Die Ilmenitlagerstatte Tellnes (sud-Norwegen): titanium between coexisting muscovite and biotite in pelitic schists from
Norsk Geologiske Tidsskrift, v. 53, p. 359-402. northwestern Maine: American Mineralogist, v. 62, p. 438-448.
, 1974, Contribution to the mineralogy of Norway, no. 57: Baddelyit von Guild, P. W., 1971, Metallogeny; A key to exploration: Mining Engineering, v. 23,
Tellnes: Norsk Geologiske Tidsskrift, v. 54, p. 193-197. p. 69-72.
Gillson, J. L., 1959, Sand deposits of titanium minerals: Mining Engineering, Gustavson, T. C., Ashley, G. M., and Boothroyd, J. C., 1975, Depositional
v. 11, p. 421-429. sequences in glaciolacustrine deltas, in Jopling, A. V., and McDonald, B. C.,
Goldberg, S. A., 1984, Geochemical relationships between anorthosite and asso- eds., Glaciofluvial and glaciolacustrine sedimentation: Society of Economic
ciated iron-rich rocks, Laramie Range, Wyoming: Contributions to Mineral- Paleontologists and Mineralogists Special Publication 23, p. 264-280.
ogy and Petrology, v. 87, p. 376-387. Haggerty, S. E., 1976a, Opaque mineral oxides in terrestrial igneous rocks, in
Goldich, S. S., 1938, A study in rock weathering: Journal of Geology, v. 46, Rumble, D., Ill, ed., Oxide minerals: Mineralogical Society of America
p. 17-58. Short Course Notes, v. 3, p. Hg 101-300.
Goldsmith, R., and Force, E. R., 1978, Distribution of rutile in metamorphic , 1976b, Oxidation of opaque mineral oxides in basalts, in Rumble, D., Ill,
rocks and implications for placer deposits: Mineralium Deposita, v. 13, ed., Oxide minerals: Mineralogical Society of America Short Course Notes,
p. 329-343. v. 3, p. Hg 1-100.
Gomes, J. M., Martinez, G. M., and Wong, M. M., 1979, Recovering by-product Hagner, A. F., 1968, The titaniferous magnetite deposit at Iron Mountain, Wyom-
heavy minerals from sand and gravel, placer gold, and industrial mineral ing, in Ridge, J. D., ed., Ore deposits of the United States, 1933-1967, v. 1:
operations: U.S. Bureau of Mines Report of Investigation 8366, 15 p. New York, American Institute of Mining, Metallurgical, and Petroleum
, 1980, Recovery of by-product heavy minerals from sand and gravel opera- Engineers, p. 666-680.
106 E. R. Force
Hahn, A. D „ and Fine, A. D., 1960, Examination of ilmenite-bearing sands in the western interior: U.S. Geological Survey Professional Paper 994 A, 29 p.
Otter Creek valley, Kiowa and Tillman Counties, Oklahoma: U.S. Bureau of Howard University Geology Field Camp, 1974, Geologic map of the Port Leyden
Mines Report of Investigation 5577, 77 p. Quadrangle: New York State Geological Survey Open-File Report, scale
Hails, J. R., 1969, The nature and occurrence of heavy minerals in three coastal 1:24,000.
areas of New South Wales: Journal of the Royal Society of New South Hubert, J. F., 1962, A zircon-tourmaline-rutile maturity index and the interde-
Wales, v. 102, p. 21-39. pendence of the composition of heavy mineral assemblages with the gross
Hails, J. R., and Hoyt, J. H., 1972, The nature and occurrence of heavy minerals composition and texture of sandstones: Journal of Sedimentary Petrology, v.
in Pleistocene and Holocene sediments of the lower Georgia coastal plain: 32, p. 440-450.
Journal of Sedimentary Petrology, v. 42, p. 646-666. Hudson, J. P., 1986, Evidence for littoral drift at Holocene low sea levels on the
Hammarbeck, E.C.I., 1976, Titanium, in Coetzee, C. B., ed., Mineral resources of south Sydney inner continental shelf, southeastern Australia: 12th Interna-
the Republic of South Africa: South African Geological Survey Handbook 7, tional Sedimentological Congress, Canberra, Abstracts, p. 146.
p. 221-226. Hunt, J. A., and Kerrick, D. M., 1977, The stability of sphene; Experimental
Hammond, P., 1952, Allard Lake ilmenite deposits: Economic Geology, v. 47, redetermination and geologic implications: Geochimica et Cosmochimica
p. 634-649. Acta, v. 41, p. 279-288.
Hand, B. M., 1967, Differentiation of beach and dune sands, using settling veloci- Hunter, R. E., 1968, Heavy minerals of the Cretaceous and Tertiary sands of
ties of light and heavy minerals: Journal of Sedimentary Petrology, v. 27, extreme southern Illinois: Illinois Geological Survey Circular 428, 22 p.
p. 514-520. Hunter, R. E., and Richmond, B. M., 1988, Daily cycles in coastal dunes: Sedi-
Harben, P., 1984, Titanium minerals in Brazil; Progress and potential: Industrial mentary Geology, v. 55, p. 43-67.
Minerals, January, p. 45-49. Industrial Minerals, 1972, By-product minerals from tin mining: June, p. 35.
Hargraves, R. B., 1962, Petrology of the Allard Lake anorthosite suite, in Engel, — , 1978, Richards Bay '78, Tapira '88: March, p. 7.
A.E.J., and others, eds., Petrologic studies: Geological Society of America , 1986, New heavy minerals supplier: May, p. 14.
Buddington Volume, p. 163-189. Ishihara, S., 1977, The magnetite-series and ilmenite-series granitic rocks: Mining
Hartley, M. E., 1971, Ultramafic and related rocks in the vicinity of Lake Geology, v. 27, p. 293-305.
Chatuge: Georgia Geological Survey Bulletin 85, 61 p. Itaya, T., and Banno, S., 1980, Paragenesis of titanium-bearing accessories in
Hartman, J. A., 1959, The titanium mineralogy of certain bauxites and their pelitic schists of the Sanbagawa metamorphic belt, central Shikoku, Japan:
parent materials: Economic Geology, v. 54, p. 1380-1405. Contributions to Mineralogy and Petrology, v. 73, p. 267-276.
Hedlund, D. C., and Olson, J. C., 1975, Geologic map of the Powderhorn Itaya, T., and Otsuki, M., 1978, Stability and paragenesis of Fe-Ti oxide minerals
Quadrangle, Gunnison and Saguache Counties, Colorado: U.S. Geological and sphene in the basis schists of the Sanbagawa metamorphic belt in central
Survey Map GQ-1178, scale 1:24,000. Shikoku, Japan: Japanese Association of Mineralogists, Penologists, and
Heinrich, E. W., 1966, The geology of carbonatites: Chicago, Rand McNally, Economic Geologists Journal, v. 73, p. 359-379.
555 p. Iwasaki, I., Smith, K. A., and Maliesi, A. S., 1982, By-product recovery from
Hershey, R. E., 1966, Mineral resources summary of the Seventeen Creek Quad- copper-nickel bearing Duluth Gabbro: Minneapolis, University of Minnesota
rangle, Tennessee: Tennessee Division of Geology Mineral Resources Sum- Mineral Resources Research Center, 283 p.
mary 21-NW (and map), 17 p. Jackson, M. L., and Sherman, G. D., 1953, Chemical weathering of minerals in
, 1968, Mineral resources summary of the Chesterfield Quadrangle, Tennes- soils, in Norman, A. G., ed., Advances in agronomy, v. 5: New York,
see: Tennessee Division of Geology Mineral Resources Summary 11-NE Academic Press, p. 217-319.
(and map), 24 p. Jacob, K., 1956, Ilmenite and garnet sands of the Chowghat, Tinnevelly, Ram-
Herz, N., 1969, The Roseland alkalic anorthosite massif, Virginia, in Isachsen, nad, and Tanjore coasts: India Geological Survey Records, v. 82,
Y. W., ed., Origin of anorthosite and related rocks: New York State Museum p. 567-602.
and Science Service Memoir 18, p. 357-367. Jones, H. A., and Davies, P. J., 1979, Preliminary studies of offshore placer de-
, 1976, Titanium deposits in alkalic igneous rocks: U.S. Geological Survey posits, eastern Australia: Marine Geology, v. 30, p. 243-268.
Professional Paper 959-E, 6 p. Karkhanavala, M. D., and Momin, A. C., 1959, The alteration of ilmenite:
