RESEARCH METHODS PAPERS
BREWER, R., 1976, Fabric and Mineral Analysis of Soils: Huntington,
New York, Robert E. Krieger, 482 p.
CAINE, N., 1967, The texture oftalus in Tasmania: Jour. Sed. Petrology,
v. 37, p. 796-803.
KRUMaEIN, W. C., 1941, Measurement and geological significance of
shape and roundness of sedimentary particles: Jour. Sed. Petrology,
v. 1 I, p. 64-72.
McSAVENEY, E. R., 1971, The surficial texture of rockfall talus [unpubl.
master's thesis]: Ohio State University, 91 p.
A NONTOXIC HEAVY LIQUID AND INEXPENSIVE
JOHN CALLAHAN
Geology Department
Appalachian State University
Boone, North Carolina 28608
INTRODUCTION
Two heavy liquids commonly used to separate minerals, bromoform
and tetrabromoethane, are both toxic (Sax and Lewis 1986; H a u f f a n d
Airey 1980). In addition, acetone, which is used to clean samples pre-
pared with bromoforrn or tetrabromoethane, is flammable and acts as
an irritant and narcotic (Sax and Lewis 1986). Sodium polytungstate,
3Na2WO4.9WO~ "H20, a nontoxic solid that can be made into a heavy
liquid (sp. gr. > 1.0-3.1) has been on the market for approximately
three years. A similar chemical, sodium metatungstate (N~[H2W,20,0]),
used for separation of beavy minerals, was originally described by Ple-
winsky and K a m p s (1984). Sodium polytungstate is available in liquid
or powdered form from Sometu, Falkenfied 4, D-1000 Berlin 33, West
Germany at DM 130 ($65, U.S.) per kilogram for the solid and at a
higher price for the liquid. According to the manufacturer, a i,250-g
solution with an approximate density of 2.90 g/cm 3 can be made by
adding 250 g of distilled water to 1,000 g of powdered sodium poly-
tungstate. The main purpose here is to report results obtained in heavy
mineral separations using sodium polytungstate and to compare the
results with those obtained from tetrabromoethane separations. In ad-
dition, several types o f filter paper were used in the separations in an
attempt to determine which one(s) could best be used in separations
with the sodium polytungstate.
METHODOLOGY
Heavy mineral separates from splits of the minus 40-mesh fraction
(< 0.420 m m ) of a beach sand from Panama and gold grains from a
placer deposit in North Carolina were concentrated using sodium poly-
tungstate and tetrabromoethane. The samples were separated in stan-
dard 250-mL separatory funnels and recovered on various types of filter
paper. In a separate test, a gold concentrate was separated using a small
laboratory-model centrifuge to test the efficiency of separating fine gold
(less than 40-mesh) with the sodium polytungstate liquid.
ANALYTICAL RESULTS
Essentially identical weights of beavy concentrates were obtained from
the Panama beach sand using the two different heavy liquids in duplicate
experiments. Twenty grams of sand were used in both experiments. The
heavy fraction represented 15 percent of the total weight of the sand
for both the sodium polytungstate and the tetrabromoethane separa-
tions. The magnetic fraction (30% by weigh0 of the concentrate was
separated using a Sepor Automagnet. Petrographic examination of the
nonmagnetic fraction of the concentrates showed that it consisted almost
entirely of hornblende with minor zircon and biotite. Other heavy sep-
Manuscript received 14 February 1986; revised 5 January 1987.
JOURNAl. OF SEDXS~rrAgY PETgOLO6Y, VOL. 57, NO. 4, JULY, 1987, P. 765-766
Copyright 1987, The Society of FEconomicPaleontologists and Mineralogists 0022--4472/87/0057-765/$03.00
765
P~REZ, F. L., 1986, Talus texture and particle morphology in a North
Andean pfiramo: Z. Geomorph, N.F., v. 30, p. 15-34.
SNEED, E. D., AND FOLK, R. L., 1958, Pebbles in the lower Colorado
River, Texas: a study in particle morphogenesis: Jour. Geology, v.
66, p. 114-150.
TURNER, F. J., AnD WEISS, L. E., 1963, Structural Analysis of Meta-
morphic Tectonites: New York, McGraw-Hill, 545 p.
Z~NGG,T., 1935, Beitrag zur Schotteranalyse: Schweiz. Mineral. Petrogr.
Mitt., v. 15, p. 39-140.
FILTERS FOR SEPARATION OF MINERAL GRAINS 1
aration tests using 0.04 g of minus 40-mesh gold placed in sodium
polytungstate and tetrabromoethane showed great differences in the
recovery rates for the gold. Only 40 percent of the gold was recovered
in heavy concentrates in the first experiment using sodium polytung-
state, and 89 percent of the gold was recovered in the tetrabromoethane
heavy concentrates. In a second experiment, great care was taken to stir
the sodium polytungstate-gold-grain mixture, and the recovery rate
improved to 83 percent and remained almost the same (88%) for the
tetrabromoethane separation.
In a separate experiment, the sodium polytungstate-gold concentrate
that was recovered from the second experiment noted above was cen-
trifuged for one minute. When the minus 40-mesh gold concentrate was
first placed on the sodium polytungstate in the centrifuge tube, most of
it floated. After one minute of centrifuging, 100 percent o f the gold sank
to the bottom. In another experiment, some gold grains in the greater
than 40-mesh range were placed in the sodium polytungstate liquid.
After stirring, 100 percent o f them were recovered in the heavy portion
of the concentrate.
One of the major problems encountered with the use of the sodium
polytungstate was the slow rate at which it passed through even one of
the more rapid laboratory grades of filter paper (Whatman #4). Several
varieties of filter paper with various sample materials and amounts in
them were tested to determine which paper allowed both heavy liquids
to pass through most rapidly. After several experiments, it was deter-
mined that both the tetrabromoethane and sodium polytungstate passed
through coffee filter paper at least twice as fast as the W h a t m a n #4. The
coffee filters are one-tenth the cost of the laboratory-grade filter paper.
In addition, the wet strength of the coffee filters appears to be better
than the laboratory-grade filter papers we used.
DISCUSSION AND CONCLUSIONS
It appears that identical results can be obtained for heavy mineral
separations using the heavy liquids tetrabromoethane and sodium poly-
tungstate. For fine-grained gold, the recovery rate may be increased by
stirring the sample repeatedly or by using a centrifuge in the separation
process. Fine gold tends to float in the sodium polytungstate-water
solution, as gold is not readily wettable by water.
The main advantage of sodium polytungstate, according to the sup-
plier, is that it is nontoxic. According to Kazantzis (1979), there appears
to be little potential for tungsten compounds to be extremely hazardous
unless inhaled or ingested. No direct reference to the toxicity of sodium
polytungstate could be found by the author, and it is not listed in the
widely quoted Registry of Toxic Effects of Chemical Substances (1983).
Other advantages are that the heavy mineral separates can be cleaned
with distilled water, and the specific gravity of the sodium polytungstate
solution can be changed by adding either distilled water or powdered
sodium polytungstate. A table is available from the supplier for making
changes in the specific gravity. According to the supplier, the m a x i m u m
766 JOHN CALLAHAN

