SP09

GEMS & GEMOLOGY
SPRING 2009
VOLUME XLV
SPRING 2009
PAGES 1 78
The French Blue and the Hope

Gray-to-Blue-to-Violet
Argyle Diamonds
VOLUME 45 NO. 1
Hackmanite
THE QUARTERLY JOURNAL OF THE GEMOLOGICAL INSTITUTE OF A MERICA
OUR EDUCATION.
YOUR WORLD OF OPPORTUNITY.
Because Public Education

Happens at the Counter.
GIA launches Retailer Support Kit and website
SEOUL 8 : 0 0 P M
LONDON
NOON
GIA alumni network at cultured pearl seminar.
GIA-trained jeweler
advises client on 5 carat solitaire.
NEW YORK
7:00 AM
Diamonds Graduate negotiates

purchase of rough parcel.
TOKYO 8 : 0 0 P M
CARLSBAD
4:00 AM
Student completes gem ID project.
MUMBAI 4 : 3 0 P M
Core gem curriculum updated

to reect new research.
Sales associate explains 4Cs to customer.
HONG KONG 7 : 0 0 P M
Manufacturing exec expands business skills online.
BANGKOK 6 : 0 0 P M
Graduate Gemologist spots treated emeralds in bulk order.
A $97.00 value, shipping

and handling extra.
GIAs Retailer Support Kit has been developed to help

Almost anywhere you go, someone is using education acquired from GIA. Our international campuses, traveling classes,
sales associates educate the public about diamonds,
corporate seminars and online courses help individuals dene and rene vital skills.
the 4Cs, and thoroughly explain a GIA grading report.
And GIA supports that learning with credentials valued throughout the gem and jewelry world.
Take full advantage of all that GIA has to offer by visiting
W W W. G I A . E D U
www.retailer.gia.edu
To order your FREE kit, log on to www.retailer.gia.edu
CARLSBAD
MOSCOW
NEW YORK
MUMBAI
LONDON
BANGKOK
ANTWERP
HONG KONG
FLORENCE
BEIJING
GABORONE
TAIPEI
SEOUL
JOHANNESBURG
OSAKA
TOKYO
GMSGG
RTGG09
Spring 2009
Volume 45, No. 1
EDITORIAL ________
1
The Dr. Edward J. Gbelin Most Valuable Article Award
FEATURE ARTICLES _____________

4
Carat Points
The French Blue and the Hope: New Data from the Discovery of a
Historical Lead Cast
Franois Farges, Scott Sucher, Herbert Horovitz, and Jean-Marc Fourcault
Computer modeling of a recently discovered lead cast of the French Blue diamond
reveals important details about this fabled gem.
pg. 5
20
Gray-to-Blue-to-Violet Hydrogen-Rich Diamonds from

The Argyle Mine, Australia
Carolyn H. van der Bogert, Christopher P. Smith, Thomas Hainschwang, and Shane F. McClure
Famed for its pink and champagne diamonds, Argyle is the only known source
of the rare type IaB hydrogen-rich diamonds colored gray to blue to violet.
NOTES AND NEW TECHNIQUES ________

38
Hackmanite/Sodalite from Myanmar and Afghanistan

David Kondo and Donna Beaton
Samples from two new sources help characterize the tenebrescent gem hackmanite.
pg. 21
RAPID COMMUNICATIONS _____________

44
Solution-Generated Pink Color Surrounding Growth Tubes and Cracks in

Blue to Blue-Green Copper-Bearing Tourmalines from Mozambique
John I. Koivula, Kevin Nagle, Andy Hsi-Tien Shen, and Philip Owens
48
pg. 39
Identification of the Endangered Pink-to-Red Stylaster Corals

By Raman Spectroscopy
Stefanos Karampelas, Emmanuel Fritsch, Benjamin Rondeau,
Aude Andouche, and Bernard Mtivier
REGULAR FEATURES ________________________________________

53
Lab Notes
Coated CZ Mineral assemblages in diamond etch channels Clarity grading
radiation stains Rare mixed type (Ia/IIb) diamond with N and B centers
Purplish pink spinel from Tajikistan
59
Gem News International

