SU99

SUMMER 1999
VOLUME 35 NO. 2
T A B L E
O F
C O N T E N T S
EDITORIAL
79
Nothing Is as Simple as It Seems

Alice S. Keller
80
L ETTERS
F EATURE A RTICLES
pg. 83
82
On the Identification of Various Emerald Filling Substances

Mary L. Johnson, Shane Elen, and Sam Muhlmeister
108
Sapphire and Garnet from Kalalani,

Tanga Province, Tanzania
Antonin V. Seifert and Jaroslav Hyrsl
122
Russian Synthetic Ametrine

Vladimir S. Balitsky, Taijin Lu, George R. Rossman,
Irina B. Makhina, Anatolii A. Marin, James E. Shigley,
Shane Elen, and Boris A. Dorogovin
REGULAR FEATURES
pg. 109
pg. 130
pg. 145
135
136
142
156
157
160
1999 Challenge Winners

Gem Trade Lab Notes
Gem News
Thank You, Donors
Book Reviews
Gemological Abstracts
ABOUT THE COVER: Emeralds typically contain numerous fractures when recovered,
and yet consumers have high expectations for their clarity as well as color. As a result,
the overwhelming majority of emeralds on the market today have undergone some form
of clarity enhancement. The proliferation of filler substances in recent years has raised
widespread concern in the trade about their identity and durability. The lead article in
this issue reviews the history of emerald clarity enhancement and the current situation
in the trade. It then focuses on possible methods of distinguishing among different clarity-enhancement substances.
The emerald cabochon in this Sabi-finish gold necklace weighs 77.68 ct. Courtesy of
Henry Dunay Designs Inc., New York.
Photo Harold & Erica Van PeltPhotographers, Los Angeles, California.
Color separations for Gems & Gemology are by Pacific Color, Carlsbad, California.
Printing is by Fry Communications Inc., Mechanicsburg, Pennsylvania
1999 Gemological Institute of America
All rights reserved.
ISSN 0016-626X
Nothing Is as
Simple as It Seems
omeone asked me recently if gemologists really used

tion for most others. Once the determination of, for examthe infrared, EDXRF, and Raman spectra that we now
ple, R.I. and S.G. has narrowed the field, microscopic feapublish in most Gems & Gemology articles. Twenty
tures often will reveal strong indications of natural or synyears ago, a good gemologist could identify almost any gem
thetic origin, or whether the stone has been treated. Yet even
material, and most enhancement processes, with straightforthe field of microscopy has changed. In many cases where
ward gemological testing refractive index, specific gravity,
standard inclusions or color zoning cannot conclusively idenoptic character, and the like. Twenty years ago, we did quite
tify flawless synthetic rubies and sapphires, growth structure
well (some tell me) without all of this advanced instrumentaanalysis can make the distinction, but this requires a new
tion. A new material like cubic zirconia could be addressed
level of technical knowledge and skill with a microscope.
quickly and efficiently. The major issues were gem localities,
The article on the identification of emerald fillers in this
cause of color, and interesting new inclusions. Twenty years
issue is a good case in point. As the authors note, there are tens
ago, we did not have commercial quantities of flux and
of thousands of potential emerald fillers. They studied 39
hydrothermal synthetic sapknown and possible filler subphires, gem-quality synthetic diastances and found that to make
monds, or bleached and impregeven basic distinctions between prenated B jade, and few questions
sumed natural oils and artificial
were asked about the filler used
resins required a combination of
in oiled emeralds.
techniques, such as microscopy and
Think about it. Ask yourself
infrared or Raman spectroscopy.
how many gem materials preEven the advanced techniques cannot
sented to you in a simple white
always resolve the challenges posed by
parcel paper you could identify
emeralds that have undergone more
conclusively the material,
than one generation of treatment (a
whether natural or synthetic, the
common practice) or have been filled
Impressive. What does it do?
presence and nature of any treatwith mixtures of substances.
ment with the same basic equipment gemologists used
Yes, the articles have become more technical and somealmost exclusively twenty years ago. The stone is green. The
times can be daunting. And, no, the 21st century gemologist
question no longer is simply whether it is emerald, peridot,
does not have to be a spectroscopist, chemist, or physicist.
tourmaline, or the like. But rather, once you know its emerHowever, he or she should develop some understanding of
ald, is it a natural stone or a flux or hydrothermal synthetic?
how sophisticated techniques in spectroscopy, chemistry, and
Could there be an overgrowth? Is it filled? How extensively?
physics can contribute to the gem identification process. As
With what? Twenty years ago, the grading of near-colorless
new materials and treatments have become more advanced
diamonds required primarily determination of color, clarity,
and difficult to detect, the means of identification have
and finish along with the plotting of inclusions. Now, it
become more complicated as well. Like all good professionmust first be determined whether it is diamond or a convincals, gemologists must know the limitations of their knowling imitation, andif diamondwhether it is natural or
edge and the outside resources that are available when they
synthetic. Then, is it fracture-filled? And eventually, perhaps,
reach these limitations. An awareness of the capabilities of
does it show evidence of lightening by the GE process?
advanced testing methods available in modern gem testing
Carefully executed standard gemological practices can still
laboratories is a must. Because, indeed, nothing is as simple
provide identification for many stones and essential informaas it seems.
Alice S. Keller
EDITOR, Akeller@gia.edu
Editorial
GEMS & GEMOLOGY
Summer 1999
79
Heat Exposure May Affect Reflectance

Testing of Synthetic Moissanite
In the Winter 1997 Gems & Gemology article by K.
Nassau et al., Synthetic Moissanite: A New Diamond
Substitute, it was noted that testing with a reflectivity
meter can also give diagnostic results for identifying
this material, provided great care is taken with the instrument (see p. 273). Recently, however, scientists at C3 Inc.,
the company that manufactures and distributes synthetic
moissanite to the gem trade, discovered that exposure of
gem-quality synthetic moissanite to high temperatures
for extended periods of time can result in thin oxide films
on the surface that may cause a shift in the measured surface reflectance. While such coatings are beneficial to
electronics applications of silicon carbide, they can create
substantial errors when the material is tested on gemstone reflectance meters, potentially causing synthetic
moissanite to be mistaken for diamond.
Most commercial reflectance testers bounce an
infrared beam off the surface to be tested and onto a
detector that records the intensity of the reflected beam. If
the surface is composed of a single (homogeneous) wellpolished material, then the light reflected can be related
to the refractive index of that material. If, however, a thin
film of another material coats the surface of the gemstone, the reflectance signal will no longer be related in a
simple way to the refractive index of the underlying bulk
material (see, e.g., E. Hecht and A. Zajac, Optics, Addison
Wesley Publishing Co., Reading, MA, 1979, p. 311).
In the case of synthetic moissanite (silicon carbide, or
SiC), a stable, uniform, thin surface film of silicon dioxide
(SiO2) is produced when the gem material is heated sufficiently in an oxygen-containing atmosphere. The resulting change in surface reflectance strongly depends on the
thickness and composition of the oxide film (figure 1). As
the thickness of the film increases, the measured surface
reflectance drops through the arbitrarily chosen
reflectance values for moissanite, diamond, cubic zirconia, and yttrium aluminum garnet (YAG). In fact, when
the oxide film reaches a certain thickness, the surface will
become antireflective and will register as zero on a
reflectance meter. The exact thickness of the oxide film
that produces the erroneous diamond reflectance measurement will vary according to the precise wavelength of
the infrared beam and the incidence angle used by the
reflectance tester.
We performed some simple experiments to confirm
the predicted reflectance behavior. We tested about 30
80
Letters
faceted synthetic moissanites on three standard

reflectance testers (Presidium, DiamondEye, and Jemeter
Digital 90) and with the C3 Tester Model 590, and verified that they measured as moissanite. (With a thermal
conductivity probe, the same samples registered as diamond.) The stones were then heated in air for varying
times using a furnace. Although the times and temperatures used were more extreme than those encountered in
typical jewelry fabrication, they were still well below
those required to cause damage (e.g., surface pitting visible at 50100 magnification) to the stones. The stones
were re-measured on the reflectance testers. Depending
on the time and temperature of heating, the treated samples of synthetic moissanite measured incorrectly as diamond, cubic zirconia, or YAG; longer exposure to the
heat treatment produced lower values of surface
reflectance as the oxide film grew thicker. The results
given by the different reflectance meters were in general
agreement. A surface reflectance measurement equivalent
to that of diamond was produced by a 40- to 60-nm-thick
film of oxide, which is very difficult to detect using commonly available means; no visible changes in the appearance of the heated stones could be seen. Note that, after
treatment, all of the faceted synthetic moissanites still
Figure 1. The surface reflectance of synthetic

moissanite is reduced by the thermal growth of an
oxide film. The measured reflectance value may
equal that of diamond or any lower-R.I. gem material, depending on the thickness of the oxide coating
(as shown here for a 45 incident beam angle).
GEMS & GEMOLOGY
Summer 1999
Comments on the Sources

of Blue Diamonds
tested as diamond with a thermal conductivity probe,

but they were correctly identified by the C3 Tester Model
590 as synthetic moissanite. Other distinguishing gemological features (e.g., doubling of the back facets) also
remained the same.
The oxide film produced by the high-temperature
treatment was extremely robust and could not be
removed by solvents, ultrasonic cleaning, or by most
acids. However, use of standard hydrofluoric-acid-based
glass etchants for silicon dioxide did restore the surface
reflectance readings to their initial values.
Testing of faceted synthetic moissanites (about 20
samples) that had been subjected to normal jewelry operations, such as the re-tipping of prongs on a ring, did not
produce any modifications to the surface reflectance
unless the samples were heated excessively. Given the
very slight amount of oxidation required to produce a substantial measurement error, it is conceivable that such a
film might be grown inadvertently during jewelry manufacturing or repair. C3 has no plans to manufacture or sell
faceted synthetic moissanite that has a modified surface
reflectance. Nevertheless, because the possibility exists
that intentional or unintentional surface reflectance
modifications may occur, C3 does not recommend
the use of reflectance testing as an independent
means for distinguishing synthetic moissanite gemstones from diamond.
I read with great interest the article Characterizing

Natural-Color Type IIb Blue Diamonds, by J. M. King et
al., which appeared in your Winter 1998 issue. The intricate scientific information on a difficult subject such as
color and electron resonance is well presented and
explained; however, some inaccuracies have crept into
the section on history and geographic origin.
Although it is customary to speak of Golconda diamonds, in actuality there are no observed mines or diamond occurrences at or near the city of Golconda. The
famous Indian diamonds, including the Koh-i-Nur, were
found in the peripheral areas of the ancient Golconda
Kingdom, in the valley of the lower course of the Krishna
River. This is a distance of 120 miles (193 km) from
Golconda. This city was the capital of the ancient kingdom, where all diamonds found in the surrounding region
were traded. The old fortress of Golconda still exists, and
is located 10 miles west of Hyderabad.
Note also, in the last paragraph on page 249, that it
should be the Helam mine near Swartsruggens, and Sierra
Leone is better described as being in West Africa.
Nonetheless, I think that this is a very authoritative paper, and it gets my vote for most valuable article of the year.
Mark Kellam, Ph.D.

Director of Technology, C3 Inc.
Morrisville, North Carolina
Dr. A. J. A. (Bram) Janse

Diamond Exploration Consultant
Perth, Western Australia
IN MEMORIAM
D.VINCENT MANSON (19361999)
Gems & Gemology mourns the loss of Vince Manson,
who passed away on July 3 following a year-long battle
with cancer. At the time of his death, at age 63, Dr.
Manson was an associate editor of Gems & Gemology (a
position he had held for 19 years) as well as
GIAs director of strategic planning. Most
recently, he was chairperson of the Third
International Gemological Symposium,
which was held in San Diego on June 2124.
Although Dr. Manson was not able to attend
the 1999 Symposium because of his illness,
he learned from his many visitors of the
overwhelming success of the event he had
spent three years planning. He originally
established the concept of the International
Gemological Symposium, and also chaired
the first and second Symposia in 1982 and
1991.
Vince Manson received his Bachelor of Science and
Masters degrees in geology from the University of
Witwatersrand in South Africa. He subsequently spent
two years at the De Beers Diamond Research Laboratory
in South Africa and six months doing geochemical
prospecting in Nova Scotia. In 1964, he earned his Ph.D.
in geology from Columbia University. For the next 12
Letters
years, Dr. Manson was curator of gems and minerals at

the American Museum of Natural History in New York.
He joined GIA in 1976 and was responsible for establishing the Research Department. Among his many other
technical accomplishments, he served as a
consultant for the widely acclaimed documentary The Time of Man, and was coproducer for both the Time-Life film Our
Dynamic Earth and the award-winning
Gems of the Americas.
In addition to his role as associate editor
of Gems & Gemology, Dr. Manson is
remembered for his comprehensive articles
on garnet classification, which he coauthored with G&G technical editor Carol
Stockton. The final article in this series, A
Proposed New Classification of Gem-Quality
Garnets, won the journals most valuable
article award for 1985.
Vince Manson is survived by his wife Averil, sons
Dirk and Scott, and four grandchildren. At the familys
request, GIA has established the D. Vincent Manson
Fund for Gemological Research in his honor. Admired for
his vision and intellect, treasured for his wit and warmhearted laughter, he will be profoundly missed.
GEMS & GEMOLOGY
Summer 1999
81
ON THE IDENTIFICATION
OF VARIOUS EMERALD
FILLING SUBSTANCES
By Mary L. Johnson, Shane Elen, and Sam Muhlmeister
Criteria for distinguishing emerald filling substances were investigated. Thirty-nine fillers
were divided into six substance categories
three presumed natural (essential oils
[including cedarwood oil], other oils, waxes)
and three artificial resin (epoxy prepolymers, other prepolymers [including UV-setting
adhesives], polymers). Regardless of their
composition, fillers with R.I.s of 1.54 or above
show flash effects in emeralds. On the basis of
Raman and infrared spectroscopy, the fillers
could be separated into five spectral groups, A
through E. Most, but not all, commonly used
artificial resins have spectra distinct from that
of cedarwood oil. However, the detection of
one substance in a fissure does not imply that
all others are absent.
ABOUT THE AUTHORS

Dr. Johnson is manager of Research and
Development, and Mr. Muhlmeister is a
research associate, at the GIA Gem Trade
Laboratory in Carlsbad, California. Mr. Elen is
a research gemologist at GIA Research in
Carlsbad.
Please see acknowledgments at the end of the
article.
Gems & Gemology, Vol. 35, No. 2, pp. 82 107
82
Emerald Filling Substances
he vast majority of emeralds on the market today

have undergone clarity enhancement; identification
laboratories estimate that about 90% of the emeralds
they see (most of which are high quality) have been clarity
enhanced in some fashion (Genis, 1997; Hnni et al., 1997).
In recent years, however, there has developed a growing
need to reach a consensus on acceptable trade practice
regarding such clarity enhancements, and to find appropriate ways to disclose this information in the marketplace.
If clarity enhancement is a cause for concern, why is it
done? For one thing, as they are recovered, most emeralds
have many fissures (see, e.g., Sinkankas, 1981; Ottaway et
al., 1994). In addition, there are many more potential emerald buyers with high expectations of color and clarity (see,
e.g., figure 1) than there are untreated gems to satisfy this
demand. Filling substances that almost match the refractive
indices of the host emerald make the fissures less noticeable, especially to the unaided eye (figure 2). Even colorless
fillers improve the color appearance: The light that would
scatter from an unfilled fissure (and thus lighten the apparent color) is no longer visible (Ringsrud, 1983).
Major colored stone trade organizationssuch as the
International Colored Gemstone Association (ICA) and the
American Gem Trade Association (AGTA)as well as the
international jewelry organization CIBJO, recommend that
gemstone treatments be disclosed at every stage of the distribution process. However, the trade continues to debate
exactly what should be disclosed and how it should be disclosed. In particular, emeralds are less salable in some markets if they are perceived to contain some filler substances
such as Opticon (an epoxy resin) rather than others such
as cedarwood oil. However, one proposal to distinguish oiling from enhancement with epoxy resins was voted down
by ICA members as unworkable (Congress votes. . . ,
GEMS & GEMOLOGY
Summer 1999
Figure 1. The desire for

beautiful emeralds has
fueled the practice of filling surface-reaching fissures with oils or resins.
This jewelry demonstrates
the irreplaceable beauty
that emeralds provide.
The 5.79 ct oval-cut emerald on the left is surrounded by 2.84 ct of diamonds,
while the 5.40 ct emerald
on the right is surrounded
by 2.59 ct of diamonds.
Rings courtesy of Henry
Dunay; photo Harold &
Erica Van Pelt.
1995). In 1997, ICA decided to delay an emerald

promotion due to lack of a consensus on the enhancement issue (Review on ruby promotion, 1997).
A number of reasons have been given for the
need to determine the nature of a particular filling
material. First, there is the growing body of anecdotal evidence that various fillers age differently,
thus changing the appearance of the enhanced
emerald over time. For instance, Ringsrud (1998)
reported that the artificial resin palm oil (also
called palma) turns white within a few months in
about 20% of the emeralds filled with this substance. Federman (1998a) noted that traditional
natural oils and resins (such as cedarwood oil and
Canada balsam) could dehydrate or leak out.
Kammerling et al. (1991) experimentally determined that Opticon treatment was somewhat more
durable than cedarwood oil and Canada balsam, but
that all three treatments were affected by routine
cleaning and jewelry-manufacturing processes. In
addition, due to the variety of growth conditions,
emeralds from different localities may not respond
similarly to the same filler. Last, different cultures
may find some fillers more acceptable than others.
The purpose of this article is to critically evaluate the effectiveness of various methods for distinguishing among clarity-enhancement substances.
The main methods currently in use are the flash
effect (usually seen with magnification), Raman
microspectrometry, and infrared spectroscopy. This
investigation is part of an ongoing study of emerald
fillers; other aspects of emerald enhancement will
be addressed in future articles.
HISTORICAL BACKGROUND
Clarity Enhancement in Emeralds. Emeralds from
all sources generally contain fissures, and emerald
processing (e.g., mining, fashioning, and jewelry
manufacturing) can add fractures as well. The filling
of such surface-reaching features (hereafter called
fissures) to make them less visible has been practiced for centuries. The first description of oiling
of green gems (probably turquoise, but possibly
emeralds) was made by Pliny the Younger in about
55 AD. At least by the 14th century, reference was
being made to the oiling of emeralds in particular
(Nassau, 1994).
The first references to emerald treatments and
their detection in the gemological literature are
much more recent. Nassau (1994) cites the sixth edition of Liddicoats Handbook of Gem Identification
(1962) as the first gemology book to mention oiling,
although it was occasionally mentioned earlier in
Gems & Gemology (Crowningshield, 1958,
19601961; Benson, 1960). Subsequently, Gems &
Gemology published two comprehensive articles
describing some of the techniques and substances
used to clarity enhance emeralds (Ringsrud, 1983;
Kammerling et al., 1991). Hnni (1988) described
the detection of filled fissures.
Disclosure. We examined records from two major
auction houses, Christies and Sothebys, and found
that disclaimers regarding filling substances in
emeralds were first published in November 1992
(Sothebys, 1992a, 1992b) and November 1993
(Christies, 1993). The clarity enhancement of a par-
GEMS & GEMOLOGY
Summer 1999
83
ticular emerald was first noted in October 1993

(Sothebys, 1993). Although various classes of possible filling materialsoil, natural resin, and artificial
(or synthetic) resinwere noted in these early disclaimers, the specific substance categories were not
disclosed for individual emeralds (see Box A for a
discussion of filler substance classification). The disclosure statements were made more explicit in
early 1997 (Federman, 1997). However, specific
types of enhancements (e.g., oil type) were first
noted in October 1997 (Christies, 1997a).
Clarity Enhancement Substances. Although there is
almost universal agreement that most emeralds
need some treatment to be salable, there is still little agreement as to which filling substances are
acceptable and why one substance is preferable to
another. A good filling substance should have certain properties: It should hide fissures, it should
flow into the fractures (i.e., be liquid, at least initially), it should hold up over time or else be easy to
restore, and it should be removable or not have any
physical properties that might later harm the stone
(e.g., during jewelry repair or repolishing). It should also
have an R.I. similar to that of the stone being treated.
Several substances have been used by the trade
to fill fissures in emeralds. M. Kostetski (pers.
comm., 1999) used linseed oil as an emerald filler in
the 1930s through 1960s, and C. Altobelli (pers.
comm., 1999) used rapeseed oil (now usually called
canola oil) from the late 1940s onwards.
Crowningshield (1958) mentioned 3-in-1 penetrating oil, whale oil, and mineral oil. Ringsrud (1983)
noted clove oil in addition to Merck cedarwood oil
and Merck Canada balsam. Hnni (1988) mentioned
also sperm whale oil, paraffin, and an unspecified
artificial resin. Kammerling et al. (1991) noted that
cedarwood oil and Canada balsam were considered
traditional filling materials, with Opticon 224
resin (generally only surface hardened) as a newer
substitute. Kennedy (1998) described clarity
enhancement of emeralds in Brazil: Although baby
oil and cedarwood oil (cedro) were mentioned,
Brazilian goods were usually treated with
Opticon. (It is important to remember that, as
Federman [1998b] cautioned, not all epoxy resins are
Opticon, which is a specific substance with a registered trade name; nevertheless, Opticon is often
misused as a generic term in the trade.) Ringsrud
(1998) wrote that a synthetic resin known as
palm oil was used heavily in Colombia from 1990
to 1996. Three widely discussed enhancement sub-
84
stances at the present time are cedarwood oil,

Opticon resin, and palm oil (Box B).
In fall 1997, we asked 33 emerald dealers worldwide to tell us which emerald treatments they
were aware of. We received 12 responses (one each
from Israel, Italy, and Switzerland, and the rest
from the U.S.). Each of the following substances
was mentioned by at least one dealer: natural and
synthetic oils and resins, Opticon 224 resin (with
or without hardener), green-colored oil, wax,
paraffin, cedarwood oil with hardener, WD
40, Joban oil, Yehuda treatment, Arthur
GroomGematrat treatment, linseed oil, epoxy
fillers, and plastic sealers.
To many, the most important distinction is that
between oils and artificial resins. As summarized by
Chalain et al. (1998): the use of artificial resins to
fill emeralds has only been significant in the last fifteen years. . . . Unlike oils, artificial resins act to
consolidate emeralds. . . . It is easy to replace oil
that oozes out, but hard to get rid of altered resins.
The Guide reported that 64% of the dealers surveyed
at Tucson in February 1997 thought that the clarity
enhancement of emeralds with colorless oil was
acceptable, while fewer than 48% found Opticon
treatment acceptable (Survey: Emerald treatments, 1997). However, no natural filler currently in use is found within emeralds in nature; all are
added during clarity enhancement (Genis, 1997).
From our discussions with members of the trade,
we also know that certain fillers are used not only
individually, but in mixtures as well. For example,
treaters have experimented with mixtures of
Canada balsam and cedarwood oil (Ringsrud 1998)
to see if an all-natural material can replace cedarwood oil as an emerald filler. Such experiments
were prompted by reports of a change in the viscosity of Merck cedarwood oil, which had been evident
since the late 1980s (as reported by Johnson and
Koivula, 1998b). It is also possible that mixtures of
oils and artificial resins have been used.
In addition, many stones have undergone several
generations of enhancement (i.e., treatment, cleaning, and re-treatmentsometimes with a new substance). Residues of earlier treatments may be left
deep within an emerald. This further complicates
the identification process (H. Hnni, pers. comm.,
1998; C. Smith, pers. comm., 1999).
Previous Work on Characterizing Fillers.
Gemological determination of clarity enhancement
in emeralds has been described by GIA researchers
GEMS & GEMOLOGY
Summer 1999
AFTER
BEFORE
Kammerling et al. (1991) and Johnson et al. (1998a),
as well as in several Gems & Gemology Gem Trade
Lab Notes (see, e.g., Fryer et al., 1984; Hurwit, 1989;
Kane, 1990; Kammerling, 1993; and Kammerling et
al., 1995). One distinguishing feature is the flash
effect seen with magnification in some filled emeralds, which is similar to that seen in fracture-filled
diamonds (Kammerling et al., 1994).
Two relatively recent technological advances
have made possible the characterization of specific
filling materials within gemstones. The first was
the development of a Raman spectrometer designed
for gemological use (the Renishaw laser Raman
microspectrometer; see, e.g., Hnni et al., 1996a).
The second was the adaptation of Fourier-transform
infrared (FTIR) reflectance spectroscopy to probe
within an emerald (see, e.g., Zecchini and Maitrallet, 1998).
In 1994, the Asian Institute of Gemmological
Sciences (AIGS; Bangkok, Thailand) was among the
first gemological laboratories to use a Raman
microspectrometer for gemological characterizations. The Raman spectra of Opticon 224 prepolymer and polymer, and several other fillers, were presented at an ICA-organized meeting in Bangkok in
1994 (K. Scarratt, pers. comm., 1999). Mr. Scarratt
currently offers filler identification services at the
AGTA Gemological Testing Center, in New York
City.
The SSEF Swiss Gemmological Institute has
been particularly open in sharing the criteria that
they use for filler identification (see, e.g., Hnni et
al., 1997; Hnni, 1998; SSEF Swiss Gemmological
Institute, 1998; Weldon, 1998b). The techniques
employed by SSEF include microscopy, observations of fluorescence, reflectance infrared spectrometry, and (since 1995) Raman microspectrometry.
Hnni et al. (1996a) described four substances that
are used to fill emeralds: cedarwood oil, an artificial
resin, universal oil, and a wax. Since 1988, SSEF
has examined at least 13 different artificial resins
(seven epoxy and other hardening resins, and six
UV-setting resins; Chalain et al., 1998).
Zecchini and Maitrallet (1998) have used
reflectance infrared spectroscopy to characterize
eight potential emerald-filling materials and two
related substances, both as loose materials and as
fillers in emeralds. This research team represents a
collaboration between the Universit de FrancheComt and the Service du Contrl des Diamants,
Pierres et Perles of the Paris Chamber of
Commerce.
R.I. = 1.500
R.I. = 1.517
R.I. = 1.531
R.I. = 1.550
R.I. = 1.570
Figure 2. These before-and-after photographs illustrate that the fissures in emeralds become less visible as the R.I. of the filling material increases.
Shown are fillers with R.I.s of 1.500, 1.517, 1.531,
1.550, and 1.570. The refractive indices of the five
emeralds ranged from 1.569 to 1.580. Photos by
Maha DeMaggio.
GEMS & GEMOLOGY
Summer 1999
85
BOX A: CLASSIFICATION OF FILLING SUBSTANCES:

OILS, ESSENTIAL OILS, RESINS, PREPOLYMERS, AND POLYMERS
Several substances have been used in the clarity
enhancement of emeralds. Some are called oils
(e.g., Cedarwood, paraffin, whale, 3-in-1, palma),
and some resins (e.g., Canada balsam, unhardened Opticon 224). Since the substances used to fill
emeralds originally come from various branches of
chemistry (from perfumes to polymers), with somewhat different definitions for each field, it is important to create a consistent terminology for use by
gemologists.
The following definitions are taken from technical and popular dictionaries, including Websters
Ninth New Collegiate Dictionary (Websters, 1987)
for everyday English; Stecher et al. (The Merck
Index, 1968), Alger (1989), and Brady and Clauser
(1986) as general chemical and technical dictionaries; Sivry (1985) for European usage; Elias (1993) and
Lee and Neville (1967) for plastics; Herout (1982) for
perfumes; and Mutton (1982) for the wood products
industry. Information on the trees from which certain
filling materials are derived is from Lincoln (1986).
are extracted from their host plants using solvents.

According to the Polymer Science Dictionary
(Alger, 1989), an essential oil is the predominantly
volatile material isolated by some physical process
from an odorous single-species botanical. Over
3000 oils have been identified. . . . Essential oils are
generally liquid at room temperature; however,
some are semisolid and several are solid. . . .
Essential oils are made up of carbon, hydrogen, and
oxygen, and occasionally nitrogen and sulfur. . . . It
is not uncommon for an essential oil to contain
over two hundred components.
The main components of many essential oils
have been synthesized (again, including cedarwood
oilsee, e.g., Stecher et al., 1968), and commercial
essential oils may be partially or wholly synthetic.
Because of their volatility, open containers (such as
an emerald) of essential oils cannot be stable over
time, since the fragrant components are continuously lost to the atmosphere (or they would not
reach our noses, to be smelled).
OIL
Oil originally meant olive oil, which is not generally used in emeralds today. Its common meaning is
one of numerous unctuous combustible substances that are liquid or at least easily liquefiable
on warming, are soluble in ether but not in water,
and leave a greasy stain on paper or cloth
(Websters, 1987). However, oil means one thing to
petroleum chemists, who use it to refer to longchain hydrocarbons of a certain weight (e.g., paraffin oil); something else to cooks and nutritionists,
who use it to refer to triglycerides (e.g., olive oil,
sesame oil); and something else again to perfume
chemists, who use it to refer to essential oils (e.g.,
cedarwood oil, clove oil; see below).
Other Oils. Because essential oils differ from

other oils (such as mineral and vegetable oils) in
many respects (volatility, viscosity, chemistry, and
infrared and Raman spectra), we have treated these
as separate categories when reporting the results of
our research.
Essential Oils. Some compoundssuch as cedarwood oilare essential oils. These compounds
MATERIALS AND METHODS

Almost every chemical product (including natural
compounds and extracts) available in the U.S. has a
Chemical Abstracts Service (CAS) registry number,
andif there is any possibility of a hazard associated with the materiala Material Safety Data Sheet
(MSDS). Details about a chemical compound can be
found in many databases if the CAS registry number is known; this is accessible on the World Wide
Web at www.cas.org. (Another useful database,
86
RESIN
Resin originally meant pine sap. Today, resin can
mean (at least) three different things: natural plant
exudates (i.e., saps), either hardened or unhardened;
hardened manufactured polymers; or the unhardened prepolymer building blocks that can be
used to make manufactured polymers (Websters,
1987; Stecher et al., 1968; Alger, 1989). Natural
resins can harden (polymerize) over time (e.g.,
amber, copal). The emerald filler Canada balsam is
generally considered a natural resin, and the essen-
ChemFinder, is available at www.chemfinder.com.)

The MSDS sometimes contains information on the
color, odor, specific gravity, and other properties of
the material.
Safety is an issue when working with any
chemical substance: The MSDS discloses hazardous constituents and their associated risks. Many
of the substances in this article pose slight-to-high
acute (short-term) health risks. For the most part,
they can affect skin, eyes, mucous membranes,
GEMS & GEMOLOGY
Summer 1999
tial oil cedarwood oil is considered a resin by the

wood pulp industry (Mutton, 1982).
Polymers and Prepolymers. A polymer is a large
molecule made up of repeating units of smaller
molecules. The smallest such unit is a monomer
(figure A-1). We prefer to use the term prepolymers
for the small units that are assembled into polymers, since they could be either monomers or a few
monomers attached together. They polymerizeor
hardenwith the use of a chemical catalyst, illumination, heat, or time.
Specific polymers and prepolymers are named
after the parts of their chemical structures that
attach these units together. Epoxy resins have two
carbons bonded to each other and to the same oxygen; bonds in this epoxide group break in order for
the prepolymers to be linked. Most artificial resins
used to fill emeralds (e.g., Opticon 224, Epon 828,
Araldite 6010) are based on one type of epoxy
molecule, diglycidyl ether of bisphenol A (DGEBA).
They are introduced into the emeralds as prepolymers, and then may be polymerized later. However,
UV-setting adhesives, which are also used to fill
gems, are based on different chemical compounds,
usually methacrylates. These also polymerize, but
their infrared and Raman spectra are different from
those of the epoxies.
Synthetic versus Artificial Resin. Chemists
use the term synthetic differently from gemologists: For a chemist, a synthetic chemical is one
that was made from other chemicals, regardless of

whether or not it has a natural analog; in gemology,
only a material with a natural analog can be called
a synthetic. Some of the natural filler materials we examined contained synthetic ingredients
(e.g., one of the cedarwood oils; see Box B), or were
stated to be completely synthetic (Sigma cinnamon
oil). We know of no nondestructive technique to
determine whether a filler with a natural composition within an emerald is completely natural.
As a consequence, we refer to the possibly natural
materials in this article as presumed natural,
since there is no way to independently guarantee
that they are completely natural.
The emerald trade frequently uses the terms
resin or synthetic resin to refer to epoxy prepolymers and polymers such as Opticon 224, Epon 828,
and Araldite 6010. As we have seen, both resin and
synthetic are ambiguous. However, artificial resin
is an unambiguous term that can be used for both
liquid prepolymers and solid polymers.
WAX
Another material sometimes encountered as a filler
(or luster enhancer) in gems is wax. This term originally meant beeswax, but it now also includes
various similar substances (Websters, 1987) with
higher molecular weights, on average, than oils.
Common waxes include: beeswax, spermaceti
(from the sperm whale), vegetable waxes, and mineral waxes (including paraffin wax; Brady and
Clauser, 1986).
Figure A-1. Snap-Lock beads illustrate the

concept of polymerization. Polymers are
large molecules (right) made from smaller
identical piecesmonomers (here, one
bead) or other prepolymers (a few linked
monomers; here on the left, two beads).
Usually, prepolymers are liquid and polymers are solid. Photos by Maha DeMaggio.
and lungs; some can cause mutations in dividing

cells; and most are flammable. Persons wishing to
experiment with any of these substances should
consult the relevant MSDS and follow recommended procedures scrupulously. Although most
emerald filling substances present some health
hazard to the experimenter, they are probably not
dangerous to the ultimate consumer. (However,
some people are allergic to cedarwood oil; Bleumink et al., 1973.)
Filling Substances. For this study, we chose fillers

that we knew were commonly used in the trade
(e.g., Baker and Merck cedarwood oils, Opticon 224
resin), some that have been used previously (e.g.,
clove oil; Ringsrud, 1983), some that were stated to
be in use now (e.g., UV-setting adhesives; C. Osorio,
pers. comm., 1998; D. Allen, pers. comm., 1998),
and some that were easily available but not known
to be in use (e.g., sesame oil, Epo Tek prepolymers).
We selected the fillers in this last set because they
GEMS & GEMOLOGY
Summer 1999
87
Box B: CEDARWOOD OIL,

OPTICON, AND PALM OIL (PALMA)
Three of the most-discussed enhancement substances at this time are cedarwood oil, Opticon 224,
and palma. Cedarwood oil from one source
(Merck) and Opticon 224 were profiled by
Kammerling et al. (1991).
WHAT IS CEDARWOOD OIL?
Cedarwood oil is one of the most widely accepted
filling substances today, both because it is perceived as a natural oil, and because it has a higher
R.I. and viscosity than other traditional oils (e.g.,
linseed, rapeseed oil). Detailed information about
this essential oil comes from the perfume industry
(see, e.g., Herout, 1982), where it is used in the production of perfumes and soaps; it is also used in
medicine and microscopy, and as an insecticide
(Stecher et al., 1968).
Natural sources. Cedarwood oil is extracted from
several coniferous trees, especially the junipers
Juniperus virginiana (Virginia cedarwood oil), J. procera (Kenya cedarwood oil), and J. mexicana (Texas
cedarwood oil), among others; however, some
cedarwood oil is still produced from the true cedar
Cedrus atlantica (Atlas cedarwood oil; Herout,
1982). Both Texas and Atlas cedarwood oils have
the same CAS registry number, 68990-83-0, while
most other cedarwood oils have the CAS registry
number 8000-27-9.
Composition. Cedarwood oil is primarily composed
of tricyclic compounds (molecules with three carbon rings), including cedrol, cedrene, and cedrenol.
All these chemicals are synthesized, and commercial cedarwood oils may contain either the synthetic or natural compounds (or both) plus other (synthetic and/or natural) ingredients (Stecher et al.,
1968; Herout, 1982).
We looked at five commercially available cedarwood oils from four different sources, all of which
were suggested to us by individuals in the emerald
trade. These were: Baker cedarwood oil for immersion, EM (Merck) cedarwood oils for immersion and
for clearing, Shemen Tov (Texas) cedarwood oil, and
aceite de cedro from Antonio Negueruela S.A.
fit into substance categories commonly used as

emerald fillers and they have R.I.s close to those of
emerald.
Rather than refer to each substance individually,
88
We inquired about the composition of each

cedarwood oil from its distributor, and received
replies from Russell Lance at Mallinckrodt Baker
Inc., Rande Klein at EM Science (Merck), Lee Saal
at Shemen Tov, and Antonio Negueruela at
Antonio Negueruela S.A. Because some of the
information we received was proprietary, we will
not describe their responses in detail. None of the
distributors manufactured the cedarwood oil themselves, and all regarded their sources as proprietary.
Three distributors stated that the material was all
natural, but the fourth acknowledged a considerable amount (but less than 50%) of synthetic ingredients. Another distributor mentioned that their
cedarwood oil also contained rosin (tree sap from
which the turpentine has been distilled, according
to Brady and Clauser, 1986) and a castor oil solubilizer. None contained Canada balsam. No distributor specified which species of tree the cedarwood
oil came from, or the extraction technique used.
Despite the differences in composition, four of the
cedarwood oils had the same CAS registry number,
8000-27-9; the Shemen Tov (Texas) cedarwood oil
had CAS registry number 68990-83-0.
We also asked the distributors to explain the difference between cedarwood oil for clearing and that
for immersion. They replied that these have slightly different formulations because they have different uses in biological microscopy. Cedarwood oil
for immersion has a higher refractive index (see
table 1 in text), and is the substance typically used
to fill emeralds.
Physical and Optical Properties. These are given in
table 1 in the text; all were somewhat variable.
Infrared spectra (figure B-1) were similar for all but
the Merck cedarwood oil for clearing, which had a
spectrum closely resembling the spectra of Norland
Optical Adhesives 63 and 65. Raman spectra (also
in figure B-1) showed small variations in detail,
with the Baker cedarwood oil the most distinctive.
In short, there are many slight differences
among cedarwood oils, as this substance is not a
single chemical compound and it can come from
several sources. Although some of these differences
we sought to establish categories of related materials (e.g., oil, resin). However, when we looked into
the definitions of these categories, we found ambiguity, both in the meanings attributed to specific
GEMS & GEMOLOGY
Summer 1999
are detectable within emeralds, we do not yet know

if any are significant.
WHAT IS OPTICON?
The term Opticon should properly refer only to the
epoxy resin Opticon 224, manufactured by Hughes
Associates, Victoria, Minnesota. This artificial
resin may be used in emeralds as a liquid prepolymer or a hardened polymer. The most common
procedure (Kammerling et al., 1991) is to fill emeralds with liquid Opticon and brush the catalyst on
the surface to provide a hardened surface layer.
Because Opticon 224 is a DGEBA (see Box A) epoxy
resin, it cannot be separated easily from other
DGEBA resins by spectroscopic means, especially
when it is within an emerald.
WHAT IS PALM OIL OR PALMA?
According to anecdotal evidence (J. Rotlewicz and
R. Giraldo, pers. comm., 1998), when the viscosity
of Merck cedarwood oil changed in the 1980s,
Colombian emerald laboratories began using

palma or palm oil instead. This substance
more closely matched the R.I.s of emerald than
cedarwood oil diddisguising the fissures well
but it was distinguished from cedarwood oil by its
flash effect (Ringsrud 1998).
At first, dealers believed from the name that
this was another natural oil. Palm oil is aceite
(or azeite) de dende in Spanish, but castor bean
oil has been translated as aceite de palma.
However, both of these substances have relatively
low refractive indices (about 1.475), so they would
not show a flash effect.
The material represented as palma is probably
an unhardened epoxy prepolymer resin. Both Araldite
6010 and Epon 828 have been proposed as the identity of palma (the former by A. Groom, pers. comm.,
1997, and by C. Osorio, J. Rotlewicz, and R. Giraldo,
pers. comms., 1998; the latter by C. Beesley, pers.
comm., 1998). The R.I.s and infrared and Raman
spectra of these substances are nearly identical.
Figure B-1. The five cedarwood oils we examined had slightly different infrared and Raman spectra.
terms and in the ways these terms are applied by

different branches of chemistry (again, see Box A).
Nevertheless, on the basis of the MSDS information, we classified the fillers we studied into the fol-
lowing six substance categories: essential oils