Herz, N., and Force, E. R„ 1987, Geology and mineral deposits of the Roseland Economic Geology, v. 54, p. 1095-1102.
district of central Virginia: U.S. Geological Survey Professional Paper 1371, Karkhanavala, M. D., Momin, A. C., and Rege, S. G., 1959, An X-ray study of
56 p. leucoxene from Quilon, India: Economic Geology, v. 54, p. 913-918.
Herz, N., and Valentine, E. B., 1970, Rutile in the Harford County, Maryland, Kildal, E. S., 1970, Geologisk karte over Norge, berggrunnskart, Maloy sheet:
serpentinite belt: U.S. Geological Survey Professional Paper 700-C, p. 43-48. Norges Geologiske Undersokelse (Norway), scale 1:250,000.
Higgs, D. V., 1954, Anorthosite and related rocks of the western San Gabriel Kinney, D. M., 1949, The Magnet Cove Rutile Company mine, Hot Springs
Mountains, southern California: University of California Publications in County, Arkansas: U.S. Geological Survey Open-File Report 19, 12 p.
Geological Sciences, v. 30, p. 171-222. Klugman, M. A., 1966, Resume of the geology of the Laramie Anorthosite mass:
Hocking, R..M., Warren, J. K., and Baxter, J. L., 1982, Shoreline geomorphol- Mountain Geologist, v. 3, p. 75-84.
ogy, in Exploitation of mineral sands: Perth, Western Australia Institute of Kolker, A., 1982, Mineralogy and geochemistry of Fe-Ti oxide and apatite (nel-
Technology, p. 81-92. sonite) deposits and evaluation of the liquid immiscibility hypothesis: Eco-
Hocq, M., 1982, Region du lac Allard: (Quebec) Ministere de Energie et des nomic Geology, v. 77, p. 1146-1158.
Ressources Rapport DPV-894, 99 p. Komar, P. D., and Wang, C., 1984, Processes of selective grain transport and the
Holbrook, D. F., 1948, Molybdenum in Magnet Cove, Arkansas: Arkansas formation of placers on beaches: Journal of Geology, v. 92, p. 637-655.
Division of Geology Bulletin 12, 16 p. Koraeliussen, A., 1980, Rutil: eklogittiske bergarter: Sunnfjord: Norges Geolo-
Hosterman, J. W., Scheid, V. E., Allen, V. T., and Sohn, I. G., 1960, Investiga- giske Undersokelse Rapport 1717/5, 29 p.
tions of some clay deposits in Washington and Idaho: U.S. Geological Sur- Korneliussen, A., and Foslie, G., 1986, Rutile-bearing eclogites in the Sunnfjord
vey Bulletin 1091, 147 p. region of western Norway: Norges Geologiske Undersokelse Bulletin 402,
Houston, R. S., and Murphy, J. F., 1962, Titaniferous black sandstone deposits of p. 65-71.
Wyoming: Geological Society of Wyoming Bulletin 49, 120 p. Korneliussen, A., Geis, H.-P., Gierth, E., Krause, H., Robins, B., and Schott, W.,
, 1977, Depositional environment of Upper Cretaceous black sandstones of 1986, Titanium ores; An introduction to a review of titaniferous magnetite,
ilmenite, and rutile deposits in Norway: Norges Geologiske Undersokelse American Institute of Mining, Metallurgical, and Petroleum Engineers
Bulletin 402, p. 7-24. Transactions, v. 217, p. 311-318.
Krause, H., and Pape, H., 1975, Mikroskopische Untersuchungen der mineral- , 1985, Titanium, in Mineral facts and problems, 1985, ed.: U.S. Bureau of
vergesellschaftung in erz und nebengestein der ilmenitlagerstâtte Storgangen Mines Bulletin 675, p. 859-880.
(süd-Norwegen): Norsk Geologisk Tidsskrift, v. 55, p. 387-422. , 1988, "Ilmenite," "Rutile," and "Titanium and titanium dioxide," in
, 1977, Untersuchungen zum geologischen und petrographischen aufbau des Mineral commodity summaries 1988: U.S. Bureau of Mines, p. 72-73,
Storgangen-ilmenitskörpers und seiner nebengesteinseinheiten (süd-Nor- 132-133, 170-171.
wegen): Norsk Geologisk Tidsskrift, v. 57, p. 263-284. Lynd, L. E„ Sigurdson, H., North, C. H., and Anderson, W. W., 1954, Character-
Krause, H., and Zeino-Mahmalat, R., 1970, Untersuchungen an erz und neben- istics of titaniferous concentrates: American Institute of Mining, Metallurgi-
gesteinen der grübe Blâfjell in SW-Norwegen: Norsk Geologisk Tidsskrift, cal, and Petroleum Engineers Transactions, v. 6, p. 817-824.
v. 50, p. 45-88. Macdonald, E. H., 1971a, Country report; Indonesia: United Nations Economic
Krause, H., Gierth, E., and Schott, W., 1986, Fe-Ti deposits in the South Roga- Commission for Asia and Far East Technical Bulletin, v. 5, p. 48-53.
land igneous complex, with special reference to the Âna-Sira anorthosite , 1971b, Country report; West Mayalsia: United Nations Economic Com-
massif: Norges Geologiske Undersokelse Bulletin 402, p. 25-38. mission for Asia and Far East Technical Bulletin, v. 5, p. 74-78.
Krogh, E. J., 1980, Geochemistry and petrology of glaucophane-bearing eclogites , 1983, Alluvial mining: London, Chapman and Hall, 508 p.
and associated rocks from Sunnfjord, western Norway: Lithos, v. 13, Mader, D., 1980, Authigener rutil im Buntsandstein des Westeifel: Neues Jah-
p. 355-380. rbuch fur Mineralogie Monatshefte, no. 3, p. 97-108.
Kudrass, H.-R., 1982, Cores of Holocene and Pleistocene sediments from the east Mallik, T. K., 1986, Micromorphology of some placer minerals from Kerala
Australian continental shelf (SO-15 cruise 1980): Geologische Jahrbuch, beach, India: Marine Geology, v. 71, p. 371-381.
v. D 56, p. 137-163. Mallik, T. K., Vasudevan, V., Verghese, P. A., and Machado, T., 1987, The black
Lang, H. D., 1970, Sekundare rutil-lagerstatten in Sierra Leone (West-Africa): sand placer deposits of Kerala Beach, southwest India: Marine Geology,
Erzmetall, v. 23, p. 179-183. v. 77, p. 129-150.
Large, P. R.,1972, Metasomatism and scheelite mineralization at Bold Head, King Mancini, A., Mancini, R., and Martinotti, G., 1979, Valorization of new titanium
Island: Australasian Institute of Mining and Metallurgy Proceedings no. 238, resource; Titaniferous eclogites: Proceedings IV-15, 10th World Mining
p. 31-45. Congress, Istanbul, 19 p.
La Roche, H., de Kern, M „ and Bolfa, J., 1962, Contribution a l'etude de Markewicz, F. J., 1969, Ilmenite deposits of New Jersey coastal plain, in Sub-
l'altération des ilmenites: Sciences de la Terre, v. 8, p. 215-248. itzky, S., ed., Geology of selected areas in New Jersey and eastern Pennsyl-
Larsen, E. S., Jr., 1942, Alkalic rocks of Iron Hill, Gunnison County, Colorado: vania: New Brunswick, New Jersey, Rutgers University Press, p. 363-382.
U.S. Geological Survey Professional Paper 197-A, 64 p. Marsh, S. P., 1979, Rutile mineralization in the White Mountain andalusite
Lawrence, L. J., and Savage, E. N., 1975, Mineralogy of the titaniferous porphyry deposit, California: U.S. Geological Survey Open-File Report 79-1622, 7 p.
copper deposts of Melanesia: Australasian Institute of Mining and Metal- Marsh, S. P., and Sheridan, D. M., 1976, Rutile in Precambrian sillimanite-quartz
lurgy Proceedings, no. 256, p. 1-14. gneiss and related rocks, east-central Front Range, Colorado: U.S. Geologi-
Layton, W., 1966, Prospects of offshore mineral deposits on the eastern seaboard cal Survey Professional Paper 959-G, 17 p.
of Australia: Mining Magazine, v. 115, p. 344-351. Martens, J.H.C., 1935, Beach sands between Charleston, South Carolina,
Lee, D. E., and Dodge, F.C.W., 1964, Accessory minerals in some granitic rocks and Miami, Florida: Geological Society of America Bulletin, v. 46,
in California and Nevada as a function of calcium content: American Miner- p. 1563-1596.
alogist, v. 79, p. 1660. Martinis, B., and Pasquare, G., 1971, Carta Geologica d'ltalia, sheet 82, Genova:
Lindberg, P. A., 1986, Fe-Ti-P mineralizations in the larvikite-lardalite complex, Servicio Geologico d'ltalia, scale 1:100,000.