specific gravity that can be attained by adding powdered sodium poly-
tungstate to distilled water is 3.1. The biggest disadvantage is that the
cost o f l kg o f the powdered sodium polytungstate solid is D M 130,
which at the present rate of exchange is $65, or $151 per liter, when
made into a solution. This is approximately i .6 times more expensive
than the purified grade of tetrabromocthane at $96 per liter. Two other
disadvantages are recognized. Sodium polytungstate passes through
standard laboratory filter paper rather slowly as a result of its higher
viscosity. This problem can be avoided by using a rapid filter paper
such as coffee filters. In addition, the manufacturer indicates that sodium
polytungstate has rather high trace element contents (Mo-2 ppm, As-5
ppm, U-10 ppm). These might interfere with chemical analysis of the
concentrates if they are not cleaned carefully with distilled water to
remove any residue. However, because essentially identical results may
be obtained using nontoxic sodium polystungstate and toxic tetrabro-
moethane, the use o f sodium polytungstate is recommended as a safe
way to carry out effective heavy mineral separations.
REFERENCES
HAUFF, P. L., AND AIREY, JOSEPH, 1980, The Handling, Hazards, and
Maintenance o f Heavy Liquids in the Geologic Laboratory: U.S. Geo-
logical Survey, Circular 827, 24 p.
KAZANTZaS,G. 1979, Tungsten, in Friberg, et al., ed., Handbook on the
Toxicology of Metals: New York, Elsevier, 709 p.
PLEWlNSKY, B., AND KAMPS, R., 1984, Sodium metatungstate, a new
medium for binary and ternary density gradient centrifugation: Die
makromolekulare Chemie, v. 185, p. 1429-1439.
SAx, N. L, AND LEWIS, R. J., 1986, Rapid Guide to Hazardous Chemicals
in the Workplace: New York, Van Nostrand Reinhold Co., 236 p.
TATKEN, R. L., AND LEWIS, R. J., SR. EDITORS, 1983, Registry o f Toxic
Effects o f Chemical Substances, 1981-82 Edition, vol. 3: U.S. De-
partment of Health and H u m a n Services, National Institute for Oc-
cupational Safety and Health, 988 p.