(including natural resins), other oils, waxes, epoxy
prepolymers, other prepolymers (i.e., which have
different chemistries, including UV-setting adhe-
GEMS & GEMOLOGY
Summer 1999
89
sives), and (hardened) polymer resins. The first three

categories contain both natural and manufactured
substances; the last three contain only manufactured substances. The two hard polymers Super
Tres and Permasafe are similar in their properties
and spectra to hardened Opticon 224 (see below), so
we suspect these are epoxy (DGEBA) polymers, at
least in part.
The 39 materials studied are listed according to
these substance categories in table 1. We examined
the liquid filling substances both as droplets and as
samples mounted between two glass microscope
slides (Corning pre-cleaned micro slides, #2927).
The solid substances were examined both as polished blocks and as KBr pellets for infrared spectroscopy. Our investigation of isolated fillers is consistent with the recommendation by Zecchini and
Maitrallet (1998) that the spectra of the substances
alone should be examined before the fillers are
investigated within emeralds. To answer key questions about filler mixtures, however, we did study
some materials within emeralds.
We observed the color of each filling substance
on a white background, using daylight-equivalent
fluorescent light. We determined the refractive
index of each liquid by placing a drop on the hemicylinder of a Duplex II refractometer equipped with
a sodium-equivalent light source; we polished small
flats on blocks of the solid materials to get their R.I.
values. We observed fluorescence in a darkened
room using a GIA GEM short-wave/long-wave
ultraviolet lamp.
We assessed the approximate specific gravity of
each filler (relative to de-ionized water at 21.5C) by
inserting a few drops (or, if solid, a small fragment)
of the substance below the meniscus of a small vial
of water, and observing whether the material sank
or floated to the surface. (Although this test is not
useful to identify fillers in emeralds, many dealers
do have access to the material that is being used to
fill their stones, and this test may help confirm the
identity of that filler.) All observations of viscosity
were made at room temperature; the flow characteristics were observed by tilting a transparent container of the material and observing whether the
surface shifted back to level on a short (like water)
or a long (like honey) time scale.
Experiments with Mixtures. For two specific experiments, we investigated some properties of three
groups of our own filler mixtures, both as isolated
substances and within emeralds. These experiments
90
were designed to answer two questions: (1) Does a

flash effect imply an artificial resin? And (2) how
well can adulterated cedarwood oil be detected
with spectroscopic techniques?
We addressed the first question with the following substances, chosen (or mixed) for their refractive
indices (table 2): the epoxy prepolymer Epo Tek 314
(R.I. = 1.500); Baker cedarwood oil (R.I. = 1.517);
Sigma clove oil (R.I. = 1.531); mixtures of Epo Tek
314 and Epo Tek 302-3M (R.I.s = 1.517, 1.531,
1.550, 1.570); and mixtures of Sigma clove oil and
Sigma cinnamon oil (R.I.s = 1.550, 1.570). We also
examined the stones filled with these substances to
determine the effect of filler R.I. on apparent clarity
and to observe the appearance of the fillers within
fissures. We addressed the second question using
mixtures of two components with very different
spectral characteristics: Baker cedarwood oil and
Araldite 6010 (table 3).
We made 50 ml batches of each filler mixture.
The proper volume of each component was determined by assuming a linear relationship between
volume and refractive index. The substances were
mixed shortly prior to filling the stones, and the R.I.
of each mixture was measured.
From a batch of about 250 natural emeralds
approximately 0.5 ct and larger, which we had
selected for durability testing, we chose 18 to be
cleaned, photographed, and filled. Each was first
emptied by Arthur GroomGematrat in New
York, using a proprietary process, and then examined with a Reichert Stereo Star Zoom microscope
to check that the previous filler material had been
removed. Both macro- and micro-photography were
performed before and after filling on the nine stones
for the first experiment; only macro-photography
was done on the other nine. Refractive indices were
measured and fluorescence was observed with the
above-listed equipment. Care was taken with the
R.I. determinations, to ensure that the R.I. liquid
was not drawn (or wicked) into the fissure. We
determined the weight of each stone before and
after filling (in carats, to five decimal places) using a
Mettler MT5 balance.
For the first experiment (table 2), we filled one
each of nine emeralds using a Mini Oiler (Zamrot
Ashalim Engineering Ltd., Ramat Gan, Israel) for
the essential-oil mixtures and a Color Stone Oiling
Unit (Jairo Vaca Camacho, Bogot) for the artificial-resin mixtures. In the Mini Oiler, samples were
held in an evacuated container (i.e., one from which
the air had been pumped out) and suspended in a
GEMS & GEMOLOGY
Summer 1999
porous basket over the oil, which was heated to

about 5060C. The stones were then plunged into
the filler, and 700 mbar overpressure (1.7 bars total
pressure) was applied overnight. In the Color Stone
Oiling Unit, the emeralds were placed in the bottom of the container and resin was poured over
them; the chamber was heated to 5060C and
then pressurized to 4000 psi (about 275 bars total
pressure) overnight. After the stones were cooled to
room temperature, each device was brought to
atmospheric pressure and the stones were removed,
cleaned with alcohol, and then weighed and rephotographed.
For the second experiment, we examined the
mixtures of Baker cedarwood oil and Araldite 6010
both as loose substances and as fillers within emeralds. The measured liquids were poured into brown
glass bottles and stirred thoroughly. The three mixtures richest in cedarwood oil and the two richest in
Araldite 6010 combined smoothly, but the four
mixtures with 30 to 60 volume percent cedarwood
oil unmixed overnight at room temperature (figure 3). The unmixed fillers were stirred again immediately prior to filling the stones. We filled nine
emeralds with these mixtures, using the Zamrot
Mini Oiler and the same conditions as above.
Infrared Spectroscopy and Raman Microspectrometry. We analyzed the nine emeralds
filled with mixtures in table 3 with a Nicolet
Magna 550 Fourier-transform (FTIR) spectrometer, using reflectance infrared spectrometry as
described by Zecchini and Maitrallet (1998). For
this techique, we adjusted the beam to focus
within the emerald, rather than at its surface, to
generate a transmission spectrum. We used a
known clean emerald standard for comparison:
a 5.01 ct crystal collected in situ from the wall of
a mine at Muzo, Colombia, by GIAs Bill
Boyajian. We also compared the results for the
emeralds filled with these mixtures with those
for emeralds we had filled with 100% Baker
cedarwood oil or 100% Araldite 6010 as part of
the durability study.
Infrared spectra of the loose fillers in tables 1 and
3 were also produced with the Nicolet Magna 550
spectrometer. Spectra were taken of thin films of
the liquid fillers made by compressing a drop of the
filler between two glass microscope slides, each
about 1 mm thick. This collection geometry was
chosen so that the strongest infrared-active absorptions would not be too intense to show any fea-
tures. The IR spectra of solid fillers were obtained

using the KBr pressed-powder technique.
Spectra were collected over the range 6000400
cm1. We chose a spectral resolution of about 4 cm1;
in general, absorption peaks in polymers are wider
than this (5 to 30 cm1), so peaks can be reported to
the nearest cm1 without loss of accuracy (Bower
and Maddams, 1989, p. 18). We also examined different resolutions for FTIR spectra of some representative fillers and found no differences in peak position
and width for 4 cm1 as compared to 1 cm1 resolution. Because the glass slides had only broad, low
background absorption in the region of interest
(32002400 cm1)that is, the region in which
emeralds are transparentinfrared absorption related to the glass itself did not interfere with the collection of important data, and we observed that the
spectra of the loose fillers were practically identical
to those taken of the same fillers in emeralds as
reported in table 3.
Raman spectra were produced of all the loose
fillers (both isolated and as mixtures) and of fillers in
selected fissures of the filled emeralds in tables 2 and
3. We used a Renishaw 2000 Ramascope laser
Raman microspectrometer, with a 514.5 nm argon
laser source and five summed 10-second scans on
each point. We placed a drop of each liquid filler on
a clean piece of aluminum foil, and then analyzed it
in the spectral range 1003500 cm1. For the loose
fillers, we used low (20) magnification to prevent
selective volatilization of the filler substances. In
the filled emeralds (tables 2 and 3), we recorded the
spectra of fillers in selected fissures at 50 magnification, in the same spectral range. Our Raman spectra had a resolution of about 1 cm1.
RESULTS
Physical and Optical Characteristics of Loose Filling
Materials. All of the substances tested were found
to have many properties in common. In some cases,
we saw slight differences that might be useful to
help determine the substance category or specific
filler used in a particular emerald. Properties are
listed for the fillers, grouped by substance category,
in table 1; significant differences are discussed
below.
Color. The Joban oil and green Opticon 224 resin
were moderate green, and the Negueruela cedarwood oil was very light green. The rest were colorless or very light, light, or moderate yellow.
GEMS & GEMOLOGY
Summer 1999
91
TABLE 1. Description and gemological properties of the emerald-filling materials examined for this study.
Substance
category
Filler a
Source b
Used in
the trade? c
PRESUMED NATURAL SUBSTANCES

N.R.
ESSENTIAL Cedarwood oil Merck
for clearing
OILS OR
NATURAL
Cedarwood oil Baker
Yes
RESINS
for immersion
Cedarwood oil Negueruela Yes
Cedarwood oil
for immersion
(Texas) cedarwood oil
Canada balsam
(natural resin)
Clove bud oil
Merck
Yes
Shemen
Tov
Sigma
Yes
Yes
Spectrum
N.R.
Clove oil
Sigma
Color
R.I.d
Long-wave UV
fluorescence
Light
yellow
Yellow
1.509
Inert
Inert
1.517
Strong
yellowish white
Moderate white
Weak yellowish
green
Inert
Weak yellowish
green
Inert
Inert
Strong yellowish white

Very weak
yellowish green
Very weak
greenish yellow
Very weak
yellowish green
Inert
Very weak
yellowish green
Weak yellowish
green
Weak yellowish
green
Weak yellowish
green
Inert
Very weak
green
Inert
Very light
green
Very light
yellow
Light
yellow
Yellow
1.518
1.520
1.520
1.521
Short-wave UV
fluorescence
Inert
Relative
viscosity e
S.G.f
Flows
sluggishly
Flows
sluggishly
Semi-solid
Floats
C, 11l
C, 12l
Floats
C, 11n
C, 12n
Floats
C, 11n
C, 12o
Floats
C, 11n
C, 12o
Floats
C, 11n
C, 12o
Floats
C, 11n
C, 12n
Sinks
D, 11p
D, 12q
Sinks
D, 11p
D, 12q
Sinks
D, 11p
D, 12q
Sinks
D, 11q
E, 12r
Sinks
D, 11p
None
Floats
B, 11f
B, 12f
Floats
B, 11f
B, 12f
Floats
B, 11g
None
Floats
B, 11g
B, 12g
Floats
B, 11g
B, 12g
Floats
B, 11g
B, 12g
Floats
B, 11g
B, 12g
Floats
B, 11h
B, 12h
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
Flows
easily
Flows
easily
Flows
easily
Flows
easily
Flows
easily
Clove stem
Spectrum
oil
Cinnamon oil Sigma
(synthetic)
Cinnamon oil Spectrum
(cassia)
N.R.
Very light 1.531

yellow
Light
1.531
yellow
Yellow
1.537
N.R.
Yellow
1.589
N.R.
Yellow
1.589
Mineral oil
Spectrum
Yes
Colorless 1.478
Inert
Inert
Paraffin oil
Schroeder
Yes
Colorless 1.478
Inert
Inert
Mineral(?) oil Isocut fluid Buehler
N.R.
Colorless 1.449
Inert
Inert
Vegetable oil Sesame oil
No
1.474
Inert
Inert
No
Light
yellow
Yellow
1.473
Moderate
greenish yellow
Very weak
greenish yellow
Flows
easily
Flows
easily
Flows
easily
Flows
easily
Flows
easily
N.R.
Colorless 1.479
Real Gems
Yes
Green
Moderate
yellowish green
Moderate
yellowish green
Very weak
yellowish green
Weak yellowish
green
Flows
easily
Flows
easily
Spectrum
Yes
Colorless About
1.52
Inert
Inert
Solid
Rarely
IR group, Raman group,

spectrum g spectrum g
OTHER OILS
Mineral oil
Health
Valley
Mavalrio
Azeite de
Dende (palm
[tree] oil)
Castor oil
Spectrum
Mineral(?)
Joban oil
oil plus vegetable(?) dye
WAX
Paraffin wax
1.478
a Fillers
in boldface type have R.I.s greater than 1.54 and should show a flash effect; fillers in italics have R.I.s less than
1.500 and should not enhance clarity as effectively as others in the table.
b Sources: Baker (Mallinckrodt Baker Inc., Phillipsburg, NJ); Buehler (Buehler Ltd., Lake Bluff, IL); Centro Gemolgico (Centro
Gemolgico para la Investigacin de la Esmeralda, Bogot, Colombia); CIBA-GEIGY Corp., Hawthorn, NY; Conservators
Emporium, Reno, NV; Epo Tek (Epoxy Technology, Billerica, MA); Health Valley Co., New York City; Hughes
(Hughes Associates, Victoria, MN); Liquid Resins (Liquid Resins International Ltd., Olney, IL); Malavrio (Indstria de Productos
Alimentcios Malavrio Ltda., Proporinha, Brazil); Merck (EM Science, Gibbstown, NJ); Negueruela (Antonio Negueruela S.A.,
Madrid, Spain); Norland (Norland Products Inc., New Brunswick, NJ); Real Gems (Real Gems Inc., New York City); Schroeder
J. Schroeder Pharm., Geneva, Switzerland); Shell (Shell Oil Co., Houston, TX); Shemen Tov (Shemen Tov Corp.,
West Orange, NJ); Sigma (Sigma Chemical Corp., St. Louis, MO); Spectrum Chemical Manufacturing Corp., Gardena, CA;
Treatment World (Treatment World Emerald Gemolgico Universal, Bogot, Colombia).
c N.R. = not reported.
d The R.I. of the filler increased (by about 0.03) on curing for the three cases we examined (Opticon 224, Epo Tek UVO114,
Norland type 65).
e Flows easily = flows like water; flows sluggishly = flows like honey. These were evaluated at room temperature (21.5C).
92
GEMS & GEMOLOGY
Summer 1999
TABLE 1 (contd). Description and gemological properties of the emerald-filling materials examined for this study.
Substance
category
Filler a
ARTIFICIAL RESINS
EPOXY PREPOLYMERS
Epo Tek 314
R.I.d
Long-wave UV
fluorescence
Weak yellowish
green
Inert
Very weak
yellowish green
Inert
Weak to moderate greenish white

Inert
Very weak
yellowish green
Inert
Weak greenish
yellow
Weak yellowish
green
Weak yellowish
green
Weak yellowish
green
Weak yellowish
green
Weak yellowish
green
Very weak
yellowish green
Inert
Short-wave UV
fluorescence
Colorless
1.500
Conservators N.R.
Emporium
Epo Tek
N.R.
Colorless
1.501
Colorless
1.538
Opticon
Resin 224
Opticon Resin
(green)
Araldite 506
Hughes
Yes
Colorless
1.550
Hughes
N.R.
Green
1.550
Sigma
N.R.
Colorless
1.551
Araldite 502
Sigma
N.R.
1.559
Araldite 6005
Sigma
N.R.
Very light
yellow
Colorless
Araldite 6010
CIBA-GEIGY
Yes
1.572
Epon 828
Shell
Yes
Epo Tek 3023M
Epo Tek
N.R.
Very light
yellow
Light
yellow
Colorless
Liquid
Resins
N.R.
Colorless
1.481
Inert
Inert
Epo Tek
N.R.
Colorless
1.527
Green
Norland
Yes
Colorless
1.501
Moderate
green
Weak greenish
yellow
Norland
N.R.
Colorless
1.519
Strong blue
Centro
Gemolgico
Treatment
World
Hughes
Yes
1.565
Yes
Very light
yellow
Light
yellow
Very light
yellow
Strong bluewhite
Strong bluish
white
Strong light
blue
Norland
Yes
Colorless
1.529
Moderate blue
Epo Tek
N.R.
Light
yellow
1.553
Strong blue
Epo Tek 301
OTHER PREPOLYMERS
Liquid Resin
(green cap
formula)
UV-setting Epo Tek
UVO114
Norland
Optical
Adhesive
type 65
Norland
Optical
Adhesive
type 63
POLYMERS
Permasafe
Super Tres
Opticon
Resin 224
(cured)
Norland Optical
Adhesive type
65 (cured)
Epo Tek
UVO114
(cured)
Epo Tek
Used in
Color
the trade? c
N.R.
HXTAL
UV-setting
Source b
Yes
1.570
1.575
1.577
1.570
1.580
Relative
viscosity e
IR group, Raman group,

spectrum g spectrum g
Sinks
C, 11m
C, 12m
Sinks
B, lli
B/C, 12i
Sinks
A, 11c
A, 12a
Sinks
A, 11c
A, 12c
Sinks
A, 11c
A, 12c
Sinks
A, 11c
A, 12c
Sinks
A, 11c
A, 12c h
Sinks
A, 11c
A, 12c
Sinks
A, 11c
A, 12c
Sinks
A, 11c
A, 12c
Sinks
A, 11b
A, 12b
Flows
easily
Sinks
C, 11o
C, 12p
Flows
sluggishly
Flows
easily
Sinks
A, 11c
A, 12a
Sinks
C, 11j
C, 12j
Inert
Flows
easily
Sinks
C, 11j
C, 12j
Weak blue
Solid
1.11
A, 11e
A, 12e
Moderate
grayish blue
Moderate
light blue
Solid
1.11
A, 11e
A, 12e
Solid
Sinks
A, 11d
A, 12d
Weak green
Solid
Sinks
C, 11k
C, 12k
Moderate blue
Solid
Sinks
A, 11a
A, 12d
Very weak
yellowish green
Strong yellow
Inert
Inert
Weak greenish
yellow
Inert
Inert
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
Flows
easily
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
Flows
sluggishly
S.G.f
S.G. is reported relative to water: floatsS.G. less than 1.0; sinksS.G. greater than 1.0. Both Permasafe and Super Tres had hydrostatically measured S.G.s of 1.11.
groups as discussed in text; individual spectra shown in figures 11 and 12. None means that the sample was too
fluorescent to get a Raman spectrum with the 514.5 nm argon laser source.
h Plus two additional Raman peaks.
g Spectral
GEMS & GEMOLOGY
Summer 1999
93
TABLE 2. Emeralds clarity enhanced for R.I. tests of presumed natural and artificial fillers.
Filler
R.I.
of filler a
Epo Tek 314 (artificial resin)

Baker cedarwood oil
(natural oil)
Mixture of Epo Tek 314 and
Epo Tek 302-3M
(artificial resins)
Sigma clove oil (natural oil)
Epo Tek 302-3M
(artificial resins)
Epo Tek 302-3M
(artificial resins)
Mixture of Sigma clove oil
and Sigma cinnamon oil
(natural oils)
Mixture of Epo Tek 314
and Epo Tek 302-3M
(artificial resins)
Mixture of Sigma clove oil
and Sigma cinnamon oil
(natural oils)
1.500
1.517
Not notable
Not notable
1.517
Not notable
1.531
1.531
1.550
aAll
1.550
1.570
1.570
Microscopic features of
filled fissures in emerald
Fluorescence to longwave
UV in emerald
Raman spectrum of
filler in emerald
Not seen
Very weak yellowgreen in fissures
Weak aliphatic feature, possibly group C

Positive match to Canada balsam or Baker
cedarwood oil; group C
Group C
Incomplete filling; white residue

Incomplete filling; flow structure
Not seen
Not identifiable as to group

Good match to spectra for both components
groups A and C
Orange and blue flash effects; incomplete

filling; iridescence in fissure; light
brownish yellow substance in fissure
Yellow-orange and blue flash effects
(evaporating rapidly); incomplete filling;
iridescence in fissure
Yellow-orange and blue flash effects;
incomplete filling; iridescence in fissure;
white residue
Flash effects (disappearing rapidly as
filler evaporates); incomplete filling;
iridescence in fissure; brownish yellow
substance in fissure; white residue
Shows features of both groups A and C,

less distinct overall
Weak unidentifiable spectrum
Good match to group A reference spectrum
Not seen
Not identifiable as to group
the emeralds for this experiment had R.I. values of 1.571 ( 0.002) to 1.577 ( 0.003).
Refractive Index. The fillers had R.I.s of 1.449 to

1.589; by comparison, the R.I. values of emerald
range from 1.564 to 1.593 (extraordinary ray), and
1.570 to 1.602 (ordinary ray), depending on their
source (Webster and Read, 1994). The UV-setting
adhesives we studied had low R.I. values relative to
emerald, 1.5011.527; however, the R.I. values
increased to 1.5291.553 as they cured. The R.I.
value of Opticon 224 also increased with curing.
For the mixtures of essential oils and those of
artificial resins, the R.I. of the mixture could be predicted from the measured volumes and R.I.s of its
components. However, the mixtures of Baker cedarwood oil with Araldite 6010 behaved nonlinearly,
so the R.I.s did not agree with their predicted values (see table 3).
white to blue fluorescence to long-wave UV, and

the same fillers fluoresced weak to moderate blue to
short-wave UV. The exception, Norland Optical
Adhesive type 65, fluoresced moderate blue to longwave UV and weak green to short-wave UV.
Fluorescence. As indicated in table 1, not all cedarwood oils fluoresce and those that do fluoresce do
not all fluoresce alike. Similarly, although most
epoxy prepolymers fluoresced very weak yellowish
green to short-wave UV, or were inert, Araldite 502
fluoresced strong yellow and Epon 828 fluoresced
weak greenish yellow.
Specific Gravity. Some essential oils floated in

water (cedarwood oils, Canada balsam), whereas
others sank (clove oils, cinnamon oils). The other
oils we studied floated, as did paraffin wax. All of
the epoxy prepolymers and other prepolymers sank,
and all of the hardened resins we studied also sank.
All but one of the polymers displayed strong bluish
94
Viscosity. Some essential oils flowed readily (like

water); however, the cedarwood oils and Canada
balsam were sluggish (like honey). The cedarwood
oil from Antonio Negueruela was nearly solid at
room temperature (similar to petroleum jelly).
The other oils all flowed easily. The paraffin wax
was a soft solid. The epoxy prepolymers flowed
sluggishly, and the UV-setting adhesives flowed easily (Norland Optical Adhesives) or sluggishly. All of
the hardened polymers were solids.
Gemological Characterization of Filling Materials

in Emeralds. Effect of Refractive Index on Apparent
GEMS & GEMOLOGY
Summer 1999
Clarity. Examination of the emeralds filled with the

nine substances in table 2 revealed that as the R.I. of
the filling material increases toward the R.I.s of
emerald, the fissures become less prominent (see,
e.g., figure 2).
Internal Appearance and Flash Effect. Although
most of the substances we used for these R.I. tests
are not common filling materials, we were able to
reproduce features seen in filled emeralds, including
brownish yellow filled voids (figure 4), flow structure (figure 5), incomplete filling (figure 6), and natural iridescence in fractures as well as white residues
(figure 7). The white residues were seen in three of
the stones (one filled with pure clove oil, one with
an Epo Tek mixture, and one with a clove oilcinnamon oil mixture), deep within the fissures at the
time of first examination.
None of the five stones filled with a substance
that had an R.I. of 1.53 or lower showed a flash
effect, whereas all four stones filled with substances
that had R.I.s of 1.55 or higher did show a flash
Figure 3. These four bottles of Baker cedarwood oil

Araldite 6010 mixtures show the unmixing of intermediate mixtures with 40% to 70% Araldite (left to
right). Photo by Maha DeMaggio.
effect (table 2 and figure 8). However, the natural

fillers were quite volatile, so often this effect was
gone within hours. We refilled one emerald with
the 1.550 R.I. natural oil mixture, and photographed
it within minutes of completing the filling process.
TABLE 3. Properties of emeralds filled with cedarwood oilAraldite 6010 mixtures for spectroscopy.
Composition of the filler
Baker cedarwood
oil (%)
Araldite
(%)
Properties
Color of the
loose filler
R.I. of the
filler
Predicted R.I.
of the filler a
100
Yellow
1.517
90
10
Yellow
1.522
1.522
80
70
60
20
30
40
Yellow
Yellow
Yellow
1.528
1.533
1.539
1.528
1.534
1.539
50
50
Yellow
1.544
1.544
40
60
Light yellow
1.545
1.550
30
70
Light yellow
1.547
1.556
20
80
Very light
yellow
1.559
1.561
10
90
Very light
yellow
1.564
1.566
100
Near-colorless
1.572
Flash effect in
stone? b
Unmixing of
fillers in bottle?
No
No
Weak blue
brightfield
No
No
Weak blue
brightfield
Very weak
orangy brown
darkfield
Weak brown
darkfield
Yellow-orange
darkfield
Yellow-orange
darkfield, blue
brightfield
Yellow-orange
darkfield
Yes
IR result
(in emerald)
Raman result
(in emerald)
No
Cedarwood oil
(reference spectrum)
Cedarwood oil
Cedarwood oil
Cedarwood oil
No
No
Yes
Cedarwood oil
Cedarwood oil
Cedarwood oil
Features of both
Features of both
Features of both
Yes
Features of both
Features of both
Yes
Features of both
Features of both
Yes
Features of both
Features of both
No
Features of both
Features of both
No
Araldite 6010
Araldite 6010
No
Araldite 6010
Araldite 6010
R.I. predicted by: R.I. = [(% cedarwood oil) 1.517 + (% Araldite) 1.572] / 100. Where the measured R.I. and predicted R.I.
differ by more than 0.002, this indicates that the two components are not mixing ideally.
b
The emeralds in this test had the same refractive indices, 1.5711.578.
GEMS & GEMOLOGY
Summer 1999
95
Weight Added. The Epo Tek mixtures added about

0.03%0.10% to the weight of the stones (i.e.,
0.00020.0008 ct for a 0.51 ct stone). The cinnamon oil mixtures were so volatile that they evaporated before the stones could be reweighed.
Figure 4. The filler (mixture of Epo Tek resins, R.I.

= 1.550) in this 0.70 ct emerald appears brownish
yellow. Photomicrograph by John I. Koivula; magnified 25.
The flash effects disappeared while we watched as

the stone was being photographed (figure 9).
Fluorescence of Filled Fissures. If enough of a fluorescent filling material is present, the fluorescence
may be visible within the stone (for example, in six
of the nine stones in table 2, we saw very weak yellow-green fluorescence to long-wave UV in the fissures; strictly speaking, we might have expected to
see this in all nine stones). In those cases where fluorescence can be observed, this property can be
especially useful to target specific fissures for
Raman analysis (see Chalain et al., 1998).
Figure 5. Flow structure is evident in the filling

material (mixture of Epo Tek resins, R.I. = 1.531) in
this 0.82 ct emerald. Photomicrograph by John I.
Koivula; magnified 20.
96
FTIR and Raman Spectra. In the infrared region,

emerald has a transmission window between about
3200 and 2400 cm1; therefore, this is the region
that we studied for the loose filling substances. The
absorption spectra of the glass slides that we used,
and of a never-filled Colombian emerald crystal, are
shown in figure 10.
The infrared and Raman spectra enabled us to
place the loose filler materials into five spectral
groupsdesignated here A, B, C, D, and Ewith 17
infrared (figure 11) and 18 Raman (figure 12) subgroups (i.e., each spectrum may represent more than
one filler). The five spectral groups are based on
assignments of peaks to functional groups (see
Discussion, below), but they can be summarized
in terms of the appearance of certain key peaks:
Group A spectra have many peaks, especially at
about 2926, 3006 (stronger in the Raman spectrum, and weak in some examples), 3058 (IR) or
3068 (Raman), andonly seen with Raman in
emeralds1609 cm1. (Recall that peak assignments can vary 4 cm1.) The 2962 cm1 peak is
usually, but not always, present. Typical group A
IR spectra (see, e.g., figure 11c) show a pattern of
three strong peaks decreasing in height from 2962
to 2926 to 2872 cm1. Typical group A Raman
spectra (figure 12c) have a distinctive pattern of
five strong peaks in the region between 2870 and
3070 cm1.
Group B spectra also have many peaks, including
those at 2853 (2848 cm1 for paraffin wax), 2882
(Raman only), 2926 (5), and 2954 cm1 (IR only);
there may be Raman peaks at 3010 and 1655 cm1.
The IR spectra of the liquids show a pattern of
two peaks, each with shoulders at a higher
wavenumber (figures 11f, g); and the Raman spectra show plateaus centered at about 2890 cm1
(figures 12f, g). In contrast, the IR spectrum of the
paraffin wax has a broad central feature (figure
11h), while the Raman spectrum is sharp (figure
12h).
Group C spectra exhibit peaks at 2870 (5), 2929
(ranging to 2942 cm1 in figure 12i), usually 2958,
and 1442 (Raman only) cm1. Sometimes Raman
peaks are seen at 1725 (figures 12j, m, and p) or
GEMS & GEMOLOGY
Summer 1999
1643 cm1 (figures 12j, n, and p). Typical IR (figure

11n) and Raman (figure 12l) spectra have an overall triangular appearance in the 28003050 cm1
region.
Group D spectra exhibit features resembling those

of group A and group B or C, with six or more
peaks in the 28003100 cm1 region (see, e.g., figures 11p and 12q).
The Group E Raman spectrum is notable for the

low intensity of the 28003100 cm1 region (only
one peak at 3064 cm1: figure 12r), compared to
three strong peaks at 1598, 1626, and 1674 cm1.
Additional details were used to place these spectra into subgroups. Although many of these details
are unlikely to be discernible in a filled emerald,
they may be important in some cases. For example,
Raman microspectrometry can distinguish Baker
cedarwood oil and Canada balsam from other cedarwood oils, but not from each other (compare figures
12n and o); these IR spectra are identical (figure
11n). In addition, the IR and Raman spectra of
Permasafe and Super Tres (figures 11e and 12e)
are slightly different from those of cured Opticon
(figures 11d and 12d), although these distinctions
(e.g., a peak at 1000 cm1 in the Raman spectrum)
might not be obvious in a filled emerald. Last, the
Norland Optical Adhesives show a distinctive
Raman peak at 2575 cm1 (figure 12j); this peak
diminishes significantly on curing (figure 12k).
The IR and Raman spectra of the cedarwood
oilAraldite 6010 mixtures as loose fillers are
shown in figure 13. These spectra contain no new
features; detection limits based on these results are
discussed below.
As noted earlier, the spectra of the fillers in
emeralds taken with reflectance IR spectroscopy
were almost identical to the IR spectra of the corresponding loose fillers (compare, e.g., figure 10 inset
with figure 11n). Consequently, we will refer to the
IR spectra of the loose fillers and mixtures throughout the discussion.
DISCUSSION
Separating Filler Materials on the Basis of Their
Optical and Physical Characteristics. The following
optical and physical characteristics of loose filler
materials may be useful in distinguishing among
filling substances within emeralds: color, refractive
index (and flash effect), fluorescence, viscosity, and
distinctive microscopic features.
Figure 6. Note the incomplete filling by clove

oil (R.I. = 1.531) in the fissures of this 0.94 ct
emerald. Photomicrograph by John I. Koivula;
magnified 15.
Color. Dyed filler materials are distinctly colored

(e.g., green Opticon and Joban oil; see table 1).
However, other features can be mistaken for dye,
including the green polishing compound chromic
oxide (Ringsrud, 1983; Hnni, 1988) and green internal reflections. If dye is suspected, the observer
should examine the distribution of color within the
stone using diffuse transmitted light; another technique is to check whether the fissures are green
from all angles and to look for color concentrations
along fractures. The yellow color of other materials
listed in table 1 is unlikely to be seen in an emerald
unless the filler is present in broad fissures or cavities (Koivula, 1999).
Refractive Index (and Flash Effect). As the refractive index of the filling material increases toward
the R.I.s of emerald, the filled fissures become less
prominent (again, see figure 2). For the emeralds we
studied, which had R.I. values between 1.569 and
1.580, we saw flash effects for all fillers with R.I.
values of 1.54 or above. Of the fillers in this study,
only those in spectral groups A, D, and E (epoxy prepolymers and polymers, and some essential oils)
have R.I.s above 1.54. Only fillers in spectral group
B (other oils) and C (the prepolymer Liquid Resin)
have R.I.s below 1.50, and these are much less
effective at disguising fissures in emerald. Mixtures
gave intermediate R.I.s.
An indication of a fillers R.I. value can be provided by the prominence of the filled fissures and
GEMS & GEMOLOGY
Summer 1999
97
Figure 7. A white residue was seen deep within a

fissure in this 0.47 ct emerald, which had been
filled with a mixture of clove and cinnamon oils
(R.I.= 1.570). Natural iridescence is also visible in
this fissure (top of the image). Photomicrograph by
John I. Koivula; magnified 20.
the presence of a flash effect. Because most materials that show flash effects in emeralds are artificial
resins, and the high-refractive-index essential oils
we examined are extremely volatile, it is likely that
an emerald showing this effect contains an artificial
resin. However, there are many artificial resins with
lower R.I. values, so the absence of a flash effect
does not mean that the substance should be considered natural. For instance, the early formulation of
Arthur GroomGematrat had a lower R.I. value
than other commonly used artificial resins, as evidenced by the greater prominence of the filled fissures and the lack of a flash effect (see, e.g., Johnson
et al., 1997).
Fluorescence. Most of the fillers we studied were
inert, or fluoresced very weak or weak yellow-togreen to long-wave UV. However, two of the five
cedarwood oils fluoresced moderate white or strong
yellowish white, and most of the cured (hardened)
polymers fluoresced strong white to blue, so strong
fluorescence in fissures may be useful for distinguishing cured and surface-hardened fillers from
liquid fillers. (Of course, one risks hardening the
UV-setting adhesives by exposing them to UV
radiation.) Recent formulations of Arthur
GroomGematrat are said to contain a fluorescent
tracer (Weldon, 1998a). We were not able to study
the Arthur GroomGematrat filler as an isolated
material; however, all five polymer samples we
examined fluoresced blue to bluish white to longwave UV, and all but the cured Norland Optical
Adhesive 65 fluoresced blue to short-wave UV
(again, see table 1).
Viscosity. All filling materials are liquid when they
are emplaced in emeralds, but some are hardened
once in the fissures (e.g., UV-setting adhesives), others are frequently surface hardened (Opticon:
Kammerling et al., 1991), and others slowly solidify
over time (e.g., Canada balsam: Kammerling et al.,
1991). It is also possible that certain internal characteristicssuch as gas bubbles, flow structures, and
incomplete filling of fracturesare related to the
viscosity of a filling material, but we have not tested this hypothesis.
Before the viscosity of a filler in an emerald can
be inferred, its presence must be established.
Johnson et al. (1998a), among others, have described
procedures for finding evidence of clarity enhancement in emeralds. Using a microscope at low magnification with reflected light will reveal the trace of
the fissure on the surface; switching to darkfield
illumination will show the extent of the fissure
inside the stone. Higher magnification is then used
Figure 8. This 0.59 ct emerald filled with a mixture of

epoxies (R.I. = 1.570) shows
orangy yellow (darkfield,
left) and blue (brightfield,
right) flash effects.
Photomicrographs by John I.
98
GEMS & GEMOLOGY
Summer 1999
Figure 9. This emerald, filled

with a mixture of the natural
essential oils clove oil and
cinnamon oil (R.I. =1.550),
also shows yellow-orange
(darkfield) and blue (brightfield) flash effects. Photomicrographs by John I.
to reveal features indicative of a filled fissure: flow

structure, gas bubbles (which may be flattened or
round), incomplete filling, or yellowish voids (sometimes shaped like crystal inclusions).
Once the presence of a filler has been established, there are at least two ways to see whether
the filler is fluid: (1) Watch the emerald as it warms
with exposure to the microscope lamp, and see if
there is any movement of the filler within the
stone; and (2) while observing carefully with the
microscope, bring a heated sharp point nearbut
not touchingthe surface of the emerald, and
watch for movement of the filler (it may even form
droplets on the surface of the stone). In either case,
use extreme care to avoid damaging the emerald by
extending fissures. A liquid filler may show motion
with these techniques, but the absence of a reaction
does not guarantee that the filler is solid, because
surface-hardened fillers might act the same (Johnson
et al., 1998a).
could also happen with other substances.)

Another possible source of fine, white surface
contamination seen in the trade is quartz dust,
which contaminates some polishing compounds
used in Bogot (R. Giraldo, pers. comm., 1999). As
volatile filling material evaporates from the surface
of an emerald, the 1.54-R.I. quartz dust left behind
would gradually become visible.
Figure 10. For reference purposes, we recorded

the FTIR spectra of (a) the glass slides we used
as windows for obtaining the spectra of filling
materials, and (b) a 5.01 ct natural (unfilled)
emerald crystal. The inset shows a detail of the
emerald spectrum in the region where filler
peaks are seen, together with the spectrum of an
emerald that had been filled with Baker cedarwood oil (in red); note that
this area of the spectrum also
contains small emerald features..
Distinctive White Cloudy Inclusions. Although this

internal feature is commonly attributed to palma
(see, e.g., Ringsrud, 1998), we did see such inclusions in emeralds filled with natural essential oils
(clove and cinnamon), and with a mixture of epoxy
prepolymers (Epo Tek 314 and Epo Tek 302-3M).
We suspect that they result from some sort of emulsification, which can be demonstratedbut not
provedby the following test: Place nearly equal
amounts of filling material (here, Araldite 6010) and
water in a bottle, then close and shake the bottle
(figure 14). The white cloudy mixture that forms is
an emulsion of small droplets of one substance in
the other substance; the droplets are prevented from
coalescing by the coating of the other substance.
Perhaps in the case of our experimentally filled
emeralds, this emulsion formed deep within the
filled fissures, where not all the water had been
removed from the emeralds before filling. (This
GEMS & GEMOLOGY
Summer 1999
99
Figure 11. On the basis of their spectra, the filling

substances examined were placed in four infrared
groups (AD) and 17 subgroups (aq). Representative spectra for each subgroup are illustrated
here. The appropriate group and spectrum for each
of the fillers studied are listed in table 1. NOA =
Norland Optical Adhesive.
FTIR and Raman Spectrometry: Complementary

Techniques, Similar Information. Peaks in infrared
and Raman spectra have the same cause: The radiation excites vibrations between atoms. In crystalline substances, IR and Raman spectra of the
100
same material may look quite different, because

the two techniques often excite different vibrations, depending on the symmetry of the material
and on quantum mechanical selection rules
(Harris and Bertolucci, 1978). However, liquids and
polymers are less ordered than crystalline solids
(i.e., they have less overall structural order and
greater variability in bond lengths and angles, etc.),
so the IR and Raman spectra of a particular liquid
or polymer look quite similar in general, with a few
differences in peak position and (especially) in relative intensity.
When it comes to examining spectral information of fillers within emeralds, FTIR and Raman
often provide similar information from distinct
regions within the emerald. With the reflectedbeam geometry, FTIR spectroscopy gathers spectral information from much of the interior of the
emerald, and it is not necessary to find the filled
fissures prior to analysis. In contrast, the Renishaw
Raman microspectrometer must be focused at or
near the surface of the stone, and the fissures to be
examined must first be located with the microscope. Because the spectral information only
comes from the near-surface area, other filler substances deeper in the stone would not be detected.
In addition, the emerald host has a different
effect on the results obtained with each technique.
With FTIR, only the limited spectral region
between 3200 and 2400 cm1 is visible in the emerald (again, see figure 10). Fortunately, this region
includes one of the two diagnostic spectral regions
for polymers. With Raman, the emerald (and some
fillers) may luminesce to the 514 nm laser beam
that induces the Raman effect; this luminescence
causes a large interfering background signal in the
spectrum that generally begins at wavenumbers
above 3000 cm1 or so (but can be a problem in
some emeralds even at 2000 cm1). With both techniques, the distinctive filler peaks are very small
compared to the features of the emerald host.
Reference infrared spectra for many of the materials we tested are available in atlases such as Pouchert
(1985) and Julian et al. (1991), software such as the
Omnic Interpretation Guide (Nicolet, 1992), and references such as those cited by the ChemFinder Web
site. The Raman peaks (especially for liquids) can be
interpreted as if they were infrared peaks. However,
the interpretation of both types of spectra requires
the terminology of organic chemistry. Functional
groups are the pieces of a molecules structure that
produce distinctive spectral features. Certain peak
GEMS & GEMOLOGY
Summer 1999
Figure 12. On the basis of their spectra, the filling

substances were placed in five Raman groups (AE)
and 18 subgroups (ar), which largely correspond to
those for the IR spectra. Representative spectra for
each subgroup are illustrated here. The appropriate
group and spectrum for each of the fillers studied are
listed in table 1. NOA = Norland Optical Adhesive.
assignments are characteristic of different functional

groups and bonding environments. The following
terms (and some associated spectral positions) are relevant to the fillers we examined:
Aliphatic hydrocarbons consist of chains of carbon atoms. Pine et al. (1980, p. 28) restrict the term
to molecules that do not contain cyclic structures.
These are often referred to as open-chain or acyclic
hydrocarbons. In contrast, alicyclic, or cyclic,
hydrocarbons are composed of carbon atoms
arranged in a ring or rings. Paraffin oil is mostly
composed of straight-chain aliphatic hydrocarbons,
and cedarwood oil contains cyclic hydrocarbons.
Both types of hydrocarbons are called alkanes.
Aromatic hydrocarbons are special groups of
cyclic compounds that usually have six-member
rings that may be visualized as having alternate
single and double bonds (i.e., benzene rings).
Cinnamon oil contains benzene rings. When an
OH (hydroxyl) group is connected to a carbon atom
of a benzene ring, the compound is known as a phenol (Pine at al., 1980, pp. 28, 47). DGEBA resins (see
Box A) contain phenol groups that are connected
together in pairs (bisphenol groups). These artificial
resins are rich in aromatic bonds (e.g., a large peak

at 3066 cm1 is seen in the Raman spectrum of
Opticon 224).
Hydrocarbons that have one or more carbon-carbon double bonds are called alkenes or olefins (Pine
at al., 1980, p. 37). For instance, a peak in the
3000 3020 cm1 region is associated with a hydrogen atom attached to one of the carbons in a carbon-
GEMS & GEMOLOGY
Summer 1999
101
Figure 13. The infrared and Raman spectra of mixtures of (Baker) cedarwood oil and Araldite 6010
show features characteristic of both; mixtures with much more of one filler than the other are hard
to distinguish from the pure compounds.
carbon double bond; this is seen in vegetable oils

and Joban oil.
Separating Presumed Natural Substances (Essential Oils/Resins, Other Oils, Waxes) from Artificial
Resins by FTIR and Raman. Peak assignments and
Interpretation. The five spectral groups (A through
E; again, see figures 11 and 12) are based on assignments of peaks to functional groups as follows:
Group A spectra exhibit strong aromatic, aliphatic, and bisphenol epoxy peaks. It is very difficult
to distinguish among these DGEBA epoxy resins
spectroscopically. However, these compounds can
vary in refractive index, viscosity, fluorescence,
and degree of polymerization (again, see table 1).
Group B spectra exhibit multiple strong aliphatic
peaks, generally producing several distinct peaks
in the 29252850 cm1 region of the IR spectrum
(or a broad overall peak in the Raman spectrum);
some spectra also exhibit traces of alkene (olefin).
The artificial resin HXTAL has a group B infrared
spectrum (figure 11i), and affinities to both groups
B and C in its Raman spectrum (figure 12i).
Group C spectra exhibit less-distinct multiple IR
peaks, but a single dominant aliphatic Raman
102
peak, with small aromatic features seen in some

compounds.
Group D spectra exhibit both aromatic and

aliphatic features.
The Group E spectrum is a Raman spectrum only.