Oslo Rift: Norges Geologiske Undersokelse Bulletin 402, p. 93-98. Mathis, J. M., and Sclar, C. B., 1980, The oxidation and titanium-enrichment
Lissiman, J. C„ and Oxenford, R. J., 1973, The Allied Minerals N. L. heavy mechanism of "altered ilmenite" grains in the Tertiary Kirkwood and Co-
minerals sand deposit at Eneabba, Western Australia: Australasian Institute hansey Formations of New Jersey: Geological Society of America Abstracts
of Mining and Metallurgy (Parkville, Victoria), Western Australia confer- with Programs, v. 12, p. 72.
ence, preprint. Maud, R. R., 1968, Quaternary geomorphology and soil formation in coastal
, 1975, Eneabba rutile-zircon-ilmenite sand deposits, W . A., in Knight, Natal: Zeitschrift Geomorphologie Supplementband 7, p. 155-199.
C. L., ed., Economic geology of Australia and Papua New Guinea: Aus- McCartan, L., Weems, R. E., and Lemon, E. M., 1990, Quaternary stratigraphy
traliasian Institute of Mining and Metallurgy Monograph 5, p. 1062-1069. in the vicinity of Charleston, South Carolina, and its relationship to local
Lister, G. F., 1966, The composition and origin of selected iron-titanium deposits: seismicity and tectonics: U.S. Geological Survey Professional Paper 1367-A
Economic Geology, v. 61, p. 275-310. (in press).
Llewellyn, T. O., and Sullivan, G. V., 1980, Recovery of rutile from a porphyry McCauley, C. K., 1960, Exploration for heavy minerals on Hilton Head Island,
copper tailings sample: U.S. Bureau of Mines Report of Investigation 8462, South Carolina: South Carolina Division of Geology Bulletin 26, 13 p.
18 p. McElroy, C. T., 1962, The geology of the Clarence-Moreton Basin: Geological
Loughman, F. C., 1969, Chemical weathering of the silicate minerals: New York, Survey of New South Wales Geological Memoir 9, p. 1-172.
American Elsevier, 154 p. Mclntyre, A., and 7 others, 1976, Glacial North Atlantic 18,000 years ago; A
Lowrie, D. C., Horwitz, R. C., and Gemüts, I., 1967, Busselton and Augusta C L I M A P reconstruction, in Clise, R. M „ and Hays, J. P., eds., C L I M A P
sheets: Geological Survey of Western Australia, scale 1:250,000. investigation of late Quaternary paleoceanography and paleoclimatology:
Lowrie, D . C „ Low, G. H., Playford, P. E., and Dedman, R „ 1973, Dongara and Geological Society of America Memoir 145, p. 43-76.
Hill River sheets: Geological Survey of Western Australia, scale 1:250,000. Mclntyre, D. D., 1959, The hydraulic equivalence and size distributions of some
Lowrie, D. C., Wilde, S. A., and Walker, I. W., 1983, Collie sheet: Geological mineral grains from a beach: Journal of Geology, v. 67, p. 278-301.
Survey of Western Australia, scale 1:250,000. McKellar, J. B., 1975, The eastern Australia rutile province, in Knight, C. L., ed.,
Lowright, R., Williams, E. G., and Dachille, F., 1972, An analysis of factors Economic geology of Australia and Papua New Guinea: Australasian Insti-
controlling deviations in the hydraulic equivalence in some modern sands: tute of Mining and Metallurgy Monograph 5, p. 1055-1061.
Journal of Sedimentary Petrology, v. 42, p. 635-645. McLaughlin, R.J.W., 1955, Geochemical changes due to weathering under vary-
Lynd, L. E., 1980, Study of the mechanism and rate of ilmenite weathering: ing climatic conditions: Geochimica et Cosmochimica Acta, v. 8, p. 109-130.
108 E. R. Force
Meilke, H., and Schreyer, W., 1972, Magnetite-rutile assemblages in metapelites Newhouse, W . H., and Hagner, A. F., 1951, Preliminary report on the titaniferous
of the Fichtelgebirge, Germany: Earth and Planetary Science Letters, v. 16, iron deposits of the Laramie Range, Wyoming: U.S. Geological Survey
p. 423-428. Open-File Report, 17 p.
Melville, G., 1984, Headlands and offshore islands as dominant controlling factors , 1957, Geologic map of anorthosite areas, southern part of Laramie Range,
during late Quaternary barrier formation in the Forster-Tuncurry area, Wyoming: U.S. Geological Survey Map MF-119, scale 1:63,360.
New South Wales, Australia: Sedimentary Geology, v. 39, p. 243-271. Nielsen, N.-A., 1972, The ilmenomagnetite-apatite deposit at Kodal, S. Norway:
Mertie, J. B„ Jr., 1979, Monazite in the granitic rocks of the southeastern Atlantic Industrial Minerals, April, p. 35-37.
States; An example of the use of heavy minerals in geologic exploration: U.S. Oakeshott, G. B., 1958, Geology and mineral deposits of San Fernando Quadran-
Geological Survey Professional Paper 1094, 79 p. gle, Los Angeles County, California: California Division of Mines Bulletin
Miller, R. L., and Zeigler, J. M., 1958, A model relating sediment pattern in the 172, 147 p.
region of shoaling waves, breaker zone, and foreshore: Journal of Geology, Oaks, R. Q., Coch, N. K., Sanders, J. E., and Hint, R. F., 1974, Post-Miocene
v. 66, p. 417-441. shorelines and sea levels, southeastern Virginia, in Oaks, R. Q., and Du Bar,
Minard, J. P., Force, E. R., and Hayes, G. W., 1976, Alluvial ilmenite placer J. R., eds., Post-Miocene stratigraphy, central and southern Atlantic coastal
deposits, central Virginia: U.S. Geological Survey Professional Paper 959-H, plain: Logan, Utah State University, p. 53-87.
15 p. Oeschger, H.,1985, The contribution of ice core studies to the understanding of
Mineragao Metalurgia, 1977, Projecto titanio: v. 41, p. 26-29. environmental processes: American Geophysical Union Geophysical Mono-
Mitchell, R. H., 1986, Kimberlites: New York, Plenum Press, 442 p. graph 33, p. 9-17.
Mohr, D. W., and Newton, R. C., 1983, Kyanite-staurolite metamorphism in Oftedahl, C., 1960, Permian rocks and structures of the Oslo region: Norges
sulfidic schists of the Anakeesta Formation, Great Smoky Mountains, North Geologiske Undersokelse (Bulletin) 208, p. 298-343.
Carolina: American Journal of Science, v. 283, p. 97-134. Opdike, N. D., Spangler, D. P., Smith, D. L., Jones, D. S„ and Lindquist, R. C.,
Moore, W . J., and Czamanske, G. K., 1973, Compositions of biotites from 1984, Origin of the epeirogenic uplift of Pliocene-Pleistocene beach ridges in
unaltered and altered monzonitic rocks in the Bingham mining district, Utah: Florida, and development of the Florida karst: Geology, v. 12, p. 226-228.
Economic Geology, v. 68, p. 269-274. Ortega-Gutierez, F., 1981, Metamorphic belts of southern Mexico and their tec-
Morad, S., 1986, SEM study of authigenic rutile, anatase, and brookite in Pro- tonic significance: Geofisica International, v. 20-3, p. 177-202.
terozoic sandstones from Sweden: Sedimentary Geology, v. 46, p. 77-89. Overstreet, W . C., Yates, R. G., and Griffitts, W. R „ 1963, Heavy minerals in the
Morad, S., and Aldahan, A. A., 1985, Leucoxene-calcite-quartz aggregates in saprolite of the crystalline rocks in the Shelby Quadrangle, North Carolina:
sandstones and the relation to decomposition of sphene: Neues Jahrbook fur U.S. Geological Survey Bulletin 1162-F, 31 p.
Mineralogie Monatshefte no. 10, p. 458-468. Owens, J. P., 1985, Sediment dispersal pattern in the emerged Atlantic coastal
, 1986, Alteration of detrital Fe-Ti oxides in sedimentary rocks: Geological plain (New Jersey-Georgia) from Early Cretaceous to Holocene; A guide to
Society of America Bulletin, v. 97, p. 567-578. potential placer deposits [abs.]: U.S. Geological Survey Circular 949,
Morley, I. W., 1981, Black sands; A history of the mineral sand mining industry in p. 39-40.