Tucson 2009 Opal from Welo, Ethiopia Rhodochrosite from China Amethyst
from Morocco Update on Bolivian ametrine Azurite/malachite from Mexico
Light yellow-green grossular from Kenya Kornerupine from Tanzania
Labradorite from Alaska Myanmar update Natural and synthetic spinels
Zoisite from Afghanistan Australian chrysoprase with dendritic inclusions
Ankangite and celsian inclusions in Brazilian quartz Paraba quartz with
copper inclusions Color-change glass Unusual dyed imitations
76
2009 Gems & Gemology Challenge
S1
Book Reviews
S4
Gemological Abstracts
pg. 64
www.gia.edu/gemsandgemology
EDITORIAL
STAFF
Editor-in-Chief
Alice S. Keller
akeller@gia.edu
Editor
Brendan M. Laurs
GIA, The Robert Mouawad Campus
5345 Armada Drive
Carlsbad, CA 92008
(760) 603-4503
blaurs@gia.edu
Managing Editor
Thomas W. Overton
tom.overton@gia.edu
Technical Editor
Emily V. Dubinsky
emily.dubinsky@gia.edu
Circulation Coordinator
Mary Navarro
(760) 603-4000, ext. 7142
mary.navarro@gia.edu
Contributing Editor
James E. Shigley
PRODUCTION
STAFF
EDITORIAL
REVIEW BOARD
Art Director
Karen Myers
Robert E. Kane
Helena, Montana
Kenneth Scarratt
Bangkok, Thailand
Edward W. Boehm
Solana Beach,
California
Jaroslav Hyr`sl
Prague, Czech Republic
Lore Kiefert
New York, New York
A. J. A. (Bram) Janse
Perth, Australia
Michael Krzemnicki
Basel, Switzerland
James E. Shigley
Carlsbad,
California
Alan Jobbins
Caterham, United
Kingdom
Thomas M. Moses
New York, New York
John Emmett
Brush Prairie,
Washington
MANUSCRIPT
SUBMISSIONS
COPYRIGHT
AND REPRINT
PERMISSIONS
THE
ABOUT
COVER
G&G Online:
gia.metapress.com
Emmanuel Fritsch
Nantes, France
Alan T. Collins
London, United
Kingdom
DATABASE
COVERAGE
Editors, Gemological Abstracts

Brendan M. Laurs
Thomas W. Overton
Shigeru Akamatsu
Tokyo, Japan
James E. Butler
Washington, DC
SUBSCRIPTIONS
Editor, Gem News International

Brendan M. Laurs
Editors, Book Reviews
Susan B. Johnson
Jana E. Miyahira-Smith
Thomas W. Overton
Associate Editor
Stuart D. Overlin
soverlin@gia.edu
Consulting Editor
Carol M. Stockton
Editors, Lab Notes

Thomas M. Moses
Shane F. McClure
Mary L. Johnson
San Diego, California
Mark Newton
Coventry, United
Kingdom
Anthony R. Kampf
Los Angeles, California
George Rossman
Pasadena, California
Christopher P. Smith
New York, New York
Christopher M.
Welbourn
Reading, United
Kingdom
Copies of the current issue may be purchased for $19.00 in the U.S., $22.00 elsewhere. Online subscriptions, and print subscriptions sent to addresses in the U.S., are $74.95 for one year (4 issues), $194.95 for three years (12 issues). Print subscriptions sent
elsewhere are $85.00 for one year, $225.00 for three years. Combination print/online subscriptions are $99.95 in the U.S. and
$110.00 elsewhere for one year, and $269.95 in the U.S. and $300.00 elsewhere for three years. Canadian subscribers should add
GST. Discounts are available for renewals, GIA alumni, and current GIA students.
To purchase subscriptions and single print issues, visit www.gia.edu/gemsandgemology or contact the Circulation Coordinator.
Electronic (PDF) versions of all articles and sections from Spring 1981 forward can be purchased at gia.metapress.com for $10
each. Full issue access can be purchased for $20.
To obtain a Japanese translation of Gems & Gemology, contact GIA Japan, Okachimachi Cy Bldg., 5-15-14 Ueno,
Taitoku, Tokyo 110, Japan. Our Canadian goods and service registration number is 126142892RT.
Gems & Gemologys impact factor is 1.227 (ranking 11th out of the 26 journals in the Mineralogy category), according to
Thomson Scientifics 2007 Journal Citation Reports (issued July 2008). Gems & Gemology is abstracted in Thompson
Scientific products (Current Contents: Physical, Chemical & Earth Sciences and Science Citation IndexExpanded, including the Web of Knowledge) and other databases. For a complete list, see www.gia.edu/gemsandgemology.
Gems & Gemology welcomes the submission of articles on all aspects of the field. Please see the Guidelines for Authors on
our Website, or contact the Managing Editor. Letters on articles published in Gems & Gemology are also welcome.
Abstracting is permitted with credit to the source. Libraries are permitted to photocopy beyond the limits of U.S. copyright law
for private use of patrons. Instructors are permitted to photocopy isolated articles for noncommercial classroom use without
fee. Copying of the photographs by any means other than traditional photocopying techniques (Xerox, etc.) is prohibited without the express permission of the photographer (where listed) or author of the article in which the photo appears (where no photographer is listed). For other copying, reprint, or republication permission, please contact the Managing Editor.
Gems & Gemology is published quarterly by the Gemological Institute of America, a nonprofit educational organization for the gem and jewelry industry, The Robert Mouawad Campus, 5345 Armada Drive, Carlsbad, CA 92008.
Postmaster: Return undeliverable copies of Gems & Gemology to GIA, The Robert Mouawad Campus, 5345 Armada Drive,
Carlsbad, CA 92008.
Any opinions expressed in signed articles are understood to be the opinions of the authors and not of the publisher.
It has long been believed that the Hope diamond was cut from the French Blue, which disappeared in 1792. A cast of the
French Blue was recently discovered in the Musum National dHistoire Naturelle in Paris. In the lead article, the authors
used the cast to create a computer model of the fabled diamond that sheds new light on the HopeFrench Blue connection.
Shown here are the diamond-set Hope, a cubic zirconia model of the French Blue, and a rendering of Louis XVs Golden
Fleece of the Colored Adornment, in which the French Blue was mounted. Drawing by Pierre-Andr Jacquemin, dated after
1749; photo of the French Blue model Franois Farges/MNHN; photo of the Hope Harold & Erica Van Pelt.
Color separations for Gems & Gemology are by Pacific Plus, Carlsbad, California.
Printing is by Allen Press, Lawrence, Kansas.
2009 Gemological Institute of America
All rights reserved.
ISSN 0016-626X
As we celebrate Gems & Gemologys