The substance (cinnamon oil) that has this spectrum is unusual in that it exhibits both aliphatic
and aromatic IR features (so its IR spectrum is in
group D), but mostly aromatic Raman features.
We observed no Raman spectrum for the other
(cassia) cinnamon oil due to its strong fluorescence.
Spectra versus Flash Effect as a Detection
Technique for Mixtures. Looking again at our experimental mixtures of Baker cedarwood oil and
Araldite 6010, we found that mixtures with up to
20% Araldite (for Raman) or 40% Araldite (for IR)
could not be distinguished from pure cedarwood oil
on the basis of their spectra. Similarly, the spectra of
the mixture with 90% Araldite10% cedarwood oil
could not be distinguished from those of pure
Araldite 6010 with either technique (again, see figure 13 and table 3).
The fractures in stones filled with these mix-
GEMS & GEMOLOGY
Summer 1999
tures became less apparent as the filler increased in

refractive index. Although a weak flash was noted
in the stone filled with the 1.522 R.I. mixture (90%
cedarwood oil10% Araldite), in general the flash
effect became noticeable (as a weak blue flash) in
the 1.539 R.I. mixture (60% cedarwood40%
Araldite), growing stronger as the R.I. increased. We
concluded that, in this case, the variation in relative
peak heights in the IR and Raman spectra was about
as sensitive as flash effect for determining the presence of the high-R.I. artificial resin in the cedarwood oil mixtures.
Comparison with Previous Published Work. In general, there is good agreement between the Raman
spectra published by Hnni et al. (1996a) and this
study. Our results are similar to Hnnis for
Permasafe (New type of epoxy resin, 1998), but
we saw an additional peak at 2928 cm1. According
to Hnni et al. (1996b), artificial resins can be distinguished from presumed natural substances by the
1250, 1606, 3008, and 3069 cm 1 peaks in the
Raman spectra of the former. In our filling substances, however, we could see (generally small)
examples of each of these peaks in the spectra of
certain cedarwood oils, other essential oils, or other
vegetable oils. Hnni et al. (1996b) also saw features
for cedarwood oil at 14401457 cm1, which we
observed in some prepolymers as well.
Chalain et al. (1998) examined several UV-setting materials with lower R.I. values than the ones
we studied; they also mentioned natural hydrocarbons in emerald, which we have not seen. They did
not report the spectra for any of these, however.
Zecchini and Maitrallet (1998) looked at the IR
spectra of several emerald fillers and related substances, including cedarwood oil, methyl methacry-
late polymer, olive oil, Israeli (paraffin?) oil,

Opticon 224, palm oil, Indian Joban oil, red-colored Indian oil, and the Arthur GroomGematrat
treatment, as well as the oil from human fingerprints. Our results agree with most, but not all, of
theirs, in the cases where we studied the same
materials. However, they interpreted the 3009 cm1
peak in Joban and red oil as a dye band, whereas we
saw this olefin band in all the vegetable oils we
examined.
Sometimes, complete characterization cannot be
done using FTIR or Raman, and only the type of
chemical class can be identified (Julian et al., 1991).
This is especially true for DGEBA epoxies, and neither FTIR nor Raman is likely to be useful in distinguishing prepolymers from hardened polymers in
emeralds (but see the discussion of fluorescence,
above, for another test).
Comparison with Substance Categories. Unfortunately, our spectral groups do not correspond precisely to the six substance categories indicated in
table 1. Essential oils are found in IR and Raman
spectral groups C, D, and E (group C also contains
artificial resins). Other oils are in IR and Raman
spectral group B; so is paraffin wax. Most epoxy prepolymers and polymers are in IR and Raman spectral group A. The other prepolymers (e.g., UV-setting adhesives) can be in group A (Epo Tek
UVO114) or group C (in general), and HXTAL has
characteristics of groups B and C.
Consequently, distinctions between artificial
and presumed natural substances can most easily
be made spectroscopically if the substance found is
in group A (artificial resins only) or group D or E
(essential oils only; but even in these cases, mixtures and multiple compounds must also be consid-
Figure 14. When waterinsoluble filling substances

(here, Araldite 6010) are
mixed with water, they
form a white cloudy emulsion. Left: Araldite 6010
and water before mixing;
right: the substances shaken together. Photos by
Maha DeMaggio.
GEMS & GEMOLOGY
Summer 1999
103
ered). Distinctions between artificial and presumed

natural substances are more problematic if the
substance shows a group B or C spectrum. The distinctions within a spectral group may not be evident in filled emeralds, although this remains to be
tested. The distinctions among groups C, D, and E
can also be difficult to discern in a filled stone.
As another complication, the presence of spectral features indicating one particular compound
does not preclude the existence of some other substance. As mentioned above, there can be almost
40% Araldite 6010 (palma) in cedarwood oiland
the IR spectrum will still look like cedarwood oil,
and not a mixture (although the Raman spectrum of
the 20% Araldite mixture shows a DGEBA peak at
3068 cm1). Similarly, there can be about 10%
cedarwood oil in Araldite before its presence is
detected.
Combining the Approaches. That different gemological laboratories already determine the nature of
filling materials is evident from auction records and
other publications. Auction records (cited below)
indicate that the identification of clarity enhancement substances by category or type has been performed by at least two laboratories: the SSEF in
Basel, Switzerland, and the American Gemological
Laboratory (AGL) in New York City.
SSEF provides identification reports, or treatment slips, which state the category of filling
material present (or the categories of fillers that are
not present). This information is based on gemological examination as well as on Raman microspectrometry and FTIR spectroscopy (see, e.g., SSEF
Swiss Gemmological Institute, 1998). The categories of fillers mentioned in SSEF treatment slips,
as quoted in auction catalogs, include: oil, natural
oil, natural resin, artificial resin, and wax
(Christies, 1997b, 1997c, 1998a). According to Dr.
Hnni (pers. comm., 1999), the categories currently in
use are oil, Canada balsam, and artificial resin.
As reported in auction catalogs, the AGL categories for types of filling materials include: oil
type, Opticon/oil type, and unidentified type
(Sothebys 1997; Christies 1998b). As of July 1997,
AGL did not have Raman microspectrometry, and
relied on the flash effect to identify the vast
majority of fillers in emeralds examined there
(Edry, 1997).
Our research has shown that the best results
come from a combination of all the testing methods
104
we have described. For instance, the conclusion that

a material showing a flash effect is probably an artificial resin (DGEBA-based epoxy) is reasonable but
not perfect; and the Raman and IR spectra of the
common epoxy fillers are easily distinguished from
those of available cedarwood oils. Raman and IR are
more effective than gemological observation for distinguishing low-R.I. artificial resins from commonly
used essential oils (all of which have no flash effect),
and high-R.I. essential oils from DGEBA epoxies (all
of which show a flash effect). However, the artificial
resins in spectral group C all have low R.I.s (and no
flash effect) and may not be distinguishable from
cedarwood oils.
CONCLUSIONS
Filler Terminology. On the basis of the findings presented here, and the discussion in Box A, we recommend the following terminology:
The term natural should be avoided in discussions of emerald fillers, as natural substances
are not currently distinguishable from their chemical equivalents synthesized in the laboratory
(with the nondestructive tests we have available).
Instead, substances such as essential oils, other
oils, natural resins, and waxes should be referred
to as presumed natural.
A distinction should be made between essential
oils and other oils, as essential oils have different
properties (e.g., they must be volatile) and a
broader range of chemical structures (e.g., they
can contain aromatic compounds). This difference may also be relevant to durability concerns.
The term synthetic should be avoided, as it does
not have the same meaning in chemistry that it
has in gemology. Instead, manufactured prepolymers and polymers should be referred to as artificial resins, as they have no natural equivalent.
Prepolymers and polymers should be distinguished wherever possible, as their different
mechanical properties are probably significant
(e.g., liquid versus solid states), and may affect
their durability in an emerald.
Artificial resins should not be called epoxies
unless they contain epoxide groupsif still liquidor were joined into polymers with these
groups (if solid; again, see Box A).
Artificial resins should not be called Opticon or
Opticon type unless they are known to be the
GEMS & GEMOLOGY
Summer 1999
specific resin Opticon 224 (manufactured by

Hughes Associates).
Because a large quantity of one filler may make
small amounts of another undetectable, we recommend restricting comments to the filler identified, and making no assurances that any others are
absent.
Separating Fillers by Categories: Possible? Of the
thousands of possible filling materialsabout 3,500
essential oils and natural resins, many other oils,
and several thousand commercially available polymers and prepolymerswe studied fewer than 40.
None of the microscopic features (e.g., flash effect,
fluorescence of the filled fractures) fully differentiates between the various categories of substances
(again, see table 1), although such features may provide important clues.
Using Raman and FTIR spectroscopy, we found
that the 39 isolated materials could be categorized
into five spectral groups. The two groups that contain the most important natural fillerscedarwood oil, Canada balsam, and vegetable oilsalso
contain materials that are not natural (Epo Tek 314
resin, UV-setting adhesives, Isocut fluid). Hence,
identification of the spectral group alone is not sufficient to prove that a filling material is natural.
(However, the most commonly used artificial resins
Opticon 224, Araldite 6010, Epon 828do occur in a
spectral group that is easily distinguished from those
in which the presumed natural fillers occur.)
Most of the isolated fillers we tested were
loose, that is, not in an emerald. A number of
factors may limit the effectiveness of spectroscopic
testing of substances after their emplacement in an
emerald. In particular, transmission IR techniques
are complicated by the fact that emerald is opaque
in one diagnostic region (below 2000 cm1). Raman
microspectrometry only samples near-surface areas
of fissures, and fluorescence in the 3000 cm 1
region may interfere with the strongest Raman
peaks. For both techniques, the filler peaks are generally much less prominent than the emerald spectral features.
On the basis of our results, we believe that the
artificial resins most commonly used as emerald
fillers can, as pure substances, be distinguished
from cedarwood oils by a combination of gemological and spectroscopic properties, but other artificial
resins also used in the trade cannot be so separated.
The UV-setting adhesives are particularly problem-
atic. At this time, we recommend the following procedure to identify individual fillers (or at least a substance category or spectral group): first, estimate the
fillers refractive index (based on the presence or
absence of a flash effect), and then obtain IR
and/or Raman data. Other properties, such as fluorescence, also may be useful.
New filling materials and mixtures will continue to be discovered or adapted from other branches
of science and technology. Competing factors make
it unlikely that one best filler exists, and filler development continues. Kammerling et al. concluded
their 1991 article with the comment: It would,
therefore, seem both inappropriate and misleading,
in describing a filled fracture, to use wording that
implies that the filling substance has been conclusively identified if in fact it has not. We must keep
these uncertainties in mind as we try to identify
specific substances.
However, the issue of filler identification is not
the only one that must be addressed to understand
fully the use of fillers in general, or of a specific filler
in a particular emerald. Key questions remain: What
is the effect of the degree of enhancement on an
emeralds appearance? How does the enhancement
change with time, normal wear, and other typical
events in the life of a piece of jewelry? Given these
questions, we are focusing on two main topics as
we continue our research: (1) determination of the
extent to which any particular emerald has been
clarity enhanced; and (2) durability testing of several
common emerald treatments.
Acknowledgments: The authors thank the following

individuals for sharing their knowledge of emerald
enhancements: Ricardo Alvarez Pinzon, Luis Ernesto
Vermudez, and others at Treatment World Emerald,
Gemolgico Universal, Bogot, Colombia; Dr. Rodrigo
Giraldo, Centro Gemolgico para la Investigacin de la
Esmeralda, Bogot; Carlos Osorio, Mineralco, Bogot;
Jaime Rotlewicz, C. I. Gemtec, Bogot; the staff of
Alagecol (Asociacin de Laboratorios Gemolgicos),
Bogot; Ray Zajicek, Equatorian Imports, Dallas,
Texas; Darrold Allen, Gemological Laboratory of Los
Angeles, California; Kenneth Scarratt, AGTA
Gemological Testing Center, New York City; Amnan
Gad, Amgad Inc., New York City; Michael and Ari
Gad, Gad International Ltd., New York City; Arthur
Groom and associates, Arthur GroomGematrat, New
York City; Morty Kostetsky, Arigad Gems, New York
City; Dr. Kumar Shah, Real Gems Inc., New York City;
GEMS & GEMOLOGY
Summer 1999
105
I. Z. Eliezri, Colgem Ltd., Ramat Gan, Israel; Ron

Ringsrud, Constellation Colombian Emeralds,
Saratoga, California; and Jess Williams and Andy
Rendle, EmeraldStone Inc., Vancouver, BC.
Dr. Henry Hnni, Dr. Lore Kiefert, and J.-P.
Chalain, SSEF Swiss Gemmological Institute, Basel,
provided hospitality and information about the identification techniques used in their laboratory. MLJ
also thanks the organizers of the First World Emerald
Congress in Bogot for their warm hospitality, and
Mark Parisi for his support. Dr. George Rossman, of
the California Institute of Technology in Pasadena,
California, supplied much useful chemical information and advice.
Many GIA colleagues provided support for this
project, including: Shane McClure, Maha DeMaggio,
John Koivula, and Dino DeGhionno, GIA Gem Trade
Laboratory (GTL), Carlsbad, California, and curator
Jo Ellen Cole, GIA Carlsbad, who determined gemological properties; Margot McLaren of GIAs Richard
T. Liddicoat Gemological Library and Information
Center, Carlsbad, who located auction records; and
GIA and GIA GTL colleagues Dr. James Shigley,
Tom Moses, Dr. Ilene Reinitz, and Phil York, who
gave suggestions that improved this manuscript.
GIA thanks Jewelers Circular-Keystone, Radnor,
Pennsylvania, and Kyocera Corp., Kyoto, Japan, for
financial support of this research project. Michael
Scott generously donated a Renishaw Raman
microspectrometer to GIA, which was used in this
research.
REFERENCES
Alger M.S.M. (1989) Polymer Science Dictionary. Elsevier
Applied Science, London.
Benson L. (1960) Highlights at the Gem Trade Lab in Los
Angeles. Gems & Gemology, Vol. 10, No. 1, pp. 36, 3031.
Bleumink E., Mitchell J.C., Nater J.P. (1973). Allergic contact dermatitis from cedar wood (Thuja plicata). British Journal of
Dermatology, Vol. 88, No. 5, pp. 499504.
Bower D.I., Maddams W.F. (1989) The Vibrational Spectroscopy
of Polymers. Cambridge University Press, Cambridge,
England, 326 pp.
Brady G.S., Clauser H.R. (1986) Materials Handbook: An
Encyclopedia for Managers, Technical Professionals,
Purchasing and Production Managers, Technicians,
Supervisors, and Foremen, 12th ed. McGraw-Hill Book Co.,
New York, 1038 pp.
Chalain J.-P., Hnni H.A., Kiefert L. (1998) Dtermination des
substances de remplissage dans les meraudes. In D. Giard,
Ed., Lmeraude: Connaisances Actuelles et Prospectives.
Association Franaise de Gemmologie, Paris, pp. 107115.
(Translations where cited are by Mary Johnson.)
Christies (1993) Magnificent Jewels. Auction catalog for
Christies Geneva, November 16 and 18, 1993.
Christies (1997a) Magnificent Jewels. Auction catalog (in two
106
parts) for Christies New York, October 2728, 1997.

Christies (1997b) Magnificent Jewels. Auction catalog for
Christies Geneva, November 20, 1997.
Christies (1997c) Important Jewellery. Auction catalog for
Christies London, December 16, 1997.
Christies (1998a) Important Jewels. Auction catalog for
Christies St. Moritz, February 2021, 1998.
Christies (1998b) Magnificent Jewels. Auction catalog (in two
parts) for Christies New York, April 67, 1998.
Congress votes yes to promotion, no to codes (1995). ICA
Gazette, August 1995, p. 9.
Crowningshield, G.R. (1958) Ascertaining the nature and extent
of damage or inherent flaws in gemstones. Gems &
Gemology, Vol. 9, No. 5, pp. 131133, 159.
Crowningshield, G.R. (1960 1961) Developments and highlights
at the Gem Trade Lab in New York. Gems & Gemology, Vol.
10, No. 4, pp. 114123.
Edry S.G. (1997) Emerald filling: More research sought. National
Jeweler, Vol. 41, No. 13, pp. 1, 64, 65.
Elias H.-G. (1993) An Introduction to Plastics. Weinheim, New
York.
Federman D. (1997) Beyond buyer beware. Modern Jeweler, Vol.
96, No. 3, pp. 50, 51, 5355.
Federman D. (1998a) Modern Jewelers Consumer Guide to
Emeralds. Supplement to Modern Jeweler, Vol. 97, No. 1, 7
pp.
Federman D. (1998b) Fair play. Modern Jeweler, Vol. 97, No. 2, p.
108.
Fryer C., Crowningshield R.C., Hurwit K., Kane R.E. (1984) Gem
trade lab notes: Emerald, oiled. Gems & Gemology, Vol. 20,
No. 1, pp. 4647.
Genis R. (1997) The emerald controversy. Rapaport Diamond
Report, Vol. 20, No. 38, pp. 69, 7273.
Hnni H.A. (1988) An oil well in your garden. Swiss Jewellery
Journal, No. 3, pp. 461464.
Hnni H.A. (1998) Fracture filling in emeralds and its detection
with laboratory methods. ICA Gazette, July/August 1998, pp.
5 7.
Hnni H.A., Kiefert L., Chalain J.-P., Wilcock I.C. (1996a) Ein
Renishaw Raman Mikroscop im gemmologischen Labor:
Erste Erhfarhrungen bei der Anwendung. Zeitschrift der
Deutschen Gemmologischen Gesellschaft, Vol. 45, No. 2, pp.
5570.
Hnni H.A., Kiefert L., Chalain J.P. (1996b) How to identify fillings in emeralds using Raman spectroscopy. Jewellery News
Asia, No. 145, pp. 154, 156.
Hnni H.A., Kiefert L., Chalain J.P. (1997) Renishaw Raman
nothing new. Rapaport Diamond Report, Vol. 20, No. 40, p.
11.
Harris D.C., Bertolucci M.D. (1978) Symmetry and Spectroscopy:
An Introduction to Vibrational and Electronic Spectroscopy.
Oxford University Press, Oxford, England, 550 pp.
Herout V. (1982) Sesquiterpene alcohols. In E.T. Theimer, Ed.,
Fragrance Chemistry: The Science of Smell. Academic Press,
New York, pp. 222263.
Hurwit K. (1989) Gem trade lab notes: Emerald, with plastic-like
filling. Gems & Gemology, Vol. 25, No. 2, p. 104.
Johnson M.L., Koivula J.I. (1997) Gem news: A new emerald
filler. Gems & Gemology, Vol. 33, No. 2, pp. 148149.
Johnson M., McClure S., Koivula J., Moses T. (1998a) How to
detect clarity enhancement in emeralds. Modern Jeweler, Vol.
97, No. 1, pp. 5861.
Johnson M.L., Koivula J.I. (1998b) Gem news: A preliminary
report from the First World Emerald Congress. Gems &
Gemology, Vol. 34, No. 1, pp. 5657.
Julian J.M., Anderson D.G., Brandau A.H., McGinn J.R., Million
A.M., Brezinski D.R. (1991). An Infrared Spectroscopy Atlas
GEMS & GEMOLOGY
Summer 1999
for the Coatings Industry, 4th ed., Vol. 1. Federation of

Societies for Coatings Technology, Blue Bell, PA.
Kammerling R.C., Koivula J.I., Kane R.E., Maddison P., Shigley
J.E., Fritsch E. (1991) Fracture filling of emeralds: Opticon and
traditional oils. Gems & Gemology, Vol. 27, No. 2, pp.
7085.
Kammerling R.C. (1993) Gem trade lab notes: Emerald, with
large filled etch channels. Gems & Gemology, Vol. 29, No. 4,
p. 280.
Kammerling R.C., McClure S.F., Johnson M.L., Koivula J.I.,
Moses T.M., Fritsch E., Shigley J.E. (1994) An update on filled
diamonds: Identification and durability. Gems & Gemology,
Vol. 30, No. 3, pp. 142177.
Kammerling R.C., Maddison P., Johnson M.L. (1995) Gem trade
lab notes: Emerald, with unusual flash-effect colors. Gems &
Gemology, Vol. 31, No. 1, p. 54.
Kane R.E. (1990) Gem trade lab notes: Emerald, with filled fractures. Gems & Gemology, Vol. 26, No. 1, pp. 9596
Kennedy H.F. (1998) Brazilian emeralds: Oiling at the source.
National Jeweler, Vol. 42, No. 10, pp. 3652 passim.
Koivula J.I. (1999) Gem trade lab notes: Emerald, the case of the
invisible filler. Gems & Gemology, Vol. 35, No. 1, pp. 4344.
Lee H., Neville K. (1967) Handbook of Epoxy Resins. McGrawHill Book Co., New York.
Liddicoat R.T. Jr.(1962) Handbook of Gem Identification, 6th ed.
Gemological Institute of America, Los Angeles, 396 pp.
Lincoln W.A. (1986) World Woods in Color. Linden Publishing,
Fresno, CA, 320 pp.
Mutton, D.B. (1982) Wood resins. In W.E. Hillis, Ed., Wood
Extractives and Their Significance to the Pulp and Paper
Industries. Academic Press, New York, pp. 331365.
Nassau K. (1994) Gemstone Enhancement: History, Science, and
State of the Art, 2nd ed. Butterworth-Heinemann, Oxford,
England, 252 pp.
New type of epoxy resin. (1998) Jewellery News Asia, No. 172, p.
60.
Nicolet (1992) Omnic Interpretation Guide: Infrared
Spectroscopy, Version 1.0, Nicolet Instrument Corp.,
Madison, Wisconsin.
Ottaway T.L., Wicks F.J., Bryndzia L.T., Kyser T.K., Spooner
E.T.C. (1994) Formation of the Muzo hydrothermal emerald
deposit in Colombia. Nature, Vol. 369, pp. 552554.
Pine S.H., Hendrickson J.B., Cram D.J., Hammond G.S. (1980)
Organic Chemistry, 4th International Student Edition.

McGraw-Hill International, Tokyo, 1039 pp.
Pouchert, C.J. (1985). The Aldrich Library of FT-IR Spectra, Vol.
1. The Aldrich Chemical Co., Milwaukee, WI.
Review on ruby promotion. (1997) Jewellery News Asia, No. 158,
p. 58.
Ringsrud R. (1983) The oil treatment of emeralds in Bogot,
Colombia. Gems & Gemology, Vol. 19, No. 3, pp. 149156.
Ringsrud (1998) Stick with cedarwood. Modern Jeweler, Vol. 97,
No. 3, p. 10.
Sinkankas J. (1981) Emeralds and Other Beryls. Chilton Book
Co., Radnor, PA, 665 pp.
Sivry B. (1985) Colles et adhdifs. Encyclopaedia Universalis,
Corpus 5, pp. 7478. (Translations where cited are by Mary
Johnson.)
Sothebys (1992a) Magnificent Jewels. Auction catalog for
Sothebys Geneva, November 18, 1992.
Sothebys (1992b) Jewels. Auction catalog for Sothebys London
(Colonnade), November 26, 1992.
Sothebys (1993) Magnificent Jewelry. Auction catalog for
Sothebys New York, October 2021, 1993.
Sothebys (1997) Magnificent Jewels. Auction catalog for
Sothebys New York, October 2930, 1997
SSEF Swiss Gemmological Institute (1998) Standards and
Applications for Diamond Report, Gemstone Report, Test
Report. SSEF, Basel, Switzerland, 118 pp.
Stecher P.G., Windholz M., Leahy D.S., Bolton D.M., Eaton L.G.
(1968) The Merck Index, 8th ed. Merck & Co., Rahway, NJ.
Survey: Emerald treatments (1997). The Guide Quarterly Report,
Vol. 16, No. 2, p. 15.
Webster R., Read P.G. (1994) Gems: Their Sources, Descriptions,
and Identifications, 5th ed. Butterworth-Heinemann, Oxford,
1026 pp.
Websters Ninth New Collegiate Dictionary (1987) MerriamWebster Inc., Springfield, MA, 1,563 pp.
Weldon R. (1998a) Treatment tracer. Professional Jeweler, Vol. 1,
No. 3, p. 50.
Weldon R. (1998b) Filling a need. Professional Jeweler, Vol. 1,
No. 5, pp. 80, 82, 84.
Zecchini P., Maitrallet P. (1998) Que peut apporter la spectrographie infrarouge dans ltude des meraudes? In D. Giard, Ed.,
Lmeraude: Connaisances Actuelles et Prospectives.
Association Franaise de Gemmologie, Paris, pp. 8196.
GEMS & GEMOLOGY
Summer 1999
107
SAPPHIRE AND GARNET

FROM K ALALANI,
TANGA PROVINCE, TANZANIA
By Antonin V. Seifert and Jaroslav Hyrsl
Located just 3 km south of the famous Umba

locality, the geologically similar Kalalani area
hosts deposits of gem-quality sapphire (reddish orange, yellow-brown, and other colors),
pyrope-almandine, and tsavorite. The gems
are mined principally from primary and eluvial sources associated with a serpentinite
body and surrounding metacalcareous rocks.
The gemological properties of sapphires from
Kalalani are very similar or identical to those
from Umba. The garnets, however, are different: Tsavorite is not known from the Umba
area, and the dark red to brownish red pyropealmandine from Kalalani is distinct from the
pink-purple rhodolite more commonly found
at Umba.
ABOUT THE AUTHORS

Dr. Seifert (seifert@cgu.cz) is a senior officer at
the Czech Geological Survey, Klarov 3, Prague
1, Czech Republic. Dr. Hyrsl is a consulting
mineralogist-gemologist based in Prague, and a
former staff member of the Faculty of Natural
Sciences at Charles University in Prague.
article.
Gems & Gemology, Vol. 35, No. 2, pp. 108120
108
Kalalani Sapphire and Garnet
he Kalalani area in northeastern Tanzania hosts several primary deposits of sapphire and the largest primary deposit of pyrope-almandine in that country. In
19891990, two significant discoveries of reddish orange
(Umba padparadscha; see figure 1) and yellow-brown sapphires were made in desilicated pegmatites that cross-cut a
small serpentinite massif. In 1994, a primary deposit of tsavorite was discovered in the surrounding graphitic gneisses,
and two pieces of gem-quality tanzanite were found in alluvial sediments adjacent to the Kalalani serpentinite massif.
This article is based on field research by the senior
author, who acted as a consultant in the exploration and
mining of the sapphire deposits in 19891990, and also
mined the large pyrope-almandine deposit in 1994. The second author visited the Kalalani area in 1993, and provided
the laboratory data on the samples.
LOCATION AND ACCESS
The Kalalani gem deposits are located approximately 100
km northwest of the town of Tanga, and 6 km (or 14 km by
dirt road) south of the Umba camp, at coordinates 4 34 S,
38 44 E (figure 2). This flat terrain is characterized by redbrown lateritic soil that is covered by thick bush vegetation
and grasslands. The mining area, called Vipigoni by locals,
lies about 5 km northwest of Kalalani village, from where it
can be accessed by a rough dirt road. During the rainy season (DecemberMarch), access to the area may be difficult.
GEOLOGY
Regional Geology. The gem-bearing districts of both the
Umba River and Kalalani areas are hosted by rocks of the
Usagaran System, a part of the Mozambique Belt. These
rocks underwent several episodes of metamorphism, plutonism, thrusting, and folding during the Late Proterozoic to
Early Paleozoic Pan-African orogenic event (Malisa and
GEMS & GEMOLOGY
Summer 1999
Figure 1. Most of the sapphires

from the Kalalani area, which
are typically fancy colors, have
been mined from primary
deposits. These reddish orange
sapphires are from the ChamShama claim. The faceted
stones range from 0.22 to 0.91
ct. Courtesy of Varujan
Arslanyan, Arslan Jewelers,
New York; photo GIA and
Tino Hammid.
Muhongo, 1990). The area was geologically surveyed

by Hartley and Moore (1965) as part of the Quarter
Degree Sheet 91 and 110, Daluni (scale 1:125,000).
The rocks consist of high-grade banded gneisses of
the granulite facies, gneisses of the amphibolite
facies, graphitic gneisses, metacalcareous rocks,
quartzites, crystalline limestones, and meta-amphibolites. These rocks surround two isolated bodies of
serpentinite (i.e., the Umba and Kalalani serpentinite
massifs; see figures 2 and 3). The geologic setting,
mineral assemblages, and presence of gem-corundum-bearing pegmatites at both localities suggest
that these massifs have a similar geologic history.
Local Geology. Corundum Deposits. Gem-quality
sapphire and ruby are confined to desilicated pegmatites that cross-cut the serpentinite (figure 3).
They generally form lenticular (lens-like) bodies
that are oriented vertically, and measure about 510
m long and up to 23 m wide. The dimensions are
quite variable, since the pegmatites commonly

pinch and swell (see, e.g., figure 4). Like those at
Umba (see Solesbury, 1967), the Kalalani pegmatites
are generally coarse grained (up to 5 cm), and contain calcic plagioclase (andesine-labradorite) but no
quartz. Other major constituents include amphibole, vermiculite, chlorite, biotite, spinel, corundum, and kyanite (see, e.g., figure 5).
The mineralized zones of the two pegmatites
that yielded reddish orange and yellow-brown sapphires measured approximately 6 m long and 2 m
wide. Although these pegmatites were only about
50 m apart, they showed distinct differences. The
reddish orange sapphire was mostly concentrated in
patches of clay minerals within golden yellow
vermiculite and was accompanied by pink spinel (of
low cabochon quality) as an important indicator.
However, the yellow-brown sapphire was found
within bluish green vermiculite, in the absence of
clay, and the tracer mineral was spinel-pleonaste.
GEMS & GEMOLOGY
Summer 1999
109
Ruby has been found at only two locations in

Kalalani. At one of these, the senior author noted
that tabular ruby was randomly distributed within
biotite in a desilicated pegmatite, and was accompanied by crystals of color-change sapphire.
Pyrope-Almandine Deposits. Along the eastern
periphery of the Kalalani serpentinite are two areas
Figure 2. The Kalalani gem deposits are associated

with the Kalalani serpentinite massif in northeastern Tanzania (inset), a few kilometers south of the
Umba area. Dirt tracks (shown here in brown) traverse both sides of the two serpentinite massifs
between the villages of Umba and Kalalani.
Figure 3. The Kalalani serpentinite is cross-cut by

desilicated pegmatites that locally contain gem
corundum. Pyrope-almandine is mined from metaamphibolites near the boundary of the serpentinite,
and tsavorite deposits are hosted by graphitic gneisses
southwest of the serpentinite. The deposits are designated here according to the name of their claim
block: 1 = original Papas pit, reopened by Ruvu
Gem Ltd. (color-change sapphire, ruby); 2 = ChamShama (reddish orange sapphire); 3 = Cham-Shama
(yellow-brown sapphire); 4 = Peter A. (an old Papas
pit; light blue and pink sapphires, ruby); 5 = Sumai
(yellow-brown sapphire); 6 = old Papas pit, reopened
by Africa-Asia Precious Stones and Mining Co. (light
blue and yellow sapphires); 7 = Peter A. (pink sapphire); 8 = Cham-Shama (pink sapphire); 9 = Peter A.
(pyrope-almandine/rhodolite); 10 = Salehe J. (pyropealmandine) 11 = Abdalah (tsavorite, main pit); 12 =
Ruvu Gem Ltd. (tsavorite); 13 = Maluki C. (tsavorite);
and 14 = Kwendo N. (alluvial garnet, sapphire, and
rare tanzanite). Mapped by A.V. Seifert, 1994.
110
GEMS & GEMOLOGY
Summer 1999
that contain meta-amphibolites (see figure 3). The

southern amphibolite body hosted an economic
deposit of pyrope-almandine (Peter A. claim), along
the faulted and sheared contact with serpentinite
(figure 6). The mineralized zone measures 100 m
long, varies from 6 to 12 m wide, and dips 45
northwest. The pyrope-almandine occurred in nodules averaging 10 cm in diameter. The nodules
were typically fractured masses that contained flattened pieces of pyrope-almandine embedded in a
fine-grained yellowish dolomite (figure 7). The bestquality stones were recovered from the northeastern part of the deposit; the material from the southwestern part was more fractured and was heavily
included.
Tsavorite Deposits. Graphitic gneisses near the
southwestern part of the serpentinite body are
interbedded with quartz lenses that are accompanied by accumulations of pure graphite, where tsavorite has formed locally. These lenses crop out
Figure 4. This geologic map (top) and cross-section
(bottom) of the Cham-Shama reddish orange sapphire deposit show the sapphire-bearing desilicated pegmatite and associated veins. Abbreviations:
a = desilicated pegmatite containing patches of
clay minerals (white) with sapphire; b = bluegreen vermiculite vein; c = golden yellow vermiculite vein; d = pyroxenite vein. Mapped by A.V.
Seifert, 1990.
Figure 5. At the Cham-Shama claim, reddish

orange sapphire crystals were recovered from desilicated pegmatites composed mainly of vermiculite
(brown flakes), calcic plagioclase (white), and
amphibole and chlorite (dark green). The prismatic
sapphire crystal in the center of the photo is about
4 cm long. Photo by Varujan Arslanyan.
locally over a distance of about 1.7 km, and each is

elongated up to 25 m in a southeasterly direction;
they vary in width from a few centimeters up to 2
m. At least four parallel zones showing surface
signs of tsavorite mineralization were traced in this
area.
Figure 6. This geologic cross-section shows the

northeastern part of the Peter A. pyrope-almandine deposit. The garnet is found as fractured
nodules within altered lenses of meta-amphibolite. The mineralized meta-amphibolites have
been faulted and sheared within the serpentinite.
Mapped by A.V. Seifert, 1994.
GEMS & GEMOLOGY
Summer 1999
111
who developed three shallow open pits to explore

the pegmatites in the 1960s. Since then, the
Kalalani area has been worked continuously by
local miners who hand sieve the eluvial soil cover
in search of fancy-color sapphires and ruby.
Organized mining of primary deposits virtually
ceased from the 1970s into the early 1980s, when
the original Papas pit, situated on the eastern border of the serpentinite (figure 3, location 1), was
reopened by Ruvu Gem Ltd. At 40 m, this is now
the deepest mine in the Kalalani area. The serpentinite rock is very soft, and mining was carried out
only by pick and shovel. A modified block mining
method was used in the open pit; that is, individual
blocks measuring about 5 m were excavated along
the pegmatite dike, leaving vertical walls in the surrounding serpentinite (figure 8). When the dike
pinched off, the miners dug a haulage level at
another 5 m depth, and the process was repeated. The
Figure 7. This nodule of pyrope-almandine (4 cm
across) from the Peter A. deposit contains gemquality fragments that are embedded in finegrained dolomite. The rhodolite nodule is attached
to its meta-amphibolite matrix. Photo by J. Hyrsl.
Figure 8. The deepest pit in the Kalalani area (40

m) was excavated for sapphire and ruby by Ruvu
Gem Ltd., using a modified block mining method.
Mining ceased here in 1994. Photo by A. V. Seifert.
Alluvial Deposits. About 1 km south of the

Kalalani serpentinite, gem-bearing alluvial sediments have been deposited among outcrops of the
crystalline dolomitic limestones. Irregularities on
the surface of the limestones have trapped the gems
at about 25 m below the surface. These stones
(sapphire, garnet, andvery rarelytanzanite) have
been derived from the Kalalani serpentinite and its
surroundings. The gems generally show a slightly
worn surface, which is consistent with what we
believe to be a relatively short transport distance.
HISTORY, MINING, AND PRODUCTION
The history of the Kalalani area is closely associated
with the nearby Umba River area, which has been
described by Solesbury (1967), Dirlam et al. (1992),
and Keller (1992). Although there are no official data
on gem production from the Kalalani area, the
senior author estimates that to date, the total production of fancy-color sapphire from the area is
about 400 kg. Thus far, less than 1% of this production has been facet grade.
Sapphire and Ruby. The first organized mining of
sapphire in the Kalalani area was by Umba Ventures
Ltd., owned by George Papas Papaeliopoulos,
112
GEMS & GEMOLOGY
Summer 1999
material from the pit was shoveled or winched to the

surface, where it was sieved and examined by hand.
During the time of greatest activity, in the mid1980s, about 30 people were involved in mining.
Some cabochon-grade color-change sapphire was
recovered, but transparent material was rare and the
faceted stones seldom exceeded 1 ct. Cabochongrade ruby (typically up to only 0.5 ct) accompanied
the sapphire in places. Although Ruvu Gem ceased
mining there in 1994, collector-quality sapphire
crystals can still be found in the mine dump.
In August 1989, an important discovery of reddish orange sapphire was found in a pegmatite dike
at the Cham-Shama claim (figure 3, location 2; see
also figure 4). The rights to develop this locality, as
well as to explore the remaining part of the claim,
were secured by V. Arslanyan of New York and H.
D. Patel of Dar es Salaam. Systematic trenching of
this claim led to the discovery of yellow-brown sapphire in another pegmatite dike (figure 3, location 3)
in March 1990.
Prior to mining the pegmatites, the overlying
eluvium was removed by a backhoe and sieved for
loose sapphire crystals. The exposed pegmatites
were then worked by pick and shovel (figure 9), or
with a backhoe after drilling and blasting. The sapphires were removed from the soft vermiculite by
hand. The mining of both deposits ceased at about
57 m depth in July 1990, and there has been no further organized activity since then.
The transparent pieces of reddish orange sapphire seldom exceeded 35 ct, although some cabochon-grade rough reached over 100 grams. Of the
approximately 80 kg of reddish orange sapphire produced, about 1.5% was of excellent gem quality. A
few of the stones were red enough to be considered
ruby (figure 10). The yellow-brown sapphire rough
ranged up to 46 ct, but clean, transparent rough
reached only 2 ct. About 100 kg of yellow-brown
sapphire were recovered, with less than 1% of gem
quality. These were the only commercially significant sapphire finds in Kalalani over the past 30
years.
In December 1989, the team briefly concentrated
on developing an old Papas pegmatite dike for
ruby and sapphire at the Peter A. claim (figure 3,
location 4). Gemmy tabular crystals of ruby up to 1
ct reportedly were found there in the early 1980s
(Peter Alfred, pers. comm., 1989). However, the
1989 effort saw no production and the locality was
abandoned. Nevertheless, small-scale mining by
primitive methods continues, to a present depth of
Figure 9. Miners work the desilicated pegmatite at

the Cham-Shama claim from which reddish orange
sapphire was recovered. The pegmatite is visible
here as patches of white and greenish blue beneath
the horizon of red lateritic soil. The hose to the left
carried compressed air for removing the excavated
material and cleaning the bottom of the pit. Photo
by A. V. Seifert.
about 30 m, and light blue and pink sapphires are

occasionally recovered in sizes ranging up to 1 ct.
During 19911994, the Sumai claims (figure 3,
location 5) were worked by the American venture
Down to Earth Co. More than 10 pegmatites were
mined to an average depth of 5 m, but none proved
economic. About 20 people were involved in the
drilling, blasting, and excavating. The pegmatitic
material was processed in a nearby crushing and
Figure 10. In addition to reddish orange sapphire
(upper left, 5.07 ct), the Cham-Shama claim has
produced a few rubies (here, 0.611.31 ct). Courtesy
of Varujan Arslanyan; photo by Maha DeMaggio.
GEMS & GEMOLOGY
Summer 1999
113
sieving apparatus (figure 11). Tens of kilograms of

yellow-brown sapphire were recovered from the
main pit. However, only a small percentage of this
was cabochon grade, and rare small (<1 ct rough)
pieces were facetable.
In 1994, an old Papas pit near the western
boundary of the serpentinite (figure 3, location 6)
was reopened briefly and mined to a depth of 25 m.
About 10 people were involved in the mining,
including Africa-Asia Precious Stones and Mining
Co. of Bangkok. Light blue and yellow sapphires
were recovered, but the facet-grade rough rarely
exceeded 1 ct.
Sapphires have also been found by locals at
many other localities in the Kalalani serpentinite
massif. Open pits for pink sapphire at the Peter A.
claim (figure 3, location 7) and at the Cham-Shama
claim (figure 3, location 8) were worked to a depth
of approximately 5 m before being abandoned.
About six claims are presently being worked in the
Kalalani area by small groups of up to about five
local miners, with a foreman appointed by the
claim owners. Currently there is little gem production from the area.
Pyrope-Almandine. The Peter A. rhodolite claim
(figure 3, location 9) has been mined since the early
1980s, and until 1989 it produced stones mostly
from the topsoil. The senior authors team began
mining the primary deposit in late 1989, using a
Figure 11. Simple crushing and sieving equipment
was used to recover sapphires at the Sumai claims
by Down to Earth Co. Photo by A. V. Seifert.
backhoe to strip the overburden. In 1990 about 40

workers selectively mined the mineralized zone
with drilling and blasting. The upper part of the
mineralized zone, to 6 m depth, was extremely
rich, and approximately 30 kg of garnet were recovered per day.
From 1991 to 1994, the deposit was worked by
local miners using only hammers and chisels. The
mineralized zone (approximately 9 m wide) was followed to 10 m depth. The best roughsome weighing over 10 gramswas found in a tunnel that followed the productive zone to a depth of 20 m.
In 19941995, the local venture Skyline
Enterprises Ltd. worked the deposit. They removed
the dangerous hanging wall, and built an access
road to the bottom of the pit. The waste rock was
removed from the open pit with the help of a backhoe, two trucks, and a bulldozer. The gem-bearing
ore was crushed and sieved at a nearby treatment
plant. The treatment plant, which had a daily
capacity of about 25 tons, yielded about 50 kg of
garnet per day. Since many larger stones were
cracked or broken during the crushing procedure,
nodules that would yield the largest, best-quality
stones were recovered directly after blasting and
carefully hand cobbed. The best stones were sold in
Idar-Oberstein, Germany, with the remainder sold
in Dar es Salaam or Bangkok.
The mine was worked to a depth of 12 m; for six
months, it produced approximately one ton of
pyrope-almandine per month. About 5% of this was
eye-clean, and 1% was microscopically clean. The
color was typically dark red to brownish red,
although some purplish red rhodolite (figure 12) was
recovered from the northeastern part of the deposit.
Gem-quality fragments averaging 12 grams were
common, with some gem pieces exceeding 10
grams. Generally, the faceted stones ranged from 2
to 10 ct, with some over 20 ct. Mechanized mining
ceased in mid-1995, and the mine recently closed
completely.
Another pyrope-almandine deposit associated
with meta-amphibolites was briefly mined on the
northeastern margin of the Kalalani serpentinite in
1991 (figure 3, location 10). The garnet was dark
red-brown and mostly translucent, but it contained
abundant black inclusions, so the pit was closed at
a depth of only 3 m.
Tsavorite. In 1994, tsavorite was discovered at the
Abdalah claim, about 200 m southwest of the
Kalalani serpentinite. About 60 people were
114
GEMS & GEMOLOGY
Summer 1999
Figure 12. Pyrope-almandine from the Peter A.

deposit is typically dark
red to brownish red (here,
2.227.05 ct). Some,
however, was the purplish red variety, rhodolite (inset, 3.1711.85 ct).
Courtesy of Varujan
Arslanyan; photo GIA
and Tino Hammid.
involved in drilling, blasting, and operating a backhoe. The main pit (figure 3, location 11) measured
100 m 50 m 20 m, and followed a mineralized
zone that was concordant with the graphitic gneisses. A single lens found at 15 m depth produced
about 6 kg of predominantly light green tsavorite.
About 1% was gem quality (see, e.g., figure 13); the
remainder showed abundant white oval inclusions
and dark impurities. The faceted stones rarely
exceeded 2 ct. Official production figures are confidential, but local miners indicated that a total of
about 20 kg of tsavorite were recovered from the
Abdalah claim in 1994. The mine closed in 1995,
and we know of no further mining activity.
Following the discovery of tsavorite at Abdalah,
exploration was launched at the adjacent Ruvu
Gem and Maluki C. claims (figure 3, locations 12
and 13, respectively) with the help of a bulldozer
and backhoe. Many signs of primary tsavorite mineralization were detected, but none of these was of
economic interest.
Tanzanite. Miners recovered two pieces of gemquality tanzanite in 1994, while searching for garnet
and sapphire in the alluvium at the Kwendo N.
claims, about 1 km south of the Kalalani serpentinite (figure 3, location 14). The larger tanzanite was
deep blue as found, and weighed over 10 ct; it was
cut into two pieces and sold in Dar es Salaam.
News of the discovery brought hundreds of local
villagers into the area. The soil covering the naturally pitted bedrock was loosened by picks, commonly
to depths of 2 5 m, and shoveled into sieve boxes at
the surface. Limited mining of the alluvial deposits
has continued, but no more tanzanite has been
recorded.
The gem material studied was collected by the
authors at the mines, and consisted of rough, polished rough, cabochons, and faceted samples of sapphire and garnet. The sapphires studied were: three
polished flattened crystals of reddish orange sapphire (0.46, 0.91, and 1.16 ct) from the Cham-Shama
claim; 10 cabochons (up to 10 ct each) and about
800 grams of rough yellow-brown sapphire from the
main pit of the Sumai claim; and four color-change
sapphire crystals (up to 5 cm long) in matrix from
the Ruvu Gem claim. About 1 kg of rough pyropealmandine and 12 faceted stones (up to 5 ct) from
GEMS & GEMOLOGY
Summer 1999
115
Figure 13. Tsavorite from the Kalalani area is typically light green, although material showing a range
of tone and saturation has been produced (shown
here, 1.193.94 ct). The tsavorite in the platinum
and diamond ring weighs 3.44 ct. Courtesy of
Varujan Arslanyan; photo GIA and Tino Hammid.
the Peter A. claim were available for the investigation. A parcel of rough containing seven polished
pieces of tsavorite (up to 1 gram each) from the
Abdalah claim was obtained later from P.
Pongtrakul of Dar es Salaam for examination.
Gemological testing was performed on all polished samples, except for five cabochons of yellowbrown sapphire which were not tested for R.I. and
S.G. Refractive indices were measured with a
German Schneider refractometer with a nearmonochromatic light source. Specific gravity was
determined hydrostatically (an average of three sets
of measurements). Fluorescence to ultraviolet radiation was observed in a darkened room using a shortand long-wave UV lamp. A polariscope was used to
observe the polarization behavior. Internal features
were examined using an immersion microscope and
a standard gemological microscope in conjunction
with brightfield, darkfield, and oblique fiber-optic
illumination, as well as polarizing filters.
The visible spectra of one reddish orange sapphire, one yellow-brown sapphire, and two pyropealmandine garnets were recorded using a PerkinElmer Lambda 9 spectrophotometer, in a range
from 350 nm to 800 nm. The chemical composition
of two polished reddish orange sapphires and three
faceted pyrope-almandines (five measurements each)
was determined quantitatively using a CAMEBAX
electron microprobe at Charles University of Prague.
TABLE 1. Properties of sapphire and garnet from Kalalani.