eastern Australia: Brisbane, University of Queensland Press, 278 p. Pasteels, P., Demaiffe, D., and Michot, J., 1979, U-Pb and Rb-Sr geochemistry of
Morse, S. A., 1982, A partisan review of Proterozoic anorthosites: American the eastern part of the south Rogaland igneous complex, southern Norway:
Mineralogist, v. 67, p. 1087-1100. Lithos, v. 12, p. 199-208.
Mukeijee, N. K., 1964, A study of the monazite-bearing ilmenite deposit from Pasteris, J. D., 1980, The significance of groundmass ilmenite and megacryst
Bothaville, Orange Free State, Union of South Africa: New Delhi, 22nd ilmenite in kimberlites: Contributions to Mineralogy and Petrology, v. 75,
International Geological Congress Report, sec. 16, p. 481-503. p. 315-325.
Nafziger, R. H., and Elger, G. W., 1987, Preparation of titanium feedstock from , 1985, Relationships between temperature and oxygen fugacity among
Minnesota ilmenite by smelting and sulfation-leaching: U.S. Bureau of Mines Fe-Ti oxides in two regions of the Duluth Complex: Canadian Mineralogist,
Report of Investigation 9065, 13 p. v. 23, p. 111-127.
Naldrett, A. J., 1979, Partitioning of Fe, Co, Ni, and Cu between sulfide liquid Patterson, S. H., Kurtz, H. F„ Olson, J. C., and Neeley, C. L., 1986, World
and basaltic melts, and the composition of Ni-Cu sulfide deposits; A reply bauxite resources: U.S. Geological Survey Professional Paper 1076-B, 151 p.
and further discussion: Economic Geology, v. 74, p. 1520-1528. Paulson, E. G., 1964, Mineralogy and origin of the titaniferous deposit at Pluma
Narayanaswami, S., and Mahadevan, T. M., 1964, Geology of Trivandrum and Hidalgo, Oaxaca, Mexico: Economic Geology, v. 59, p. 753-767.
Madras areas: 22nd International Geological Congress, New Delhi, Guide to Peers, R., 1975a, Leeuwin block, in Geology of Western Australia: Geological
excursion C-23, 13 p. Survey of Western Australia Memoir 2, p. 102-104.
Nathan, H. D., 1969, The geology of a portion of the Duluth Complex, Cook , 1975b, Northampton block, in Geology of Western Australia: Geological
County [Ph.D. thesis]: Minneapolis, University of Minnesota, 198 p. Survey of Western Australia Memoir 2, p. 104-106.
National Materials Advisory Board, 1983, Titanium; Past, present, and future: Petersen, J. S., 1978, Structure of the larvikite-lardalite complex, Oslo region,
U.S. National Academy Press Publication NMAB-392, 209 p. Norway, and its evolution: Geologische Rundschau, v. 67, p. 330-342.
Nedelcu, L., 1986, The significance of titanium minerals in some crystalline schists Peterson, C. D., Komar, P. D., and Scheidegger, K. F.,1985, Distribution, geome-
from Romania, in Mineral parageneses: Athens, Theophrastus, p. 623-645. try, and origin of heavy mineral placer deposits on Oregon beaches: Journal
Neiheisel, J., 1958, Origin of the dune system on the Isle of Palms, S.C.: South of Sedimentary Petrology, v. 56, p. 67-77.
Carolina Division of Geology Mineral Industries Laboratory Bulletin, v. 2, Pettijohn, F. J., 1957, Sedimentary rocks: New York, Harper and Row, 718 p.
p. 46-51. Philpotts, A. R., 1967, Origin of certain iron-titanium oxide and apatite rocks:
, 1962, Heavy-mineral investigations of Recent and Pleistocene sands of the Economic Geology, v. 62, p. 303-315.
lower coastal plain of Georgia: Geological Society of America Bulletin, v. 73, Pinnell, D. B., and Marsh, J. A., 1954, Summary geological report of the titanif-
p. 365-374. erous iron deposits of the Laramie Range, Albany County: Wyoming Mines
, 1976, Heavy minerals in aeroradioactive high areas of the Savannah River Magazine, v. 44, 30 p.
flood plain and deltaic plain: South Carolina Division of Geology Geologic Pirkle, E. C., and Yoho, W . H„ 1970, The heavy mineral ore body of Trail Ridge,
Notes, v. 20, p. 45-51. Florida: Economic Geology, v. 65, p. 17-30.
Nesbitt, B. E., and Kelly, W . C., 1980, Metamorphic zonation of sulfides, oxides, Pirkle, E. C., Yoho, W . H., and Hendry, C. W., 1970, Ancient sea level stands in
and graphite in and around the orebodies at Ducktown, Tennessee: Eco- Florida: Florida Bureau of Geology Geological Bulletin 52, 61 p.
nomic Geology, v. 75, p. 1010-1021. , 1971, North Florida setting for 15th field conference, Southeastern Geo-
logical society, in Geological review of some north Florida mineral re- regional metamorphism in south-central Massachusetts: Connecticut Geolog-
sources: Tallahassee, Southeastern Geological Society, p. 1-15. ical and Natural History Survey Guidebook 5, p. 289-340.
Pirkle, E. C., Pirkle, W . A., and Yoho, W . H„ 1974, The Green Cove Springs and Robson, D. F., and Sampath, N., 1977, Geophysical response of heavy-mineral
Boulougne heavy-mineral sand deposits of Florida: Economic Geology, sand deposits at Jerusalem Creek, New South Wales: (Australia) Bureau of
v. 67, p. 1129-1137. Mineral Resources Journal of Geology and Geophysics, v. 2, p. 149-154.
, 1977, The Highland heavy-mineral sand deposit on Trail Ridge in north- Roehler, H. W., 1989, Origin and distribution of six heavy mineral placer deposits
ern peninsular Florida: Florida Bureau of Geology Report of Investigation in coastal-marine sandstones in the upper Cretaceous McCourt Sandstone
84, 50 p. tongue of the Rock Springs Formation, southwest Wyoming: U.S. Geologi-
Pirkle, E. C„ Pirkle, F. L., Pirkle, W . A., and Stayert, P. R., 1984, The Yulee cal Survey Bulletin 1867, 34 p.
heavy mineral sand deposits of northeastern Florida: Economic Geology, Romey, W . D., 1968, An evaluation of some differences between anorthosites in
v. 79, p. 725-737. massifs and in layered complexes: Lithos, v. 1, p. 230-241.
Pirkle, F. L., 1975, Evaluation of possible source regions of Trail Ridge sands: Rose, C. K., and Shannon, S. S., Jr., 1960, Cebolla Creek titaniferous iron
Southeastern Geology, v. 17, p. 93-114. deposits, Gunnison County, Colorado: U.S. Bureau of Mines Report of
Pirkle, F. L., and Czel, L. J., 1983, Marine fossils from region of Trail Ridge, a Investigation 5679, 30 p.
Georgia-Florida landform: Southeastern Geology, v. 24, p. 31-38. Rose, E. R., 1969, Geology of titanium and titaniferous deposits of Canada:
Poulose, K. V., 1972, Heavy mineral and glass sand deposits of Kerala coast: Canada Geological Survey Economic Geology Report 25, 177 p.
Indian Minerals, v. 26, p. 118-124. Ross, C. S., 1941, Occurrence and origin of the titanium deposits of Nelson and
Prakash, T. N., and Verghese, P. A., 1987, Seasonal beach changes along Quilon Amherst Counties, Virginia: U.S. Geological Survey Professional Paper 198,
district coast: Journal of the Geological Society of India, v. 29, p. 390-398. 59 p.
Pryor, W . A., 1960, Cretaceous sedimentation in upper Mississippi embayment: Roy, P. S., 1977, Does the Hunter River supply sand to the New South Wales
American Association of Petroleum Geologists Bulletin, v. 44, p. 1473-1504. coast today? Journal of the Royal Society of New South Wales, v. 110,
Puffer, J. H., and Cousminer, H. L., 1982, Factors controlling the accumulation of p. 17-24.
titanium-iron oxide-rich sands in the Cohansey Formation, Lakehurst area, — — , 1982, Regional geology of the central and northern New South Wales
New Jersey: Economic Geology, v. 77, p. 379-391. coast: Geologische Jarhbuch, v. D 56, p. 25-35.