75th anniversary in 2009, were pleased to start
another year by announcing the winners of the annual
Dr. Edward J. Gbelin Most Valuable Article Award. By
recognizing excellence in feature articles, these awards
carry on the journals mission, as set forth in the January
1934 premier issue: to give our readers accurate and upto-date information concerning gemstones.
This year marks the first time we opened the competition
to online voting, and we received almost 250 ballots from
subscribers around the world. We extend our sincerest
thanks to everyone who participated.
The first-place article was Copper-Bearing (Parabatype) Tourmaline from Mozambique (Spring 2008),
which described the geology, mining, and properties of
this sought-after gem. Placing second was Color
Grading D-To-Z Diamonds at the GIA Laboratory
(Winter 2008), which examined the background and
methodology of GIAs system for color grading colorless
to light yellow polished diamonds. Third place went to
A History of Diamond Treatments (Spring 2008), a
review of the history, development, and identification of
diamond color and clarity enhancement techniques.
COPPER-BEARING (PARABA-TYPE) TOURMALINE

FROM MOZAMBIQUE
Brendan M. Laurs, J. C. (Hanco) Zwaan, Christopher M. Breeding, William B. Skip
Simmons, Donna Beaton, Kenneth F. Rijsdijk, Riccardo Befi, and Alexander U. Falster
Brendan M. Laurs is editor of Gems & Gemology and its Gem News International section. A widely published author, he has explored numerous gem
localities in Africa, Pakistan, and Brazil. Mr. Laurs holds a masters degree in
geology from Oregon State University. J. C. Hanco Zwaan is curator at the
National Museum of Natural History (Naturalis) and director of the Netherlands Gemmological Laboratory in Leiden. Dr. Zwaan has a PhD in geology
from the Free University in Amsterdam. Christopher M. Breeding is a
research scientist for the GIA Laboratory in Carlsbad, where he investigates
origin of color in diamond and other gems. Dr. Breeding holds a PhD in geology from Yale University. William B. Skip Simmons is director of the
Mineralogy, Petrology, and Pegmatology (MP 2) Research Group in the
Department of Earth and Environmental Sciences at the University of New
Orleans, and an adjunct professor at the University of Michigan. Dr. Simmons
received his PhD from the University of Michigan and has over three decades
MOST VALUABLE ARTICLE AWARD
GEMS & GEMOLOGY
Brendan M. Laurs
J. C. Hanco Zwaan
Christopher M. Breeding
William B. Skip Simmons
SPRING 2009
Donna Beaton
Kenneth F. Rijsdijk
Riccardo Befi
Alexander U. Falster
of research experience in mineralogy and petrology. Donna Beaton is manager of