Sapphire
Reddish
orangea
Property
Refractive index
Birefringence
Specific gravity
Pleochroism
UV luminescence
Absorption
spectrum
Inclusions
aElectron
116
Garnet
Yellow-brown
1.7681.777
0.009
3.98
Distinct reddish
orange to yellow
None
Lines at 700, 693,
557, and 490 nm;
bands at 450, 388,
and 376 nm
1.761.77 (spot)
Not determined
4.004.05
Weak yellow-brown
to yellow
None
Lines at 550, 453, and
421 nm
Isotropic irregular
plates oriented
parallel to the
basal plane
Fractures, red-brown and

yellow-brown flakes, rutile
needles, zircon crystals,
rare two-phase fingerprints
Brownish red
(pyrope-almandine)
Green
(tsavorite)
1.763 1.770
None
3.863.90
None
1.7371.740
None
3.633.65
None
None
Bands at 571, 523,
508, 460, 425, and
366 nm
None
Not determined
Apatite, rutile,
fingerprints
Anisotropic white particles

and negative crystallites in
rows, two- and multi-phase
inclusions, black plates,
rounded colorless crystals
microprobe analyses of two samples showed 0.9 wt.% Fe2O3 and 0.15 wt.% Cr2O3 ; other trace elements were below the detection limit.
GEMS & GEMOLOGY
Summer 1999
The mineralogical identification of some surfacereaching inclusions was made using an X-ray
diffractometer (DRON II) at Charles University.
RESULTS
The properties of the sapphires and garnets are
reported in table 1. None of the sapphires showed
evidence of heat treatment.
Corundum. The reddish orange Cham-Shama sapphire that we studied formed tabular crystals with
no pyramidal faces; only prism and basal pinacoid
faces were present. The dichroism was reddish
orange (parallel to the c-axis) and bright yellow. The
greatest range in the R.I. values was 1.768 and
1.777, yielding a birefringence of 0.009. The average
S.G. of the three samples was 3.98. Electron microprobe analyses showed the presence of 0.9 wt.%
Fe2O3 and 0.15 wt.% Cr2O3; other trace elements
were below the detection limit. The absorption
spectrum showed lines at 700, 693, and 557 nm;
bands at 450, 388, and 376 nm; and a fine line at 490
nm. The samples did not exhibit any fluorescence.
Microscopic observation revealed inclusions of very
small (up to 0.05 mm), pale, isotropic irregular plates,
oriented parallel to the basal plane {0001}. Their R.I.
was higher than that of corundum, as verified by the
Becke line. Some of these plates had an incomplete
hexagonal outline, and some had twin lines.
The yellow-brown sapphire crystals were opaque
to translucent, and tabular to rhombohedral. They
showed weak pleochroism, with a notable orange
tint seen parallel to the c-axis and yellow perpendicular to the c-axis. The R.I. values were 1.761.77
(spot method), and the S.G. varied from 4.00 to 4.05.
No fluorescence to UV radiation was observed. The
absorption spectrum showed two major lines at 550
and 453 nm, and a minor line at 421 nm. The samples contained many cracks and dark inclusions.
The most common inclusions were very thin redbrown and yellow-brown flakes (figure 14) oriented
parallel to the basal plane, which produced a schiller
effect in the samples. These flakes are probably
hematite, as seen by Hnni (1987) in Umba sapphire. Similar-appearing inclusions in some samples
were identified as vermiculite by X-ray diffraction
analysis. Some stones showed a network of rutile
needles oriented at 120 to one another. Also noted
were tiny zircon crystals and very rare fingerprints
consisting of two-phase inclusions (liquid and gas).
The samples of color-change sapphire were only
slightly translucent. They consisted of a hexagonal
Figure 14. The thin flakes in this yellow-brown

sapphire cabochon are probably hematite; vermiculite might also be present. Photomicrograph by
J. Hyrsl; oblique illumination, magnified 6.
prism {1120} that was terminated by a basal plane
{0001} or rhombohedron {1011}. The samples

showed a distinct change of color from blue or grayish blue in daylight to purple in incandescent light.
The sapphire fluoresced weak red to long-wave UV
and was inert to short-wave UV.
Pyrope-Almandine. The pyrope-almandine samples
were brownish red. R.I. ranged from 1.763 to 1.770,
and S.G. from 3.86 to 3.90. They showed moderate
anomalous double refraction in the polariscope. The
absorption spectra had three characteristic bands
centered at 571, 523, and 508 nm, and other bands
at 460, 425, and 366 nm. The most common inclusions were anisotropic rounded crystals of apatite
and needle-like rutile (figure 15). Also prominent
were partially healed fractures (figure 16) that
appeared similar to the fingerprints commonly
Figure 15. Rounded apatite crystals and needle-like
rutile crystals were commonly seen in pyropealmandine from the Peter A. claim. Photomicrograph by J. Hyrsl; brightfield illumination,
magnified 15.
GEMS & GEMOLOGY
Summer 1999
117
Figure 16. Some of the pyrope-almandine samples

had fingerprintsactually, partially healed fracturesthat formed attractive patterns. Photomicrograph by J. Hyrsl; brightfield illumination,
magnified 15.
seen in sapphire or spessartine, as illustrated by

Gbelin and Koivula (1986, p. 293). The substance
(gas or liquid) within the fingerprints is mostly
isotropic, but some portions are anisotropic, with an
R.I. lower than that of the host garnet (as seen with
the Becke line). The electron microprobe data
showed that each sample was chemically homogeneous (see standard deviations in table 2). As
expected, those samples with greater iron content
had higher R.I. and S.G. values (table 2).
Mineralogically, the garnet is referred to as an ironrich pyrope; gemologically, it is pyrope-almandine
or, if there is a purple component, rhodolite
(Stockton and Manson, 1985).
TABLE 2. Physical and chemical properties of pyropealmandine from the Peter A. claim.
Sample
Property
Refractive index
Specific gravity
Oxides (wt.%)a
SiO2
TiO2
Al2O3
FeO
MnO
Total
1.64 ct
1.28 ct
0.76 ct
1.763
3.86
1.767
3.87
1.770
3.90
39.23 0.99
0.06 0.01
22.20 0.55
21.66 0.37
0.80 0.02
100.12
38.12 1.36
na
22.73 0.40
23.90 0.18
0.64 0.05
99.74
39.63 0.72
nab
23.53 0.27
19.05 0.24
0.81 0.08
101.03
Figure 17. The central elongate inclusion (0.35 mm

long) in this tsavorite is partially filled with a
fluid that contains a single bubble and three
isotropic daughter minerals. Such multiphase
inclusions are rare in tsavorite. Photomicrograph
by J. Hyrsl; brightfield illumination.
Tsavorite. The tsavorite samples revealed R.I. values of 1.7371.740, and S.G.s of 3.633.65.
Moderate anomalous double refraction was seen
with the polariscope. The samples were inert to UV
radiation and appeared pink with the Chelsea filter.
Microscopic observation revealed planes of irregular
anisotropic white inclusions and negative crystallites, often in parallel rows. Several inclusions consisted of a flaky anisotropic substance together with
a small bubble. One sample contained an unusual
multiphase inclusion that was mostly filled by a
fluid in which floated a small bubble and three
isotropic crystals (cubes?); all of these were attached
Figure 18. Rarely, opaque black plates were
observed as inclusions in the Kalalani tsavorite;
they were probably graphite. This crystal measures 0.5 mm in diameter, and appears light colored due to reflections off its shiny surface.
Photomicrograph by J. Hyrsl; oblique brightfield
illumination.
aFor each sample, the average of five analyses (by electron microprobe) is provided together with the standard deviation.
bna = not analyzed.
118
GEMS & GEMOLOGY
Summer 1999
to an anisotropic material (figure 17). Rare, opaque

black plates were probably graphite (figure 18).
Small rounded colorless crystals with a high relief
(probably apatite) were common in some samples.
All of the anisotropic inclusions described above
have a lower R.I. than tsavorite, as determined by
Becke line observations.
Four multiphase inclusions in tsavorite from
Kalalani were studied by Petra Neumanova of
Charles University at Prague. They were interpreted
as primary fluid inclusions because of their large
size, relative isolation, and similar phase composition and phase ratios. The mineral phase within the
fluid inclusions was anisotropic, and these crystals
apparently were daughter minerals.
DISCUSSION
Gemstones from Kalalani are similar to those from
Umba, which is well known for the production of
sapphires in an unusually broad range of fancy colorsparticularly the color-change and Umba padparadscha sapphiresas well as for light blue sapphire and ruby. The garnet group at Umba is represented by pink-purple rhodolite, malaia garnet
(called also Umbalite), color-change garnet, and
pale pink grossular. Cr-tourmaline, spinel, scapolite,
apatite, zircon, Cr-diopside, kyanite, and turquoise
have also been reported from the Umba area
(Dirlam et al., 1992). The second author recently
identified a mineral new to Umba, gemmy brown
enstatite. While most Umba gems have been recovered from alluvium, gems at Kalalani are mined
almost entirely from primary deposits.
The color range and characteristic inclusions of
reddish orange sapphires from Kalalani (figure 19)
are very similar to those from Umba, as well as to
orange sapphires from the Tunduru region in southern Tanzania (Henn and Milisenda, 1997) and to
orange sapphires from Malawi (Henn et al., 1990).
Although the traditional padparadscha sapphire
from Sri Lanka shows more pink and less orange,
some reddish orange sapphires from Kalalani have a
prominent pinkish tint. The visible absorption spectrum of the reddish orange sapphire from Kalalani is
identical that of similar material from Umba (as
recorded by Hnni, 1987), and confirms the presence of Fe3+ and Cr3+ as the main coloring agents.
On the basis of the studies of similar sapphires from
Umba (e.g., Schmetzer et al., 1982), the reddish
orange sapphires from Kalalani probably do not
respond favorably to heat treatment, because color
centers apparently are not present.
Figure 19. The Kalalani area has yielded sapphires

in colors similar to those seen in sapphires from
Umba and Tunduru. These yellow to reddish
orange sapphires from the Cham-Shama claim
range from 1.53 to 2.29 ct. Courtesy of Varujan
Arslanyan; photo GIA and Tino Hammid.
The Kalalani garnets differ from their Umba

counterparts. Malaia garnet, color-change garnet,
and pale pink grossular are not known from
Kalalani, and tsavorite has not been found at Umba.
The Kalalani pyrope-almandine is typically dark red
to brownish red, whereas that from Umba is pinkpurple (i.e., rhodolite). Tsavorite from Kalalani has
the same R.I. and S.G. as most tsavorite from Kenya
or Tanzania (see, e.g., Keller, 1992). Tsavorite from
the Merelani Hills in Tanzania (see Kane et al.,
1990) can be distinguished easily from Kalalani tsavorite due to its negative reaction to the Chelsea filter. Green grossular from Mali, which has a similar
pink to red response to the Chelsea filter, shows a
higher R.I. than Kalalani tsavorite.
GEMS & GEMOLOGY
Summer 1999
119
Curiously, the parcel of rough tsavorite from

Kalalani that we obtained for study was also found
to contain minor amounts of green to dark green
Cr-tourmaline and light green kornerupine, as confirmed by X-ray diffraction analyses. These gem
minerals can be easily distinguished from tsavorite
by their anisotropy, pleochroism, and particularly
their fluorescence to UV radiation. Chrome tourmaline fluoresces light yellow to short-wave, and is
inert to long-wave, UV radiation; whereas kornerupine shows strong yellow fluorescence to both
short- and long-wave UV. When viewed with a
Chelsea filter, Cr-tourmaline appears pink to red
and kornerupine remains green.
initial investment in prospecting and mining would

be necessary. Although a source of tanzanite has
not been found in the Kalalani area, favorable geologic potential is indicated by the presence of crystalline dolomitic limestones and graphitic gneisses
with quartz veins and pure graphite, together with
widespread hydrothermal alteration.
The Kalalani area also has potential for other
gem materials. The senior author has found green
tourmaline in small pegmatite lenses hosted by
quartz-feldspar gneisses southwest of the Kalalani
massif. Also, small crystals of transparent green
diopside have been recovered from fissures inside
the Kalalani massif.
FUTURE POTENTIAL
The Kalalani area has never been systematically
prospected, but there is some evidence that additional sapphire-rich pegmatites exist. Except for the
old Papas pits, mining generally has extended to
an average depth of only 57 m. Below this depth,
there is a high potential for the discovery of additional sapphires, and efforts should be made to
reach the deeper zones of abandoned deposits that
were formerly productive.
The remaining reserves of pyrope-almandine ore
at the Peter A. deposit may be roughly estimated at
up to 100 tons, to a depth of 25 m. Also, the abundant surface signs of primary tsavorite mineralization southwest of the Kalalani serpentinite should
be explored more thoroughly. In both cases, a large
Acknowledgments: The principal author appreciates

the friendship and assistance of J. P. Giordano of New
York, who was part of the Kalalani team in 1989 1990.
H. D. Patel of Dar es Salaam and V. Arslanyan of New
York are thanked for their friendship and generous
support during the fieldwork. Mr. Arslanyan also kindly provided additional information, mining photos,
and gem material for photography. We thank Dr. C.
Milisenda of the Gemmological Institute of IdarOberstein for providing spectrophotometric analyses,
Dr. J. Hovorka of Charles University for the electron
microprobe analyses of pyrope-almandine, and E.
Zitova of the Geological Survey in Prague for drafting
the maps. Rough samples of tsavorite, as well as recent
locality and production information, were kindly provided by P. Pongtrakul of Dar es Salaam.
REFERENCES
Dirlam D.M., Misiorowski E.B., Tozer R., Stark K.B., Basset A.M.
(1992) Gem wealth of Tanzania. Gems & Gemology, Vol. 28,
No. 2, pp. 80102.
Gbelin E.J., Koivula J. (1986) Photoatlas of Inclusions in
Gemstones. ABC Edition, Zurich, Switzerland.
Hnni H.A. (1987) On corundum from Umba Valley, Tanzania.
Journal of Gemmology, Vol. 20, No. 5, pp. 278284.
Hartley E.W., Moore W.R. (1965) Quarter Degree Sheet 91 and
110, Daluni (scale 1:125,000), Geological Survey of Tanzania,
Dodoma.
Henn U., Bank H., Bank F.H. (1990) Red and orange corundum
(ruby and padparadscha) from Malawi. Journal of
Gemmology, Vol. 22, No. 2, pp. 8389.
Henn U., Milisenda C.C. (1997) Neue Edelsteinvorkommen in
Tansania: Die Region TunduruSongea. Zeitschrift der
2943.
120
Kane R.E., Kampf A.R., Krupp H. (1990) Well-formed tsavorite

gem crystals from Tanzania. Gems & Gemology, Vol. 26, No.
2, pp. 142 148.
Keller P.C. (1992) Gemstones of East Africa, Geoscience Press,
Tucson, AZ, pp. 3241.
Malisa E., Muhongo S. (1990) Tectonic setting of gemstone mineralization in the Proterozoic metamorphic terrane of the
Mozambique Belt in Tanzania. Precambrian Research, Vol.
46, Nos. 12, pp. 167176.
Schmetzer K., Bosshart G., Hnni H. (1982) Naturfarbene und
behandelte gelbe und orange-braune Sapphire. Zeitschrift der
265279.
Solesbury F.W. (1967) Gem corundum pegmatites in NE
Tanganyika. Economic Geology, Vol. 62, No. 7, pp. 983991.
Stockton C.M., Manson D.V. (1985) A proposed new classification for gem-quality garnets. Gems & Gemology, Vol. 21, No.
4, pp. 205218.
GEMS & GEMOLOGY
Summer 1999
RUSSIAN
SYNTHETIC AMETRINE
By Vladimir S. Balitsky, Taijin Lu, George R. Rossman, Irina B. Makhina,
Anatolii A. Marin, James E. Shigley, Shane Elen, and Boris A. Dorogovin
Gem-quality synthetic ametrine has been produced commercially in Russia since 1994, by
hydrothermal growth from alkaline solutions.
Faceted synthetic ametrine has many similarities to its natural counterpart from Bolivia. For
the most part, however, the synthetic ametrine
obtained for this study could be identified by a
combination of characteristics, including
growth features such as twinning and color
zoning. EDXRF chemical analyses revealed
higher concentrations of K, Mn, Fe, and Zn
than in natural ametrine. IR spectra of the synthetic citrine portions showed more intense
absorption in the 37002500 cm1 range compared to natural ametrine; the synthetic
amethyst zones showed a weak diagnostic
peak at 3543 cm 1 .
ABOUT THE AUTHORS

Dr. Balitsky (balvlad@iem.ac.ru) is head of the
Laboratory of Synthesis and Modification of
Minerals at the Institute of Experimental
Mineralogy, Russian Academy of Science (IEM
RAS), in Chernogolovka, Moscow District,
Russia. Dr. Lu is a research scientist, Dr. Shigley
is director, and Mr. Elen is a research gemologist at GIA Research, Carlsbad, California. Dr.
Rossman is professor of mineralogy at the
California Institute of Technology (Caltech) in
Pasadena, California. Dr. Makhina is project
leader, Dr. Marin is head of the Department for
Experimental Mineralogy, and Dr. Dorogovin is a
director at the Russian Research Institute for
Material Synthesis (VNIISIMS) in Alexandrov,
Vladimir District, Russia.
article.
Gems & Gemology, Vol. 35, No. 2, pp. 122134
122
methyst is one of the most popular colored gemstones because of its color, availability, and
affordability. Citrine is less often encountered in
nature, although large quantities are produced by the heat
treatment of amethyst. In the 1970s a spectacular combination of amethyst and citrine, commonly referred to as
ametrine, became available (Hehar, 1980; Koivula, 1980;
Vargas and Vargas, 1980). Although there was early speculation (stimulated in part by observations of Nassau, 1981)
that such ametrine was synthetic or the effect was produced
by treatment, the source of natural, untreated ametrine was
eventually revealed to be the Anah mine in Bolivia (Jones,
1993; Vasconcelos et al., 1994). In fact, efforts were underway in Russia in 1977 to understand the fundamental mechanisms of synthesizing ametrine (Balitsky and Balitskaya,
1986). By 1994, the growth technology was successfully
established, and the first commercial batch (on the order of
100 kg) of synthetic ametrine crystals was produced.
Currently, a few kilograms per month of synthetic
ametrine crystals and polished slabs are being sold. Faceted
synthetic ametrine is also available (see, e.g., figure 1). Batch
capacity has now grown to more than 300 kg. If necessary to
meet demand, several tons of synthetic ametrine a year
could be produced by a single laboratory (the Russian
Research Institute for Material Synthesis [VNIISIMS] of the
Ministry of Geology of the Russian Federation, in
Alexandrov, Vladimir District). Most of the synthetic
ametrine is sold in Russia. Recently, however, synthetic
ametrine (of unknown origin) has appeared in the Bolivian
market (Rivero, 1999).
Although at least three morphological types of synthetic
ametrine crystals can be grown, this article reports on the
most abundant commercial type. The cause of color, growth
conditions, crystal morphology, gemological properties,
internal features, spectra, and chemical composition are
described, and key tests to separate this synthetic ametrine
from its natural counterpart are summarized.
GEMS & GEMOLOGY
Summer 1999
Figure 1. Synthetic ametrine

hydrothermally grown in Russia
became commercially available
in the mid-1990s. Here, faceted
synthetic ametrine (the three
samples on the left, 11.3013.38
ct) is shown together with natural ametrine from the Anah
mine in Bolivia. The carving of
natural ametrine in the center
weighs 35.15 ct and was executed by Michael Dyber; the
faceted stones on the right
weigh 4.7122.31 ct. Photo
Harold & Erica Van Pelt.
HISTORICAL BACKGROUND
Synthetic quartz with the distinctive amethyst-citrine bi-coloration was first mentioned in the
Russian scientific literature in the late 1950s and
early 1960s. In particular, Tsinober and Chentsova
(1959) reported that, after X-ray irradiation, certain
crystals of brown synthetic quartz developed purple
(amethyst) coloring in the rhombohedral growth
sectors. This treatment phenomenon was revisited
in the course of research into synthetic brown
quartz and synthetic amethyst (e.g., Balakirev et al.,
1979; Balitsky, 1980; Balitsky and Lisitsina, 1981;
Nassau, 1981).
The first investigations aimed at producing synthetic ametrine on a commercial scale began in
1982 at the Institute of Experimental Mineralogy of
the Russian Academy of Science (IEM RAS) in
Chernogolovka, Moscow District. These experiments investigated how physical, chemical, and
growth factors affected the capture and distribution
of color-causing iron impurities during the growth
of synthetic ametrine (as well as other iron-bearing
varieties of quartz; see Balitsky and Balitskaya,
1986). In 19931994, the research was transferred to
VNIISIMS, also near Moscow, although colleagues
at both institutions continued to collaborate. Later,
VNIISIMS organized a special company named
AMeT p N H (Ametrine) to market synthetic ametrine. Today, almost all synthetic ametrine is produced at VNIISIMS.
SCIENTIFIC FOUNDATIONS FOR THE
GROWTH OF SYNTHETIC AMETRINE
Iron Impurities in Colored Quartz. As previously
established by several researchers (e.g., Hutton,
1964; Zaitov et al., 1974; Lehmann, 1975; Balakirev

et al., 1979; Balitsky, 1980; and Rossman, 1994), the
various colors of iron-bearing quartz (e.g., amethyst,
citrine, green, and brown quartz) are related to different forms of iron (Fe) in the crystal structure.
Only a small quantity of iron (e.g., about 0.010.05
wt.% Fe for amethyst) is needed to produce coloration in quartz. Even though the element iron and
the mineral quartz are both common in nature, the
occurrence of ametrine is quite rare. This is due to
the specific physical, chemical, and geological conditions necessary for ametrine formation (Balitsky
and Balitskaya, 1986; Vasconcelos et al., 1994).
The basic unit of the quartz structure is a tetrahedron consisting of a silicon (Si) atom and four
oxygen (O) atoms. Trivalent iron (Fe3+) can substitute for Si4+ in a tetrahedral site, or it can fill interstitial sites (as extra atoms in voids) between neighboring tetrahedra. At least two different interstitial
positions exist in the quartz structure (Rossman,
1994, p. 434).
The amethyst coloration develops in a two-step
process. First, the substitution of Si4+ by Fe3+ during
growth produces a precursor center (Fe3+ substitutional center, [FeO4] ). The resulting deficiency of
positive charge is compensated by the introduction
of alkali metal ions (such as Li+ or Na+) or protons
(H+) into the quartz. This precursor center absorbs
so little light that the material is almost colorless.
The purple color develops only when the precursor
center is converted to the amethyst color center
([FeO4] 0) by exposure to ionizing radiation. For synthetic amethyst, a dose of 56 Mrads of cobalt-60
gamma-ray radiation is commonly used, although
electron irradiation is employed occasionally.
GEMS & GEMOLOGY
Summer 1999
123
They did this with the help of a calculated diagram

(known as an Eh-pH diagramsee e.g., Garrels and
Christ, 1965; figure 2), which shows the stability of
ionic and molecular forms of iron in aqueous solution as a function of the oxidation potential (Eh) and
the acidity (pH). From this diagram, it can be seen
that there is a limited regime under which all the
iron will be in the trivalent state (i.e., region 1 in figure 2). This observation was fundamental to the
development of technologies for growing both synthetic amethyst and ametrine (Balitsky et al., 1970;
Balitsky, 1980). The diagram further shows that
there are conditions under which synthetic
amethyst, citrine, or ametrineas well as bi-colored
amethyst-brown quartz and amethyst-green
quartzcan be grown in acidic, neutral, or alkaline
solutions.
Figure 2. This diagram relates the Eh (oxidation

potential) and pH (acidity) of iron (Fe)-containing
aqueous solutions to the chemical form in which
the iron exists in the solution. The regions corresponding to the various color varieties of iron-containing quartz are indicated by purple (amethyst),
yellow (citrine), brown (brown quartz), and green
(green quartz). 1 = region of amethyst, citrine, and
ametrine (purple-and-yellow striped) formation;
2 = narrow region of brown quartz formation (along
line), with adjacent green-and-brown quartz and
amethyst-and-brown quartz regions; and 3 = region
of green quartz formation. After Melnik et al. (1974).
Citrine coloration in synthetic quartz is also

related to Fe3+. However, in contrast to the color
mechanism for amethyst, the iron does not enter
the crystal structure. Rather, it is captured on
growth faces of the quartz crystalpossibly as a
small aggregate of a hydrated iron hydroxidein the
form of a so-called nonstructural impurity.
Various colors can result, depending on the oxidation state (valence) of the nonstructural iron: Green
coloration is caused by the presence of Fe2+ impurities, and the simultaneous capture of both Fe2+ and
Fe3+ gives rise to brown quartz.
Because the color of iron-bearing quartz depends
on the valence state of the iron, the Russian
researchers first had to determine the conditions
under which the desired form of iron ions would
exist in the quartz-forming hydrothermal solutions.
124
Growth Rate and Coloration. The incorporation of

iron impuritiesand thus the development of the
specific ametrine colorationis dependent on the
growth rate and surface structure of the different
growth sectors (as explained by Balitsky and
Balitskaya, 1986; see figure 3). The growth conditions must be closely regulated so that the
amethyst- and citrine-forming iron impurities are
captured simultaneously by different growth sectors
in a crystal.
Commercial Growth and Irradiation Processes. The
technology for synthesizing ametrine is similar to
that for synthetic amethyst and synthetic citrine.
The crystals are grown hydrothermally in concentrated alkaline (K2CO3) solutions at temperatures of
330 370C and pressures in the range of 1,200 to
1,500 atmospheres, in autoclaves ranging from
1,000 to 1,500 liters in volume (for photos of the factory, visit the Web site http://minerals.gps.caltech.
edu/ametrine). The crushed silica used to grow the
synthetic quartz is derived from natural or synthetic
quartz with a very low aluminum content (10100
ppm). To facilitate the incorporation of Fe 3+ into the
growing crystal structure, manganese nitrate,
Mn(NO3)2, is used as an oxidizer.
The synthetic ametrine crystals are grown on
seeds, prepared from colorless synthetic quartz crystals, that have a specified crystallographic orientation, shape, and size. Two types of seeds are used to
produce the tabular synthetic ametrine crystals: (1)
seeds cut parallel to the positive and negative rhombohedra; and, more commonly, (2) seeds cut parallel
to the basal pinacoid. In both cases, the crystals are
GEMS & GEMOLOGY
Summer 1999
grown at rates slightly below the critical rates (see

figure 3) to produce the optimum distribution and
intensity of color. The basal surfaces of crystals
grown on seeds cut parallel to the basal pinacoid
grow at about 0.8 1.0 mm/day. Rhombohedral
faces, however, grow much slower (0.30.4 mm/day
for the negative rhombohedral face, and 0.080.1
mm/day for the positive rhombohedral face). The
rhombohedral r and z sectors are colorless or pale
yellow as grown; subsequent gamma-ray irradiation,
as described below, produces the purple color in
these sectors (figure 4). The pinacoidal c sectors are
yellow to orange as grown. Their primary citrine
color is unaffected by the irradiation, although in a
Figure 3. The growth rate of synthetic ametrine

strongly influences the coloration of the r and z
sectors (see figure 4 for an example of the coloration of the r, z, and c growth sectors). This
schematic diagram shows the effect of growth rate
on the capture and distribution of amethyst- and
citrine-forming impurities in these two sectors.
The color intensity is directly related to the quanr and V z are the
tity of captured iron impurities. Vcr
cr
critical growth rates for the r and z faces, respectively (i.e., the growth rate at which the maximum
color intensity will be produced). Above these critical values, the color intensity decreases, and the r
and z sectors begin to capture citrine-forming
impurities. In the diagram, the color of each line
indicates the particular color-forming impurities
that are incorporated into the r and z sectors at
various growth rates. At higher growth rates, the r
and z faces incorporate only citrine-forming impurities (i.e., in regions III, IV, and V for r and region
V for z). To grow synthetic ametrine crystals with
r and z sectors that will show an intense purple
with irradiation, the growth is carefully maintained at just below the critical rate. The c faces
capture only the citrine-forming impurities,
regardless of the growth rate.
Figure 4. Commercially important synthetic

ametrine is produced using rectangular seeds cut
parallel to the basal pinacoid. The crystal on the
upper left is shown beforeand the crystal on the
bottom (1.1 kg) aftergamma-ray irradiation. The
slice on the upper right is a cross-section of an irradiated crystal; about 20 such slabs can be cut from
a typical crystal. Within the slice, the colorless
band is the seed plate, the yellow regions on either
side of the seed are the c sectors, and the dark purple zones are the r and z sectors (the r sectors are
darker purple). Photo by Maha DeMaggio.
small proportion of crystals these sectors turn

brownish orange due to the presence of a minor
amethyst component.
By changing the shape, size, and orientation of
the seed, and using other proprietary techniques, it
is possible to grow synthetic ametrine crystals
showing various habits, from tabular to prismatic
(both symmetric and asymmetric), with a complex
distribution of amethyst- and citrine-colored internal growth sectors. Most commercial production
employs rectangular seed plates that are cut parallel
to the basal pinacoid and elongated in the trigonal
prism x [1120] direction (known as the ZX-cut). All

of the synthetic ametrine samples examined for this
study were grown using these seeds, which are typically 200300 mm long, 2050 mm wide, and
1.52.0 mm thick. Their use results in tabular crystals with the best color contrast between the different growth sectors. Very small quantities of crystals
have been grown with the rectangular basal pinacoid seeds elongated in the hexagonal prism m
[1010] direction (the ZY-cut), to get a more varied

distribution of color.
To study the influence of seed orientation on the
development of the synthetic amethyst and citrine
portions, the Russian researchers are growing some
crystals on seeds oriented at 1530 to the basal
pinacoid face. Synthetic ametrine crystals with a
GEMS & GEMOLOGY
Summer 1999
125
about three hours, resulting in a dose of about 5

Mrads. To fulfill customer requests, occasionally
the crystals are heat treated (at 330350C for one
hour) to reduce the color intensity of the synthetic
citrine portions.
Figure 5. Each slab of synthetic ametrine typically

yields 810 faceted samples. The largest faceted
synthetic ametrine shown here weighs 20 ct. Photo
by Maha DeMaggio.
prismatic morphology are also grown; they require

special seeds and a proprietary morphology-controlling technique. Currently, these crystals are being
produced for commercial use at a rate of only 23
kg per month. A detailed gemological study of this
material is underway by the current authors.
The largest synthetic ametrine crystal grown
thus far by the Russian authors weighed 4.5 kg.
Crystals typically range from 0.45 to 1.2 kg, and
measure 20 6 2 cm to 21 8 4 cm. For maximum amethyst color saturation, the crystals are
irradiated with a cobalt-60 gamma-ray source for
Figure 6. These faceted synthetic ametrines
(7.53 46.44 ct) represent some of the samples
examined for this study. Photo by Maha DeMaggio.
126
CUTTING
The elongate, tabular crystals of synthetic ametrine
are first cut perpendicular to their length into several slabs (see, e.g., figures 4 and 5). A 200-mm-long
tabular crystal typically yields 20 slabs. The slabs
contain both amethyst- and citrine-colored portions, except for those cut from near the ends of the
crystals. Each slab yields about 810 faceted gems
with an average weight of 1015 ct (see, e.g., figure
5), although gems exceeding 50 ct can be cut. Some
of the faceted gems will be almost entirely purple or
yellow, depending on the portion of the slab from
which they were cut. Rectangular, cushion, and
oval shapes are most common.
Approximately 180 faceted stones are obtained
from a typical rough crystal, for an overall yield of
about 50%. No noticeable differences in physical
properties, such as brittleness and hardness, have
been reported for synthetic and natural ametrine
during cutting and polishing.
All of the synthetic ametrine samples included in
this study were grown on tabular ZX-cut seeds.
About 300 synthetic ametrine crystals (80 20
10 mm to 250 80 50 mm) and 150 faceted gems
(140 ct) were examined by the Russian authors and
their staff. The GIA and Caltech authors studied
seven crystals, six polished crystal fragments, and
10 faceted samples of synthetic ametrine. The
largest crystal studied at GIA measured 207.1
64.5 30.2 mm and weighed 1.1 kg. The faceted
samples examined at GIA were cut in oval and various fancy shapes, and ranged from 7.53 ct to 46.44
ct (figure 6). To explore the dependence of physical
and gemological properties on crystallographic orientation, we had 12 rough samples cut and polished
in specific orientations (e.g., parallel and perpendicular to the optic axis). The resulting slabs weighed
10.3065.80 ct and were studied by authors from all
three institutions.
In addition, the GIA and Caltech researchers
documented the gemological properties of the following samples of natural ametrine from the Anah
mine: seven crystals (5056 grams), three slabs cut
parallel to the optic axis (25.8848.55 ct), three slabs
GEMS & GEMOLOGY
Summer 1999
cut perpendicular to the optic axis (26.00168.90 ct),

and 30 faceted samples (3.90 ct140.39 ct; see, e.g.,
figure 7) in various shapes.
Standard gem-testing equipment and methods
were used to characterize all 40 of the faceted samples examined at GIA. The equipment included: a
GIA GEM Duplex II refractometer with a nearmonochromatic, sodium-equivalent light source; a
polariscope, a calcite dichroscope, a Chelsea color
filter, a GIA GEM combination long- and shortwave ultraviolet lamp unit (4 watts), a Beck deskmodel prism spectroscope mounted on an illumination base, and a binocular gemological microscope. Twinning was observed with an immersion
cell and cross-polarized light, at 7 45 magnification. Specific gravity values were hydrostatically
determined for three synthetic and five natural
samples using a Mettler AM100 electronic balance.
Absorption spectra in the ultraviolet-visible-near
infrared (UV-Vis-NIR) were obtained at GIA with
two instruments (a Hitachi U-4001 and a PyeUnicam 8800) for some of the natural ametrine (two
crystals, three slabs, and 15 faceted stones) and synthetic ametrine (five crystals, three slabs, and 10
faceted samples). UV-Vis-NIR absorption spectra
were also collected at Caltech, with a diode array
microspectrometer. Four slabs and one cube of natural ametrine, and four slabs of synthetic ametrine,
were chosen for detailed study on this system.
Infrared spectra were recorded at GIA (on the same
samples as for UV-Vis-NIR spectroscopy) with a
Nicolet Magna-550 spectrophotometer, and at
Caltech (four slabs each of natural and synthetic
ametrine) with a Nicolet 860 instrument.
Additional data for 20 crystals, 30 slabs, and 20
faceted synthetic ametrine samples were collected
by the Russian authors using a SPECORD M40 spectrometer (visible range) and a Perkin-Elmer 983 spectrophotometer (infrared range). Equivalent instruments at the three locations gave comparable results.
Qualitative chemical analyses of 15 natural and
15 synthetic ametrine samples were obtained at
GIA by energy-dispersive X-ray fluorescence
(EDXRF) spectrometry, using a Tracor-Northern
Spectrace TN 5000 instrument. Semi-quantitative
chemical analyses of five natural (two slabs, three
faceted) and four synthetic (two slabs, two faceted)
ametrine samples were also performed at Caltech,
using a Kevex-IXRF 7000 system in conjunction
with the Fundamental Parameters program (supplied with the instruments software) and silicametal oxide mixtures as confirmatory standards.
Figure 7. Natural ametrine was obtained for comparison with the synthetic material. The largest
faceted ametrine shown here weighs 21.93 ct.
Photo by Maha DeMaggio.
Raman spectra were collected using a Renishaw

2000 laser Raman microspectrometer at GIA on 10
natural and 10 synthetic faceted samples. Each was
analyzed in both the citrine and amethyst regions
without any change in the orientation of the sample. In addition, Raman spectra of doubly polished
slabs of both natural and synthetic ametrine were
obtained at Caltech using a Kaiser Optics
Holoprobe instrument with an EIC probe head.
Because the initial results obtained by both instruments suggested that Raman spectra were not particularly useful for identification purposes, a comprehensive Raman study was not undertaken.
To evaluate the color stability of the synthetic
ametrine, the Russian authors heated 30 rough crystals, 20 slabs, and 50 crystal fragments in air, at
temperatures up to 700C. After the amethyst color
faded in a particular temperature range, the faded
samples were re-irradiated with gamma rays to
investigate color restoration.
RESULTS AND DISCUSSION
Morphology of Natural and Synthetic Ametrine. As
seen in most natural amethyst, natural ametrine
crystals typically have a prismatic habit, elongated
parallel to the optic axis (c-axis), and are terminated
by the rhombohedral faces r {1011} and z {0111} (figure 8). In some crystals, both of these rhombohedra
are equally developed, so that the terminations
resemble hexagonal pyramids; a small, rough c
(0001) face is occasionally present (Vasconcelos et
al., 1994). Prism faces m {1010} are commonly

developed, and are striated at right angles to the caxis. Rarely, other faces are developed, such as the
trigonal prism x {1120} and trigonal dipyramid s
{1121}; if present, they are very small.
GEMS & GEMOLOGY
Summer 1999
127
Figure 8. Natural quartz crystals commonly show a

simple arrangement of prism (m) and rhombohedral (r and z) faces. Ametrine crystals are typically
prismatic, and are terminated by r and z faces of
variable size. In natural ametrine crystal A, the
rhombohedral faces are equally developed, so that
the terminations resemble hexagonal pyramids. In
most ametrine crystals, however, the terminations
resemble those shown by B and C.
Unlike natural ametrine, synthetic ametrine

crystals are usually tabular, because a seed plate is
used to start the growth. As noted earlier, the crystal morphology varies depending on the characteristics of the seed plate and the growth conditions.
The crystals we examined for this study, all grown
on ZX-cut seeds, were bounded mainly by m and r
faces (figure 9); z faces were smaller, and the +x and
x faces were seldom present. A short steel wire
used to suspend the seed plate in the autoclave
was usually present at one end of the crystals.
Color and Color Distribution. The color distribution and internal features of the synthetic ametrine
Figure 9. This schematic diagram illustrates a synthetic ametrine crystal that has been grown on a
ZX-cut seed. The crystal is elongated along the
seed (indicated here by dashed lines), and is
bounded by r, z, m, and +x faces. The +x faces are
not visible in this diagram, because they are located on the back of the crystal, behind the m faces.
crystals are best seen when the they are sliced parallel to the x and m faces (figure 10). As noted earlier,
the c sectors are synthetic citrine (orangy yellow to
moderate orange), and the r and z sectors are synthetic amethyst (moderate to strong purple). The
purple coloration is somewhat more intense in the r
sectors than the z sectors. The samples exhibited
moderate dichroism (violet purple to violet) in the
amethyst-colored portion, and very weak to weak
dichroism (orangy yellow to orange) in the citrinecolored portion.
In the synthetic ametrine samples we examined,
the synthetic citrine portions were often more
intensely colored than those in natural ametrine of
similar thickness (see, e.g., figure 1). In general, we
observed that the synthetic amethyst portions in
Figure 10. Several internal features in synthetic

ametrine grown on tabular ZX-cut seeds are visible in this slab (6.0 cm long) and the corresponding schematic diagram. The slab was sliced parallel to the x faces; it shows amethyst coloration in
the r and z sectors, and citrine coloration in the c
sectors. Color zoning is present in both the synthetic citrine and amethyst regions, and faint
streaks consisting of liquid-gas inclusions are
elongated parallel to the optic axis in the citrine
region. The citrine color zoning is oriented parallel to the colorless seed (and perpendicular to the
optic axis), and the amethyst color zoning is oriented parallel to the rhombohedral crystal faces.
128
GEMS & GEMOLOGY
Summer 1999
our samples were also more intensely colored than

their natural amethyst counterparts.
The amethyst-citrine color boundary in synthetic ametrine (i.e., the boundary between the c sectors
and the adjacent r and/or z sectors) is oriented
roughly parallel to the rhombohedral faces, or at
about 51 to the optic axis for r sector amethyst and
23 to the optic axis for z sector amethyst. In contrast, the color boundary in natural ametrine (i.e.,
the boundary between the amethyst r sectors and
the citrine z sectors) is oriented roughly parallel to
the optic axis.
In addition, to obtain maximum cutting yield
from the slabs, some synthetic stones may show a
sharp bend or angle in the amethyst-citrine color
boundary (see, e.g., figure 5). This angle varied from
2248 in the synthetic slabs and faceted samples
examined at GIA. When natural ametrine is faceted
to show the maximum color contrast between
amethyst and citrine, the color boundary is usually
straight (see, e.g., figure 7).
Color Stability. The citrine color in the synthetic
ametrine remained stable during heating (at temperatures up to 700C). The amethyst color was stable
below 400C. In the temperature range of
400450C, the purple faded completely within
two hours of exposure, but it was fully restored by
subsequent gamma-ray irradiation. However, when
the synthetic ametrine samples were heated to
higher temperatures, especially above the phase
transformation of quartz at 573C, the amethyst
color centers in nearly all samples were destroyed
irreversibly (i.e., the purple could not be restored
with gamma-ray irradiation), and a milky turbidity
obscured the color and transparency of both the
amethyst and citrine zones. As the temperature was
increased from 550 to 700C, in particular, the
intensity of milkiness increased, cracks appeared,
and the orange component of the citrine color
decreased.
Microscopic Characteristics. Amethyst and citrine
color zoning, stream-like structures, polysynthetic
Brazil-law twinning, and two-phase (liquid-gas)
inclusions were the principal microscopic features
seen in the synthetic ametrine. The slab in figure 10
shows most of these features, and the corresponding
schematic diagram illustrates their interrelationships and their orientation relative to the optic axis.
The citrine portions (c sectors) of the faceted
synthetic ametrine commonly showed color zon-
Figure 11. Stream-like structures are visible in the

citrine portion of this 4.25 ct faceted synthetic
ametrine. Photo by Taijin Lu.
ing, oriented perpendicular to the optic axis. In the

amethyst portion (r and z sectors), color zoning was
generally observed, and was always oriented parallel
to either the r or the z crystal faces, or at about 50
to the optic axis. Rarely, stream-like structures were
observed in the synthetic citrine portions. With
oblique lighting, these appeared as wavy stripes of
color (figure 11) that were subparallel to the optic
axis. They result from surface irregularities caused
by growth hillocks on the basal pinacoid as the crystal grew.
No twinning was observed in the citrine portion
of the faceted synthetic ametrine. Brazil twins (see,
e.g., figure 12, left) were observed by the Russian
authors in the synthetic amethyst r sectors of only a
few of the 150 faceted samples. As observed with
crossed polarizers, Brazil twins in the synthetic
amethyst portions formed a distinctive polysynthetic twinning pattern: a group of straight, thin, parallel lamellae showing various interference colors. In
contrast, the amethyst sectors in all of the natural
ametrine studied showed Brazil twinning. Brazil
twin boundaries in natural amethyst show distinctive curved or mottled patterns that are sometimes
accompanied by linear areas of extinction called
Brewster fringes (Brewster, 1823; Crowningshield et
al., 1986; Lu and Sunagawa, 1990; Vasconcelos et
al., 1994; see figure 12, center and right). Brewster
fringes were not observed in the synthetic ametrine.
Two-phase (liquid-gas) inclusions were observed
only rarely in the synthetic citrine portions (figure
13), and very rarely in the synthetic amethyst portions. These inclusions were oval or needle shaped,
and were elongated parallel to the optic axis and the
stream-like structures; in a few instances, they
were visible to the unaided eye with strong light.
GEMS & GEMOLOGY
Summer 1999
129
Figure 12. Brazil twins were observed only rarely in the amethyst portion of the faceted synthetic ametrines.
They were composed of polysynthetic twin lamellae (thin, straight parallel lamellae showing various interference colors), as shown in the 12.42 ct faceted synthetic amethyst on the left (photomicrograph by John I.
Koivula, magnified 15). In contrast, the Brazil twins in the amethyst portion of the 11.47 ct natural ametrine
in the center (photomicrograph by Taijin Lu) form complex curved patterns; the citrine portion is demarcated
by the untwinned area on the far left side of the stone. Although the twinning in these two examples appears
somewhat similar, the striated appearance of the twinning in the synthetic sample is diagnostic. Some natural
samples show distinctive curved classic Brazil twins that may be accompanied by the darker Brewster
fringes, as shown by the amethyst on the far right (photomicrograph by Shane McClure).
Occasionally, small dark brown-to-black particles

were seen, sometimes with a needle-like morphology, along the interface between the seed plate and
the citrine portions of the synthetic ametrine.
Although they were not identified, the growth conditions suggest that they could be hydrous Fe-oxides
or Mn-oxides.
According to J. Koivula (pers. comm., 1999),
other primary fluid inclusions showing a flattened
morphology are occasionally observed along the
synthetic amethyst-citrine color boundaries. Such
inclusions have also been noted along the length of
the seed plate interface, and near the location of the
Figure 13. Needle-like two phase (liquid-gas) inclusions were seen only rarely in the citrine portion of
the synthetic ametrines. The inclusions are elongated parallel to the optic axis. Photo by Taijin Lu;
magnified 10.
130
wire used to suspend the crystals during growth.