Purucker, M., 1983, Time of formation of soft iron ore on the Gunflint and Roy, P. S., and Crawford, E. A., 1977, Significance of sediment distribution in
Mesabi ranges: Economic Geology, v. 78, p. 502-506. major coastal rivers, northern New South Wales: 3rd Australian Conference
Quirk, R., and Eilertson, N. A., 1963, Methods and costs of exploration and pilot on Coastal and Ocean Engineering, Institution of Engineers, Melbourne,
plant testing of ilmenite-bearing sands, Lakehurst mine, The Glidden Co., p. 177-184.
Ocean County, N.J.: U.S. Bureau of Mines Information Circular 8197,64 p. Roy, P. S., and Thom, B. G., 1981, Late Quaternary marine deposition in New
Rachele, L. D., 1976, Palynology of the Legier Lignite; A deposit in the Tertiary South Wales and southern Queensland; An evolutionary model: Geological
Cohansey Formation of New Jersey, U.S.A.: Reviews in Paleobotany and Society of Australia Journal, v. 28, p. 471-489.
Palynology, v. 22, p. 225-252. Rubey, W . W., 1933, The size distribution of heavy minerals within a waterlaid
Raedeke, L. D., 1982, Mineralogy and petrology of layered intrusions; A review: sandstone: Journal of Sedimentary Petrology, v. 3, p. 3-29.
Lunar and Planetary Institute Report 82-01, p. 128-134. Ruddiman, W . F., 1985, Climate studies in ocean cores, in Hecht, A. D., ed.,
Ramberg, H., 1948, Titanic iron ore formed by dissociation of silicates in granulite Paleoclimate analysis and modeling: New York, Wiley-Interscience,
facies [Greenland]: Economic Geology, v. 43, p. 553-569. p. 197-253.
— — , 1952, The origin of metamorphic and metasomatic rocks: Chicago, Illi- Rumble, D., Ill, 1973, Fe-Ti oxide minerals from regionally metamorphosed
nois, University of Chicago Press, 317 p. quartzites of western New Hampshire: Contributions to Mineralogy and
Rampacek, C., 1982, An overview of mining and mineral processing waste as a Petrology, v. 42, p. 181-195.
resource: Resources and Conservation, v. 9, p. 75-86. , 1976, Oxide minerals in metamorphic rocks, in Rumble, D., Ill, ed.,
Raufuss, W., 1973, Structur, schwermineralführung, genese, and bergbau der Oxide minerals: Mineralogical Society of America Short Course Notes, v. 3,
sedimentären Rutil Lagerstätten in Sierra Leone (Westafrika): Geologische p. Rl-24.
Jahrbuch Reihe D, no. 5, p. 3-52. Russell, E. E., 1967, Mineral resources summary of the Stantonville Quadrangle,
Reed, D. F., 1949a, Investigation of Christy titanium deposit, Hot Spring County, Tennessee: Tennessee Division of Geology of Mineral Resources Summary
Ark.: U.S. Bureau of Mines Report of Investigation 4592, 10 p. MRS13-NW (and map), 18 p.
— , 1949b, Investigation of Magnet Cove rutile deposit, Hot Spring County, , 1975, Stratigraphy of the outcropping upper Cretaceous in western Ten-
Arkansas: U.S. Bureau of Mines Report of Investigation 4593, 9 p. nessee: Tennessee Division of Geology Bulletin 75, p. A1-A65.
Reich, V., Kudrass, H.-R., and Wiedicke, M „ 1982, Heavy minerals of the east Russell, R. D., 1937, Mineral composition of Mississippi River sands: Geological
Australian shelf sediments between Newcastle and Fraser Island: Geologische Society of America Bulletin, v. 48, p. 1307-1348.
Jahrbuch, v. D 56, p. 179-195. Rust, B. R., and Jones, B. G., 1987, The Hawkesbury Sandstone south of Sydney,
Reynolds, R. L., and Goldhaber, M. B., 1978, Origin of south Texas roll-type Australia; Triassic analogue for the deposit of a large braided river: Journal
uranium deposits; 1, Alteration of iron-titanium oxide minerals: Economic of Sedimentary Petrology, v. 57, p. 222-233.
Geology, v. 73, p. 1677-1689. Saggerson, E. P., and Turner, L. M., 1978, Metamorphic map of Africa:
Rich, F. L., 1985, Palynology and paleoecology of a lignitic peat from Trail UNESCO Commission for the Geological Map of the World, scale
Ridge, Florida: Florida Bureau of Geology Information Circular 100, 15 p. 1:10,000,000.
Riezebos, P. A., 1979, Compositional downstream variation of opaque and Sainsbury, C. L., 1968, Tin and beryllium deposits of the central York Mountains,
translucent heavy residues in some modern Rio Magdalena sands (Colum- western Seward Peninsula, Alaska, in Ridge, J. D., ed., Ore deposits of the
bia): Sedimentary Geology, v. 24, p. 197-225. United States, 1933-1967, v. 2: New York, American Institute of Mining,
Rittenhouse, G., 1943, Transportation and deposition of heavy minerals: Geologi- Metallurgical, and Petroleum Engineers, p. 1555-1572.
cal Society of America Bulletin, v. 54, p. 1725-1780. Sallenger, A. H., 1979, Inverse grading and hydraulic equivalence in grain-flow
Robinson, P., and Tracy, R. J., 1977, Sulfide-silicate-oxide equilibria in deposits: Journal of Sedimentary Petrology, v. 49, p. 553-562.
sillimanite-K feldspar grade pelitic schists [abs.]: EOS Transactions of the Schluter, H. U., 1982, Results of a reflection seismic survey in shallow water areas
American Geophysical Union, v. 58, p. 524. off east Australia, Yamba to Tweed Heads: Geologische Jahrbuch, v. D 56,
Robinson, P., Hollocher, K. J., Tracy, R. J., and Dietsch, C. J., 1982, High grade p. 77-95.
110 E. R. Force
Schmidt, R. G., 1985, High-alumina hydrothermal systems in volcanic rocks and Soman, K., 1985, Origin and geologic significance of the Chavara placer deposit,
their significance to mineral prospecting in the Carolina slate belt: U.S. Kerala: Current Science, v. 54, p. 280-281.
Geological Survey Bulletin 1562, 59 p. Sorensen, H., 1974, Alkali syenites, feldspathoidal syenites, and related lavas, in
Scholle, P. A., 1979, Constituents, textures, cements, and porosities of sandstones Sorensen, H., ed., The alkaline rocks: London, Wiley, p. 22-52.
and associated rocks: American Association of Petroleum Geologists Memoir Southwick, D. L., 1968, Mineralogy of a rutile- and apatite-bearing ultramafic
28, 201 p. chlorite rock, Harford County, Maryland: U.S. Geological Survey Profes-
Scholtz, D. L., 1936, The magmatic nickeliferous ore deposits of East Griqualand sional Paper 600-C, p. 38-44.
and Pondoland: Transactions of the Geological Society of South Africa, Spear, F. S., 1981, An experimental study of hornblende stability and composi-
v. 39, p. 81-210. tional variability in amphibolite: American Journal of Science, v. 281,
Schumm, S. A., Mosley, M. P., and Weaver, W . E., 1987, Experimental fluvial p. 697-734.
geomorphology: New York, Wiley-lnterscience, 413 p. Spencer, R. V., 1948, Titanium minerals in Trail Ridge, Florida: U.S. Bureau of
Scull, B. J., 1958, Origin and occurrence of barite in Arkansas: Arkansas Geolog- Mines Report of Investigation 4208, 21 p.
ical and Conservation Commission Information Circular 18, 101 p. Spencer, R. V., and Williams, F. R., 1967, The development of a rutile mining
Seki, Y., Onuki, H., Oba, T., and Mori, R., 1971, Sambagawa metamorphism in industry in Sierra Leone: Proceedings—General, 8th Commonwealth Min-
the central Kii Peninsula, Japan: Japanese Journal of Geology and Geog- ing and Metallurgical Congress, Australia and New Zealand, 1965, Institute
raphy, v. 41, p. 65-78. of Mining and Metallurgy, London, v. 6, p. 1337-1341.
Semeniuk, V., and Johnson, D. P., 1982, Recent and Pleistocene beach/dune Spencer, S., 1964, African creeks I have been up: New York, McKay, 236 p.
sequences, Western Australia: Sedimentary Geology, v. 32, p. 301-328. Starmer, I. C., 1985a, The evolution of the south Norwegian Proterozoic as