IMAGE ANALYSIS OF CATHODOLUMINESCENT-ZONED CALCITE CEMENTS ~

S. L. DOROBEK, 2 J. F. READ, AND J. M. NIEMANN 3

Department of Geological Sciences
Virginia Polytechnic Institute and State University
Btacksburg, Virginia 24061
T. C. PONG AND R. M. HARALICK
Spatial Data Analysis Laboratory
Virginia Polytechnic Institute and State University
Blacksburg, Virginia 24061

INTRODUCTION Cathodoluminescent zonation in carbonate cement minerals most

Most studies of carbonate diagenesis have relied on time-consuming
point counting to determine abundance o f grains versus cement with
little quantitative assessment of volumetric abundance of specific ce-
ment generations. Computer-assisted image analysis provides a rapid
quantitative method for determining grain size, shape, and orientation,
cement abundance, and a m o u n t o f porosity in sedimentary rocks. This
paper describes how one such image analysis software system (the Gen-
eral Image Processing System or GIPSY; Haralick 1983), has been used
to determine abundance of grains and total cement in carbonate rocks.
More importantly, GIPSY can be used to document quantitatively spe-
cific stages of calcite cementation in carbonate rocks, as defined by the
cathodoluminescent zonation present in the cements. A step-by-step
procedure is explained for analyzing cathodoluminescent-zoned calcite
cements. This procedure should work with any combination o f image
analysis software and hardware which meet the following criteria:
1) The system is capable of high resolution analysis of digitized images.
This is necessary to resolve and analyze very thin cathodolumines-
cent zones in cements.
2) The system has a large range of gray-level selection. This allows
differentiation of subtle differences in cathodoluminescence between
various cement zones.
3) The system can manipulate and store at least t h r ~ images simul-
taneously in virtual memory. This is essential for performing the
image "multiplication" described below.
4) The system can perform various arithmetic operations described
below on binary images stored in virtual memory.
' Manuscript received 14 January 1986; revised 23 September 1986.
2 Current address: Department of Geology, Washington State Uni-
versity, Pullman, Washington 99164-2812.
3 Current address: Chevron, P.O. Box 61990, 127 Elk Place, New
Orleans, Louisiana 70161.
likely is related to the presence of trace concentrations of manganese
and iron (Pierson 1981; Frank t al. 1982; Fairchild 1983). In order for
these elements to substitute extensively for calcium in the calcite crystal
lattice, they must be in reduced (divalent) states. Therefore, incorpo-
ration of manganese and iron in carbonate cements is largely a function
o f the redox potential o f pore waters from which the cements precipi-
tated (Oglesby 1976). Cathodoluminescence in carbonates commonly
is described as nonluminescence, bright luminescence, or dull lumines-
cence (Fig. 1), and ccathodoluminescent zonation in calcite cements in
ancient carbonates has been attributed to precipitation of cements in
different diagenefic environments (Meyers 1974, 1978; Grover and Read
1983; Dorobek 1984, 1987). Image analysis of cathodolurninescent-
zoned cements can then provide rapidly determined abundances of
specific cement zones deposited in various diagenetic environments.
Therfore, this technique allows rapid quantification of the a m o u n t of
porosity loss during specific cementation stages (defined by eathodolu-
minescent zones), which may prove useful in understanding and pre-
dicting reservoir characteristics.
THE GIPSY SYSTEM
GIPSY is a general interactive image processing software system (Har-
alick 1983) that can be rapidly learned, used, modified, and transported
from one computer to another. It is a coordinated system o f over 275
programs written in the programming language RATFOR. These pro-
grams can be used to enhance, manipulate, and analyze digital images.
GIPSY was developed by R. M. Haralick and students at Virginia
Polytechnic Institute and State University. Licences for the source code
o f GIPSY are available from the Spatial Data Analysis Laboratory at
the above-named institution, and parties interested in the software should
correspond with Haralick.
With GIPSY, bands of an image can be designated numerically or
symbolically. Numeric hands contain gray-tone values that can be ma-
nipulated arithmetieally, an important feature for the application dis-
cussed in this paper. The results of all user interactions on each image
are kept in storage in user records.

JOURNALOF SEDIMENTARYPETROLOGY, VOL. 57, NO. 4, JoLv, 1987, r. 766---770

Copyright 1987, The Society of Economic Paleontologists and Mineralogists 0022-4472/g7/0057-766/$03.0(~