Both the synthetic amethyst and citrine portions
also may have partially healed fractures composed
of numerous tiny voids containing both liquid and
gas phases. Unlike some of the primary fluid inclusions described above, these secondary fingerprintlike fluid inclusions are not oriented in any specific
direction.
Spectroscopic Features. The absorption spectra in
the UV-Vis-NIR regions were identical for natural
and synthetic ametrine. In the amethyst portions of
both types of ametrine, we saw characteristic
absorption bands at 270, 350, 540, and 930 nm. In
the citrine portions, an absorption edge rose abruptly in the 400500 nm range, and showed weak
bands superimposed at 540 and 930 nm.
Although the infrared spectra of the natural and
synthetic ametrines were similar, we noted some
distinctive differences. The synthetic citrine portions showed a broad, strong region of absorption
from about 3700 to 2500 cm1 (figure 14A), which is
due to water. A similar broad band occurs in the citrine portion of natural ametrine (Vasconcelos et al.,
1994), but it is much weaker. Weak OH peaks are
superimposed on the broad band in the 37003500
cm1 region of both materials, but the natural citrine has a weak peak at 3595 cm1, which is not seen
in the synthetic material (figure 14B). In addition,
one dark sample of synthetic ametrine showed weak
peaks at 3555 and 3528 cm1 in the citrine portion.
The infrared spectra of the amethyst zones in
both the synthetic and natural ametrine show more
GEMS & GEMOLOGY
Summer 1999
intense OH peaks, superimposed on a broad

absorption band at 3435 cm1 (figure 14C). The
main distinction between the synthetic and natural
amethyst portions is a weak band in most of the
synthetic amethyst at 3543 cm1 (figure 14D). Other
differences, such as a weak band at 3595 cm1 in the
natural material, are probably too subtle to use for
identification purposes.
The Raman spectra of synthetic ametrine are typical for quartz. When a comparison of the spectra of
oriented samples was made for the different color
zones in natural and synthetic ametrine, there was
only a very slight difference: A weak 1046 cm1 peak
observed in both natural and synthetic ametrine was
somewhat weaker in the natural ametrine.
Chemical Analyses. Qualitative EDXRF spectra of

the synthetic ametrine showed the presence of iron
(Fe, all 15 samples) potassium (K, 12 samples), manganese (Mn, 12 samples), and zinc (Zn, 11 samples)
as trace elements. In general, higher concentrations
of these elements were present in the citrine zones
than in the amethyst zones. Traces of nickel (Ni)
and chromium (Cr) were also observed in the citrine
and amethyst portions, respectively, of single synthetic samples. We were not surprised to find these
elements, since Fe is the chromophore, K and Mn
are used in the growth process, and Ni, Cr, and Zn
are components of the steel used in the autoclaves.
Semi-quantitative EDXRF analyses of two slabs
of synthetic ametrine revealed large variations in
Figure 14. Some differences were noted in the mid-infrared spectra of the natural and synthetic ametrine.
Compared to natural material, the synthetic citrine zones exhibit a more intense, broad absorption band in the
37002500 cm1 region (A); weak peaks at 3555 and 3528 cm1 were diagnostic features in one darker synthetic
citrine section (B). The spectra of the amethyst zones in both synthetic and natural ametrine are similar (C),
but a weak 3543 cm1 peak is usually seen in the synthetic material (D). The spectra are presented normalized
to 1.0 mm thickness.
GEMS & GEMOLOGY
Summer 1999
131
the concentrations of the above-mentioned trace

elements which could be correlated to the amount
of inclusions present. The two faceted samples analyzed, which were eye-clean, had lower impurity
contents and showed less-pronounced chemical
variations.
The quantitative (electron microprobe) analyses
of natural ametrine reported by Vasconcelos et al.
(1994) showed higher concentrations of iron (68125
ppm) in the citrine zones than in the amethyst
zones (1940 ppm); data for other trace elements
were not reported. In our samples of natural
ametrine, we detected traces of Fe, Zn, K, Mn, and
Al. The semi-quantitative data we obtained provided only relative amounts of the elements detected,
so we compared the concentrations of K, Mn, Fe,
and Zn in both natural and synthetic ametrine by
normalizing the raw data to 1000 silicon counts per
second. We found that the synthetic ametrine had
higher contents of Zn (2), Fe (2 to 3), Mn (4
to 6), and K (4) than the natural ametrine; aluminum was present at similar (low) levels in both
materials. It appears that a combination of higher K,
Mn, Fe, and Zn is a distinctive feature of synthetic
ametrine.
SEPARATING NATURAL FROM
SYNTHETIC AMETRINE
Table 1 compares some distinguishing properties of
natural ametrine and the Russian synthetic
ametrine grown on tabular ZX-cut seeds. Note that
R.I., birefringence, S.G., UV fluorescence, and UVVis-NIR absorption spectra were identical in the
natural and synthetic samples.
Separating the crystals is straightforward, since
their morphology is distinctive. Natural ametrine
usually forms as prismatic crystals with well-developed rhombohedral faces, whereas the commercially important synthetic ametrine crystals display a
tabular habit bounded mainly by m and r faces. The
synthetic crystals also commonly contain a short
steel wire at one of the end.
Faceted stones are more difficult to distinguish,
but a number of properties provide indications:
1. The citrine color in synthetic ametrine may span
a wide range, from pale yellow to moderate
orange. In the synthetic samples we studied, the
citrine portions were often more intensely colored than those in natural ametrine of similar
thickness (again, see figure 1). The amethyst por-
132
2.
3.
4.
5.
tions of our samples were also more intensely

colored than those in natural ametrine of similar
thickness.
Faceted natural ametrine commonly displays a
sharp, straight boundary between the citrine and
amethyst portions. Synthetic ametrine, especially when cut in fancy shapes, sometimes displays
a sharp bend in the color boundary (angles of
2248 were measured in the GIA samples).
While the amethyst-citrine color boundary in the
synthetic ametrine examined for this study is
oriented roughly parallel to the rhombohedral
faces (or at about 51 to the optic axis for r sector
amethyst and 23 to the optic axis for z sector
amethyst), in natural ametrine it is oriented
roughly parallel to the optic axis.
The crystallographic orientations of the color
zones and the rare stream-like structures in the
synthetic citrine portions are distinctive. The
color zones are oriented perpendicular to the
optic axis, and the stream-like structures are oriented parallel to the optic axis. We recommend
using cross-polarized light to locate the optic
axis in faceted samples, in order to provide a reference for checking the orientation of color
zones and stream-like structures. In natural
ametrine, the color zoning in both the amethyst
and citrine portions is oriented parallel to the
rhombohedral faces, and the bands usually are
spaced irregularly.
In the amethyst portions of natural ametrine,
Brazil-law twinning is almost always present,
and Brewster fringes are often observed.
However, Brazil twinning was seen only rarely in
the amethyst portions of synthetic ametrine, in
the form of subtle parallel twin lamellae. Due to
the variety in forms and patterns displayed by
Brazil twinning in both natural and synthetic
amethyst (see, e.g., Koivula and Fritsch, 1989),
this feature should not be used alone to identify a
stone as synthetic. However, if a sample does
show the ideal curved Brazil twins with
Brewster fringes (again, see figure 12, right), it can
be identified as natural.
Irregular planes of two-phase (liquid-gas) inclusions are commonly observed in both color portions of natural ametrine (Vasconcelos et al.,
1994). In synthetic ametrine, elongate two-phase
(liquid-gas) inclusions were seen only rarely.
However, these inclusions would be diagnostic
GEMS & GEMOLOGY
Summer 1999
only for those who are very experienced with

quartz inclusions.
6. Natural and synthetic ametrine can usually be
separated by their infrared spectra. The synthetic
citrine portion has more intense absorption in
the 3700 2500 cm1 range of the infrared spectrum, and weak peaks at 3555 and 3528 cm1 in
dark material may be diagnostic. The synthetic
amethyst showed a weak band at 3543 cm1 that
has not been observed in the natural amethyst
portions.
7. As determined by EDXRF chemical analysis, synthetic ametrine will typically show higher contents of K, Mn, Fe, and Zn than occur in natural
ametrine.
Even with advanced testing, the identification of
natural and synthetic ametrine should rely on mul-
tiple criteria, since some of the distinguishing characteristics may be absent, and some features overlap. Positive identification may not be possible in
all cases.
CONCLUSIONS
By 1994, the technology was available to produce
synthetic ametrine commercially. To date, several
hundred kilograms of this material have been grown
and are being distributed in the jewelry marketplace.
Although there are many similarities between
natural ametrine and the hydrothermal synthetic
ametrine grown in Russia that was examined for
this study, most faceted samples can be separated
by a combination of standard gemological methods,
especially by observation of internal features such
as color zoning and twinning. Advanced techniques
TABLE 1. Distinguishing properties of natural ametrine and the Russian synthetic ametrine obtained for this study.a
Property
Crystal morphology
Color
Color boundary
Growth and
color zoning
Twinning
Inclusions
IR absorption
spectrum
Trace elements
detected by EDXRF
Synthetic
Natural
Tabular habit, normally elongated in the trigonal

prism x direction. Prism m and rhombohedral r
faces are the best developed. Seed plate is oriented
perpendicular to the c-axis.
Amethyst portion: moderate to strong purple in the r
and z sectors. Citrine portion: orangy yellow to
orange in the c sectors. Typically more intensely
colored than natural ametrine in both portions.
Usually a sharp straight boundary, oriented at 51 or
23 to the optic axis for the r and z sectors, respectively; may show a sharp bend.
Amethyst portion: color zoning oriented parallel to
the r or z crystal faces, or at about 50 to the optic
axis. Citrine portion: color zoning oriented parallel
parallel to the seed plate, or perpendicular to the optic axis; rarely, stream-like structures observed subparallel to the optic axis.
Polysynthetic Brazil twins are observed very rarely in
the amethyst portion. No twins in the citrine portion.
Rare two-phase (liquid-gas) inclusions in the citrine
portion, elongated parallel to the optic axis.
Amethyst portion: usually a weak peak at 3543 cm1.
Citrine portion: a more intense, broad absorption in
the 3700 2500 cm1 region; weak peaks at 3555
and 3528 cm1 present in one darker sample.
K, Mn, Fe, and Zn significantly higher than in natural
samples; also traces of Al detected.
Prismatic habit consisting of rhombohedral faces r

and z, and prism faces m.
Amethyst portion: pale purple to intense violet-purple

in r sectors. Citrine portion: light yellow to orangeyellow in z sectors.
Usually a sharp straight boundary, oriented parallel
to the optic axis.
Color zoning in both the amethyst and citrine portions
oriented parallel to the rhombohedral faces.
Brazil twins (sometimes with Brewster fringes) in the

amethyst portion, typically with distinctive curved or
mottled patterns. No twins in the citrine portion.
Irregular planes of two-phase inclusions often observed along cracks in both portions.
Amethyst portion: no peak at 3543 cm1.
Citrine portion: weaker broad absorption in the
37002500 cm1 region; no 3528 cm1 peak.
Fe and Zn; K, Mn, and Al detected in some samples
at low concentrations.
aBoth
materials share the following properties: R.I.nw=1.5401.541, ne=1.550; birefringence0.0090.010; S.G.2.65; UV fluorescenceInert to
both short- and long-wave UV; UV-Vis-NIR absorption spectrumamethyst portion: bands at 270, 350, 540,and 930 nm; citrine portion: very strong
absorption at 400500 nm and weak lines at 540 and 930 nm.
GEMS & GEMOLOGY
Summer 1999
133
(infrared spectroscopy and EDXRF analysis) can provide further useful data for the separation of these
two materials. Note, however, that the diagnostic
features presented here are specific to ametrine and
are not necessarily applicable to either amethyst or
citrine in general.
It must also be noted that the technology of synthetic crystal growth is constantly evolving.
Prismatic synthetic ametrine crystals are now being
commercially produced, and these are very similar
in crystal morphology and color distribution to natural ametrine. It is, therefore, especially important
REFERENCES
Balakirev V.G., Keivlenko E.Y., Nikolskaya L.V., Samoilovich
M.I., Khadzhi V.E., Tsinober L.I. (1979) Mineralogy and
Crystal Physics of Gem Varieties of Silica. Nedra Publishers,
Moscow. [in Russian].
Balitsky V.S. (1980) Synthetic amethyst: Its history, methods of
growing, morphology and peculiar features. Zeitschrift der
516.
Balitsky V.S., Khetchikov L.N., Dorogovin B.A. (1970) Some specific geochemical conditions of amethyst formation. In V.P.
Butusov, Ed., Trudy VNIISIMS, Synthesis and Experimental
Investigation, Vol. 13, Nedra Publishers, Moscow, pp. 7582
[in Russian].
Balitsky V.S., Lisitsina E.E. (1981) Synthetic Analogues and
Imitations of Natural Gemstones. Nedra Publishers, Moscow
[in Russian].
Balitsky V.S., Balitskaya O.V. (1986) The amethyst-citrine
dichromatism in quartz and its origin. Physics and Chemistry
of Minerals, Vol. 13, pp. 415421.
Brewster D. (1823) On circular polarization, as exhibited in the
optical structure of the amethyst, with remarks on the distribution of the coloring matter in that mineral. Transactions of
the Royal Society of Edinburgh, Vol. 9, pp. 139152.
Crowningshield R., Hurlbut C., Fryer C.W. (1986) A simple procedure to separate natural from synthetic amethyst on the
basis of twinning. Gems & Gemology, Vol. 22, No. 3, pp.
130139.
Garrels R.M., Christ C.L. (1965) Solutions, Minerals and
Equilibria. Freeman, Cooper & Co., San Francisco, CA, 450 pp.
Hehar W.C. (1980) The discovery of golden amethyst. Lapidary
Journal, Vol. 34, pp. 582583.
Hutton D.R. (1964) Paramagnetic resonance of Fe3+ in amethyst
and citrine quartz. Physics Letters, Vol. 4, pp. 310311.
Jones B. (1993) Is ametrine for real? Rock & Gem, Vol. 23, No. 9,
pp. 4652.
134
that the gemologist continue to use a multitude of

tests to make a conclusive identification.
Acknowledgments: The authors thank Dr. Ilene
Reinitz, John Koivula, Shane McClure, and Dr. Mary
Johnson of the GIA Gem Trade Laboratory for their
useful comments on the manuscript, and Julia
Goreva of Caltech for her assistance with the translation of technical articles from Russian to English.
Funding for research at Caltech was provided by the
White Rose Foundation. Funding for research in
Russia was provided by the Russian Basic
Research Foundation (grant # 97-05-64805).
Koivula J.I. (1980) Citrine-amethyst quartz. Gems & Gemology,

Vol. 16, No. 9, pp. 290293.
Koivula J.I., Fritsch E. (1989) The growth of Brazil-twinned synthetic quartz and the potential for synthetic amethyst
twinned on the Brazil law. Gems & Gemology, Vol. 25, No. 3,
pp. 159164.
Lehmann G. (1975) On the color centers of iron in amethyst and
synthetic quartz: A discussion. American Mineralogist, Vol.
60, No. 34, pp. 335337.
Lu T., Sunagawa I. (1990) Structure of Brazil twin boundaries in
amethyst showing Brewster fringes. Physics and Chemistry of
Minerals, Vol. 17, pp. 207211.
Melnik Y.P., Drozdovskaya A.A., Vorobeva K.A. (1974) Physicochemical analysis of the dissolution, migration and deposition
of iron in modern volcanic regions. In S. Naboko, Ed., Modern
Volcanic Systems, Nauka publishers, Novosibirsk, pp. 119126 [in Russian].
Nassau K. (1981) Artificially induced color in amethyst-citrine
quartz. Gems & Gemology, Vol. 17, No. 1, pp. 3739.
Rivero R. (1999) Synthetic ametrine and amethyst infiltrates
Bolivian market. ICA Gazette, May/June/July issue, p. 15.
Rossman G.R. (1994) The colored varieties of silica. In P.J.
Haney, Ed., Silica, Reviews in Mineralogy, Vol. 29,
Mineralogical Society of America, Washington, DC, pp.
433468.
Tsinober L.I., Chentsova L.G. (1959) Synthetic quartz with
amethyst color. Kristallografiya, Vol. 4, pp. 633635 [in
Russian].
Vargas G., Vargas M. (1980) A new quartz gem material.
Lapidary Journal, Vol. 34, No. 7, pp. 15041506.
Vasconcelos P., Wenk H.R., Rossman G.R. (1994) The Anah
ametrine mine, Bolivia. Gems & Gemology, Vol. 30, No. 1,
pp. 423.
Zaitov M.M., Zaripov M.M., Samoilovich M.I. (1974) The Fe3+
EPR spectrum in irradiated quartz. Kristallografiya, Vol. 19,
pp. 10901093 [in Russian].
GEMS & GEMOLOGY
Summer 1999
ongratulations to the many readers from all over the world who demonstrated their dedication to education by participating in the 1999 Gems & Gemology Challenge, which tested their knowledge of articles
published in the 1998 volume year. Participants who earned a score of 75% or better will receive a GIA
Continuing Education certificate. Those who scored a perfect 100% are also listed below.
UNITED STATES Alabama Gadsden: Jerome Denson Thomas Arizona Fountain Hills: Hank T. wodynski. Oro
U
Valley: Geraldine Alex Towns. Sun city: Anita R. Wilde. Tucson: Luella Dykhuis, David D. Arens California
Beverly Hills: Sina A. Mozafarian. Carlsbad: Lori Burdo, Carl Chilstrom. Diane Flora, Brian I. Genstel, William J.
Herberts, Mark Johnson, Doug Kennedy, Jan Luree Lombardi, Roxana J. Lucas, Wendi M. Mayaerson, Catherine
McIntyre, Jana Miyahaara-Smith, Lynn L. Myers, Duncan L. Pay, Diane H. Saito, Laura Small, Abba R. Steinfeld, Ric
Taylor, James Viall, Melissa Watson-Lafond, Philip G. Yourk. Chino Hills: Virgilio M. Garcia, Jr. Huntington Beach:
Barr L. Doty. Laguna Niguel: Shawn Anne Shannon. Los Angeles: Veronica Clark-Hudson, Margaret Shikibu. Mill
Valley: Susan F. Bickford. Oceanside: Sarah A. Horst, Laverne M. Larson. Orcutt: Sabra Rounds. Pacifica: Diana L.
Gamez. San Diego: Tracy Nuzzo, San Jose: Wendy L. Bilodeau Santa Cruz: Tony Averill, Walnut Creek: Ying Ying
Chow. West Hills: Bradley A. Partington Colorado Colorado Springs: Molly K. Knox. Denver: Kyle Hain, Alan J.
Winterscheidt Connecticut Simsbury: Jeffrey A. Adams. Westport: William A. Jeffery Florida Cape Coral: Harold
E. Holzer. Clearwater: Timothy D. Schuler. DeLand: Sue Angevine Guess. Miami Beach: Pinchas Schechter.
Plantation: Garrett Walker. Satellite Beach: Consuelo Schnaderbeck. Sunny Isles Beach: Fabio S. Pinto Idaho
Victor: Deborah L. Mock Illinois Genoa: Dale L. Johnson. Oaklawn: Eileen S. Barone. St. Charles: Lori M. Mesa
Indiana Indianapolis: Mary Campbell Wright Iowa Iowa City: Gary R. Dutton. West Des Moines: Franklin
Herman Louisiana Baton Rouge: Harold Dupuy. New Orleans: Jon Randall Duryee, Duncan Parham Maryland
Potomac: Alfred L. Hirschman Massachustts Adams: Marc Krutiak. Braintree: Alan R. Howarth. Boxford: Sharon
Kendall Heller. Brookline: Martin D. Haske. Lynnfield: John A. Caruso. Uxbridge: Bernard M. Stachura Michigan
Plymouth: Sarah K. Bukowski Nevada Las Vegas: Deborah A. Helbling. Reno: Terence E. Terras New Jersey Long
Valley: Lorraine Lopezzo New Mexico Corrales: Susan Gaspar Wilson New York New York: Joseph V. Mochulski
North Carolina Hendersonville: Robert C. Fisher. Kernerville: Jean A. Marr. Manteo: Eileen Alexanian Ohio
Columbus: John A. Schwab. Dayton: Michael Williams. Highland Heights: Michael B. Saxon. North Royalton:
Christine M. Blankenship. Steubenville: Vincent A. Restifo Oregon Beaverton: Robert H. Burns. Portland: Geri
Jeter. Salem: Donald Lee Toney Pennsylvania Hamburg: Janet L. Steinmetz. Yardley: Peter R. Stadelmeier Rhode
Island Providence: Mark W. Burns South Carolina Sumter: James S. Markides Tennessee Germantown: Charles L.
Rose Texas Corpus Christi: Warren A. Rees, Jr. Flower Mound: Elizabeth M. Roach. Houston: Karen L. Jensen,
Christine Sehnaderbeck. Tomball: Carroll Joseph Kiefer Virginia Hampton: Edward A. Goodman Washington
Lakebay: Karen Lynn Geiger. Lacey: John H. Vivian Wisconsin Beaver Dam: Thomas G. Wendt West Viginia:
AUSTRALIA Coogee, Western Australia: Helen Judith Haddy. Slacks Creek,
Charleston: Randall R. Sims A
Queensland: Ken Hunter. Success, Western Australia: Tereena Tobias. Sydney, New South Wales: Barbara
BELGIUM Diegem: Guy Lalous. Diksmuide: Honor Loeters.
AUSTRIA Vienna: Eberhard Layr B
Wodecki A
BRAZIL So Paulo: Alejandro B. Ferreyra, Maria Amelia
Hemiksem: Danial De Maeght. Ruiselede: Lucette Nols B
CANADA Bobcaygeon, Ortario: David R. Lindsay, Calgary, Alberta: Diane Koke.
Franco, Ana Flavia Pantalena C
Cowansville, Quebec: Alain Deschamps, Montreal, Quebec: Christiane Beauregard, Prince George British
Columbia: Gerald E. Cooke. St. Catharines, Ontario: Alice J. Christianson. Vancouver, British Columbia: Michael
ENGLAND Tenterden, Kent: Linda
CYPRUS Nicosia: George Stephanides E
J.P. Cavanagh, John Mattinson C
FINLAND Kajaani: Petri Tuovinen IINONESIA Ciputat, Jakarta: Warli Latumena IITALY
Anne Bateley F
Agrate: Zambone Ginafranco, Caltanissetta, Sicily: Francesco Natale, Malnate, Varese: Gabriele Trali. Lucca: Roberto
NETHERFilippi. Porto Azzurro: Diego Giuseppe Trainini. Rome: Andrea Damiani. Valenza: Rossella Conti N
PHILIPPINES Mandaluyong
LANDS Rotterdam: Joop G. Heetman, E. van Velzen. Wassenaar: Jane M. Orsak P
POLAND Lublin: Marek A. Prus P
PORTUGAL Figueira,
City, Metro Manila: Mark Alexander B. Velayo P
RUSSIA Chernogolovka, Noginsk, Moscow Province: Vadim I.
Vilado Bispo, Algarve: Johanne Cardin Jack R
SPAIN Madrid: Maria Isabel Cerijo Hierro, Shahrazad
SCOTLAND Edinburgh: James W. M. Heatlie S
Prygov S
SWEDEN Jarfalla: Thomas Larsson S
SWITZERLAND
Krewi de Urquijo, Valencia: Monika Bergel-Becker S
Basel: Christina Dahlstroem. Geneva: Jean-Marie Duroc-Danner. Lucerne: Marisa Zachovay. Zollikon: Adrian
VIRGIN ISLANDS Christiansted: Karin Ann Exposito
Meister V
Answers (See page 61 of the Spring 1999 issue for the questions):
(1) b, (2) a, (3) d, (4) a, (5) c, (6) b, (7) c, (8) c, (9) d, (10) b, (11) c, (12) d, (13) a,
(14) c, (15) d, (16) a, (17) d, (18) b, (19) b, (20) a, (21) b, (22) c, (23) c, (24) b, (25) b
GEMS & GEMOLOGY
Summer 1999
135
BASTNSITE,
A Rare Faceted Example
We seldom see rare-earth carbonates
fashioned as gems, for several reasons:
These minerals typically do not
occur in large transparent crystals,
they are usually brown, and they have
low hardness. Recently, staff members in the West Coast laboratory
examined an unusual 19.64 ct orangy
brown oval modified brilliant (figure
1), which the client represented as the
rare-earth carbonate bastnsite. The
following gemological properties were
determined by senior staff gemologist
Dino DeGhionno and staff gemologist
Phil Owens: diaphaneitytransparent; refractive indices1.722 and
greater than 1.81 (over the limits of
our standard refractometer); pleochroismlight and dark orangy brown;
optic characteruniaxial positive;
specific gravity (measured hydrostatically)5.16; inert to both long- and
short-wave UV radiation. These properties were consistent with those
Figure 2. A large negative crystal
was present in the bastnsite
shown in figure 1. Magnified 15.
136
Lab Notes
Figure 1. This 19.64 ct bastnsite

(15.79 11.38 10.66 mm) is
the first faceted example of this
mineral that has been seen in
the Gem Trade Laboratory.
reported for bastnsite by W. L.
Roberts et al. (Encyclopedia of
Minerals, 2nd ed., Van Nostrand
Reinhold, New York, 1990, pp. 7374).
The handheld spectroscope
revealed a rare earth spectrum: general absorption below 490 nm, and
Figure 3. The bastnsite also
contained numerous needle-like
inclusions. Note the optical doubling of most of the inclusions.
Magnified 40 .
bands at 500525, 530, 560585, 610,

660, and 690 nm. With magnification,
we could see negative crystals (e.g.,
figure 2), needles (figure 3), and a
partially healed fracture containing
fluid inclusions (figure 4). As bastnsite is quite soft (Roberts et al. list a
hardness of 4 to 41/2), we would not
expect it to be in common use as a
gem material.
According to Roberts et al., the formula for bastnsite is (Ce,La)(CO3)F;
energy-dispersive X-ray fluorescence
(EDXRF) analysis revealed rare-earth
elements (lanthanum, cerium, and
neodymium) and yttrium, but we
were not able to measure these quantitatively with our system. Raman
spectroscopy revealed peaks at 164,
258, 304, 351, 394, 733, 1095 (strong),
1437, and 1738 cm1 , with a good
match to the bastnsite spectrum in
our reference library. An X-ray powder diffraction pattern matched the
spectrum in the International Centre
Figure 4. This partially healed

fracture, or feather, in the bastnsite showed interesting geometric patterns. Magnified 25.
GEMS & GEMOLOGY
Summer 1999
Figure 5. This dense tangle of natural

etch channels was typical of the
numerous etch features observed in a
25.02 ct brownish yellow round brilliant diamond. Magnified 17.
Figure 6. In this view of the

25.02 ct diamond, several etch
channels have converged to
form a rough-surfaced, undulating cavern. Magnified 20.
Figure 7. Note the width variations within this single etch channel, which abruptly narrows and
turns back on itself, looping in
tight curves. Magnified 33.
for Diffraction Data (formerly JCPDS)

reference library. The infrared spectrum included an absorption edge at
about 3000 cm1 and peaks at 5828,
5070, 4660, 4464, 4292, about 4000
(strong), 3758, 3590, 3421, 3229, 3179,
and 3154 cm1.
As bastnsite is an uncommon
gem material, it is difficult to say
whether it might have been identified
without the advanced testing methods. In this case, as all the evidence
pointed to bastnsite for this stone,
we were ultimately confident of the
identification .
When it was time to write up the
conclusion, we had some slight problems concerning what to put on the
report. In Fleischers Glossary of
Mineral Species 1999 (J. A. Mandarino, Mineralogical Record Inc.,
Tucson, AZ), this material is identified as bastnsite-(Ce). (There are
also lanthanum and yttrium end
members.) However, because we were
not able to obtain a quantitative
chemical analysis, we could not confirm that a predominance of the rareearth sites were occupied by cerium
(as the -Ce notation implies), so we
left off this suffix. Also, because our
report printing process does not
accommodate umlauts, we substitut-
ed the archaic ae for . Therefore,

our report describes this unusual gem
as bastnaesite.
MLJ
mixture of blue and green fluorescence to strong visible light (transmission luminescence). These properties indicate two optical centers in
this diamondthe N3 and H3
which are composed of three nitrogen
atoms, and two nitrogen atoms with a
neutral vacancy, respectively (A. T.
Collins, Colour centres in diamond,
Journal of Gemmology, Vol. 18, No. 1,
1982, pp. 3775).
The diamond also showed medium-to-strong phantom graining, with
pale brown color along two sets of
graining planes; such colored graining
is believed to form from dislocations
between octahedral planes in the diamond (again, see the A. T. Collins reference given above). However, all colored diamonds contain some crystal
defects, and many show defect combinations similar to those observed in
this diamond, but relatively few show
any etch tubes. The presence of such
extensive etching suggests that the
original crystal (1) was full of dislocations and other weaknesses, and (2)
had an unusually prolonged exposure
to the caustic geologic fluids that can
chemically oxidize diamond.
The etch features in this diamond
formed in a number of interesting
shapes (see, e.g., figures 5 and 6), as
well as in a range of sizes (see, e.g., figure 7). They serve as a reminder of the
many differences between laser drill
holes and natural etch tubes or channels (see, e.g., M. L. Johnson et al.,
When a drill hole isnt, Rapaport
Editors note: The initials at the end of each item

identify the editor(s) or contributing editor(s) who
provided that item.
Gems & Gemology, Vol. 35, No. 2, pp. 136-141
Lab Notes
DIAMOND, with Abundant

Etch Channels
This past spring, staff members at the
East Coast laboratory had the opportunity to examine a 25.02 ct brownish yellow round brilliant diamond
with a multitude of etch tubes.
Figure 5 shows an example of these
densely intertwined channels in one
area of the girdle. We have previously
reported on etch tubes and channels
(see, e.g., Gem Trade Lab Notes, Fall
1992, pp. 193194; Winter 1992, pp.
262263; Summer 1994, p. 115), but
this stone showed much more internal etching than any of us had ever
seen in a gem diamond.
Etchingwhether it forms tubes,
channels, grooves, pits, or other structuresoccurs along crystal defects in
the diamond (Yu. L. Orlov, The Mineralogy of the Diamond, John Wiley &
Sons, 1973, pp. 82, 9899). Like other
naturally colored diamonds of similar
hue, this round brilliant owes its
colorand its distinctive gemological
propertiesto a variety of crystal
defects. We saw medium-to-strong
lines in the desk-model spectroscope
at 415 nm, 494 nm, and 503 nm. The
diamond fluoresced with medium
strength in a mixture of blue and yellow colors to both long- and shortwave UV radiation, and it showed a
GEMS & GEMOLOGY
Summer 1999
137
Figure 8. When this 0.29 ct black

round brilliant was submitted
for identification of its color origin, it was found to be a synthetic diamond.
Diamond Report, Vol. 21, No. 45,
December 1998, p. 1). Although the
individual tubes in this diamond were
rather small in diameter, they showed
angular outlines, rather than the
round or oval outlines formed by laser
drilling. Drill holes are usually
straight, but etch tubes rarely are.
Most tellingly, drill holes tend to lead
to some inclusion, such as a void from
which a dark crystal was removed, or
a feather, which may be fracture filled.
In contrast, etch tubes follow atomsize crystal defects, and thus appear to
meander randomly through the diamond.
IR
Figure 9. Magnification and fiberoptic lighting revealed both a

dark blue color and a rod-shaped
inclusion (typical of a metallic
flux) in the synthetic diamond.
Magnified 40 .
SYNTHETIC DIAMOND
Black
As the use of black diamonds in jewelry has increased, we have seen more of
them at both laboratories for color-origin determination. In most black diamonds, the overall color appearance
is caused by a large number of dark
inclusions. We routinely examine
black diamonds near the girdle with
pinpoint fiber-optic illumination to
look for areas of partial transparency
(which are typical of natural color) or
for a dark green color (which is an indicator of treatment). Such greenish
Figure 10. These two synthetic diamonds (0.59 and 0.63 ct) illustrate the range of color, from orangy yellow to greenish yellow, that
was displayed by a group of 18 synthetic diamonds submitted to
the laboratory at one time.
138
Lab Notes
black color results from a high dose of

radiation, usually in a nuclear reactor,
and diamonds treated by this method
may be radioactive (see, e.g., Summer
1992 Lab Notes, pp. 124125).
We were, therefore, somewhat
surprised when the strong light
revealed a blue color in a 0.29 ct
black round brilliant (figure 8) that
was examined last winter in the East
Coast lab. The strong electrical conductivity of the sample (100:135 to
120:135) indicated that the blue color
was caused by boron impurities (see J.
M. King et al., Characterizing natural-color type IIb blue diamonds,
Gems & Gemology, Vol. 34, No. 4,
1998, pp. 246268). The piece was
inert to both long- and short-wave UV
radiation. It also was heavily included, with only a few relatively transparent areas. One of those areas contained a dark, rod-shaped inclusion
(figure 9), which is a common morphology for flux inclusions in synthetic diamond. This prompted us to
test the sample for magnetic attraction by suspending it from a thread
and bringing a powerful magnet near
it. The samples strong attraction to
the magnet provided another indication that it was synthetic.
To confirm this identification and
further characterize this unusual
specimen, we examined the sample
with a De Beers DiamondView (see,
e.g., C. M. Welbourn et al., De Beers
natural versus synthetic diamond verification instruments, Gems &
Gemology, Vol. 32, No. 3, 1996, pp.
156169). We hoped that the greater
intensity of this source would stimulate visible fluorescence. Although
the reaction was weak and severely
interrupted by the many inclusions
(which did not fluoresce), we observed
a fluorescence pattern that was hexagonal (i.e., reflecting the faces of an
octahedron modified by a cube; cubic
growth is characteristic of synthetic
diamonds), rather than the concentric
rectangular pattern typical of natural
diamonds. The sample luminesced
more strongly to X-rays, revealing a
blue, angular pattern typical of synthetic growth.
IR
GEMS & GEMOLOGY
Summer 1999
Group of 18 Synthetic Diamonds

Submitted at One Time
A regular client submitted a group of
18 round-brilliant and square-modified-brilliant diamonds to the East
Coast lab for identification and colororigin determination. They ranged
from 0.10 to 0.71 ct and exhibited saturated colors that varied from orangy
yellow to greenish yellow, as represented by the two samples in figure
10. Standard gemological testing
quickly revealed, however, that all 18
samples were synthetic diamonds. All
were purchased in Russia, where they
were represented as natural, so the
client was surprised to hear the
results.
Six of the samples contained inclusions that were larger than pinpoints,
in the rod- and droplet-shaped morphologies that are typically seen in
synthetic diamond. These six also
showed obvious attraction to a strong
magnet, while the other samples had a
very weak or no attraction. We performed EDXRF analyses on two samples, one with large inclusions and
one without; both revealed Fe and Ni,
the principal elements in the flux that
is used to facilitate synthetic diamond
growth. All 18 samples exhibited
some unevenness of body color, but
many showed weak color zoning in
which the pattern was difficult to discern. However, their reaction to UV
radiation was vividly diagnostic: All
showed a yellowish green cross pattern. This pattern was most clearly
seen through the crown, but it was
strong enough in some samples to be
evident through the pavilion as well.
Although some natural-color and
especially treated-color diamonds fluoresce yellowish green, the cross pattern is diagnostic of synthetic origin
(see, e.g., J. E. Shigley et al., A chart
for the separation of natural and synthetic diamonds, Gems & Gemology,
Vol. 31, No. 4, 1995, pp. 256264).
The fluorescence ranged from
moderate to strong for long-wave UV
radiation and from weak to strong for
short-wave UV. This general trend of
a stronger reaction to long-wave than
short-wave UV is unlike that seen in
Lab Notes
Figure 11. The strong pleochroism of this cobalt-doped synthetic

forsterite makes it a convincing tanzanite imitation. The largest cushion mixed cut weighs 6.15 ct and measures 12.29 9.85 7.06 mm.
Courtesy of Tom Chatham.
most of the earlier yellow synthetics
GIA described (J. E. Shigley et al.,
The gemological properties of
Russian gem-quality synthetic yellow
diamonds, Gems & Gemology, Vol.
24, No. 4, 1993, pp. 228248). It suggests that some of the nitrogen impurities were aggregated at high temperature after growth, forming a type Ia
component (A. T. Collins and M.
Stanley, Absorption and luminescence studies of synthetic diamond in
which the nitrogen has been aggregated, Journal of Physics D: Applied
Physics, Vol. 18, 1985, pp. 25372545.).
Unfortunately, these synthetic diamonds were not available long
enough for us to perform the infrared
spectroscopy necessary to confirm the
diamond type.
IR
SYNTHETIC FORSTERITE,
A New Tanzanite Imitation
At the April American Gem Society
Conclave in New Orleans, Tom
Chatham of Chatham Created
Gemstones, San Francisco, California, showed contributing editor MLJ
four fashioned samples and two rough
pieces (figure 11) of a material grown
in Russia that had been represented to
him as synthetic tanzanite. As this

was clearly a pulled product, and
tanzanite is a variety of the hydrous
mineral zoisite, we were skeptical of
this claim and agreed to study the
material further.
The four fashioned samples
included a 6.15 ct cushion mixed cut,
a 3.27 ct oval modified brilliant, and
two (2.81 and 2.10 ct) modified triangular mixed cuts. The two pieces of
rough weighed 5.75 ct and 3.50 ct;
portions of these samples showed a
smooth curved surface, resembling
part of a boule. The fashioned samples had the following properties:
colorviolet; optic characterbiaxial
positive; pleochroismstrong, in blue
and purplish pink (figure 12);
(Chelsea) color-filter reactionnone;
refractive indices1.635 to 1.637,
1.650 (beta), and 1.670 to 1.671; birefringence0.034 to 0.035; specific
gravity (measured hydrostatically)
3.23 to 3.24. The samples fluoresced a
moderately chalky, but very weak,
orangy yellow to long-wave UV radiation, and a weak greenish yellow to
short-wave UV; at both wavelengths,
the color was evenly distributed.
When viewed with a handheld spectroscope, they showed a 460470 nm
GEMS & GEMOLOGY
Summer 1999
139
also seen treated-color blue beryl represented as tanzanite (Winter 1997

Lab Notes, p. 293). The pronounced
pleochroism of synthetic forsterite
makes it visually far more convincing
as a tanzanite simulant than any we
have seen before, although it can be
easily distinguished from tanzanite by
its refractive indices (1.6351.670 versus 1.691.70 for tanzanite).
MLJ and SFM
Figure 12. The two pleochroic colors displayed by the synthetic

forsterite, as seen here with polarized light, were blue and purplish
pink. Magnified 10.
PEARL,
band, a 490 nm band, diffuse lines at
510 and 520 nm, and a 570580 nm
band. With magnification, we saw
indistinct white inclusions, pinpoint
inclusions (possibly gas bubbles), tiny
needles, and stringers (figure 13).
On the basis of the refractive
indices and strong pleochroism, in
particular, we suspected that the
material was synthetic forsterite,
Mg 2SiO4. An EDXRF analysis performed by GTL research associate
Sam Muhlmeister on the 6.15 ct
cushion mixed cut confirmed this
identification. EDXRF revealed major
amounts of magnesium and silicon,
and traces of cobalt, vanadium, and
iron, with the cobalt much more
prominent than the other trace elements. A Raman spectrum showed
strong orientation effects (i.e., spectra
Figure 13. The 6.15 ct synthetic
forsterite contained indistinct
white inclusions that resemble
dust particles, as well as pinpoints, tiny needles, and
stringers. Magnified 11.
140
Lab Notes
taken from different angular positions

had large variations in peak height),
with the strongest peaks at 857 and
825 cm1 and others at 965, 919, 607,
591, 547, 435, 307, and 227 ( 2) cm1.
An infrared spectrum showed a cutoff
at 2000 cm 1, and small peaks at
2497, 2623, and 2985 cm1; there was
no evidence of any significant water
present in the sample.
Synthetic forsterite has been
described before (see, e.g., K. Nassau,
Synthetic forsterite and synthetic
peridot, Gems & Gemology, Vol. 30,
No. 2, 1994, pp. 102108); and in an
October 1996 personal communication, Dr. Henry Hnni of the SSEF
Swiss Gemmological Institute called
our attention to Czochralski-pulled
synthetic forsterite that had been
grown by Solix in Minsk, Russia. In
both of these instances, chromium
was the only dopant. Dr. Hnni mentioned colorless, blue-green, and
olive green samples, and Dr.
Nassau reported that there was also
pink and purple material, in which
the chromium could be in various
oxidation states. The material we
examined, however, contained no
chromium, but rather was colored by
cobalt (with a possible contribution
from vanadium).
A comprehensive article by L.
Kiefert and S. Schmidt (Some
tanzanite imitations, Gems &
270276) listed manufactured glass,
YAG, and synthetic corundum as tanzanite simulants. Since then, we have
Update on Non-nacreous Pearls

In the Winter 1998 Lab Notes section
(p. 288), we reported on several banded, light pink and white, non-nacreous pearls with a sheen effect.
Subsequently, the West Coast lab
received samples of the shells (figure
14) that had hosted some of these
concretions. Although it was previously thought that the host might be
an oyster, we could readily see that
the appearance was more like a scallop, of the type commonly known in
some areas of the North American
Pacific Coast as lions paw shells.
The characteristic physiognomy, with
seven coarsely nodular ribs and two
uneven ears, allowed us to identify
the genus as Chlamys nodosus. KH
Figure 14. These shells, measuring 6 cm across, are from

Chlamys nodosus, a scallop
that has produced nonnacreous pearls with color
banding and a sheen effect
(see inset, 14 12 mm).
GEMS & GEMOLOGY
Summer 1999
SYNTHETIC SAPPHIRE
Orangy Pink Padparadscha
A narrow range of pinkish orange to
orangy pink sapphires are commonly
referred to by the trade term padparadscha. Although such sapphires
have long been highly prized by collectors, the GIA Gem Trade Laboratory does not issue a report using
this trade description because of the
subjectivity of the term (see R.
Crowningshield, Padparadscha:
Whats in a name? Gems &
30 36).
A client recently submitted the
approximately 9 ct slightly orangy
pink cushion mixed cut shown in figure 15 for identification and color
description. The client was disappointed that we could not verify if the
stone qualified for the term padparadscha, but agreed to an identification
report that would give an appropriate
color description.
The R.I.s of 1.761.77, uniaxial
optic character, and absorption spectrum (three lines in the red due to Cr,
and a doublet at 460470 nm) verified
that the stone was indeed corundum.
In most cases, we can easily determine whether a fancy-colored sapphire is natural or synthetic by the
internal features (e.g., the presence of
straight or curved color banding).
However, this attractive stone was
Figure 15. This approximately 9
ct orangy pink cushion mixed
cut, which might be called padparadscha by the trade, was
identified as a synthetic sapphire.
Lab Notes
virtually flawless except for a tiny

feather near the girdle. Immersion in
methylene iodide did not reveal any
other internal features; specifically,
we could not resolve any color zoning, either straight or curved. The
sample fluoresced an equally strong
red to both long- and short-wave UV
radiation, which also provided no
indication as to its origin.
However, EDXRF chemical analysis of the stone showed a significant
amount of Ni, in addition to Fe and
Cr. Nickel is not found in natural sapphire, but it has been reported in
some hydrothermally grown synthetic corundum from Russia (see, e.g., V.
Thomas et al., Tairus hydrothermal
synthetic sapphires doped with nickel
and chromium, Gems & Gemology,
Vol. 33, No. 3, 1997, pp. 188202).
Therefore, we concluded that this
sample was a synthetic sapphire,
although whether it represents
Russian hydrothermal growth remains uncertain.
KH
Early Blue Synthetic Sapphire?
The white metal ring shown in figure
16 is characteristic of jewelry made in
the early 1900s. As we have mentioned previously in this section,
although it is possible to replicate virtually any style of jewelry, there are
often certain features that suggest
that a piece is indeed an original
and was made at the time the style
and design indicate. The ring shown
here was one such example: Its overall style and the wear, specifically the
burnishing of the metal around the
stones, suggest that the stones had
been set in the ring for some time. (A
burnish implies that the metal has
been rubbed over time; that is, it has a
polished glossiness that is also somewhat dark from the incorporation of
foreign substances. The effect, like
that of seasoning a cast-iron skillet,
cannot readily be imitated.) Even the
synthetic sapphire in the center is
consistent with material available in
the early part of this century.
Synthetic blue sapphire was first
Figure 16. This ring appears to

have been manufactured in the
early part of this century. The
Verneuil synthetic sapphire in
the center shows eye-visible concentric circles of color banding.
produced by Verneuil in about 1910
(see, e.g., K. Nassau, Gems Made by
Man, Chilton Book Co., Radnor, PA,
1980). Consequently, the large cabochon in this ring may well date from
that time, as its gemological properties were typical for a Verneuil synthetic.
The noteworthy feature of this
cabochon was the eye-visible curved
growth banding that was present as
concentric circles. Normally, synthetic corundum boules grown by the
flame-fusion method are split lengthwise in preparation for fashioning.
Once the boule is split, the curved
growth bandsor striaeare visible
only as a group of nested curves, and
not as full circles. One possible explanation for the circular growth banding
in this cabochon is that it was cut
from a small button boule that was
not split after growth (see, e.g., D.
Elwell, Man-Made Gemstones, John
Wiley & Sons, New York, 1979).
TM and GRC
PHOTO CREDITS
Vincent Cracco took photos 5, 6, and 7. Maha
DeMaggio photographed figures 1, 8, 10, 11, 14,
and 15. John King provided figure 16. John
Koivula was the photographer for figures 2, 3, and
4. Shane McClure photographed figures 9, 12,
and 13.
GEMS & GEMOLOGY
Summer 1999
141
New editorial policy for Gem News. Individual bylines

will be published for specific entries prepared primarily
by specific contributing editors or other individuals.*
Contributing editors will be identified by their initials;
all others will be identified by their full names and affiliations. When referencing such entries, please cite them
by the name of the person(s) listed in the byline, for
example: Schmetzer K. (1999) Gem news: Twelve-rayed
star sapphire from Madagascar. Gems & Gemology, Vol.
35, No. 2, pp. 146147. All entries without bylines are
provided by the section editors, and should be cited as
being written by Johnson, Koivula, McClure, and DeGhionno.
DIAMONDS
Diamond presentations at the PDAC conference. Since
1992, new developments in diamond exploration have
been presented at the annual conferences of the
Prospectors and Developers Association of Canada.
Consulting geologist A. J. A. (Bram) Janse, of Perth,
Australia, attended the 1999 meeting and sent in the following report. This year the meeting was held in
Toronto, on March 1417. Attendance was high (6,200),
despite two bad years for the metals and mining industry
as a whole. The mood was cautiously optimistic, and diamonds were one of the most popular subjects. The first
diamond mine in Canada, Ekati, began operations in
October 1998 (see Winter 1998 Gem News, pp. 290292);
and mining at a second deposit, Diavik, is expected to
start in 2002.
This years conference included one technical session
on diamonds, and informal diamond talks in the
Investors Exchange forum and an open forum. Mike
Jones of Aber Resources, Toronto, presented a design for
*Anyone interested in contributing a Gem News entry should

follow the procedures listed in the Guidelines for Authors,
published on pp. 7778 of the Spring 1999 issue of Gems &
Gemology. A copy can also be obtained by contacting Gems &
Gemology senior editor Brendan Laurs at blaurs@gia.edu
(e-mail), 760-603-4503 (phone), or 760-603-4595 (fax).
142
Gem News
barrier dams for the Diavik project at Lac de Gras. Roy