Semeniuk, V., and Searle, D. J., 1986, Variability of Holocene sealevel history revealed by the major and mega-tectonics of the Kongsberg and Bamble
along the southwestern coast of Australia; Evidence for the effect of significant sectors, in Tobi, A. C., and Touret, J.L.R., eds., The deep Proterozoic crust in
local tectonism: Marine Geology, v. 72, p. 47-58. the North Atlantic provinces: Hingham, Massachusetts, Reidel, p. 259-290.
Shepherd, M. S., 1986, Australian heavy mineral reserves and world trends, in , 1985b, Geological map of the Bamble sector, south Norway, 3 sheets:
Australia; A world source of ilmenite, rutile, monazite, and zircon: Perth, NATO Advanced Studies Institute Excursion guide, scale 1:100,000.
Australasian Institute of Mining and Metallurgy, p. 61-68. Stephens, A. W., 1982a, Quaternary coastal sediments of the southeast Queens-
, 1990, Eneabba heavy mineral sand placers, Western Australia, in Geology land: Geologische Jahrbuch, v. D 56, p. 37-47.
and Mineral Deposits of Australia and Papua New Guinea: Parkville, Aus- , 1982b, Surficial sediments of the southern Queensland shelf; Southport-
tralasian Institute of Mining and Technology (in press). Point Lookout and Fraser Island areas: Geologische Jahrbuch, v. D 56,
Sheridan, D. M., Taylor, R. B., and Marsh, S. P., 1968, Rutile and topaz in p. 125-135.
Precambrian gneiss, Jefferson and Clear Creek Counties, Colorado: U.S. Stephenson, R. C., 1945, Titaniferous magnetite deposits of the Lake Sanford
Geological Survey Circular 567, 7 p. area, New York: New York State Museum Bulletin 340, 78 p.
Sherman, G. D., 1952, The titanium content of Hawaiian soils and its significance: Stone, B. D „ and Force, E. R., 1980, The Port Leyden, New York, heavy mineral
Soil Science Society of America Proceedings, v. 16, p. 15-18. deposit: New York State Museum Bulletin 436, p. 57-61.
Shideler, G. L., and Smith, K. P., 1984, Regional variability of beach and fore- , 1983, Coarse-grained glaciolacustrine deltaic sedimentation in ice-
dune characteristics along the Texas Gulf Coast barrier system: Journal of marginal and ice-channel environments: Geological Society of America Ab-
Sedimentary Petrology, v. 54, p. 507-526. stracts with programs, v. 15, p. 126.
Short, A. D., 1987, Modes, timing, and volume of Holocene cross-shore and Storch, R. H., and Holt, D. C., 1963, Titanium placer deposits of Idaho: U.S.
aeolian sand transport, southern Australia, in Coastal sediments '87: New Bureau of Mines Report of Investigation 6319, 69 p.
York, American Society of Civil Engineers, v. 2, p. 1925-1937. Subrahmanyam, N. P., and Rao, G.V.U., 1980, Niobium and tantalum in the
, 1988, Holocene coastal dune formation in southern Australia; A case rutile of Karala beach sands: Geological Society of India Journal, v. 21,
study: Sedimentary Geology, v. 55, p. 121-142. p. 623-626.
Short, N. M., 1961, Geochemical variations in four residual soils: Journal of Subrahmanyam, N. P., and 5 others, 1982, Alteration of beach sand ilmenite from
Geology, v. 69, p. 534-571. Manavalakurichi, Tamil Nadu, India: Geological Society of India Journal,
Siever, R., and Woodford, N., 1979, Dissolution kinetics and the weathering of v. 23, p. 168-174.
mafic minerals: Geochimica et Cosmochimica Acta, v. 43, p. 717-724. Sullivan, G. V., and Llewellyn, T. O., 1981, Occurrence and recovery of rutile
Sillitoe, R. H., 1983, Unconventional metals in porphyry deposits, in Shanks, from western copper mill tailings: Society of Mining Engineers Preprint
W . C. Ill, ed., Cameron volume on unconventional mineral deposits: New 81-333, 7 p.
York, Society of Mining Engineers, p. 207-221. Sundquist, E. T., and Broecker, W . S., 1985, The carbon cycle and atmospheric
Sinha, R. K., 1967, A treatise on industrial minerals of India: Bombay, Allied C 0 2 ; Natural variations, Archean to present: American Geophysical Union
Publishers, 513 p. Geophysical Monograph 32, 627 p.
Skjerlie, F. J., and Pringle, I. R., 1978, A Rb/Sr whole-rock isochron date from Swift, D.J.P., 1968, Coastal erosion and transgressive stratigraphy: Journal of
the lowermost gneiss complex of the Gaular area, west Norway and its Geology, v. 76, p. 444-456.
regional implications: Norsk Geologisk Tidsskrift, v. 58, p. 259-265. Taylor, H. P., Jr., and Coleman, R. G., 1968, 0 l 8 / 0 1 6 ratios of coexisting
Slingerland, R. L., 1977, The effects of entrapment on the hydraulic equivalence minerals in glaucophane-bearing metamorphic rocks: Geological Society of
relationships of light and heavy minerals in sands: Journal of Sedimentary America Bulletin, v. 79, p. 1727-1756.
Petrology, v. 47, p. 753-770. Temple, A. K., 1966, The alteration of ilmenite: Economic Geology, v. 61,
, 1984, Role of hydraulic sorting in the origin of fluvial placers: Journal of p. 695-714.
Sedimentary Petrology, v. 54, p. 137-150. Temple, A. K., and Grogan, R. M., 1965, Carbonatite and related alkalic rocks at
Slingerland, R., and Smith, N. D., 1986, Occurrence and formation of water-land Powderhorn, Colorado: Economic Geology, v. 60, p. 672-692.
placers: Annual Review of Earth and Planetary Science, v. 14, p. 113-147. Teufer, G., and Temple, A. K., 1966, Pseudorutile; A new mineral intermediate
Smith, N. D., and Minter, W.E.L., 1980, Sedimentological control of gold and between ilmenite and rutile in the natural alteration of ilmenite: Nature,
uranium in two Witwatersrand placers: Economic Geology, v. 75, p. 1-14. v. 211, p. 179-181.
Smithson, S. B., and Hodge, D. S., 1972, Field relations and gravity interpretation Thom, B. G., 1983, Transgressive and regressive stratigraphies of coastal sand
in the Laramie Anorthosite complex: Laramie, University of Wyoming Con- bodies in southeast Australia: Marine Geology, v. 56, p. 137-158.
tributions to Geology, v. 11, p. 43-59. Thom, B. G., and Roy, P. S., 1985, Relative sea levels and coastal sedimentation
in southeastern Australia in the Holocene: Journal of Sedimentary Petrology, Walker, T. R., 1967, Formation of red beds in modern and ancient deserts:
v. 55, p. 257. Geological Society of America Bulletin, v. 78, p. 353-368.
Thorn, B. G., Adams, R. P., Cazeau, C. J., and Heron, S. D., 1972, Aspects of the Wall Street Journal, 1976, Buttes says studies on Colorado prospect indicate
texture and mineralogy of surficial sediments, Horry and Marion Counties, major deposit of titanium ore: Wall Street Journal, v. 187, no. 37, p. 6
South Carolina: Southeastern Geology, v. 14, p. 39-58. (Feb. 24).
Thorn, B. G., Bowman, G. M., and Roy, P. S., 1981, Late Quaternary evolution of Ward, W . T., 1977, Sand movement on Fraser Island; A response to changing
coastal sand barriers, Port Stephens-Myall Lakes area, central New South climates: University of Queensland Department of Anthropology Papers,
Wales, Australia: Quaternary Research, v. 15, p. 345-364. v. 8, p. 113-126.
Thomas, B. I., and Berryhill, R. V., 1962, Reconnaissance studies of Alaskan , 1978, Notes on the origin of Stradbroke Island: University of Queensland
beach sands, eastern Gulf of Alaska: U.S. Bureau of Mines Report of Investi- Department of Geology Papers, v. 8, p. 97-104.
gation 5986, 40 p. Ward, W . T., Little, I. P., and Thompson, C. H., 1979, Stratigraphy of two
Thompson, C. H., and Bowman, G. M., 1984, Subaerial denudation and weather- sandrocks at Rainbow Beach, Queensland, Australia, and a note on humate