Spencer, vice president of Denver-based Archangel
Diamond Corporation (ADC), outlined the difficulties of
operating in Russia, drawing from ADCs experiences
with the Grib kimberlite pipe, discovered in February
1996, where development is now at a standstill. John
Auston, president of Vancouver-based Ashton Mining of
Canada, gave an update on their recent investigations in
the Buffalo Hills kimberlite field in Alberta, Canada (see
Summer 1998 Gem News, pp. 134135), where exploration continues although a commercial diamond deposit
has not yet been proved.
Ian McGeorge, of MPH Consulting in Toronto, presented a summary of activities in Botswana: Debswana is
developing the Martins Drift kimberlite dike prospect,
on the border with South Africas Northern Province (formerly Transvaal), and has applied for a mining license
over the Gope 25 prospect in central Botswana.
According to Mr. McGeorge, recent research has indicated that the Orapa kimberlite pipe is not located on the
Archean basement of the Kalahari Craton, but actually
occurs within the surrounding mobile belt of paleoProterozoic age. If substantiated, this would upset previously held concepts on the distribution patterns of economic kimberlites (see, e.g., M. B. Kirkley et al., Gems &
Gemology, Spring 1991, pp. 2 25; and A. J. A. Janse,
Gems & Gemology, Winter 1995, pp. 228255, and
Spring 1996, pp. 230).
Chris Jennings, president of SouthernEra, Toronto,
gave one of the most popular presentations, an update on
activities at the Klipspringer project (figure 1), which is
located near Potgietersrus in South Africas Northern
Province. SouthernEra has discovered a 20 mile (about 33
km) long dike swarm that contains several near-parallel
dikes (colloquially called fissures); the main ones are
the Leopard and Sugarbird fissures, each of which contains one or more blows (small pipes). The company
experienced a setback in April 1998, when the heirs to
the Marsfontein farm, which is centrally located in the
dike swarm and contains the M-1 pipe, sold their titles to
De Beers. This dispute has now been resolved (see
GEMS & GEMOLOGY
Summer 1999
Happy families, Mining Journal, London, June 19,

1998, pp. 469 470); SouthernEra retained 40% of
Marsfontein, while it continues to hold 100% of all other
relevant farms. Jennings described the recent development of the Sugarbird blow, for which the payback period for capital expenditure was not the two years usually
considered adequate, but only one week.
In the Investors Exchange forum, John Kaiser of the
Kaiser Bottom-Fishing Report, spoke on understanding
the diamond exploration cycle. He explained the lesser
significance of microdiamonds recovered from small
mineral grain-size fractions, as compared to macrodiamonds from bulk and mini-bulk samples, and the need
to restrain enthusiasm until results of all samples are
released. Ongoing activities worldwide were highlighted
in short talks and displays about diamond exploration
Figure 1. This aerial view (looking east) of part of the
Klipsinger project shows the Leopard fissure. The
dike is about 1 m wide, and is visible over a distance of approximately 1 km in this photo; drilling
has shown it to be 3 km long. The first underground
adit at this fissure is visible in the foreground. Photo
courtesy of SouthernEra.
Figure 2. Octahedral diamond crystals are featured in

this jewelry, which was designed by Andrew Jordan
and manufactured by Steve Fong (both of Vancouver,
British Columbia, Canada). The 18k gold ring contains a 2.36 ct diamond crystal and two 0.06 ct
faceted yellow diamonds. Each diamond crystal in
the earrings weighs about 1.25 ct. Courtesy of Betty
Sue King; photo by Maha DeMaggio.
and mining in Botswana, Brazil, Canada, Finland,

Guinea, Lesotho, Namibia (offshore), and South Africa.
A. J. A. (Bram) Janse
Archon Exploration
Carine, Western Australia
Diamond octahedra in jewelry. In their earliest uses in
jewelry, diamonds were left in their distinctive natural
shapes, such as octahedra. At the Tucson show this year,
Betty Sue King of Kings Ransom, Sausalito, California,
exhibited contemporary jewelry featuring octahedral diamond crystals from Africa (figure 2). These designs illustrated an interesting use of diamonds that are already
attractive as natural crystals. Ms. King later reported that
she is expanding this line of jewelry to include pieces featuring flat triangular twinned diamond crystals (macles).
Jo Ellen Cole, GIA
First diamonds from the Merlin project, Australia.
Ashton Mining has generated several press releases (see,
e.g., those dated October 30 and December 17, 1998, and
February 18, 1999) concerning its wholly owned Merlin
project in the Northern Territory. Merlin yielded the first
production of diamonds in mid-February 1999, including
more than 720 carats from one days production at its
Excaliber and Sacramore pipes. This sample, which was
taken from low-grade ore at the surface of the two pipes,
contained a large selection of diamonds 1 ct and larger.
A second shipment of Sacramore concentrate yielded a
14.76 ct white octahedron (figure 3). Although ore
grades have not been specified, Ashton forecasts produc-
Gem News
GEMS & GEMOLOGY
Summer 1999
143
tion of 200,000 carats from 500,000 tonnes of ore in the

first year (grade 0.4 ct/ton). A commercial-scale trial mining and processing operation is now underway, and will
involve taking ore from nine of the 12 pipes on the
Merlin property. (For information on the geology of this
deposit, see D. C. Lee et al., The Merlin kimberlites,
Northern Territory, Australia, Proceedings of the Sixth
International Kimberlite Conference, Russian Geology
and Geophysics, Vol. 38, No. 1, 1997, pp. 8296.)
The Merlin diamonds will be sold through Argyles
European sales office in Antwerp, as the production is
similar to high-end Argyle material and is anticipated to
appeal to Argyles established range of customers.
Ashton is also mining diamonds at Cuango, in northeastern Angola; more than 61,000 carats from the first three
months production were sold in December to the De
Beers Central Selling Organisation (CSO) in Luanda. By
agreement, all of the Cuango production will be sold
through the CSO.
A review of GE-processed diamonds. On March 1,
1999, Lazare Kaplan International Inc. (LKI) sent a press
release to the financial press, stating that its new subsidiary, Pegasus Overseas Limited (POL), would be the
exclusive agent for selling natural diamonds that had
undergone a new process developed by General Electric
(GE). According to LKI, the process was permanent and
irreversible, and was designed to improve the color, brilliance, and brightness of qualifying diamonds, which
they said were a small fraction of the overall diamond
market. Both round and fancy-shaped processed diamonds were to be marketed, each with an accompanying
grading report from a major laboratory. The release
stressed the all-natural content of the processed diamonds, and stated that the process did not involve irradiation, laser drilling, surface coating, or fracture filling.
In a March 19 press release, GIAs president Bill
Boyajian called for information from the parties involved
Figure 3. This 14.76 ct diamond octahedron was
recovered recently from the Sacramore pipe at the
Merlin project in Northern Territory, Australia.
Photo courtesy of Ashton Mining.
144
Gem News
in the process. The release further noted that GIA

researchers had investigated a small sample of diamonds
they believed had undergone the GE process, and that
they had noted unusual gemological features in some of
the diamonds. No specific information as to the nature of
these features was provided, and GIA has yet to release
any conclusive evidence that will identify that a diamond has undergone the GE process.
After extensive discussions with GE and LKI, GIA
announced the following in a press release dated April
28: LKI will laser-inscribe all GE-processed diamonds on
the girdle with the inscription GE POL (figure 4). GIA
will make the following comment on grading reports for
all GE-processed diamonds sent to GIA from POL: GE
POL is present on the girdle. Pegasus Overseas Limited
(POL) states that this diamond has been processed to
improve its appearance by General Electric (GE). GIA
will move forward with a research project to understand
the nature of the process and explore any identifying
characteristics of these processed diamonds. It will use
its HORIZON laboratory operations and information
management system to track and build data on every
GE-processed diamond submitted to GIA for grading.
Ultimately, GIA intends to publish its findings in Gems
& Gemology. LKI began selling the GE-processed diamondsall with GE POL inscribed on the girdle
through POL in Antwerp in late May.
During a special June 23 session at the International
Gemological Symposium in San Diego, California, Bill
Boyajian, GIA Gem Trade Laboratory chief executive
officer Tom Yonelunas, GIA Gem Trade Laboratory vice
president of identification services Tom Moses, and GIA
director of research Jim Shigley provided a Trade
Update on the GE-processed Pegasus diamonds.
Although they began by explaining that there was still
no black box that could identify these diamonds, they
did reveal some information based on study of several
hundred such diamonds that POL had submitted to the
laboratory for grading reports. A summary of their findings includes: weight0.30 to 7 ct, with most in the 13
ct range); shapesmost cut in fancy shapes, but some
rounds; clarityIF to I 2, with most IF to SI1; colorD to
light yellow, with most D to H (the majority of all colors
are brownish or grayish). The overwhelming majority
of diamonds were type IIs, but some type Is were seen.
Mr. Yonelunas, who presented these data, cautioned that
the lab had yet to see the full range of goods. The best
hope to understand this process requires examination of
diamonds before and after processing, but these have not
yet been available for study.
Dr. Shigley reported that GIA has been conducting its
own experiments on treating diamonds using high pressure and high temperature (HPHT) to help develop identification criteria. He noted that one could add or subtract color to diamonds with this technique. Mr. Moses
pointed out that HPHT treatment of diamonds has been
known in the scientific community since the early
GEMS & GEMOLOGY
Summer 1999
1970s. Although GE has not provided details of the process being used, many people have expressed their opinion that HPHT is involved (see, e.g., K. Schmetzer,
Behandlung natrlicher Diamanten zur Reduzierung
der Gelb- oder Braunsttigung (Treatment of natural
diamonds in order to reduce the yellow or brown color),
Goldschmiede Zeitung, Vol. 97, No. 5, May 1999, pp.
4748 [in German]), so it is clearly important to know
what the starting features of such treated diamonds are.
Mr. Yonelunas stated the belief that at least a portion
of the GE-processed diamonds will be recognizable
gemologically. In fact, a parcel of GE-processed diamonds
from which the inscriptions had been either partially or
completely removed had already been resubmitted to the
Gem Trade Laboratory, and the diamonds had been
noticed using archived data from the laboratorys HORIZON system. A GIA press release dated July 6 indicated
how such diamonds would be handled: The company
submitting the diamond will be asked to immediately
authorize re-inscription of that diamond. Should the
company choose otherwise, GIA would be obligated to
report this to the appropriate authorities, including the
Jewelers Vigilance Committee (JVC).
At the World Federation of Diamond Bourses meeting in Moscow in early July, a resolution was proposed
regarding the GE-processed diamonds. As quoted on the
Internet in Rapaport News (and, at our press time, scheduled to appear in the July 30 issue of the Rapaport
Diamond Report), the resolution states the following:
1. If a diamond has been treated or processed in order to
alter or enhance its color, other than by generally
accepted procedures of cutting and polishing, this fact
must be disclosed in writing when such a diamond is
offered for sale or submitted for certification.
2. The removal of a lasered inscription which identifies
a diamond as having been treated or processed as
above, shall be considered a deceptive process.
3. Any violation of Articles 1 or 2 above shall be regarded as fraudulent and shall be referred to the applicable
Bourse for disciplinary action as the Bourse sees fit.
4. If such a treated or processed diamond is sold without
disclosure in breach of the above rule, even in good
faith, the buyer shall be entitled to cancel the sale,
return the diamond and obtain a refund of the purchase price.
COLORED STONES AND ORGANIC MATERIALS
International Colored Gemstone Association Congress.
More than 20 countries were represented by 173 participants at ICAs eighth biennial Congress, which was held
in Abano Terme, Italy, May 1619. Dona Dirlam,
Director of GIAs Library and Information Center, provided the following report.
The keynote speaker was Rashmikant Durlabhji of
India, a founding organizer of ICA, who spoke on the role
of colored gemstones in the 21st century. He reminded
the audience that ICA is the only organization devoted to
Gem News
Figure 4. GE POL has been inscribed on the girdle

of this diamond, which Pegasus Overseas Ltd. stated
had been processed by GE. Photomicrograph by
Shane McClure; magnified 40 .
colored stones on an international basis, and he pointed
out the need to continue to bridge the gaps from miners
to cutters to wholesale dealers. He suggested several
ideas for promoting jewelry, including the development
of a line of jewels for people who wear uniforms. He also
encouraged ICA members to be ready to embrace the
new paradigm in trading, the Internet.
Professor Jurgen Kleiber-Wurm of the University of
Innsbruck, Austria, spoke on the global marketing of colored stones. What is fundamentally new is the speed at
which information travels. Also, new markets will be
found by focusing on the concept of selling desire.
An update on the history, mining techniques, and
future prospects of the Paraba tourmaline mines in
Brazil was provided by Marcelo Bernardes. He noted that
currently there are three groups working the area, but
there has been no recent production. Mr. Bernardes estimated that only 7.65 kg of gem-quality Paraba tourmaline has reached the trade in 17 years of production, making it one of the rarest of the commercial gemstones.
Mehul Durlabhji, the ICA ambassador to India, gave
an update on the commercial ruby localities in Indias
Orissa and Madhya Pradesh states, on rhodolite garnet
mining in Orissa, and on tanzanite mining in Tanzania.
In particular, he noted that decreased production combined with steady demand have resulted in sharp price
increases for tanzanite. He concluded by suggesting that
future mining for colored stones will require more
sophisticated techniques and equipment.
Particularly memorable this year was the unveiling of
a silver, gold, and platinum sculpture by Ninni Verga
called The Gate of the 2000 Gems (figure 5). Decorated
with 2,000 diamonds and colored gems, the sculpture
was presented to the Israeli Ambassador to Italy, Yehuda
Millo, by outgoing ICA president Paolo Valentini, as a
symbol of peace and hope for the next millennium. After
touring countries throughout the world to promote ICA,
colored gemstones, and peace, it will be presented to the
City of Jerusalem.
A proposal has been submitted to hold the next ICA
GEMS & GEMOLOGY
Summer 1999
145
a recent article (B. S. Wilson and W. Wight, Gem andradite garnet from Black Lake, Quebec, Canadian
Gemmologist, March 1999, pp. 18 19) and include: R.I.
of 1.880; S.G. of 3.833.94; no horsetail inclusions; and
low chromium content (less than 0.01 wt.% Cr2O3),
based on microprobe analysis of a sample for which the
color was not specified.
Figure 5. The Gate of the 2000 Gems, an artistic replica of the Lions Gate of the old city of Jerusalem, symbolizes the crossing from the second to the third millennium. Designed by Ninni Verga, it is made of gold, silver, and platinum and set with 2,000 colored gems
weighing 253.13 ct. The sculpture, which stands 36 cm
tall, has a 49 37 cm base. Photo courtesy of ICA.
Congress in 2001 in Sydney, Australia. The final location
will be decided later this year.
Dona Mary Dirlam, GIA
Andradite (including demantoid) from Canada. At the
1999 Tucson show, Brad Wilson, of the Kingston,
Ontario, office of Coast-to-Coast Rare Stones, showed
G&G senior editor Brendan Laurs a 5.8-mm-diameter
(0.78 ct) round brilliant andradite garnet that was reportedly from Black Lake, Quebec. Because the color was in
the yellow-to-green range, and possibly achieved the
green hue necessary to be called demantoid, Mr. Laurs
requested the loan of two additional pieces so that their
colors could be observed under more controlled conditions. These samples (figure 6) were examined with daylight-equivalent fluorescent lighting in a MacBeth Judge
II viewing environment. Using Munsell color chips as
comparators (see, e.g., J. M. King et al., Color grading of
colored diamonds in the GIA Gem Trade Laboratory,
Gems & Gemology, Vol. 30, No. 4, 1994, pp. 220242),
we found that one was yellowish green (green enough to
be demantoid), but the other was greenish yellow and so
not demantoid.
Limited quantities of this andradite were found in
1995 and April 1998 in an asbestos mine at Black Lake,
Quebec; this locality was described in the 1940s (J. D. H.
Donnay and C. Faessler, Trisoctahedral garnet from the
Black Lake region, Quebec, University of Toronto
Studies, Geological Series, No. 46, 1941, pp. 1924).
Although most of the stones cut thus far are smaller than
0.5 ct, larger ones have been faceted (e.g., 1.08, 1.38, and
2.74 ct). The properties of this material were reported in
146
Gem News
Trapiche cats-eye emeralds. Most gemologists and jewelers are familiar with trapiche emeralds and their sixspoke, wagon-wheel appearance. What is less well
known, however, is that the pie-shaped sections in some
of these emeralds may lend themselves to the fashioning
of chatoyant gems.
Chatoyancy in emeralds is typically the result of light
reflection from fine parallel growth tubes elongated in
the c-axis direction. Less commonly, the chatoyancy
results from the accidental parallel to subparallel bunching of fibrous mineral inclusions such as amphiboles.
Instead of being aligned parallel to the c-axis, as are
growth tubes, the fine bunching structure that produces chatoyancy in trapiche emeralds is oriented at 90
to the c-axis. This results in cats-eye cabochons with the
optic axis direction oriented through the dome (i.e., perpendicular to the girdle), rather than through the long
side, as would be the case in a cats-eye created by reflection from growth tubes.
Although fine-scale structure has been recognized as
a characteristic of trapiche emeralds for many years (see,
e.g., K. Nassau and K. A. Jackson, Trapiche emeralds
from Chivor and Muzo, Colombia, American
Mineralogist, Vol. 55, 1970, pp. 416427; Spring 1981
Gem Trade Lab Notes, pp. 4344), the Gem News editors could find no mention of chatoyancy in such gems.
At the 1999 Tucson show, however, Michael Gray of
Coast-to-Coast Rare Stones, Missoula, Montana, showed
Figure 6. The 0.39 ct pear shaped andradite garnet
from Black Lake, Quebec, Canada, is yellowish
green, and therefore the variety demantoid; the 0.34
ct oval from the same locality is greenish yellow,
and should be referred to simply as andradite.
Courtesy of Brad Wilson; photo by Maha DeMaggio.
GEMS & GEMOLOGY
Summer 1999
Figure 7. Bright, chatoyant bands are visible near

the apexes of several wedge-shaped sections of this
18.37 ct (16.2815.88 9.82 mm) trapiche emerald
cabochon. Courtesy of Coast-to-Coast Rare Stones;
photo by Maha DeMaggio.
us an 18.37 ct trapiche emerald cabochon (figure 7) that
had a chatoyant band in each of the six wedge-shaped
sections, near the apex of the cabochon, when viewed
with a penlight. Magnification revealed a fibrous-looking
structure that was oriented perpendicular to the c-axis.
From this observation, it seemed logical that individual
emerald cats-eyes could be cut from the trapiche crystal
if the fibrous structure was fine enough, and if the rough
was properly oriented during cutting.
As a coincidence to this earlier observation, Roxanne
Kremer of Collectors in Rosemead, California, subsequently loaned GIA two cats-eye emerald cabochons
(2.05 and 2.15 ct) that reportedly had been cut from
trapiche rough from the Muzo mine (figure 8).
Examination with magnification showed that the structure of these two cabochons was identical to that of the
much larger, complete trapiche emerald cabochon we
had examined previously. The internal structure became
even more apparent when the cabochons were examined
with polarized light (figure 9). It was also apparent that
the fibers did not extend radially from the core (i.e., like
the spokes on a bicycle wheel); instead, they maintained
a parallel orientation within each trapiche section.
Opaque black fringe-like extensions protruding from the
spokes also pointed in the same direction as the fibrous
structure in the emerald sections.
Figure 8. Both of these free-form double cabochons,

which were cut from a single trapiche emerald crystal, show chatoyancy. These 2.05 and 2.15 ct samples measure 10.79 7.78 3.32 and 10.98 8.01
3.33 mm, respectively. Courtesy of Collectors;
lagoon at Penrhyn was noted for the small natural pearls
found there, which continue to be harvested today.
Penrhyn is located 650 miles (about 1,045 km) north of
Rarotonga; its lagoon, which is about 50 miles (80 km) in
circumference, is the home of the species Pinctada maculata, also known as the Polynesian Pipi Shell. This
mollusk, which averages 4 cm in diameter, occasionally
produces natural pearls (figure 10). P. maculata mollusks
cluster underwater, from near the surface to about 2 m
deep, on massive coral heads that occur randomly
throughout the lagoon. The natural pearls are harvested
year-round, mostly by local women who use small outboard-powered boats to reach the pearl beds. Some divers
employ a unique technique to determine selection of the
shell: They actually look into the naturally gaping shells
to see if there is a pearl inside. This approach requires
considerable finesse, as the mollusk will close if it
Figure 9. When the two cabochons in figure 8 were
viewed in the optic axis direction with cross-polarized light, the parallel fibrous structure that causes
the chatoyancy was readily apparent. Notice how the
fringed edges of the black spokes taper in the same
direction as the fibrous structure. Photomicrograph
by John I. Koivula.
Natural pearls from the northern Cook Islands. In

October 1998, GIA Extension Education manager Eddie
Buscher visited the Cook Islands and met with Ben
Bergman of Bergman & Sons (on Rarotonga, the capital
island), who provided the following information about
natural pearls from Penrhyn Island, in the northern Cook
Islands.
According to Mr. Bergman, commercial farming of
South Sea black (cultured) pearls was established in the
Cook Islands 27 years ago. Before then, the enormous
Gem News
GEMS & GEMOLOGY
Summer 1999
147
Figure 10. These natural pearls (4.55.0 mm) and polished shells (2.53.8 cm) are typical of Pinctada maculata, a small mollusk native to Penrhyn Lagoon in the
Cook Islands. Pearls courtesy of Bergman & Sons,
and shells courtesy of the Beachcomber, Rarotonga;
detects the presence of the diver. In the last decade,

Penrhyn has been affected by some unusually low tides
(attributed to El Nio), which have left portions of the pearl
beds exposed on the surface. The resulting mortality of the
mollusks has restricted the availability of these pearls.
According to Mr. Bergman, the pearls produced by P.
maculata average 45 mm in diameter. Although the
color can vary considerably, the pearls are predominantly
a soft golden hue. Rarely, pearls are found that could
be considered deep gold, green, copper, black, gray,
pink to orange, or white. Common shapes include
baroque, semi-baroque, button, and occasionally round.
Blister pearls are referred to locally as puku. The population of P. maculata is insufficient to justify commercial exploitation, and local law forbids foreigners from
harvesting them. As a consequence, no official production statistics have been kept. These natural pearls are
usually sold to local buyers for use in jewelry manufacturing.
Mr. Bergman noted that Penrhyn and the neighboring
island Manihiki also have a large population of wild P.
margaritifera mollusks (about 5 million currently),
which have been harvested for the last 100 years. Local
legend alludes to fabulous collections of natural pearls
from the P. margaritifera. Although Mr. Bergman was
not able to confirm this, he has seen some notable natural pearls from this mollusk, including a pair of silver
white egg-shaped pearls about 10 mm in diameter (both
said to have come from the same shell) and a single dropshaped silver white pearl that measured 10.2 mm.
Eddie Buscher, GIA
are surrounded by milky white to intense blue growth

sectors, as described by D. Schwarz and J. Kanis
(Madagaskar: Korund aus Nord und Sd, extraLapis,
No. 15, 1998, pp. 6063). In general, the six-rayed star
sapphires are cut to remove this colorless core, or to place
it at the rim of the cabochon. The latter was the case for a
71.18 ct 12-rayed star sapphire (figure 11) that was recently examined by contributing editor Karl Schmetzer. The
stone was reportedly recovered from Antsira n
ana
Province in northern Madagascar, a large region from
which blue and yellow sapphiresincluding six-rayed
star sapphireshave been recovered since about 1997
(see, e.g., Summer 1998 Gem News, pp. 140141).
The flat back of this cabochon revealed that the original crystal had indeed grown around a colorless transparent core (some of which remained on one edge of the
cabochon). This core was surrounded by three milky
white growth sectors, which were found to be hexagonal
dipyramidsprobably {14 14 28 3}. Between two of

these milky white growth sectors were narrower, trans
parent, intense blue rhombohedral r {1011} growth zones.
These forms are consistent with those of sapphire seen
from northern Madagascar in recent years; as such, they
provided additional evidence that the stone was natural.
The 12-rayed star actually consisted of two six-rayed
stars with the same axis on which the rays were offset at
Figure 11. This 12-rayed star sapphire is from
Antsiran
ana Province, in northern Madagascar;
the 71.18 ct cabochon measures 26.2 18.9 mm.
Twelve-rayed star sapphire from Madagascar. Star sapphires from northern Madagascar are frequently fashioned from crystals with colorless transparent cores that
148
Gem News
GEMS & GEMOLOGY
Summer 1999
30. The arms of the most prominent star were oriented

perpendicular to the hexagonal dipyramids. (This is the
usual orientation of six-rayed star cabochons from northern Madagascar.) The rays of the other star were oriented
parallel to the hexagonal dipyramids, between the arms
of the first star. The cause of the double starthat is, the
identity of the elongated mineral inclusions in two orientations within this stonehas not yet been determined.
KS
Blue, pink, and purple sapphires from Ilakaka,
Madagascar. . . . Tom Cushman of Allerton Cushman
and Co., Sun Valley, Idaho, recently showed us several
sapphires from new production at Ilakaka, in southern
Madagascar. According to Mr. Cushman, 30 kg of gem
rough were legally exported in NovemberDecember
1998; however, the miners probably recovered a great deal
more than that, and the rate of production is increasing.
He considers several kilograms per week a reasonable estimate, with the majority of the sapphires in pink-to-purple
colors. The stones (all alluvial) are being mined principally
in two areas (one near the town of Ilakaka) that are about
100 km apart; recently, however, stones also have been
recovered from the region between these two areas. Pieces
as large as several grams have been found.
The sapphires generally are heat treated after export,
and untreated fashioned stones are comparatively rare. Mr.
Figure 12. These sapphires came from Ilakaka in

southern Madagascar. The long blue oval (bottom
right) showed evidence of heat treatment. Courtesy of
Tom Cushman; photo by Maha DeMaggio.
Cushman reports that more than 90% of the Madagascar
sapphires he is marketing have been heat treated.
We examined five faceted stones (figure 12), ranging
from blue to purple to pink. All but the 1.54 ct stone, a
modified octagonal brilliant, were oval mixed cuts. The
gemological properties are given in table 1. Four of the
stones contained tiny birefringent transparent crystals,
TABLE 1. Properties of five sapphires from Ilakaka, Madagascar.

Properties
Orangy pink
Purplish pink
Pinkish purple
Blue
Bluea
Weight (ct)
Pleochroism
1.99
Purple pink
Orange pink
Reddish orange
1.54
Purplish pink
Orangy pink
Reddish orange
2.12
Purple
Orange
Red
1.72
Violet-blue
Blue
Weak pink
2.03
Violetish blue
Greenish blue
None
4.00
1.760 1.768
3.99
1.7601.768
4.00
3.99
3.98
1.7611.769
1.7601.768
1.7611.769
Moderate to strong
orange
Very weak orange
Moderate red-orange Moderate red-orange
Very weak red
Inert
Very weak orange
Very weak orange
Inert
Visible
luminescence
Absorption
spectrum
None
None
None
Weak red
Chalky, uneven,
mod. yellowish green
None
Chromium lines (ruby

spectrum)
Weak ruby spectrum
Ruby spectrum +
450 Fe line
Inclusions
Stringers and clusters

of tiny transparent
near-colorless birefingent crystals, tension
cracks around crystals
Tiny individual transparent near-colorless

birefringent crystals,
tension cracks, partially healed fractures
Tiny transparent
near-colorless birefringent crystals,
stress haloes, lamellar
twinning, residues
in surface cavities
Very weak ruby

spectrum + 450
Fe band
Tiny transparent
near-colorless birefringent crystals,
irregular silk, lamellar twinning, fingerprints
Color filter
reaction
Specific gravity
Refractive indices
UV fluorescence
Long-wave
Short-wave
Weak 450 Fe band
Melted crystals,
discoid fractures,
cloud, fingerprints
Internal features showed evidence of heat treatment.
Gem News
GEMS & GEMOLOGY
Summer 1999
149
Gem News section (pp. 133134). These similarities suggest that the gems may have been derived from a geologically similar source that existed while Madagascar was
still attached to the African continent (see, e.g., C. B.
Dissanayake and R. Chandrajith, Sri LankaMadagascar
Gondwana linkage: Evidence for a Pan-African mineral
belt, Journal of Geology, Vol. 107, 1999, pp. 223235,
which includes a map showing a possible reconstruction
of East Africa, Madagascar, and Sri Lanka).
HAH
Figure 13. This lot of mixed gem rough was recovered

from alluvial deposits in the Ilakaka region of
Madagascar. The largest pieces measure about 1 cm.
Photo courtesy of Henry Hnni.
possibly zircon. Although we were told that two of these
samples had been heat treated, only the 2.03 ct blue oval
mixed cut showed evidence of this, in the form of melted
crystals surrounded by discoid fractures.
Tanzanite fluorescence: a seldom-noted property. Bill

Vance, a Graduate Gemologist and gem dealer from
Waldport, Oregon, recently showed the Gem News editors a transparent, 3.46 ct pear-shaped mixed-cut tanzanite that fluoresced to long-wave UV radiation. Since UV
fluorescence has not been reported previously for tanzanite (see, e.g., pp. 387 388 of R. Websters Gems, 5th ed.,
Butterworth-Heinemann, 1994: There is no noticeable
luminescence under ultra-violet light), we examined the
stone in more detail.
The tanzanite was slightly purplish blue face-up, and
Figure 14. Miners emerge from a tunnel at the
Merelani mining area in northeast Tanzania. Photo
by Angelique Crown.
. . . and other gems from near Ilakaka. Blue, pink, and

purple sapphires are not the only gem materials being
found in the Ilakaka area. Contributing editor Henry
Hnni examined a 1 kg parcel of gem rough obtained by
Werner Spaltenstein of Chanthaburi, Thailand, from the
Ilakaka district. The 150-km-long alluvial deposit begins
about 50 km beyond the city of Tulear (Toliara) on the
southwest coast of Madagascar, along the road to
Antananarivo; the village of Ilakaka is currently at the
center of the excavations.
The colorful mix of slightly to distinctly rounded
pebbles (figure 13) was analyzed with Raman spectroscopy at the SSEF Swiss Gemmological Institute in
January 1999. The following gem minerals were identified: blue, pink-to-purple, and brown sapphire; blue, violet, and purple spinel; yellow-to-brown grossular garnets,
including orange hessonite; pyralspite garnets, including
almandine, rhodolite, violet-blue and malaia-type garnets; colorless-to-yellow, greenish yellow, or brownish
yellow chrysoberyl, with some cats-eye material and
some alexandrite; colorless, light yellow, and light blue
topaz; brown and red tourmaline; andalusite; yellow,
brown, and green zircon; grayish blue platy kyanite; and
citrine and amethyst, as well as colorless quartz. The parcel was also found to contain one light green pebble of
manufactured glass.
It is interesting to note that these gem pebbles were
very similar in species and shape to the rough in the parcel from Puchapucha, in the Tunduru area of southern
Tanzania, that Dr. Hnni described in the Summer 1995
150
Gem News
GEMS & GEMOLOGY
Summer 1999
showed rather typical pleochroism in shades of reddish

purple, brownish yellow, and grayish to greenish blue.
The refractive indices were 1.691, 1.693, and 1.701, with
a birefringence of 0.010 and a biaxial positive optic character; a partial interference figure could also be resolved
through the pavilion of the stone. The specific gravity,
determined hydrostatically, averaged 3.36 over three
measurements. No absorption spectrum was seen
through either a diffraction-grating or a prism spectroscope. With magnification, we saw several small scratches and surface abrasions, and a single small, iridescent
cleavage extending from the girdle.
The stone showed no reaction to short-wave UV radiation, but it did fluoresce a weak yet definite slightly
chalky bluish green to long-wave UV. Three other
faceted tanzanitesranging in tone from very light to
very dark purplish bluewere used as comparison
stones; all three were inert, and their presence in the
ultraviolet testing box made the reaction of the 3.46 ct
stone all the more obvious. Since UV fluorescence has
apparently not been recognized in tanzanite until now,
we do not know if this particular stone represents an
anomaly or, because the reaction is weak, if similar fluorescence may have been overlooked in the past.
Regardless, this example shows the importance of using
darkroom conditions when testing gems for their
potential reaction to UV radiation.
Tanzanite mining update. GIA instructor Angelique
Crown visited the Merelani mining area in northeast
Tanzania in late April 1999 (figure 14) and provided the
following update. Blocks A, B, and D were being actively
mined. Although no activity was noted at Block C (the
largest mining area), this Block was acquired in February
1999 by a Tanzanian/South African joint venture from
Graphtan International, which had been mining graphite
there. African Gem Mining Resources (AFGEM) purchased the Block through its subsidiary, Mererani
Mining Ltd. According to Michael Nunn, chief executive
officer of AFGEM, 5% of the equity of the mining company will be placed into the Merelani Development
Trust, which will be operated for the benefit of the community of Merelani. This company, which is investing
$10 million in the project in the short term, is now conducting a comprehensive feasibility study and plans to
begin production by the second quarter of 2000.
Meanwhile, the American Gem Trade Association
(AGTA) relief program has raised approximately $35,000
to help the tanzanite miners recover from the April 1998
mine disaster (see Summer 1998 Gem News, p. 145), and
a shelter for the miners is being built. During the
authors visit, Philip Zahm, chairman of the AGTA project, was at the mining site with three professional
paramedic and rescue training experts (Jeff Schneider,
Larry Bingham, and Rod Shimamoto) from the former
Samaritan Training Center based in Vacaville, California.
At least 15 miners completed three days of emergency
Gem News
Figure 15. These stones (approximately 0.50 2.52 ct)

are representative of the color range we observed in
faceted tsavorites from the new deposit in southern
Tanzania. Photo by Maha DeMaggio.
medical and rescue training. Assistance has also been
provided to the Arusha Fire Brigade, the emergency
responder.
Angelique Crown, GIA
A new source for tsavorite. Dr. Horst Krupp, technical
director of Wahuwa Mining Ltd. of Tanzania, recently
informed us of a major new primary deposit for tsavorite
garnet in southern Tanzania that may radically change
the availability of finecolor material. The new locality
is about 15 km from the town of Ruangwa in Lindi
Province. Word of this new discovery has spread quickly,
and miners are flocking to the area. According to Dr.
Krupp, mining at Tunduru, about 150 km west of this
new deposit, has virtually ceased as a result.
Dr. Krupp told us that the tsavorite is found as
cracked masses in pods that may reach as large as 8 kg.
The masses yield approximately 10%15% gem-quality
rough, most of which will cut small stones. Dr. Krupp
showed us approximately one kilo of gem tsavorite rough
and about 40 faceted stones (some of which are shown in
figure 15), which were among the largest tsavorites cut to
date (most larger than 1 ct, the largest being 3.65 ct). All
were a medium dark green with very little or no yellow
component, and many were equal in quality to the finest
material on the market. According to Dr. Krupp, the new
locality is producing this quality of tsavorite by the kilo,
with the overall color of the rough being very consistent.
The gemological characteristics of these stones are
virtually identical to those of material from other localities in Tanzania. The R.I. is very consistent at 1.74, there
were no distinct lines in the spectroscope, and none of
the samples fluoresced to UV radiation. Even the inclusions were the same as those in tsavorites from the older
deposits, consisting of sugary-appearing fingerprints,
curved or angled needles or growth tubes, andidentified
on the basis of morphologyplaty crystals of opaque
black graphite and transparent dark brown mica, probably biotite. We also noted some unusual graining in one
of the stones. This consisted of fine parallel curved lines
that veered off in several directions (figure 16) and was
GEMS & GEMOLOGY
Summer 1999
151
Figure 16. This unusual curved parallel graining, which

was observed in one of the faceted tsavorites from the
new Tanzanian deposit, is reminiscent of flow lines
seen in some types of glass. Photomicrograph by Shane
F. McClure; magnified 20.
reminiscent of the swirled striae that are seen in glass.

Indeed, if the stone had not had natural inclusions we
would have been suspicious of its origin.
We obtained qualitative chemical analyses of three of
the tsavorites using EDXRF spectrometry. The chemistry
of these samples was consistent with tsavorite from
other localities, with the main chromophore being vanadium. Traces of chromium may have been present, but
those peaks would be concealed by overlap with the
vanadium spectral peaks.
This deposit appears to be located within the same
geologic belt (the Mozambique belt) as the well-known
tsavorite mines at Komolo, Tanzania. However, it is
almost 800 km to the south, and much of the region
between has yet to be thoroughly explored.
SFM
SYNTHETICS AND SIMULANTS
Synthetic beryl showing zoned pleochroism. Russian laboratories have been growing hydrothermal synthetic
beryl in a variety of colors for more than a decade (see,

e.g., Winter 1988 Gem News, pp. 252253; and articles
by J. I. Koivula et al. [pp. 3239] and K. Schmetzer [pp.
4043] in the Spring 1996 Gems & Gemology). Although
most of the Russian hydrothermal synthetic beryl is
green (i.e., synthetic emerald), colored by chromium,
more exotic colors also are produced. These include:
rich turquoise blue, colored by copper; intense pink,
colored by manganese; orangy red, resulting from a trace
of cobalt; and purple, colored by traces of chromium and
manganese in combination. Recently, the Gem News
editors received for examination an interesting hydrothermal synthetic beryl crystal with a coloration not previously encountered.
The crystal, grown on a colorless seed plate as is typical for such material, measured 45.7 10.5 9.3 mm
and weighed 42.50 ct. Thomas Hunn of Thomas Hunn
Co. in Grand Junction, Colorado, purchased it from the
Fersman Mineralogical Museum (Moscow) as a purple
hydrothermal (synthetic) beryl. The crystal was obviously bi-colored when examined through the side in a
direction parallel to the plane of the seed plate: The outer
rim was dark purplish blue, and the center sections next
to the seed plate were intense purplish pink (figure 17,
left). When the crystal was examined in a direction perpendicular to the seed plate, or in the optic-axis direction, the mixture of these two hues resulted in an overall
purplish color.
When viewed with the spectroscope, the crystal
showed a sharp line at 678 nm, weaker lines at 648 and
630 nm, general absorption in the yellow-green to orange
region from approximately 540 to 580 nm, and a band in
the blue region between 433 and 434 nm. The spectrum
did not change as the orientation of the specimen was
changed, provided the beam did not pass through the
seed plate. The R.I. values obtained on a relatively flat
and smooth crystal face were 1.576 1.582 (typical for
beryl), yielding a birefringence of 0.006. The crystal was
inert to both long- and short-wave UV radiation. Since
Figure 17. This 9.3-mm-wide crystal

of hydrothermal synthetic beryl
shows distinct color zoning, with a
dark purplish blue rim and a purplish
pink central area (left); these are the
colors normally seen when observing
the crystal. The colorless layer in the
middle is the seed plate; the dark horizontal cracks are due to strain. A 90
rotation of the analyzer (right) produces a dramatic shift in the
pleochroic colors, resulting in a reddish orange rim and an orangy red
center. Photomicrographs by
John I. Koivula.
152
Gem News
GEMS & GEMOLOGY
Summer 1999
the crystal had metal wires embedded at both ends, no

attempt was made to obtain the specific gravity.
When we viewed the sample with a microscope, we
saw numerous two-phase fluid inclusions. Most of these
were conical- to needle-shaped growth tubes that extend
in the c-axis direction. There were a few partially healed
fingerprint-type fluid inclusions, particularly around the
suspension wires at either end of the crystal. The
pleochroism related to the color zonesperhaps the
most striking aspect of this crystalwas also most evident with magnification. We observed the sample
through the microscope using only the analyzing plate
(that is, not the polarizer as well). The colors in each
zone shifted dramatically as the plate was rotated
through 90: from dark purplish blue to reddish orange in
the rim, and from purplish pink to orangy red in the center section (figure 17).
GIA Gem Trade Laboratory research associate Sam
Muhlmeister obtained qualitative chemistry on the crystal using a Tracor Spectrace 5000 EDXRF system. He
analyzed two separate surfaces, one parallel and the other
perpendicular to the seed plate (i.e., in directions that
maximized information from the bulk and the rim of the
sample). The major elements detected were aluminum
and silicon (light elements such as beryllium cannot be
detected using this instrumentation). Phosphorus
appeared to be present in minor amounts, and traces of
calcium, chromium, copper, gallium, iron, manganese,
nickel, sulfur, and zinc were also noted. The dark purplish blue rim contained more chromium and copper
than the purplish pink central area. The exact cause of
color in the crystal was not determined, but it is logical
to speculate that the purplish pink central zone is colored primarily by manganese, as are natural morganite
and red beryl, and additional chromium and copper are
responsible for the coloration in the rim.
Polymer-impregnated synthetic fire opal from Kyocera.
At the 1998 Basel Fair, the Kyocera Corp. of Kyoto,
Japan, was offering a new product: synthetic fire opal.
Kyocera had already test marketed several colors of polymer-impregnated synthetic opal (see, e.g., Summer 1995
Gem News, pp. 137139), including orangy pink, reddish
orange, and orangy yellow. However, the product presented in Basel looked somewhat different from these in
that it showed green and orange play-of-color against a
bright orange bodycolor, similar in appearance to some
Mexican fire opal.
A 24.55 ct block of this bright orange synthetic opal
(figure 18) was recently studied by contributing editor
Emmanuel Fritsch. The columnar structure typical of
synthetic opal was clearly visible, with the cells elongated perpendicular to the top and bottom surfaces of the
sample. The overall appearance was reminiscent of the
atypical Gilson synthetic opal described in the Summer
1985 Gem Trade Lab Notes (pp. 110111). The R.I. of the
block was about 1.46; the S.G. (estimated by volumetric
Gem News
Figure 18. This 24.55 ct block of Kyocera polymerimpregnated synthetic fire opal, introduced at the
April 1998 Basel Fair, has properties that are significantly different from those of natural fire opal. The
block measures about 16.5 mm in maximum dimension. Photo by A. Cossard.
means) was about 1.8, which is low for natural gem-quality opal. The sample fluoresced strong yellow-orange to
long-wave UV radiation and very strong orangy yellow to
short-wave UV; the orange appearance of this luminescence might be influenced by the samples bodycolor.
The luminescence was evenly distributed (whereas some
natural opals have uneven luminescence) and was not
followed by phosphorescence (many natural and some
synthetic opals, including fire opals, do phosphoresce).
With magnification, many small fractures perpendicular to the top and bottom surfaces were visible; these
extended about 1 mm into the block. The resulting small
brownish patches (figure 19) on the top and bottom sur-
Figure 19. A typical chicken wire structure was

seen within some brown patches on the surface of
the synthetic opal shown in figure 18.
Photomicrograph by E. Fritsch; magnified 60.
GEMS & GEMOLOGY
Summer 1999
153
years of experimentation; it debuted in August last year

at Expo Prestige 98 (the annual jewelry show of the
Quebec Jewellers Corp.). Mr. Couture holds U.S. and
Canadian patents on this cut, and a French patent is
pending. He plans to produce butterflies from ruby,
sapphire, and emerald in 1999, and from diamond for the
new millennium.
Yves Morrier
Les Gemmologistes Associs OCYM
Montreal, Quebec, Canada
Figure 20. These citrine, tourmaline, and amethyst

gemstones (39.46 ct, 6.46 ct, and 4.24 ct, respectively)
were fashioned by Guy Couture as Butterfly cuts.
Courtesy of Yves Morrier; photo by Maha DeMaggio.
faces showed the chicken wire structure that is typical
for synthetic opal.
Mid-infrared absorption spectroscopy revealed
absorption maxima at about 5950, 5875, 5815, 5800,
5625 (with a shoulder at 5125), 4371, and 4675 cm1. The
sample was opaque at lower wavenumbers due to bulk
absorption. These features are reminiscent of those seen
in Kyoceras previous generation of colored impregnated
synthetic opalsindicating the presence of polymers
but the differences in band positions suggest that a different polymer is being used for this material. A significant
concentration of low-S.G. organic materials could
account for the low specific gravity value.
Although this synthetic opal appears to be somewhat
different from Kyoceras previous material, it still can be
readily identified by standard gemological properties (i.e.,
columnar texture, chicken wire pattern, UV luminescence, and low S.G.).
EF
MISCELLANEOUS
A new cut for colored stones. Gemologist Yves Morrier
of Les Gemmologistes Associs OCYM Inc., Montreal,
Quebec, Canada has provided the following information
about a new cut for colored stones. The Butterfly cut
features 43 facets (figure 20). Gems fashioned so far with
this cut have ranged from 0.27 to 44 ct (with some calibrated goods) and include the quartz varieties amethyst
and citrine, tanzanite, rhodolite and almandine garnets,
fire opal, topaz, tourmaline, scapolite, and cubic zirconia.
Mr. Morrier examined more than one hundred samples
that revealed fairly consistent proportions, including an
average length-to-width ratio of 1.38:1; an average total
depth of 62%; an average crown height-to-pavilion depth
ratio of 1: 3; and excellent balance for all profile views.
The Butterfly cut was created by jewelry designer
Guy Couture of Beauport, Quebec, Canada, after two
154
Gem News
Pearl Branding: Utopia cultured pearls. One of the latest entries in the branding of gem materials is the
Utopia pearl. According to Claudio Pagani, international promotion manager for the South Sea Pearl
Consortium, each cultured pearl that is branded and marketed as a Utopia pearl has not been either treated or
polished, and is considered the finest-quality white
South Sea cultured pearl. In the patented system, a
minute (1 mm 1 mm) alphanumeric code (figure 21) is
placed on each pearl by a gold vaporization method.
Although a special sealant is laid over the code for protection, the customer can still remove the code without
damaging the nacre. All Utopia pearls are accompanied by a certificate that includes a digital photo as well
as the dimensions, weight, and form. To date, Mr. Pagani
reports, approximately 10,000 cultured pearls in virtually
all sizes have been branded Utopia pearls.
Noted U.S. designer Henry Dunay of New York is
creating a special jewelry collection of rings, pendants,
earrings, and bracelets fashioned from Utopia pearls.
This collection will be introduced nationwide at the
Neiman Marcus department stores this fall.
ANNOUNCEMENTS
Hong Kong Jewellery & Watch Fair 99. This show will
be held from September 23 to 27 at the Hong Kong
Convention and Exhibition Centre (HKCEC). Specialized
Figure 21. Magnification reveals the alphanumeric
code that has been placed on this Utopia pearl by a
gold vaporization method. Courtesy of Henry Dunay.
GEMS & GEMOLOGY
Summer 1999
pavilions highlighting new jewelry designs, fine designs,

antique and estate jewelry, equipment and packaging
products, and design competitions will be featured. Two
major pearl auctions, by Paspaley and Robert Wan, will
also be held at the HKCEC just before the fair. There will
also be professional seminars and workshops. For details,
contact Miller Freeman Asia Ltd. at 852-2827-6211
(phone) or 852-2564-5496 (fax), or visit http://www.jewellery-net-asia.com/fair9jw.htm.
Cartier at the Field Museum. The exhibition Cartier:
19001939 will be held at the Field Museum in Chicago
from October 2, 1999 to January 16, 2000. Over 200
works will be displayed, including jewelry, clocks,
watches, and ornamental accessories. Client order books,
idea sketches, and original plaster casts of jewels will
provide a behind-the-scenes look into Cartiers creative
process during this innovative period. For more information, call the Field Museum at 312-922-9410, or visit the
Web site http://www.fmnh.org/exhibits/spec_exhibits.htm.
Gem Conference 99. The 10th annual Canadian
Gemmological Association conference will take place
October 2324 at the University of Toronto. Lectures
and workshops will cover inclusions, synthetic moissanite, jewelry design and manufacturing, and appraisals. For
Gem News
more information, contact Mary Ellis at 416-785-0962

(phone) or 416-785-9043 (fax).
Jewellery Arabia 99. The 8th Jewellery Arabia exhibition
will take place November 913, 1999 at the Bahrain
International Exhibition Centre. The event is intended to
serve jewelry manufacturers, importers, wholesalers, and
retailers throughout the Middle East. Visitors will be
admitted on the basis of a business card or invitation. For
further information, contact the event organizers (based
in London) at 44 171-862-2041 (phone), 44 171-862-2049
(fax), or via e-mail at jewellery@montnet.com.
CORRECTIONS:
1. On page 19 of the Schmetzer and Peretti article
Some Diagnostic Features of Russian Hydrothermal
Synthetic Rubies and Sapphires, published in the
Spring 1999 issue, the photographer of the main
photo in figure 3 was Maha DeMaggio.
2. In the Spring 1999 Gem News entry Cats-eye
andradite from San Benito County, California, the
large cats-eye andradite in figure 4 (p. 48) and the
two orangy andradites in figure 5 (p. 49) were courtesy of Rick Kennedy.
GEMS & GEMOLOGY
Summer 1999
155
Thank You, Donors

The Treasured Gifts Council, chaired by Richard Greenwood, was established to encourage individual and
corporate gifts-in-kind of stones, library materials, and other non-cash assets that can be used directly in
GIA's educational and research activities. Gifts-in-kind help GIA serve the gem and jewelry industry worldwide while offering donors significant philanthropic and tax benefits. Treasured Gifts Awards are presented
to those who have given gifts valued at $10,000 or more. We extend a sincere thank you to all those who
contributed to the Treasured Gifts is Council in 1998.
$100,000 or more
Ramsey Gem Imports, Inc.
$50,000 to $99,999
The Bell Group, Inc.
Japan Pearl Exporters'Association
Dr. Duane Lennartson
Rouben & Mary Varozian
Zvi & Rachel Wertheimer
$10,000 to $49,999
Mr. Gary L. Alexander
Mr. George Brooks
Mr. Martin T. Coeroli
Color Change Empire
G. Robert Crowningshield, G.G.
August & Margaret Hansch
Gary & Sang Leonard
Mr. James Leonard
Nancy B & Company
Mr. Rubin Snyder
Mr. Edward R. Swoboda
$5,000 to $9,999
American Ariana Trading Corp.
Daniel & Bo Banks
Broken Hill Propriety Minerals
Classic Colors
Golay Buchel USA, Ltd.
Susan G. Goldstein, G.G.
H. G. Kazanjian & Sons, Inc.
Mariora Australia Pry. Ltd.
Stephen M. Neely, M.D.
Oro America
Pacific Perles
$1,000 to $4,999
Ms. Audrey Afholter
American Gem Trade Association
Mr. German Ayubi
AZCO Mining, Inc.
Dr. Vladimir S. Balitsky
Ben Bridge Jewelers
C3 Inc.
Chatham Created Gems

Colgem EL 97, Ltd.
Color Masters Gem Corporation
George C. Houston, Inc.
House of Onyx
Jostens
K R Gems & Diamonds
International
Ms. Betsy Ross Marcinkus
Marcus Jewelers & Gemologists
Mr. Nick Michailidies
Stone Flower Co.
$500 to $999
All That Glitters
Ashland Press, Inc.
William E. Boyajian, G.G., C.G.
Carlsbad City Library
Constellation Colombian
Emerald Co.
Eddie Buscher, G.G.
Frances & Roger Drezek
The Golden Cache
Archie Henderson, G.G.
Henderson Jewelry
Institute of Experimental
Mineralogy
Mr. Robert E. LaPrad
Mr. William F. Larson
William Lerner, III, G.G.
Pala International, Inc.
D. Poynter & Co., Inc.
Provockative Gems, Inc.
Mr. William C. Reiman
Ronald H. Ringsrud, G.G.
Mr. Till Schoeffel
Ms. Cri Cri Solak-Eastin
Thomas Hunn Co.
$100 to $499
Christopher Amo, G.G., F.G.A.
Antique Gas & Steam Engine
Museum
Mr. K. C. Bell
Mr. Charles L. Bennett

Bergman and Sons
C.P. S. Gems
Donald Clary, Gemologist
Colorado Gem & Mineral Co.
Mr. Ralph T. Davenport
Martin B. de Silva, G.G.
Michael T. Evans, G.G.
Fortunoff Fine Jewelry
Mr. David Frizell
Ms. Carole Johnson
K & K International
Bernard H. Kaplan, G.G.
Mr. George R. Kaplan
Mr. John S. Kerr
John Koivula, G.G., F.G.A.
Gail Brett Levine, G.G.
Mason-Kay, Inc.
Joseph Mayer, G.J.G.
Yianni B. Melas, G.G.
Professor Milan Novk
Mr. Carlos D. Parsons
Mr. Jeffrey E. Patterson
Mr. Gary H. Peterson
Phoenix Gems
Prime Art Jewelry
Tish & Wes Rankin
Ms. Joan Rolls-Gragg
Mr. Henry Shawah
Lewis & Alice Silverberg
John Sinkankas, G.G.
Dr. Iouliia P. Solodova
Ms. Helen L. Stanley
H. Stern Jewellers, Inc.
Richard E. Taylor, G.G.
The World of Science
USAA Alliance Services Co.
Valalan's Jewelers
Mr. Patrick M. Weglarz
E. John Wells, IV
World Wide Golf
Mr. Louis Zara
Ms. Barbara Anne Zinke
Under $100
Ms. Jo Anna Agle
Allerton Cushman & Co.
Boxlee - Applachian Gems &
Minerals
Brundage Jewelers
Capalion Enterprises
The Charles Cecil Collection
Jo Ellen Cole, G.G., F.G.A.
Companhia Real Dc Metals
Mr. Vincent Dahlgren
Ms. Jan David
Del Time
Nun Durucu, G.G.
Exclusive Merchandisers, Inc.
Fine Jewelry, LLC
Gemming International Co.
Houston Imperial Crystals
Mr. Richard M. Knox
Mr. Nickolai B. Kuznetsov
Mr. Richard J. LaForge
Brendan Laurs, G.G.
Lawrence Welk Country Club
Mr. Donald B. Lee
Glen Lehrer, G.G.
MB Enterprises
Nebula Stone
Ms. Rene Newman
Ms. Susanne S. Patch
Pearl Museum
Mr. David L. Peterson
Mr. Bill Pinch
Charles R. Richmond Fine Gems
Howard Rubin, G.G.
Ms. Elizabeth Scherer
Jim Shigley, Ph.D.
Mr. Neil A. Sims
Dr. David R. Soller
Elisabeth Strack, G.G., F.G.A.
Tahiti Perles
Thai Lanka Trading Ltd.
Mr. Randy Thill
Daniel E. Weddle, G.G.
In its efforts to serve the gem and jewelry industry, GIA can use a wide variety of gifts. These include natural untreated and treated
stones, as well as synthetics and simulants, media resources for the Richard T. Liddicoat Library, and equipment and instruments for
ongoing research support. If you are interested in making a donation, and receiving tax deduction information, please call
(800) 421-7250, ext. 4155. From outside the U.S. call (760) 603-4155, fax (760) 603-4199. Every effort has been made to avoid
errors in this listing. If we have accidentally omitted or misprinted your name, please notify us at one of the above numbers.
156
Thank You Donors
GEMS & GEMOLOGY
Summer 1999
Book Reviews
Susan B. Johnson & Jana E. Miyahira-Smith, Editors
THE COMPLETE
HANDBOOK FOR
GEMSTONE WEIGHT
ESTIMATION
By Charles I. Carmona, 427 pp.,
illus., publ. by Gemania Publishing, Los Angeles, CA, 1998.
US$59.95* (softbound)
The Complete Handbook for Gemstone Weight Estimation is a compendium of formulas, tables, and
other pertinent data for estimating
the weight of mounted gemstones of
all sizes, shapes, and typesfrom the
mundane to the arcane.
The author has devised a clever
methodology for organizing and presenting his approach to weight estimation. He first assigned gemstones
into eight categories based on specific gravity. This premise allowed for
the presentation of voluminous
information in a relatively compact
set of tables that are conveniently
indexed by the 24 most common
shapes encountered in the trade
today. Weight calculation formulas
for an additional 48 less-common
shapes are also provided. These formulas, combined with the authors
explanation of applicable correction
factors, allow the reader to estimate
weights for even the most esoteric
gemstone shapes. Because diamond
cutting criteria and pearl shapes are
unique among gem materials, separate sections detail weight estimation information for diamonds and
pearls.
The value of this handbook, however, extends far beyond a rote system of tables, charts, and formulas, as
the author provides special insight
into the effect of gemstone shape on
weight variations. His premise is that
gemstone shape is not limited to a
Book Reviews
simple finite set of profiles; rather, it

is a continuum, which he illustrates
by showing how weight correction
factors vary as a gems shape transitions from round to oval to square
and so on.
Whether you appraise, buy, or sell
mounted gems, this handbook is a
valuable resource. Tabular data are
presented in a concise and easy-toaccess format. Ancillary discussions
are straightforward and well illustrated. An expansive list of formulas
and an extensive glossary of gemstone cuts (which includes more
than 200 trade and proprietary
names) complete this repository of
weight estimation information. In
short, the Complete Handbook for
Gemstone Weight Estimation delivers on the authors stated promise
to guide both novice and professional in estimating the weight of
mounted gemstones with a high
degree of accuracy.
SHARON WAKEFIELD
Northwest Gem Lab
Boise, Idaho
GEMMOLOGISTS
COMPENDIUM,
7th Edition
By Robert Webster, revised by
E. Alan Jobbins, 240 pp., illus.,
publ. by N.A.G. Press, London,
US$35.00*
The seventh edition of the Gemmologists Compendium continues
the excellent tradition of earlier editions by enabling quick and easy
access to a vast amount of valuable
gemstone data and information that
even the most experienced gemologist will appreciate.
The first part of the Compendium
is an extensive alphabetical glossary
of names and terms that covers several subject areas rarely offered in other
compilations, such as testing liquids
and plastic gem simulants. Minor
problems include the fact that some
minor, past, or non-gem localities are
listed, while other very important
current localities (e.g., Afghanistan
for spodumene and tourmaline) are
conspicuously absent. Also, the book
describes tourmaline species according to their color: Brown tourmaline
must be dravite, and green, blue, yellow-green, honey yellow, or palecolored tourmaline must be elbaite.
What about liddicoatite with its
palette of colors, or uvite in red and
green?
In addition to standard tables such
as refractive index and specific
gravity, the second part of the
Compendium contains information
on such subjects as transparency of
gemstones to X-rays, liquids for
refractive index determination, and
the important current topics of gemstone treatment and disclosure. One
page covers country name changes.
The 32 color plates show microscopic features of various gems, as well as
timely illustrations of Russian
hydrothermal synthetic emerald,
heat-treated Mong Hsu ruby, diffusion-treated sapphire, heat-treated
Thai ruby, and fracture-filled diamonds. These are complimented by
old favorites such as byssolite in
demantoid and lily-pad inclusions
in peridot. Mr. Jobbins took many of
these excellent photos specifically for
this edition.
*This book is available for purchase through

the GIA Bookstore, 5345 Armada Drive,
Carlsbad, CA 92008. Telephone: (800)
421-7250, ext. 4200; outside the U.S. (760)
603-4200. Fax: (760) 603-4266.
GEMS & GEMOLOGY
Summer 1999
157
Despite some discrepancies and

omissions, the Gemmologists Compendium is highly useful for gemologists, bench jewelers, and lapidaries.
For gemologists trained in North
America, it provides valuable exposure to British gemology at a reasonable price.
MICHAEL T. EVANS
Gemological Institute of America
Carlsbad, California
CHRISTMAS JEWELRY
By Mary Morrison, 158 pp., illus.,
publ. by Schiffer Books for
Collectors, Atglen, PA, 1998.
US$18.95.
Ms. Morrison provides an intimate
journey into the pleasure of small
details that only a dedicated collector
of Christmas jewelry would notice.
Page after page of crisp photos cover
an unbelievable array of holiday
themes. On the surface it looks like
jumbled kitsch, but nine orderly chapters soon dispel that apprehension.
A chronicle of costume jewelrys
little-known history, which started
in the 1940s, provides unexpected
depth. The chapter on signed Christmas tree jewelry is particularly helpful, as it places these pieces in the
context of the early years, when costume jewelry was in its heyday. The
thumbnail biographical sketches of
major and minor manufacturers,
however, leave you wanting more.
The conversational tone, personal
notations, and observations within
the body of the explanatory paragraphs introduce the reader to characteristics that make these pieces collectible. Small details are given to
help differentiate a pieces time period and manufacturer. Unfortunately,
jewelry items in the photos are not
numbered to coordinate with the
text, which created minor frustration. A glossary of costume jewelry
terminology would be immensely
helpfulI still dont know what
japanned means.
The prices cited are, in the words
158
Book Reviews
of the author, a reflection of real

prices paid for these pieces in the
mid-1990s in the United States. As
a gemologist I was aghast, but independent research and interviews
revealed that even though we are
talking about rhinestones and nonprecious metals, Christmas jewelry is
highly sought after: It is a viable market that fully supports the prices and
ranges given.
GAIL BRETT LEVINE, G.G.
Auction Market Resource
Rego Park, New York
GLOSSARY OF MINERAL
SYNONYMS
By Jeffrey de Fourestier, 442 pp.,
illus., Special Publication 2 of The
Canadian Mineralogist, publ. by the
Mineralogical Association of
Canada, Ottawa, Canada, 1999.
US$50.00
With this book, the author has succeeded in compiling a glossary of
undeniable utility for his intended
audience of private collectors, museum curators, researchers, and those in
the gem trade. The more than
35,000 entries include an eclectic mix
of synonyms, variety names, names
of discredited minerals, names for
synthetics, trade names, and morea
much-appreciated resource for those
of us who have had to deal with cryptic mineral or gem names that eluded
all of our deciphering efforts.
Entries are organized in alphabetical order. In most cases, the synonym
is linked to the currently accepted
mineral species name, occasionally
with brief modifying or explanatory
information. The author has made a
conscientious effort to follow the
nomenclature guidelines of the
Commission on New Minerals and
Mineral Names (CNMMN) of the
International Mineralogical Association (IMA). The book is well printed
and bound, and nicely (though sparingly) decorated with attractive
black-and-white drawings of miner-
als. It is evident that the text has

been carefully screened for typographical errors.
The compilation is surprisingly
inclusive. The bibliography lists
many of the publications that the
author scoured for names, which date
from 1260 to 1998 AD; besides
English, it also include texts in
German, French, Spanish, and Latin,
among other languages. When a
name has been applied to several different materials, all are listed, and
virtually every variation in spelling of
a name is provided.
Overall, this book is a marvelous
resource, but it does have a few shortcomings. The conciseness of each
entry makes it an excellent quick reference, but it often leaves the reader
in want of more information. In particular, the significance and full
meaning of some terms is not provided. Amethyst, for example, is equated
with quartz, but with no mention of
color. Amethyst is also noted as
being a synonym for corundum and
beryl, but there is no mention of the
fact that this usage is old and obsolete. Although I wanted to learn more
about this, unfortunately no bibliographic references are provided for
specific entries. It would also have
been useful to include all accepted
mineral species names as separate
entries, regardless of whether they
have synonyms. Thus, there is no
entry for painite, an accepted name
for a mineral species (and rare gem
material) without a synonym.
In reviewing this book, I was
reminded of a joke I heard many
years ago. A young man, asked about
the last book he read, responds,
There wasnt much plot, but the
cast of characters was tremendous.
Although he was actually talking
about a telephone directory, he might
just as well have been referring to the
Glossary of Mineral Synonyms.
However, you can probably get by
without a telephone directorythere
is always directory assistancebut
anyone interested in gems and min-
GEMS & GEMOLOGY
Summer 1999
erals will quickly come to rely on

this book as an essential reference.
ANTHONY R. KAMPF
Los Angeles County Museum of
Natural History
Los Angeles, California
THE PHOTO-ATLAS OF
MINERALS [CD-ROM]
Developed by Anthony R. Kampf
and George Gerhold, produced by
The Gem & Mineral Council, Los
Angeles County Museum of Natural
History, Los Angeles, 1998.
US$49.95.*
The Photo-Atlas of Minerals is a
thoughtful, sophisticated product, in
CD-ROM format, that uses the capabilities of computer technology to
provide exciting features. To put the
Photo-Atlas in perspective, I spent
$40 on the Encyclopedia of Minerals
by W. L. Roberts, G. R. Rapp, Jr., and
J. Weber (Van Nostrand Reinhold,
New York) when it was first published in 1974. It was an extravagance for a struggling graduate student, but I bought it for two reasons:
First, it had entries for 2,200 mineral
species, the most comprehensive reference then available. More importantly, the Encyclopedia was the first
standard reference in mineralogy to
be extensively illustrated in color,
with almost 1,000 color photographs
of an unusually wide variety of
species. Now, a generation later, The
Photo-Atlas of Minerals serves the
same purpose even better, and at a
much cheaper price.
The Photo-Atlas is easy and fun
to use. Clear instructions are provided, but they really are not needed.
The core of the product is 6,500 color
images of more than 800 mineral
species. Most of these are the work of
Wendell E. Wilson and Louis Perloff,
two of the very best mineral photographers. Some species, especially the
gem minerals, are represented by
dozens of illustrations. There are
occasional images of gemstones, as
Book Reviews
well as inclusions. These are all

linked to a database of mineral properties that covers 3,600 species and
an extensive glossary. The user can
search and view minerals by name,
chemistry, locality, crystal class, or
physical properties. For users who are
wired for sound, click on the name of
a mineral and it is pronounced! The
CD also can be used to play an identification game or create a slide show.
The Photo-Atlas of Minerals provides an unparalleled collection of
excellent mineral images with supporting text, and a search capability
that can be used and enjoyed by a
wide audience.
CARL FRANCIS
Harvard Mineralogical Museum
Cambridge, Massachusetts
pare directly this translation to the

originals in native languages which
are conveniently placed in parallel
[Bengali, Hindi, and Sanskrit].
In this reprint, a number of original lithographic illustrations are
poorly reproduced, but the text is
sharp and easily legible. Although the
binding leaves much to be desired,
nevertheless the volumes should
hold together and wear well. It is
interesting to note the meteoric rise
in price of an original Mani-Mala,
beginning with an auction sale at $85
in 1971, to $4,000 at Christies in
1987, and an astonishing $8,625 paid
in the Swann auction of 1994. Given
those astronomical sums, perhaps
$200 isnt too much after all.
JOHN SINKANKAS
Peri-Lithon
San Diego, California
OTHER BOOKS
RECEIVED
Edelsteine: Symbole der Schnheit

und der Macht (Gemstones: Symbols
of Beauty and Power), by Edward J.
Gbelin and Franz-Xaver Erni, 240
pp., illus., publ. by Verlag Hans
Schner GmbH, Knigsbach-Stein,
1999 [in German], US$43.00. Forming the core of this large (30.5
24.5 cm), richly illustrated volume
are more than 110 pages devoted to
26 different gemstones. Discussions
of each gemstone are accompanied by
the photographs of Harold and Erica
Van Pelt, among others. Remaining
chapters contain general overviews of
origins and characteristics, symbolism, production, processing, inclusions, and synthesis of gemstones.
Tables, charts, and other graphics
supplement the text throughout.
Edelsteine closes with a glossary and
a bibliography. An English version is
forthcoming.
Mani-Mala, or A Treatise on Gems,

by Sourindro Mohun Tagore, reprinted by Nandishkor R. Barot, 1,046
pp., illus., publ. by Mehul Barot,
Philadelphia, PA, 1996, US$200. It is
always a cause for celebration whenever a much-esteemed historical
gemological book, hitherto inaccessible to most persons because of rarity,
is reprinted. The basic worth of
Tagores treatise, first published in
1879, remains unchanged from my
evaluation in Gemology: An Annotated Bibliography (Scarecrow Press,
Metuchen, New Jersey, 1993), entry
no. 6475: On all counts one of the
most remarkable and important
works on gems ever published, particularly so because its Indian author
stresses the knowledge of gems and
their cultural significance in the
Indian lands. For information on
Indian gems, one must ultimately
seek out and consult this splendid
work. It is the product of many years
research in locating and extracting the
ancient literature of India. . . .[Readers
are given] the opportunity to com-
STUART D. OVERLIN
Gemological Institute of America
Carlsbad, California
GEMS & GEMOLOGY
Summer 1999
159
Gemological
Abstracts
EDITOR
A. A. Levinson
University of Calgary, Calgary,
Alberta, Canada
REVIEW BOARD
Troy Blodgett
GIA Gem Trade Laboratory, Carlsbad
Anne M. Blumer
Bloomington, Illinois
Peter R. Buerki
GIA Research, Carlsbad
Jo Ellen Cole
GIA Museum Services, Carlsbad
Maha DeMaggio
Michael Gray
Coast-to-Coast, Missoula, Montana
R. A. Howie
Royal Holloway, University of London
Mary L. Johnson
Jeff Lewis
San Diego, California
Margot McLaren
Richard T. Liddicoat Library, Carlsbad
COLORED STONES AND

ORGANIC MATERIALS
Amber specimen yields another link to the past. C.
Fenelle, National Jeweler, Vol. 42, No. 20, October
16, 1998, p. 20.
A new species of mayfly estimated to be more than 25
million years old was found in a piece of Mexican amber
recently unveiled at Britains Natural History Museum.
Mayflies are rarely found in amber, because the insects
only live above water for a few hours and die shortly after
mating. Other insect and plant remains contained in
amber are spiders, ants, gnats, flies, wasps, bees, flowers,
mushroom caps, seeds, leaves, and stems. The most common plant remains found in amber are pine needles and
pine cones.
MM
Crystal, quartz, or what? S. Frazier and A. Frazier,
Lapidary Journal, Vol. 52, No. 6, September 1998,
pp. 2631, 62.
Fourteen different meanings of the word crystal are presented in this article. More common uses of the term
include any mineral with natural faces, transparent
quartz, clear cut glass of superior quality, the colorless
cover of a watch face, or any material with an orderly
internal arrangement of atoms. Crystal may also refer to
an early diamond grading term corresponding to GIA
color grade J, a clarity grade for opal, or any material that
Jana E. Miyahira-Smith
GIA Education, Carlsbad
Carol M. Stockton
Alexandria, Virginia
Rolf Tatje
Duisburg University, Germany
Sharon Wakefield
Northwest Gem Lab, Boise, Idaho
June York
160
This section is designed to provide as complete a record as practical of the recent literature on gems and gemology. Articles are
selected for abstracting solely at the discretion of the section editor and his reviewers, and space limitations may require that we
include only those articles that we feel will be of greatest interest
to our readership.
Requests for reprints of articles abstracted must be addressed to
the author or publisher of the original material.
The reviewer of each article is identified by his or her initials at the
end of each abstract. Guest reviewers are identified by their full
names. Opinions expressed in an abstract belong to the abstracter and in no way reflect the position of Gems & Gemology or GIA.
GEMS & GEMOLOGY
Summer 1999
has particular transparency or purity. However it is used,

the word crystal stems from the Greek krystallos, or
clear ice. For 2,000 years, the main source for quartz
crystals was cavities in rocks in the higher elevations of
the Alps. The cavities also contained clear ice crystals
during the cold months, so that, at times, quartz crystals
would appear to be growing out of the ice. Quartz was
considered permanently frozen ice; thus, krystallos also
became synonymous with clear quartz. The various
other meanings of the word crystal evolved over time as
a result of such factors as the apparent similarity of
quartz to glass and other minerals, and the adaptation of
the word through the Latin, French, and Middle English
languages.
MD
Demantoid garnet enjoys market transformation. ICA
Gazette, NovemberDecember 1998, pp. 1, 7.
Demantoid (andradite) garnet was first discovered in 1854
in Russias Central Ural Mountains, where it is still being
mined. The rarest of all gem garnet varieties, it was found
as a by-product of platinum mining in the Bobrovka
River. Demantoid was originally thought to be peridot,
based on its color resemblance. In 1878, the stone was
identified as a garnet and named demantoid (from the
Dutch demant, meaning diamond) because of its
extremely high dispersion. A second deposit of vivid
green demantoid garnet was discovered at the Chusovaya
River near Ekaterinburg in 1879. Demantoid was popular
in the early part of the 20th century, particularly among
royalty, and was frequently used in Peter Carl Fabergs
intricately crafted objects.
More recently, demantoid has enjoyed limited popularity because of the relatively small quantities and
faceted sizes available. In 1997, a new deposit of demantoid was discovered at Usakos in central Namibia. The
Namibian material shows a beautiful brilliance, and
occurs in larger sizes, but it does not contain the characteristic horse-tail inclusions found in the Russian
material. Production is expected to increase in both
Russia and Namibia, perhaps enabling demantoid to
regain its popularity with jewelers, designers, and gem
collectors.
MM
The gemological identification of emeralds and blue sapphires. J. E. Shigley, CIM [Canadian Institute of
Mining and Metallurgy] Bulletin, Vol. 91, No. 1025,
Nov. 1, 1998, pp. 9196.
The gemological properties of natural emeralds and sapphires, and ways of differentiating them from other gemstones and from their synthetic counterparts, are
reviewed for a geologic audience. In addition to tables of
gemological properties and distinctive features observed
with a gemological microscope, the following topics are
discussed for both emeralds and sapphires: types of geologic occurrence, methods of synthesis, treatments, and
means of identifying synthetic and treated stones using
both standard (e.g., microscopic) and advanced (e.g.,

infrared and Raman spectroscopic) techniques.
An underlying theme of the article is that ethical business practices on the part of the entire jewelry industry
are essential to maintain consumer confidence. Thus,
synthetic stones must be identified and treatments disclosed. Ethical failures at the jewelry counter, for example, eventually will have negative consequences throughout the industry. One possible result could be reduced
exploration and mining of natural stones.
MD
Investigation of the intracrystalline Cr3+ distribution in
natural and synthetic alexandrites. H. Rager, A.
Bakhshandeh-Khiri, and K. Schmetzer, Neues
Jahrbuch fr Mineralogie Monatshefte, Vol. 12,
December 1998, pp. 545557.
The substitution of Cr for Al in chrysoberyl (ideally
BeAl2O4) produces the phenomenal gem, alexandrite.
The intensity of the color-change effect in alexandrite
is directly proportional to Cr content. Electron Paramagnetic Resonance (EPR) was used on eight samples
(five natural and three synthetic) to establish which of
two different Al crystallographic lattice sites (designated
AlI and AlII) controls Cr distribution in alexandrite.
The data suggest that Cr is mostly concentrated in the
larger of the two Al sites (AlII), contrary to the conclusions of several previous studies. The influence of Fe on
the Cr distribution was minimal. The amount of Cr in
the larger sites increases with overall Cr content, but the
Cr fraction in the smaller sites either remains constant or
increases minutely. Therefore, it was concluded that the
larger Al site is host to the Cr that causes the color
change in alexandrite.
JL
Investigation of the stability conditions and structuralchemical transformations of Baikal lazurite. V. L.
Tauson, V. V. Akimov, A. N. Sapozhnikov, and K. E.
Kuznetsov, Geochemistry International, Vol. 36,
No. 8, 1998, pp. 717733.
Lazurite, the main constituent of lapis lazuli, is part of
the sodalite mineral group (along with sodalite, nosean,
and hayne). This in-depth geochemical study of lazurite
from the Baikal region of Siberia was undertaken to
establish the causes of structural modulations at the
crystal lattice-atomic level. Samples were annealed according to three different experimental methods. The
first two were dry runs that used gold or quartz ampoules to hold ~50 mg of lazurite powder and chemical
reaction indicators. The third group of experiments was
designed to mimic hydrothermal conditions, using aqueous solutions at higher pressures. On the basis of the
results, the authors constructed a crystallization phase
diagram.
Since the Baikal lazurite is inferred to have formed
over the narrow temperature range of about 500600C,
the experimental data suggest that the microstructural
GEMS & GEMOLOGY
Summer 1999
161
modulations depend more on the partial pressure (fugacity) of the sulfur dioxide (i.e., SO2) gas phase than on the
temperature at the time of crystallization. In the experiments, the lazurite was transformed into complex sulfate
and oxide phases at high SO2. At low SO2, sodalite was
formed. The lazurite that formed at moderately high
SO2 showed a modulated structure. The blue lazurite
color is attributed to the fugacities of sulfur and hydrogen
gas phases during formation, as well as to the size of the
crystals.
JL
Is opal . . . ? S. Frazier and A. Frazier, Lapidary Journal,
Vol. 52, No. 5, August 1998, pp. 1625.
This article, which answers 12 commonly asked questions about opal, establishes the distinction between the
different types, including precious (has play-of-color),
common (lacks play-of-color), boulder (cut with matrix as
part of the gem), and fire (has a fiery body color). The
distinction between play-of-color (a flash of different colors) and opalescence (a milky body color) is explained.
The formation of fire opal requires an unusual combination of geologic conditions, hence the rarity of fine specimens, with the most valuable coming from Quertaro,
Mexico. Although Australia is the best-known source of
precious opal, gem-quality material is also found in
Brazil, Honduras, Mexico, the U.S. (Nevada, Idaho, and
Oregon), Canada (British Columbia), and Slovakia.
Particular attention is paid to Brazil as a source: Six types
of opal (precious, green, black, hyalite, greenish yellow
cats-eye, and fire) are mined there, though none in large
quantities. Historical aspects of opal, such as its appeal to
the Aztecs and Mayans, and practical aspects, such as
crazing of opal from some localities, also are noted.
Like other gemstones, opal has simulants (e.g., plastics introduced in the 1970s, which are easily detected). In
1964, John Slocum produced what is known as Slocum
Stone, a glass showing apparent play-of-color that imitates precious opal. It is harder, tougher, and much more
stable than natural opal. Pierre Gilson of France perfected
the manufacture of material marketed as (Gilson) synthetic opal. Today, synthetic opals are produced in Japan
and Russia.
MD
Marketing high quality Chinese freshwater pearls. B.
Sheung, Jewellery News Asia, No. 169, September
1998, pp. 124, 126.
Of the companies that supply Chinese freshwater cultured pearls, a few now seek to upgrade the products
image by focusing on quality rather than quantity. By
concentrating on pearls that have been cultivated over a
longer time period and selecting only the largest, highest
quality specimens, these dealers are able to sell loose
pearls and strands at prices closer to those usually seen
for Akoya and South Sea cultured pearls.
Chinese freshwater cultured pearls can be as large as
15 mm, but the ones sold by high-end dealers typically
range from 9 to 11 mm. Among these are treated black
162
freshwater cultured pearls that offer an affordable alternative to natural-color Tahitian cultured pearls. Natural
colors in Chinese freshwater cultured pearls include
white, champagne, orange, and purple. The article
includes several company profiles; some Chinese freshwater cultured pearl processors in Hong Kong and Japan
focus on the favored pinkish tint rather than a bleached
yellowish white.
Doug Fiske
Mined in America: Five stones to watch (Part Two). D.
Yonick, American Jewelry Manufacturer, Vol. 43,
No. 12, December 1998, pp. 1819, 21.
Part two of this series focuses on Arizona Anthill garnet and Montana sapphire. (Part one, published in October 1998, featured Oregon sunstone, California benitoite,
and Utah red beryl.) Red pyrope garnet is mined from an
alluvial deposit on Navajo land in northeast Arizona,
with 90% of the material marketed by Tucker Gems of
Homedale, Idaho. The mine could produce about 1 million carats of rough a year. However, there is a limited
supply of cut material over 1 ct. Meanwhile, American
Gem Corp. (AGC) of Helena, Montana is expected to
achieve full-scale production of Montana blue and fancycolored sapphires from its alluvial deposits in southwestern Montana in 1999. Once plagued by financial difficulties, environmental liabilities, and internal problems, the
company restructured in 1997. By the end of 1998, AGC
had processed about 2 million carats of rough, which is
expected to yield approximately 200,000 carats of cut
stones.
MM
Models of corundum origin from alkali basaltic terrains:
A reappraisal. F. L. Sutherland, P. W. O. Hoskin,
C. M. Fanning, and R. R. Coenraads, Contributions
to Mineralogy and Petrology, Vol. 133, 1998, pp.
356372.
Corundum from basalt fields, particularly in Australia
and Southeast Asia, has been the subject of numerous
geologic studies. Two types, magmatic and metamorphic, are recognized based on their trace-element characteristics, mineral inclusions, and likely geologic origin
within the earths crust and mantle. This paper is concerned only with magmatic corundum, designated as
BGY for the dominant blue, green, and yellow colors.
BGY corundum shows distinctive trace-element contents
(up to 0.04 wt.% Ga2O3 and low Cr/Ga and Ti/Ga ratios);
20 different primary mineral inclusions, including several reported here for the first time, have been identified
within them.
Detailed examination was made of the crystallization
characteristics (e.g., temperature of formation and intergrowth relationships) of several primary inclusions, particularly zircon, columbite, and feldspar. The authors
evaluated these data in light of the tectonic setting of the
deposits, and performed thermodynamic modeling of the
crystallization of igneous melts of various compositions,
to reach the following conclusion: The BGY corundum
GEMS & GEMOLOGY
Summer 1999
crystallized from a low-volume silicate melt derived from

amphibole-bearing mantle assemblages at temperatures
of 720880C and pressures of 0.71.1 GPa (about 2540
km depths). The corundum was subsequently brought to
the surface as xenocrysts in alkali basalts. The seven previously published origins that have been suggested for
magmatic corundum from alkali basalts are reviewed in
light of the conclusions reached in this paper.
AAL
Prevalence and origin of birefringence in 48 garnets from
the pyrope-almandine-grossularite-spessartine quaternary. A. M. Hofmeister, R. B. Schaal, K. R.
Campbell, S. L. Berry, and T. J. Fagan. American
Mineralogist, Vol. 83, No. 1112, Part 1, NovemberDecember 1998, pp. 12931301.
Of 48 samples of garnet with pyrope-almandine-grossular-spessartine chemical compositions, 40 were anisotropic and the other eight samples were too thin to determine anisotropy. The birefringence () of the garnets
ranged from 0.0001 to 0.0006 (26 were only marginally
birefringent, with 0.0002), and was generally undulatory in appearance. Hence, weak and undulatory birefringence is apparently widespread in garnets. A relationship
between birefringence and geologic setting was recognized. Specifically, there is an inverse relationship between the degree of birefringence and the stress encountered during a garnets geologic history. Thus, the greatest
birefringence occurs in garnets from xenocrysts of mantle
origin and in garnets from eclogite xenoliths in kimberlite pipes, as these are farthest from their conditions of
formation.
It is suggested that most of the optical anomalies (i.e.,
birefringence and undulatory extinction) result from
internal strain within the garnets, although other possible origins are acknowledged for a few specimens, such as
those with sector twinning. Retention of internal strain is
compounded by a mismatch in size between Ca2+ and
Mg2+ within the crystal structure. Four previous models
that explain the origin of birefringence are reviewed.
Kyaw Soe Moe
World jade resources. F. Ward, Arts of Asia, Vol. 29, No.
1, 1999, pp. 6871.
There is more jade available today than at any other
time in history. This is due primarily to three newly discovered deposits in the Western Hemisphere: two are
nephrite (in Canada and the U.S.), and one is jadeite (in
Guatemala). The Polar mine in northern British
Columbia, Canada, produces an attractive translucent,
bright, intense green nephrite that is reportedly harder
than any other nephrite. It has found a market of its own
because of its distinctive characteristics and beauty. The
Blue Moon mine in California produces two beautiful
varieties of botryoidal nephrite (Mint in Snow and
Blossom Jade) that are highly sought by collectors and
jewelry manufacturers. Galactic Gold is the name
given to a newly discovered black jadeite from Guatemala
that has bright metallic inclusions of undetermined mineralogy (they may be pyrite). Worldwide, nephrite is
much more abundant than jadeite.
The history and uses of jade, and the differences
between nephrite and jadeite, are reviewed. Value factors
for the different colors and physical properties (e.g., ability to take a good polish) of both types of jade are noted,
along with the high esteem in which jade is held in Asia.
A brief market overview is presented for all major
nephrite and jadeite producing countries (China, Russia,
Canada, U.S., Australia, New Zealand, Guatemala, and
Myanmar).
It is encouraging that in this age of ever-diminishing
deposits and reserves, a gem of such historical importance and renown continues to be discovered in exciting
new forms.
JEC
DIAMONDS
Of diamonds, dinosaurs and diastrophism: 150 million
years of landscape evolution in southern Africa.
T. C. Partridge, South African Journal of Geology,
Vol. 101, No. 3, September 1998, pp. 167184.
This article chronicles and interrelates the geomorphologic and geologic history of southern Africa starting with
the breakup of Gondwanaland 150 million years ago,
which led to the separation of Africa, South America,
Antarctica, Australia, and India. At that time, southern
Africa had a high elevation, but 13 km of the surface was
subsequently eroded. Also, from 150 million years ago to
the present, southern Africa has experienced extreme
and/or rapid climatic changes (ranging from desert to
tropical rainforest to grassland), fluctuations in sea level,
volcanism (including two episodes of diamondiferous
kimberlite emplacement at approximately 120 and 9060
million years ago), and the enormous meteorite impact at
the CretaceousTertiary boundary that caused major
extinctions of plant and animal life (most prominently
the dinosaurs). During this time, the major rivers in
southern Africa (Orange and Vaal) carried gem diamonds
from their kimberlite hosts in the central part of South
Africa to the sea.
Using details from all these geomorphic, geologic, climatic, and biological events, the continually changing
landscape of southern Africa is reconstructed. One aim of
this review is to explain and predict the locations of the
important secondary (alluvial, beach, marine) diamond
deposits of South Africa and Namibia.
AAL
Diamonds from kimberlite diatreme 50, Liaoning
Province, China: Microtextural, mineralogical, geochemical, and genetic characteristics. S. F. Vinokurov, A. I. Gorshkov, Y. N. Bao, I. D. Ryabchikov,
L. V. Bershov, and M. I. Lapina, Geochemistry
International, Vol. 36, No. 8, 1998, pp. 676683.
Diamonds from pipe 50 in Liaoning Province show several textural, mineralogical, and geochemical features
GEMS & GEMOLOGY
Summer 1999
163
related to their primary crystallization conditions and

later explosive events that occurred in the lower part of
the kimberlite diatreme. Of special interest are unique
inverse coated diamonds, which have a transparent
outer zone and a black inclusion-rich inner zone; normally, coated diamonds have the black inclusion zone on
the outside. The black inclusion-rich zone contains a
diverse population of submicron-size inclusions, including native metals, alloys, sulfides, oxides, chlorides, and
the like. On the basis of the texture, composition, and
inclusion distribution within the zones of an inverse
coated diamond, the authors suggest that these diamonds most likely crystallized via secretion (i.e., from
the outside inward).
TB
Eclogitic diamond formation at Jwaneng: No room for a
recycled component. P. Cartigny, J. W. Harris, and
M. Javoy, Science, Vol. 280, No. 5368, May 29,
1998, pp. 14211424.
Geologists recognize two major groups of diamonds based
on their distinctive mineral inclusions: peridotitic (Ptype) and eclogitic (E-type). Diamonds of both types are
generally believed to form in the earths mantle, the Ptype from mantle materials and the E-type as a consequence of the transformation of subducted biospheric
carbon (i.e., a recycled organic component from the
earths surface). Support for the different origins is found
in the carbon isotopic compositions (13C) of diamonds.
P-type diamonds have a narrow mantle range, (13C =
about 5), whereas E-type diamonds have a broad range
(13C = 34.5 to +2.7) that is characteristic of carbon
compounds in the earths near-surface environment.
Diamonds from Jwaneng, Botswana, contain small
amounts of nitrogen (average 380 ppm) as is characteristic of E-type diamonds; based on a study of their nitrogen
isotopic compositions (15N), the authors question the
theory outlined above, at least for E-type diamonds from
this particular locality. Specifically, the mean 15N value
(5.5) for Jwaneng E-type diamonds is essentially identical to that of P-type diamonds worldwide, but the
authors believe that this value should be more positive if
these diamonds crystallized from recycled carbon that
originated from organic matter (which has a small nitrogen component).
The authors suggest that the carbon in E-type diamonds, like that in P-type diamonds, also comes from the
mantle (and not from a recycled surface component).
They believe that the broad range of 13C values in E-type
diamonds is the result of a fractionation process of carbonatitic melts (involving such processes as the escape of
CO2), which undergo different evolutions before diamond
precipitation when percolating through a peridotite or
eclogite in the mantle. These different evolutions depend
mainly on the presence or absence of olivine (in peridotite
and eclogite, respectively), and can account for the respective 13C distributions.
AAL
164
The Merlin kimberlites, Northern Territory, Australia.