ing of vegetated coastal dunes in eastern Australia, in Thorn, B. G., ed., composition: Palaeogeography, Palaeoclimatology, and Palaeoecology, v. 26,
Coastal geomorphology in Australia: Sydney, Academic Press, p. 263-290. p. 305-316.
Thompson, C. H., and Ward, W . T., 1975, Soil landscapes of North Stradbroke Watson, T. L., 1912, Kragerite, a rutile-bearing rock from Kragere, Norway:
Island: Proceedings of the Royal Society of Queensland, v. 86, p. 9-14. American Journal of Science, v. 34, p. 509-514.
Thompson, J. V., 1987, Titanium resource in Colorado equals all other U.S. Watson, T. L., and Taber, S., 1913, Geology of the titanium and apatite deposits
deposits: Engineering and Mining Journal, v. 188, p. 27-30. of Virginia: Virginia Geological Survey Bulletin 3A, 308 p.
Thorman, C. H., 1977, Geologic map of the Monrovia Quadrangle, Liberia: U.S. Weiblen, P. W., and Morey, G. B., 1980, A summary of the stratigraphy, petrol-
Geologial Survey Map I-775-D, scale 1:250,000. ogy, and structure of the Duluth Complex: American Journal of Science,
Tipper, G. H., 1914, The monazite sands of Travancore: Geological Survey of v. 280A, p. 88-133.
India Records, v. 44, p. 186-196. Welch, B. K., 1964, The ilmenite deposits of Geographe Bay: Australasian Insti-
Toewe, E. C., Schatz, R. W „ Carmichael, R. L., Gonser, B. W., and Goldberger, tute of Mining and Metallurgy Proceedings, no. 211, p. 25-48.
W. M., 1971, Evaluation of Columbian-bearing rutile deposits, Magnet Welch, B. K., Sofoulis, J., and Fitzgerald, A.C.F., 1975, Mineral sand deposits of
Cove, Arkansas: U.S. Department of the Interior Contract Report 14-01- the Capel area, West Australia, in Knight, C. L., ed., Economic geology of
001-1738, 134 p. Australia and Papua New Guinea: Australasian Institute of Mining and
Tourtelot, H. A., 1968, Hydraulic equivalence of grains of quartz and heavier Metallurgy Monograph 5, p. 1070-1087.
minerals, and implications for the study of placers: U.S. Geological Survey Wells, N., 1960, Total elements in topsoils from igneous rocks; An extension of
Professional Paper 594-F, p. F1-F13. geochemistry: Journal of Soil Science, v. 11, p. 409-424.
, 1983, Continental aluminous weathering sequences and their climatic White, J. R., and Williams, E. G., 1967, The nature of a fluvial process as defined
implications in the United States: U.S. Geological Survey Circular 822, by settling velocities of heavy and light minerals: Journal of Sedimentary
p. 1-5. Petrology, v. 37, p. 530-539.
Towner, R. R., Gray, J. M., and Porter, L. M., 1988, International strategic Whitworth, H. F., 1959, The zircon-rutile deposits on the beaches of the east coast
minerals inventory summary report; Titanium: U.S. Geological Survey Cir- of Australia with special reference to their mode of occurrence and the origin
cular 930-G, 58 p. of the minerals: New South Wales Department of Mines Technical Report 4,
Troger, W . E., 1928, Alkaligesteine aus der Sierra do Salitre in westlichen Minas p. 7-60.
Gerais, Brazilien: Zentralblatt für Mineralogie, Geologie, and Paleontologie, Wiedner, J. R., 1982, Iron-oxide magmas in the system Fe-C-O: Canadian Min-
1928A, p. 202-204. eralogist, v. 20, p. 555-566.
, 1935, Spezielle Petrographie der Eruptivgesteine: Berlin, Verlag Deutsch Wilcox, J. T., 1971, Preliminary investigation of heavy minerals in the McNairy
Mineralogischen Gesellschaft, Berlin, p. 279-282. sand of west Tennessee: Tennessee Division of Geology Report of Investiga-
Turekian, K. K., 1977, Geochemical distribution of elements, in McGraw-Hill tions 31, 11 p.
encyclopedia of science of technology: New York, McGraw-Hill, Williams, L., 1967, Heavy minerals in South Carolina: South Carolina Division
p. 627,630. of Geology Bulletin 35, 35 p.
Turner, F. J., 1968, Metamorphic petrology; Mineralogical and field aspects: New Williams, S. A., and Cesbron, F. P., 1977, Rutile and apatite; Useful prospecting
York, McGraw-Hill, 403 p. guides for porphyry copper deposits: Mineralogical Magazine, v. 41,
Turner, R., 1986, Brazilian titanium: Engineering and Mining Journal, v. 187, p. 288-292.
p. 40-42. Wilmart, E., and Duchesne, J. C., 1987, Geothermobarometry of igneous and
Udubasa, G., 1982, Rutile of post-magmatic mineral formation, in Ore genesis; metamorphic rocks around the Ana-Sira anorthosite massif; Implications for
G. C. Amstutz, ed., The state of the art: Berlin, Springer-Verlag, p. 784-793. the depth of emplacement of the South Norwegian anorthosites: Norsk
Ulbrich, H.H.G.J., and Gomes, C. B., 1981, Alkaline rocks from continental Geologisk Tidsskrift, v. 67, p. 185-196.
Brazil: Earth Science Reviews, v. 17, p. 135-154. Wilson, A. F., 1964, The penological features and structural setting of Australian
United States Bureau of Mines, 1986, Domestic consumption trends, 1972-82, granulites and charnockites: New Delhi, 22nd International Geological Con-
and forecasts to 1993 for twelve major metals: U.S. Bureau of Mines Open- gress Report, sec. 13, p. 21-44.
File Report 27-86, 156 p. , 1969, Granulitic terrains and their tectonic setting and relationship to
Ussing, N. V., 1911, Geology of the country around Julianehaab, Greenland: associated metamorphic rocks in Australia: Geological Society of Australia
Meddelelserom Grönland, v. 38, p. 1-376. Special Publication 2, p. 243-258.
Valeton, I., 1972, Bauxites: New York, American Elsevier, 226 p. Winward, K., and Nicholson, D . A., 1974, Quaternary coastal sediments, in
van Andel, Tj. H., 1950, Provenance, transport, and deposition of Rhine sedi- Markham, N. L., and Basden, H., eds., The mineral deposits of New South
ments: Wageningen, Veenman und Zonen, 129 p. Wales: Geological Survey of New South Wales, p. 597-621.
Vos, R. G., and Hobday, D. K., 1977, Storm beach deposits in the late Paleozoic Wise, W . S., 1975, The origin of the assemblage quartz + Al-silicate + rutile +
Ecca Group of South Africa: Sedimentary Geology, v. 19, p. 217-232. Al-phosphate: Fortschrift Mineralogie, v. 52, p. 151-159.
Wahlstrom, E. E., 1948, Pre-Fountain and Recent weathering on Flagstaff Moun- , 1977, Mineralogy of the Champion mine, White Mountains, California:
tain near Boulder, Colorado: Geological Society of America Bulletin, v. 59, Mineralogical Record, Nov.-Dec., p. 478-486.
p. 1173-1190. Wright, T. L., and Peck, D. L., 1978, Crystallization and differentiation of the
112 E. R. Force
Alae magma', Alae lava lake, Hawaii: U.S. Geological Survey Professional Yudin, B. A., and Zak, S. I., 1971, Titanium deposits of northwestern USSR
Paper 935 C, 20 p. (eastern part of Baltic Shield): International Geology Review, v. 13,
Wynn, J. C., and Grosz, A. E., 1986, Application of the induced polarization p. 864-872.
method to offshore placer resource exploration: Offshore technology confer- Zen, E-an, 1960, Metamorphism of lower Paleozoic rocks in the vicinity of the
ence, 18th Houston, Preprint OTC 5199, p. 395-398. Taconic range in west-central Vermont: American Mineralogist, v. 45,
Wynn, J. C., Grosz, A. E., and Foscz, V. M., 1985, Induced polarization and p. 129-175.
magnetic response of titanium-bearing placer deposits in the southeastern
United States: U.S. Geological Survey Open-File Report 85-756, 25 p. MANUSCRIPT ACCEPTED BY THE SOCIETY J U L Y 6, 1 9 9 0
Typeset by WESType Publishing Services, Inc., Boulder, Colorado

Printed in U.S.A. by JL Printing, Loveland, Colorado
Geological Society of America Special Papers

Geology of Titanium-Mineral Deposits
Eric R. Force
Geological Society of America Special Papers 1991;259; 1-112

doi:10.1130/SPE259-p1
E-mail alerting services click www.gsapubs.org/cgi/alerts to receive free e-mail alerts when new
articles cite this article
Subscribe click www.gsapubs.org/subscriptions to subscribe to Geological Society of

America Special Papers
Permission request click www.geosociety.org/pubs/copyrt.htm#gsa to contact GSA.
Copyright not claimed on content prepared wholly by U.S. government employees within scope
of their employment. Individual scientists are hereby granted permission, without fees or further
requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in
subsequent works and to make unlimited copies of items in GSA's journals for noncommercial
use in classrooms to further education and science. This file may not be posted to any Web site,
but authors may post the abstracts only of their articles on their own or their organization's Web
site providing the posting includes a reference to the article's full citation. GSA provides this and
other forums for the presentation of diverse opinions and positions by scientists worldwide,
regardless of their race, citizenship, gender, religion, or political viewpoint. Opinions presented in
this publication do not reflect official positions of the Society.
Notes
© 1991 Geological Society of America

Geology of Titanium-Mineral Deposit Force1991

Uploaded by

Copyright:

Available Formats

Geology of Titanium-Mineral Deposit Force1991

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Geology of Titanium-Mineral Deposit Force1991

Uploaded by

Copyright:

Available Formats

Downloaded from specialpapers.gsapubs.

org on March 16, 2016

listed in Table 2. The partitioning of titanium between oxide and

economic geology of titanium than it is for most other mineral Si6AI2022(0H,F)2

Class Type Typical Importance* Example

1. Metamorphic a. Eclogite Rutile B Piampaludo, Italy

Metamorphlc deposits (Chapter 2)

State District or Typet Rutile + Altered Low-TlC^ Perovskite

Alabama Sand and gravel 4 100

Totals by deposit types 1b 1,000

Total by mineral 14,700 33,500 35,000 20,000

Total all minerals 103,200

'Modified from Force and Lynd (1984); none reported or negligible.

Byproduct-coproduct relations Titanium From shoreline placer From alkalic deposits:

Occurrence and deposits of titanium oxide minerals in metamorphic rocks

Greenschist facies A m p h i b o l i t e facies

Sunnfjord region, Norway

Titanium oxide minerals in anorthosite-ferrodiorite massifs

Figure 13. Photomicrograph of ilmenite, magnetite, and apatite enclosing

CONTACT-METASOMATIC RUTILE DEPOSITS

Although magmatic ilmenite deposits commercially are the

Plunge of fold The following descriptions of anorthosite-ferrodiorite-

Sanford Lake (Tahawus, Maclntyre) district, New York

Tellnes district, Norway

Geology of titanium-mineral deposits 29

In contrast to the Tellnes body, the Storgangen deposit is

Figure 26. Layering in Storgangen ore. Plagioclase and orthopyroxene in

Figure 27. Photomicrograph of interstitial texture of oxide minerals rela-

Figure 29. Elliptical domains of felsic minerals among cumulate ortho-

Laramie Range, Wyoming

Figure 31. Photomicrographs of textural relations of oxide and sulfide

The other type of resource, of economic interest for titanium

The most valuable magmatic ilmenite deposits occur in an-

Titanium oxide minerals in alkalic igneous rocks

Deposit Deposits In igneous rocks Contact- Weathered

M A J O R DEPOSITS Pyroxenite consists of diopsidic augite, locally titaniferous,

The anatase deposits in weathered overburden of alkalic

Magnet Cove district, Arkansas

Geology of titanium-mineral deposits 52

Titanium-oxide mineral deposits in alkalic rocks suffer from

Titanium oxide minerals in other igneous suites

Rock Suite TÌO2 Range Primary Mineralogy^ Exsolution Pseudomorphic

Rhyolitic rocks 0.2-0.6 P P ilm (R) hem (R)

*From Haggerty, 1976a, and Force, 1976a.

Basaltic rocks In general, titanium-oxide mineral accumulation in layered

Biotite + sulfur •phlogopitic biotite + pyrite + rutile

Hornblende + sulfur - - ••actinolitic hornblende + pyrite + rutile + alumina

Titaniferous magnetite + sulfur • l e s s e r magnetite + pyrite + rutile

llmenite + sulfur • p y r i t e + rutile

ECONOMIC IMPORTANCE THE BEHAVIOR OF TITANIUM IN THE

Igneous and metamorphic deposits

Postdepositional weathering of sedimentary placer deposits

Predepositional weathering of detrital grains of placer deposits

Residual enrichment of titanium oxide minerals

Ilmenite (+ hematite) concentrate

Average Gain (%) 109 >140 7 -66 3 41 -10

'From Mertie (1979) and my own observations.

acids. Last, we know that humate can also precipitate in heavy-

Climate change and weathering