D. C. Lee, H. J. Milledge, T. H. Reddicliffe, B. H.
S. Smith, W. R. Taylor, and L. M. Ward. In Proceedings of the Sixth International Kimberlite
Conference, Russian Geology and Geophysics,
Vol. 38, No. 1, 1997, pp. 8296.
The Merlin field (10 5 km) consists of 11 diamondiferous kimberlites, which are located in four clusters in the
Batten region of Australias Northern Territory (about
800 km southeast of Darwin). The kimberlites were
emplaced about 367 million years ago. They are all small
(the largest is only 125 m in diameter), and all have a
constant diameter to a depth exceeding 100 m. None of
the kimberlites was exposed at the surface at the time of
discovery.
The extensively altered kimberlite rocks are mainly
diatreme-facies (that part of the pipe filled with angular
broken fragments resulting from gaseous explosions),
with some hypabyssal rocks (nonbroken kimberlite
characteristic of the root zone). The trace-element
chemistry resembles micaceous kimberlites and also
some phlogopite lamprophyres (dark-colored, mafic
intrusive igneous rocks) from the Kimberley region of
Western Australia; their trace-element patterns are different from those of South African kimberlites or of the
Argyle lamproite.
Chrome spinel, pyrope, chrome clinopyroxene (diopside), apatite, and sulfides were identified in heavy mineral concentrates of the kimberlites. Only limited information was available on the macrodiamonds, which
occur in dodecahedral and resorbed shapes in about equal
proportions. Most showed no visible evidence of plastic
deformation. Microdiamonds (0.10.8 mm) were abundant; from a study of 40 samples, at least three different
populations were defined based on nitrogen content and
aggregation state: (1) predominantly type IaA and type
IaAB, (2) type IIa (two microdiamonds), and (3) partially
type Ib (one microdiamond). Three nitrogen-rich (>1,500
ppm N) microdiamonds had exceptionally high hydrogen contents (actual H content not stated), while the rest
had no hydrogen. Carbon and nitrogen isotopes indicate
that the microdiamonds originated by mixing between a
mantle source of these elements and either a SiC-derived
or a recycled crustal source. The abundance of twins and
aggregates in the microdiamonds indicate that there was
a plentiful supply of carbon in the source region. The diamonds were interpreted to have crystallized 1.65 billion
years ago.
[Editors note: The 1998 Annual Report of Ashton
Mining Ltd. (p. 10) states that mining operations at the
Merlin Diamond Project commenced in November 1998,
at the southern cluster of four kimberlites with an
inferred resource base of 0.4 ct/tonne. The diamonds have
a weighted average value of approximately US$75 per
carat.]
MLJ
GEMS & GEMOLOGY
Summer 1999
The problem of the diamond potential of the northwestern Russian Plate. A. A. Konstantinovskii and T. E.
Shcherbakova, Lithology and Mineral Resources,
Vol. 33, No. 3, 1998, pp. 226 234.
A potential new diamond-producing region has been
recognized in Russia. Located in the Latvian Saddle,
on the northwestern part of the Russian Plate, it is centered in the vicinity of Borovichi (about halfway
between St. Petersburg and Moscow). Although kimberlite pipes have not been located, a few octahedral diamonds and numerous kimberlite indicator minerals
mainly pyrope garnet, but also chrome spinel and
picroilmenitehave been found in alluvium and stream
sediments in the region.
The diamond indicator minerals have been traced to
two specific Devonian and Lower Cretaceous sedimentary formations (secondary sources), in which they
occur as accessory heavy minerals. These heavy minerals arrived at their present sites as a result of the weathering of a primary (presumably kimberlite) source that
has yet to be found. By plotting the distribution of the
indicator minerals in alluvium and in stream sediments
eroded from the sedimentary formations, and determining that these minerals did not travel far from either
their primary or secondary sources, geologists have
identified a relatively small potential kimberlite area
(the Borovichi Uplift). The favorable potential of this
area is confirmed by geophysical data (e.g., the earths
crust is thick in this area, and basement faulting is present) and geologic features (e.g., an Archean-age basement, regional folding, and local dome-shaped uplift).
AAL
Rare and unusual mineral inclusions in diamonds from
Mwadui, Tanzania. T. Stachel, J. W. Harris, and
G. P. Brey, Contributions to Mineralogy and Petrology, Vol. 132, No. 1, 1998, pp. 3447.
Diamonds form in the earths mantle and have a complex
history of crystallization. Mineral inclusions in diamonds
can give insight into the conditions under which the diamonds formed. One hundred diamonds from the Mwadui
mine in Tanzania were crushed, and 91 of these yielded
inclusions that were analyzed for geochemical data. Most
of the inclusions were minerals common in the upper
mantle, such as olivine, garnet, and pyroxenes. However,
rare mineral inclusions also were found, such as ilmenite
with an unusual chromium content, magnetite with a silicate exsolution phase, native iron, and dolomite.
The associations of minerals that formed under different conditions in the same diamond suggest to the
authors that strong and variable geochemical compositional gradients were present during diamond formation
at Mwadui. Different oxidation states of coexisting iron
minerals (magnetite, wstite, and native iron) imply that
the diamonds grew along oxidation boundaries. On the
basis of these data, the authors conclude that the Mwadui
diamonds formed from carbon that was released from

methane-rich gas phases during mantle recrystallization
and mineralogic changes.
JL
A rift model for the sedimentary diamond deposits of
Brazil. R. Fleischer, Mineralium Deposita, Vol. 33,
1998, pp. 238254.
Essentially all the diamonds found in Brazil are secondary
in origin, obtained from either diamond-bearing conglomerates or recent alluvials derived from them; their
primary sources have never been found. Although several hundred kimberlites have been discovered within the
Brazilian diamond districts in recent years, almost all are
barren and thus cannot be the primary diamond source.
In this article, a new rift model proposes where the diamond-bearing rocks (presumably kimberlites) may have
been emplaced and why they have not been found.
The model is based on the observation that although
the diamond-bearing conglomerates are of four vastly different ages (Proterozoic, Devonian, Carboniferous, and
Cretaceous), they all have geologic features in common
that have recurred through time. First, the diamond-bearing conglomerates are always found in small graben-like
troughs (basins) formed during successive rifting events.
Second, a short transport distance is inferred for the conglomerates, and the diamonds occur above the basal
member of each sequence (i.e., diamond deposition took
place after each basin formed). Therefore, the volcanic
events are adjacent to the faults that form the basins.
Third, volcanic debris in the sedimentary sequences indicates that, for the most part, volcanic events (including
possible emplacement of kimberlites or lamproites) occurred after the rifting.
These general observations link diamond-bearing volcanic sources to rift basins. The diamond-bearing kimberlites proximal to the basins supplied diamonds to the
conglomerates within the basins and, in the process, have
been eroded or buried by post-rift material. Those
noneconomic kimberlites that have been found are farther from the basins and probably represent different volcanic events.
AAL
Unusual upper mantle beneath Guaniamo, Guyana
shield, Venezuela: Evidence from diamond inclusions. N. V. Sobolev, E. S. Yefimova, D. M. DeR.
Channer, P. F. N. Anderson, and K. M. Barron,
Geology, Vol. 26, No. 11, 1998, pp. 971974.
The mineral inclusions in industrial-quality diamonds
from alluvial and primary deposits of the Guaniamo
region of Venezuela were examined to study the environments of diamond genesis. Diamonds form deep
beneath the continental lithosphere in the mantle,
which has different zones with distinct mineralogies.
Populations of diamonds mined from any one region will
have a dominant proportion of crystals from one mantle
source over another. Most of the diamond occurrences in
GEMS & GEMOLOGY
Summer 1999
165
the world are associated with a peridotite mineralogy,

which includes such minerals as olivine and pyroxenes.
In some locales, mineral inclusions and source rock
characteristics suggest an eclogite source, with a mineralogy that includes omphacite (a clinopyroxene) and
pyrope garnet.
During this study, the existence of transparent garnet
and clinopyroxene inclusions, as well as opaque sulfides
and other minerals, was microscopically established in
329 alluvial and 53 kimberlite diamonds. The chemistry
of inclusions extracted from 124 alluvial and 53 kimberlite diamonds was measured by electron microprobe
analysis. The authors determined that 99% of the diamonds were eclogitic. The coexistence of coesite (a highpressure quartz polymorph) and corundum, combined
with the variability of the garnet and clinopyroxene
compositions, suggests a highly variable source composition. The unique characteristics of the Guaniamo diamonds and associated rocks imply that the mantle
source under the Guyana shield in Venezuela has an
unusual composition.
JL
Why pink diamonds are special. R. Weldon, Professional
Jeweler, Vol. 1, No. 11, December 1998, p. 22.
Although Australias Argyle mine is the most consistent
producer of pink and red diamonds, these gems are by no
means plentiful. Only some 700 carats (rough) out of the
millions of carats mined each year from Argyle qualify as
pinks or reds. Fancy pink diamonds were once
reserved for royalty, but now anyone who can afford the
pricesome polished stones sell for hundreds of thousands of dollars per carathas the opportunity to acquire
one. The origin of the pink and red colors is a mystery,
although in the book The Nature of Diamonds, Dr.
Emmanuel Fritsch suggests that the color is due totally
or in part to a color center resulting from deformation.
The supply of pink diamonds is decreasing, in spite of the
recent decision to continue mining at Argyle.
MM
GEM LOCALITIES
Arizona andradite debuts. Jewelers Circular Keystone,
Vol. 169, No. 12, December 1998, p. 28.
Fine-quality andradite garnet, including a limited amount
of demantoid, is being mined near San Carlos, Arizona.
The recently discovered deposits are located near the
well-known peridot mines; they are mined by hand, as
the gems are found near the surface. The andradite ranges
from golden yellow to greenish yellow to reddish brown.
Compared to the highly regarded Russian demantoid, the
Arizona material has lower saturation, attributed to its
lower chromium content, but also higher dispersion.
Similarly, the other colors of andradite from this locality
are not overly saturated and also are highly dispersive.
Thus far, the San Carlos production has yielded only
small faceted gems that have been used as side stones to
accent larger gems.
MM
166
City of pearls. L. Werner, Aramco World, Vol. 49, No. 5,

September/October 1998, pp. 10 19.
Founded in 1591, Hyderabad has been Indias pearl-trading capital for more than 400 years. The city developed
and sustained a robust jewelry industry to support the formidable demands of the wealthy dynastic rulers. From
the early 18th century until 1948, when the last ruler was
forcibly removed from power by the Indian army, untold
quantities of gems and pearls passed through Hyderabads
jewel shops on Patthargatti Road. Although the formerly
important diamond trading center at Golconda is only 12
miles (20 km) away, pearls have left the boldest mark on
Hyderabadi culture and trade over the years. Today it is
the citys pearl merchants who occupy premier positions
in the jewelry industry.
Hyderabad is an unlikely candidate to be at the center
of the pearl trade. The ocean is 200 miles distant and,
commercially speaking, the city is less advanced than,
say, Bombay or Bangalore. One local jeweler attributes
Hyderabads position to the high quality and low cost of
labor. Indeed, behind almost every door on Patthargatti
Road there are pearl sorters, drillers, and stringers, each
with hundreds of years of family experience. This cumulative experience in the pearl trade has firmly established
Hyderabad as Indias city of pearls.
SW
Contribuio ao estudo mineralgico dos cristais de
esmeralda do distrito mineiro de Campos Verdes,
Estado de Gois [Contribution to the mineralogical
study of emerald crystals from the Campos Verdes
mining district, Gois State]. G. M. Pulz, T. M. M.
Brum, P. L. Juchem, L. J. H. DeR. Silva, L. B. Neto,
and P. Barreto, Pesquisas, Vol. 25, No. 2, 1998, pp.
1119 [in Portuguese with English abstract].
Five emerald crystals from the Campos Verdes mining
district in Brazil were analyzed by inductively coupled
plasma mass spectrometry (ICP-MS) and Fourier-transform infrared (FTIR) spectroscopy. The samples contained relatively high Fe (e.g., 1.77 wt.% Fe2O3 and 0.25
wt.% FeO) and high weight loss on ignition (2.924.62
wt.%), and low alkali contents (with Na>>Cs>Li). The
FTIR spectra indicate type-II H2O vibrations. By combining the chemical data with optical and physical properties, and comparing these to published literature, the
authors were able to distinguish these Campos Verdes
emeralds from those of the Swat district (Pakistan),
Franqueira (Spain), Itabira (Brazil), and Colombia, as well
as from synthetic emeralds.
RAH
Compositional variation of tourmaline in the granitic pegmatite dykes of the Cruzeiro mine, Minas Gerais,
Brazil. M. Federico, G. B. Andreozzi, S. Lucchesi,
G. Graziani and J. Csar-Mendes, Canadian Mineralogist, Vol. 36, Part 2, 1998, pp. 415431.
The complex granitic pegmatite dikes of the Cruzeiro
mine contain several species of tourmaline (schorl,
dravite, and elbaite) in well-defined zones. Because of its
GEMS & GEMOLOGY
Summer 1999
sensitivity to the concentrations of Fe, Mg, Al, and Mn in

the melt, tourmaline can be used as a recorder of the internal evolution of granitic pegmatites. Thus, differences in
the chemical compositions of the tourmalines found at
Cruzeiro enabled the authors to reconstruct the various
stages in the crystallization of the pegmatite dikes.
The evolutionary features shown by the compositions
of these tourmalines suggest the following: (1) Al was
abundant in the melt; (2) Mg, Fe, Zn, Mn, and Li each
reached their maximum concentrations at different
stages of pegmatite evolution; and (3) there was a limited
amount of F in the volatile phase and in the melt. The
removal of large amounts of boron from the melt is
responsible for the formation of tourmaline-rich pockets
and the late crystallization of OH-rich tourmaline.
TB
Gem corundums and zircons from the Primorye placers.
S. A. Ananyev, T. A. Ananyeva, V. K. Garanin, and
G. P. Kudryavtseva, Proceedings of the Russian
Mineralogical Society, Vol. 127, No. 4, 1998, pp.
120124 [in Russian with English abstract].
Descriptions are given of gem-quality sapphire and
hyacinth zirconand the associated chrome spinel,
magnetite, and enstatitefound at Primorye in heavymineral concentrates that were ultimately derived from
alkali basalts. Transparent blue, blue-green, green, and
occasional multicolored sapphires are found as rounded
pebbles (2 to 30 mm in size); some star sapphires are
recovered. Chemical analyses and visible absorption
spectra show that the sapphire coloration is related to the
presence of Fe (0.721.75 wt.% FeO) and Ti (0.020.20
wt.% TiO2). Inclusions of hercynite and ferrocolumbite
were identified in some of these corundums.
The zircon also is found as rounded pebbles (1 to 25
mm in size). They are transparent orange, orange-yellow,
or brown-red, and become colorless on heating. Their
chemical compositions correspond to those of the rare
zircon variety naegite, with 4.596.14 wt.% Ta2O5,
3.223.59 wt.% Nb2O5, and 3.984.09 wt.% ThO2.
[Editors note: The terms hyacinth and naegite are not
used in most modern gemology and mineralogy references in English.]
RAH
Gemstones of New England. G. B. Webb and F. L.
Sutherland, Australian Journal of Mineralogy, Vol.
4, No. 2, December 1998, pp. 115121.
Brief details are given of the occurrence of diamond, sapphire, beryl, topaz, quartz, garnet, zircon, feldspar,
rhodonite, nephrite, peridot, and fluorite from this area of
northeastern New South Wales. More than 500,000
carats of diamonds have been produced, and two distinct
categories are recognized: (1) yellow, rounded crystals
with a high, uniform nitrogen content; and (2) white,
internally complex crystals with low, variable nitrogen
content. Most of the diamonds have been substantially
resorbed; with cathodoluminescence, one showed octa-
hedral growth zones on a resorbed surface. Fine blue sapphires are produced with a medium to dark hue, but
greenish blue, yellow, and green varieties also are found;
color zoning down the c-axis is a prominent feature. Beryl
with emerald stripes perpendicular to the c-axis is
illustrated. Colorless (rarely blue) topaz occurs in alluvium derived from the granitic rocks of the area. Zircon
in yellow, brown, and pinkis derived from basalts and
often is encountered in the sapphire workings.
RAH
A note on Buttala and Okkampitiya gem fields in Sri
Lanka. R. Chandrajith, C. B. Dissanayake, and H. J.
Tobschall, Zeitschrift der Deutschen Gemmologischen Gesellschaft, Vol. 47, No. 2, June 1998,
pp. 6976 (in English with German abstract).
The adjacent Buttala and Okkampitiya gem-producing
areas in southeastern Sri Lanka are relatively small and
have received little geologic or mineralogic attention.
Geologically, they are located near the boundary
between two metamorphic terrains, the HighlandSouthwest Complex (which contains most of the gem
deposits in Sri Lanka) and the Vijayan Complex. This
preliminary report indicates their present gem production and future potential.
Both areas produce a wide variety of fine-quality gems,
including blue, yellow, and pink sapphires, as well as
spinel, hessonite and other garnets, and tourmaline.
There is also some geuda (heat-treatable) sapphire. The
gems are mined from residual deposits that are characterized by layers of alternating sands, clays, and laterites
containing large boulders of marble. The authors note
that expansion of mining activities for such residual
deposits has increased the damage to forest reserves on
the island.
MM
Philippine pearl report: The Philippines gathers strength
for 21st century. B. Sheung, Jewellery News Asia,
No. 169, September 1998, pp. 112122 passim.
Mastery of oyster hatchery techniques should allow
Philippine cultured pearl producers to increase their
annual volume from the present estimate of 100150 kan
to more than 200 kan by the year 2000 [1 kan = 1,000
momme; 1 momme = 3.75 grams]. Of the 23 pearl farms
in the Philippines, most of which are in Palawan
Province, 15 are of commercial scale. Neither the number
of farms nor the number of pearls cultivated is restricted
by the government.
The cultivation period for Philippine South Sea pearls
ranges from 18 to 24 months. Sizes range from 9 to 16
mm, with most falling between 10 and 12 mm. Colors
vary from silver, white, cream, champagne to golden
for pearls from white-lipped oysters, and blue to black
for pearls from black-lipped oysters.
A prominent pearl industry figure predicts that South
Sea cultured pearls, including those produced in the
Philippines, will play a major role in the jewelry scene in
the 21st century.
Doug Fiske
GEMS & GEMOLOGY
Summer 1999
167
Polar jade. F. Ward, Lapidary Journal, Vol. 52, No. 8,

November 1998, pp. 2226.
The Canadian province of British Columbia is the worlds
largest supplier of nephrite, and it has huge reserves.
Geologically, British Columbias nephrite is found in
association with serpentinite. At present, the two major
producing mines are Kutcho and Polar, located about 25
miles (40 km) apart east of Dease Lake in the northern
part of the province. Current annual production from
each mine is about 80 tons. Nephrite from each mine is
characteristic, with that from the Polar mine being particularly desirable because it is brighter green (highly saturated) in color, more translucent, and somewhat harder than Kutcho nephrite. Boulders as large as 20 tons
have been found recently at the Polar mine.
Both mines are owned by Kirk Makepeace, who currently produces about 85% of the worlds gem-quality
nephrite. Since entering the business in about 1980, he
has sold approximately 2,400 tons of material. During the
last decade, the main market for nephrite has shifted
from Taiwan to China, where it is used for souvenir carvings and inexpensive jewelry. Rough nephrite currently
sells for a few dollars a pound.
MM
Ruby-bearing mineralization in Sljudjansky complex of
the south-western Baikal region. V. I. Levitsky and
L. A. Pavlova, Proceedings of the Russian
Mineralogical Society, Vol. 127, No. 6, 1998, pp.
84 88 [in Russian with English abstract].
Ruby-bearing rocks are hosted by dolomite marbles in
two separate areas of the Sljudjansky complex in Siberia.
Near the town of Sljudjanka, corundum has been found
locally within zoned magnesian skarns, where it is associated with forsterite, spinel, phlogopite, amphibole, and
calcite. Two varieties of corundum are recognized: (1)
translucent pink and grayish pink, and (2) water-clear
pink and ruby-red. In the Tunkinsky district (Buryat
Republic), corundum-bearing rocks occur as vein-like and
pocket-like bodies with rare disseminated forsterite, red
and pink spinel, and water-clear pink and red corundum. At both localities, the spinel is characterized by
low amounts of iron and appreciable Cr2O3 and V2O3
contents. Electron microprobe analyses are reported
for 10 spinels, five rubies (Cr2O3 0.010.22 wt.%, TiO2
0.010.08 wt.%, V2O5 0.010.15 wt.%), and a single phlogopite sample. In general, the ruby mineralization is
related to postmetamorphic alteration of the precursor
skarns and calc-silicate rocks by deep (mantle) fluids.
RAH
INSTRUMENTS AND TECHNIQUES

Fourier transform-Raman spectroscopy of ivory: A nondestructive diagnostic technique. H. G. M. Edwards,
D. W. Farwell, J. M. Holder, and E. E. Lawson, Studies
in Conservation, Vol. 43, No. 1, 1998, pp. 9-16.
Because of the declining elephant population and the
resultant worldwide ban on trade in elephant ivory, it is
168
essential to distinguish nondestructively between true

elephant ivory and non-elephant ivory. Fourier transformRaman (FT-R) spectra obtained from seven African elephant tusks seized by British Customs in London were
identical, regardless of the color of the ivory or the age of
the elephant from which the tusks were obtained (i.e., the
spectra are not subject to intra- or inter-tusk variations).
Spectra were also obtained for ivory from an Asian elephant and from the extinct woolly mammoth, as well as
for non-ivory samples of human and animal origin (including an adult incisor tooth and an animal bone).
Characteristic vibrational features were found that
enabled each ivory and non-ivory sample to be identified
by its FT-R spectrum. Thus, FT-R spectroscopy is a viable
and nondestructive technique for distinguishing genuine
elephant ivory from specific regions (Africa and Asia) from
other types of ivory (woolly mammoth) and from imitations. At present, this type of identification is not achievable by any other method. For large samples that cannot
be accommodated in the FT-R instrument, a remote sensing probe (1 m long) was adapted successfully.
MM
JEWELRY HISTORY
Fabulous fakes. L. Grant, Antiques & Collecting, Vol.
103, No. 12, February 1999, pp. 38-43.
The use of fakes or costume jewelry as inexpensive alternatives to fine jewelry dates back to the ancient
Egyptians, who first coated glass beads to make them
resemble precious stones. Subsequent milestones include
the production of glass beads in Czechoslovakia in the
14th century and the advent of a clear glass known as
cristallo (crystal) in the 15th century, at a locality near
Venice. In 18th century France, George Stras developed a
high-quality lead glass known as paste or rhinestones, which he cut and set into spectacular jewelry
creations. Late in the 19th century, Swarovski produced
small brilliant-cut crystal stones of exceptional quality; the company remains the premier manufacturer of
crystal used by costume jewelers today. During the same
era, Ren Lalique created imaginative pieces using colored stones and a variety of unusual materials, including
plastic.
The golden era of costume jewelry was between 1930
and 1970, with fashion designers such as Schiaparelli and
Coco Chanel developing their own styles. Whereas
Schiaparelli captured the market with unexpected and
humorous pieces (including earrings shaped like telephones and bracelets made with porcelain vegetables),
Chanels fashions reflected timelessness and elegance.
She broke all conventions by mixing real gem materials
with fake ones and made it acceptable to wear jewelry on
informal occasions. Today, Coco Chanels stylish designs
continue to look contemporary, and they are widely
sought by collectors. Other designers who have embraced
costume jewelry include Miriam Haskell, Cartier, Trifari,
Christian Dior, Yves Saint Laurent, and Donna Karan.
While costume jewelry was first accepted in Europe, it
GEMS & GEMOLOGY
Summer 1999
became a thriving industry in the U.S., where it was sold

at boutiques and department stores. Vintage costume
jewelry pieces produced during the 20th century are now
collectibles, although early examples of Schiaparelli,
Chanel, and Haskell are difficult to find.
MD
Ren Lalique, Art Nouveau genius. C. L. E. Benesh,
Ornament, Vol. 22, No. 2, Winter 1998, pp. 7174.
Ren Lalique (18601945) was one of the worlds greatest
Art Nouveau jewelry designers. This article chronicles
the development of Laliques career, as well as the Art
Nouveau artistic movement. This design style of the late
19th century was characterized especially by winding
lines and foliate forms.
After an apprenticeship to a Paris goldsmith at the age
of 16 and additional studies in London, Lalique returned
to Paris as a designer and freelance illustrator. His creativity exploded in 1885, when he purchased his own studio and progressed from design to jewelry making. Within
a few years, his creationswhich emphasized nature and
femininity were being avidly collected by wealthy aristocrats and celebrities. Laliques stature as a great artist
was sealed after an astonishing display at the Universal
Exhibition in Paris in 1900. Because his pieces emphasized the beauty of the design over the intrinsic value of
the materials in which the design was executed, he championed the use of inexpensive gem materials such as
amber, ivory, horn, opal, and carnelian. He also made
exquisite use of glass and enamel. Art Nouveau achieved
prominence in the 15-year period from 1895 to 1910. In
1912, Lalique abandoned jewelry altogether and devoted
himself entirely to glass.
MM
JEWELRY MANUFACTURING
Thermal and microchemical characterisations of CaSO4SiO2 investment materials for casting jewellery alloys. G. M. Ingo, G. Chiozzini, V. Faccenda,
E. Bemporad, and C. Riccucci, Thermochimica
Acta, Vol. 321, No. 1/2, 1998, pp. 175183.
The decomposition of calcium sulfate in the commonly
used calcium sulfate-silica (CaSO4-SiO2) investment (an
investment is a material, such as plaster of paris, that
is used in casting gold-based alloys in jewelry) method
can cause what is known as the gas porosity defect. This
defect, seen in gold alloy jewelry as tiny surface pores, can
reduce the value of the piece. The authors simulated
numerous temperature and atmosphere (e.g., air, argon)
casting conditions to assess the potential for CaSO4
decomposition.
The decomposition temperature of pure anhydrous
CaSO4 determined in this study ranged from 1240 to
1450C. In conjunction with silica, the decomposition
temperature was lowered to a range of 990 to 1260C.
The decomposition temperature is lowered further in
reducing or inert atmospheres, or with oxides of other
alloy metals, especially Cu2O and ZnO. Since the melting
temperature of commercial gold alloys ranges from 960

to 1200C, the authors suggest that decomposition of
CaSO4 in the CaSO4-SiO2 investment method is a valid
concern.
JL
JEWELRY RETAILING
Consumers seek gems that change colors. C. Fenelle,
National Jeweler, Vol. 42, No. 22, November 16,
1998, pp. 16, 18.
Color-change gemstones are experiencing an increase in
popularity, especially as color-change synthetics become
more available. The most popular are alexandrite, sapphire, and garnet. The most valuable color-change stones
are those in which the change is to a completely unrelated color, such as from red to blue or green, when
viewed with fluorescent (sun) versus incandescent light.
The attractiveness and saturation of each color are also
factors that should be considered. Many gemstones commonly show a slight difference in color when examined
under different lighting conditions, an effect known as
metamerism. Such stones have been labeled as colorshift, but they do not justify a premium over their regular gem counterparts.
MD
Designing a successful business. M. Z. Epstein, American
Jewelry Manufacturer, Vol. 43, No. 6, June 1998,
pp. 4043, 4548.
Jewelry designers may get noticed for their artistic talent,
but their companies grow only when they combine artistic skills with financial planning and acumen. This article presents seven ways to turn a design shop into a successful business:
1. Be nice to people. The importance of interactions
with manufacturers, other designers, and retailers is often
underestimated. As much as their jewelry, it is the artists
themselves who are on display anywhere they meet other
professionals or the public.
2. Do your market research. Knowing what customers
want is critical. A large component of market research is
determining customer demographics. Such information
can help you choose the materials for designs, and possibly the way you design.
3. Pinpoint profitable price points. Jewelry has to be
sold at a profit, which means that the pricing must be correct. This requires accurate knowledge of material, manufacturing, and overhead costs. The suggested retail selling price of custom jewelry is 1.52.5 times total costs;
for designer jewelry, the factor is 2.2 3.0.
4. Give your jewelry wow power. Educating retailers
to be enthusiastic about your jewelry is necessary to
increase sales. This is accomplished in several ways, such
as using a combination of rare stones in designs, or creating pieces with cultural significance.
5. Build a cohesive image. Name recognition is crucial
for developing a cohesive image. Press releases are a good
first step. Establishing a working relationship with editors of trade and consumer publications, attending trade
GEMS & GEMOLOGY
Summer 1999
169
shows, and using direct mail for niche markets are three
more ways to build an image.
6. Hire and keep good employees. Hiring the best people, treating them well, and giving them the skills they
need (whether it be order filling or selling) will be profitable in the long term.
7. Develop a business plan. Financial consultants can be
of great value in assisting with a business plan. There is
no substitute for professional advice from someone experienced in business, especially your business.
Lorrie Ames
Fixed flaws. D. A. Yonick, American Jewelry Manufacturer, Vol. 43, No. 9, September 1998, pp. 6269,
71, 73.
Todays gemstone treatments enhance clarity, increase
durability, reduce porosity, improve and stabilize color,
and produce colors not found in nature. There is nothing
wrong with using treatments to improve on or fix what
nature has produced, provided consumers are advised of
the enhancements and their stability/durability characteristics. This article admonishes jewelry manufacturers
to require full disclosure of treatments from their suppliers and to pass this information along to retailers, who
ultimately inform the consumers.
To help the manufacturer understand the most common treatments being used today, this article explains
the following terms: heat treatment, irradiation, diffusion
treatment, fracture filling, infilling, stabilization, oiling,
impregnation, infusion, lasering, bleaching, dyeing, and
coating. Examples of each process are given, along with
the gemstones that are likely to be enhanced by that
process, the improvement that usually results, and the
stability or durability of each enhancement.
MD
Technology and the retail jeweller. J. Lawrence, Diamond
International, No. 55, September/October 1998,
pp. 6768.
Many consumers are confused by the subtle differences in
the 4Cs that affect the price of diamonds. To explain
these differences and instill confidence in the jewelers
explanations, affordable instrumentation is readily available to jewelers. In the hands of trained staff, these instruments are very effective.
The microscope is the essential instrument, and
recently it has become especially important for reading
laser-inscribed certificate numbers. A TV camera attachment enables the retailer to explain clarity, at least in
qualitative terms. Effective for unmounted stones are TV
image processing systems, such as the Brilliant-Eye
produced by Sarin, which measure the proportions of a
polished stone and compute its weight to within 0.01 ct
(1 pt). Some of them also have the ability to print a customized certificate with a proposed cut grade. Color also
can be graded by attaching a colorimeter to a port on some
proportion machines. It is reported that such systems are
selling at a rate of more than one a day in the U.S.
170
These instruments appear to have great appeal to the

public because seeing is believing. From the jewelers
perspective, they can support or verify a certificate at the
retail counter.
AAL
12 hot tips: The inside scoop on what consumers want to
see. R. Weldon, Professional Jeweler, Vol. 1, No. 11,
December 1998, p. 87.
A prominent New York jeweler predicts increased
demand for: (1) yellow gold jewelry; (2) unusual colored
stones (e.g., fire opals and brown- or gray-colored gems to
harmonize with clothing fashions); (3) fancy colored diamondsespecially pinks and yellows; (4) bold and geometric shapesparticularly chunky pieces in yellow
gold; (5) fancy cut diamonds (e.g., oval, emerald, princess,
star); (6) solitaire diamond jewelry; (7) pearlsespecially
cultured Tahitian black pearls and pearl solitaires; (8)
antique appearance in jewelryspecifically, delicate and
feminine pieces in platinum and white gold set with diamonds and colored gems; (9) romantic and colorful
nature-inspired pieces (e.g., dragonflies and butterflies);
(10) well-made charms accented with diamonds; (11) diamonds, pearls and other gems set in white or yellow gold,
and suspended from clear invisible filaments; and (12)
wrap necklaces with exotic colored gems (e.g., apatite and
tanzanite).
MM
PRECIOUS METALS
A sterling moment for silver. C. Soucy, National Jeweler,
Vol. 42, No. 13, July 1, 1998, pp. 28, 30, 3233.
The growing popularity of white metals has stimulated
the sales of platinum, white gold, and sterling silver. Sixty
million pieces of silver jewelry were sold in 1997. A longtime favorite in boutiques and fashion-oriented retail
stores, sterling silver jewelry now appears alongside platinum and gold jewelry at the high-end jewelry stores.
Lower-priced pieces targeted to the teen market are also
selling well. The hottest item in white metals is neckwearespecially chokersfollowed by earrings, rings,
and bracelets. Silver is easily worked into intricate filigree
and cut-out designs, and it offers the option of using its
oxidation characteristics in the creation of designs.
Because it is lighter than platinum or gold, larger pieces
can be designed (especially earrings), without the added
weight.
MD
SYNTHETICS AND SIMULANTS

China company produces hydrothermal rubies. Jewellery
News Asia, No. 169, September 1998, p. 50.
Hydrothermal synthetic rubies marketed under the name
JL synthetic rubies are available from Guangxi Jewelry
Co. Ltd. in Guilin, China (Guangxi Province). The material ranges from red to orange-red in color, with some displaying purple hues. The synthetic rubies fluoresce
strong red to long-wave ultraviolet radiation; the fluores-
GEMS & GEMOLOGY
Summer 1999
cence is less intense to short-wave UV. Colorless netlike inclusions are reported. The synthetic rubies have
R.I. values of 1.7591.769, and an S.G. of 4.01. They are
incorporated into gold rings, pendants, earrings, and
bracelets.
MM
Growth conditions and real structure of synthetic diamond crystals. Y. N. Palyanov, A. F. Khokhryakov,
Y. M. Borzdov, A. G. Sokol, V. A. Gusev, G. M.
Rylov, and N. V. Sobolev, Russian Geology and
Geophysics, Vol. 38, No. 5, 1997, pp. 920945.
This comprehensive article describes the growth of synthetic diamond crystals at the Institute of Mineralogy and
Petrography in Novosibirsk, Russia. The authors are leading researchers in the area of diamond synthesis at high
temperatures and pressures using the multianvil apparatus (also known as the split sphere or BARS apparatus). Synthetic diamond crystals have been produced at
this institute since the early 1990s.
The article begins with a review of the ways to grow
single-crystal synthetic diamonds, and explains details of
the growth of crystals from a molten metal alloy flux by
the temperature-gradient method. The types of inclusions and other kinds of growth defects, as well as the
crystal shapes and surface features of these synthetic diamonds, are described and photographically documented.
Last, the article reviews the physical conditions under
which the various diamond types (e.g., type Ib, type IIa,
etc.) are produced. An extensive reference list is provided.
James E. Shigley
TREATMENTS
Behandlung natrlicher Diamanten zur Reduzierung der
Gelb- oder Braunsttigung (Treatment of natural
diamonds in order to reduce the yellow or brown
color saturation). K. Schmetzer, Goldschmiede
Zeitung, Vol. 97, No. 5, May 1999, pp. 4748 [in
German].
This brief article, written for a German trade journal,
identifies nine patents that involve changing the color of
diamond. These U.S. patents (4,124,690; 4,174,380;
4,399,364; 5,523,071), as well as German (2 732 793),
European (0 014 528; 0 638 670; 0 668 377), and Japanese
(63-291895) published patent applications, all available
publicly, are from the period 19781996, and involve
research work on both natural and synthetic diamonds.
They were filed by General Electric Company, De Beers
Industrial Diamond Division (Pty.) Ltd., and Sumitomo
Electric Industries Ltd. The General Electric patents
describe treatments at high pressure and high temperature, whereas the De Beers and the Sumitomo patents
describe two-step treatment processes: (1) irradiation and
(2) high-pressure and high-temperature annealing. The
experimental conditions outlined in these documents
show that atomically dispersed nitrogen atoms (i.e., type
Ib) are forced to migrate and form clusters or aggregates
(i.e., type Ia), thereby reducing the influence of the colorcausing (yellow or brownish yellow) type Ib nitrogen
atoms. Although other decoloration mechanisms (e.g., for
brown diamonds) are imaginable, they are not discussed
in these patent documents.
This article is a response to the recent announcement
by Lazare Kaplan International Inc. of a new General
Electric process that improves the appearance of certain (presumably brown) natural diamonds. This
announcement, along with uncertainties regarding the
identification and disclosure of GE-processed diamonds,
have produced great concerns in the jewelry trade. While
GE has not revealed the details of their process, the
patents cited here suggest that various kinds of color
change can be produced by treating natural diamonds at
high temperatures and pressures.
James E. Shigley
Belgium diamond report: Enhancing low quality coloured
diamonds. Jewellery News Asia, No. 169, September 1998, pp. 320, 322.
Russian scientists are using BARS-autoclave technology
to enhance the color of low-quality, fancy-color diamonds. The treatment involves hyperbaric heating (i.e.,
at pressures high enough to prevent conversion from diamond to graphite). The effect of the heating on color
depends on two factors: the original color of the diamond
and the duration of treatment, which determines the
number of former color centers that are destroyed and
replaced by new ones.
Characteristics of these treated fancy-colored diamonds include: a greenish overcast effect visible in daylight; a strong green fluorescence related to growth lines;
color zoning; greenish, yellowish, or brownish body color;
prominent graining; cracks and cleavages at the rim of the
table and around the girdle; and corrosion marks, particularly around the girdle. Photos of treated diamonds
showing many of these characteristics are provided.
MM
Ein neuer Typ farbbehandelter Diamanten (A new type of
colour-treated diamonds). U. Henn and C. C. Milisenda, Zeitschrift der Deutschen Gemmologischen
Gesellschaft, Vol. 48, No. 1, 1999, pp. 43-45 [in
German with English abstract].
A new type of color-treated greenish yellow to brownish
yellow fancy-color diamond, with a saturated yellow
body color and a strong green fluorescence partly related
to growth lines, is described. Other identifying features
seen under magnification include strong graining, fine
cracks and cleavages under the table and at the girdle, as
well as burn marks. The absorption spectrum shows lines
at 415, 495, 503, and 985 nm, and a broad band between
450 and 500 nm; this spectrum is distinctly different
from natural colored or irradiated diamonds of these colors. An important feature of these heat-treated diamonds
is their fluorescence to long-wave ultraviolet radiation.
Natural-color greenish yellow diamonds have a chalky
GEMS & GEMOLOGY
Summer 1999
171
fluorescence, whereas these diamonds show a mixture of

greenish yellow and blue fluorescence, with distinct
green fluorescing zones following growth lines.
These diamonds were initially offered as Russian treated synthetic diamonds, but a closer investigation revealed
that they are natural diamonds that obtained their
enhanced color through a special heat treatment process
at high temperature and pressure that uses the Russian
BARS system (a belt-type apparatus normally used in the
production of synthetic diamond). The raw material consists mainly of low-quality diamonds from the Argyle
mine (Australia), but diamonds from Russian and Central
African sources are used as well. [Editors note: Similar
color-treated yellow diamonds were described in the
Summer 1997 Gem Trade Lab Notes, p. 136.]
PRB
Observation of the H2 defect in gem-quality type Ia diamond. P. R. Buerki, I. M. Reinitz, S. Muhlmeister,
and S. Elen, Diamond and Related Materials, Vol.
8, No. 6, 1999, pp. 10611066.
The authors define H2 diamonds as diamonds with a
986 nm absorption band. They typically have a greenish
yellow, yellow, or brownish orangy yellow color and
show strong green luminescence to visible light, associated with brownish yellow graining, and burns on the
surface and in fractures. Other absorption bands include
H3 (503 nm) and H4 (496 nm). The vast majority of H2
diamonds are treated, having been produced from natural
type Ia diamonds (probably brownish hues) by a new
process that most likely consists of irradiation with
high-energy electrons, followed by high-temperature
(about 1,400C) heating in air, possibly at a stabilizing
pressure of about 5 GPa. About 50 such diamonds were
received at the GIA Gem Trade Laboratory (GTL) in a
one-month period in 1996, and since then another 100
have been seen and characterized at GTL. Natural type Ia
diamonds with the same or similar colors and luminescence and a 986 nm (H2) band are suspected to be very
rare (only 11 had been identified by GTL in the previous
12 years) and are possibly formed when type Ia diamonds
are subjected to natural radiation and temperatures
above 1,000C. The distinction between diamond with
natural H2 characteristics from those described herein is
extremely difficult, and is dependent on such features as
a very low intensity 986 nm band, or evidence of laboratory treatment (e.g., burn marks, certain IR absorption
bands).
Out of a total of 510 greenish yellow diamonds with
green luminescence examined at GTL in the last 12 years,
90 showed all the typical characteristics of H2 diamonds
except the H2 band. This raises the question: Does a
172
method exist to destroy the H2 center after its formation?

If possible, such a method would have major implications
for determining whether any specific diamond owes its
color to natural or artificial processes. [Editors note:
Many questions about color-treated greenish yellow diamonds remain unanswered. The previous two abstracts
report on similar (probably identical) material and illustrate the importance of this topic.]
AAL
MISCELLANEOUS
Reach for the sky. M. Z. Epstein, American Jewelry
Manufacturer, Vol. 43, No. 10, October 1998, pp.
5765.
Jewelry manufacturers, like other workers subject to
repetitive stress injuries, are aggressively taking steps to
apply the science of ergonomics (the study of human abilities and characteristics) to eliminate workplace injuries.
This article describes nine steps to improve the ergonomic efficiency of the jewelry manufacturing environment,
where there is high risk for repetitive stress injuries:
acknowledge the problem(s), evaluate work positions,
analyze tool design, improve workspace layout, check
environmental conditions, develop ergonomic solutions,
train workers in safety, perform ergonomic checkups and,
if necessary, seek out ergonomics experts. Companies
that have taken proactive measures have benefited from
fewer injuries and illnesses. For instance, one company
found that simply providing employees with $5 wrist
supports reduced the incidence of carpal tunnel syndrome. A list of body parts and the repetitive motions
that can injure them is included.
JEM-S
The right image. G. Dawson, American Jewelry
Manufacturer, Vol. 43, No. 5, May 1998, pp. 9296.
Photographing jewelry and gemstones is especially difficult because of their highly reflective surfaces and the
problems associated with capturing subtle colors. To overcome these challenges, the author (a goldsmith-photographer) has developed several photographic techniques. He
recommends using interchangeable lenses and a camera
body that allows complete manual operation. A long
macro lens permits a greater working distance, which is
useful in avoiding reflections from metals or gemstones. A
diffusion tent can also be helpful in this regard. Another
suggestion is to use an off-camera flash with a warming
filter to bring out the richness of gold. A light meter is
essential, and several exposures bracketing the meter reading in half-stop increments should be taken. A multitude
of shots of the same object, with subtle variations in lighting or angle, is recommended for important pieces. MD
GEMS & GEMOLOGY
Summer 1999

SU99

Uploaded by

Copyright:

Available Formats

SU99

Uploaded by

Document Information

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

SU99

Uploaded by

Copyright:

Available Formats

SUMMER 1999

Nothing Is as Simple as It Seems

On the Identification of Various Emerald Filling Substances

Sapphire and Garnet from Kalalani,

Russian Synthetic Ametrine

1999 Challenge Winners

All rights reserved.

omeone asked me recently if gemologists really used

GEMS & GEMOLOGY

Heat Exposure May Affect Reflectance

faceted synthetic moissanites on three standard

Figure 1. The surface reflectance of synthetic

GEMS & GEMOLOGY

Comments on the Sources

tested as diamond with a thermal conductivity probe,

I read with great interest the article Characterizing

Mark Kellam, Ph.D.

Dr. A. J. A. (Bram) Janse

years, Dr. Manson was curator of gems and minerals at

GEMS & GEMOLOGY

ABOUT THE AUTHORS

Emerald Filling Substances

he vast majority of emeralds on the market today

GEMS & GEMOLOGY

Figure 1. The desire for

1995). In 1997, ICA decided to delay an emerald

Emerald Filling Substances

GEMS & GEMOLOGY

ticular emerald was first noted in October 1993

Emerald Filling Substances

stances at the present time are cedarwood oil,

GEMS & GEMOLOGY

Emerald Filling Substances

GEMS & GEMOLOGY

BOX A: CLASSIFICATION OF FILLING SUBSTANCES:

are extracted from their host plants using solvents.

Other Oils. Because essential oils differ from

MATERIALS AND METHODS

Emerald Filling Substances

ChemFinder, is available at www.chemfinder.com.)

GEMS & GEMOLOGY

tial oil cedarwood oil is considered a resin by the

that was made from other chemicals, regardless of

Figure A-1. Snap-Lock beads illustrate the

and lungs; some can cause mutations in dividing

Emerald Filling Substances

Filling Substances. For this study, we chose fillers

GEMS & GEMOLOGY

Box B: CEDARWOOD OIL,

fit into substance categories commonly used as

Emerald Filling Substances

We inquired about the composition of each

GEMS & GEMOLOGY

are detectable within emeralds, we do not yet know

Colombian emerald laboratories began using

terms and in the ways these terms are applied by

Emerald Filling Substances

lowing six substance categories: essential oils

GEMS & GEMOLOGY

sives), and (hardened) polymer resins. The first three