colored stone services at the GIA Laboratory in New York. She has a masters
from Columbia University. Her background includes jewelry retail, auction, and
appraisal work. Kenneth F. Rijsdijk is a geoscientist at the National Museum of
Natural History (Naturalis) and coordinator of its Dodo Research Programme. Dr.
Rijsdijk has a PhD in physical geography from the University of Wales, Swansea.
Riccardo Befi is a staff gemologist at the GIA Laboratory in New York. A graduate of the University of Siena, Italy, Mr. Befi has more than 20 years of experience
in diamond grading and gem identification. Alexander U. Falster is a scientific
research technologist in the Department of Earth and Environmental Sciences at
the University of New Orleans. He is part of the MP 2 Research Group and specializes in pegmatites and their minerals.
COLOR GRADING D-TO-Z DIAMONDS
AT THE
GIA LABORATORY
John M. King, Ron H. Geurts, Al M. Gilbertson, and James E. Shigley
John M. King
Ron H. Geurts
Al M. Gilbertson
James E. Shigley
John M. King is chief quality officer at the GIA Laboratory in New York and the
editor of Gems & Gemology in Review: Colored Diamonds. Mr. King, who is
also a noted artist, received his Master of Fine Arts degree from Hunter College,
City University of New York. Ron H. Geurts is research and development manager at GIA Belgium in Antwerp. Formerly with the HRD Laboratory in
Antwerp, Mr. Geurts has implemented new technology in GIAs diamond grading process for the last decade and contributed to the development of the
Institutes cut grading system. Al M. Gilbertson is a research associate at the
GIA Laboratory in Carlsbad. A former cutter and appraiser, he has spent years
studying the influence of proportions on the appearance of round-brilliant and
fancy-shaped diamonds. Mr. Gilbertson is the author of American Cut: The First
100 Years (2007). James E. Shigley is distinguished research fellow at the GIA
Laboratory in Carlsbad. The editor of the Gems & Gemology in Review series
and contributing editor to the journal, he received his doctorate in geology from
Stanford University.
A HISTORY OF DIAMOND TREATMENTS

Thomas W. Overton and James E. Shigley
Thomas W. Overton
Thomas W. Overton is managing editor of Gems & Gemology. An attorney and former nuclear engineer in the U.S. Navy, he currently serves as president of the
Association of Earth Science Editors. Mr. Overton holds a B.A. in English from the
University of Southern California and a law degree from UCLA. James E. Shigley
was profiled in the second-place entry.
Congratulations to Rui Galopim de Carvalho of Sintra, Portugal,

whose ballot was drawn from the many entries to win a three-year subscription to
GEMS & GEMOLOGY, along with all three GEMS & GEMOLOGY IN REVIEW volumes:
TREATED DIAMONDS, COLORED DIAMONDS, and SYNTHETIC DIAMONDS.
MOST VALUABLE ARTICLE AWARD
GEMS & GEMOLOGY
SpringWinter 2008
Spring 2004
Identification of CVD-Grown Synthetic Diamonds
Cultured Pearls from the Gulf of California, Mexico
X-Ray Fingerprinting Routine for Cut Diamonds
Summer 2004
Gem Treatment Disclosure and U.S. Law
Lab-Grown Colored Diamonds from Chatham
The 3543 cm-1 Band in Amethyst Identification
Fall 2004
Grading Cut Quality of Round Brilliant Diamonds
Amethyst from Four Peaks, Arizona
Winter 2004
Creation of a Suite of Peridot Jewelry: From the
Himalayas to Fifth Avenue
An Updated Chart on HPHT-Grown Synthetic
Diamonds
A New Method for Detecting Beryllium Diffusion
Treated Sapphires (LIBS)
Spring 2005
Treated-Color Pink-to-Red Diamonds from Lucent
Diamonds Inc.
A Gemological Study of Chameleon Diamonds
Coated Pink Diamond: A Cautionary Tale
Summer 2005
Characterization and Grading of Natural-Color
Yellow Diamonds
Emeralds from the Kafubu Area, Zambia
Mt. Mica: A Renaissance in Maines Gem Tourmaline
Production
Fall 2005
A Review of the Political and Economic Forces
Shaping Todays Diamond Industry
Experimental CVD Synthetic Diamonds from
LIMHP-CNRS, France
Inclusions in Transparent Gem Rhodonite
Winter 2005
A Gemological Pioneer: Dr. Edward J. Gbelin
Characterization of the New Malossi Hydrothermal
Synthetic Emerald
Spring 2006
Paraba-type Tourmaline from Brazil, Nigeria,
and Mozambique: Chemical Fingerprinting by
LA-ICP-MS
Identification and Durability of Lead GlassFilled
Rubies
Characterization of Tortoise Shell and Its Imitations
Summer 2006
Applications of LA-ICP-MS to Gemology
The Cullinan Diamond Centennial
The Effects of Heat Treatment on Zircon Inclusions
in Madagascar Sapphires
Faceting Transparent Rhodonite from New South
Wales, Australia
Fall 2006Special Issue
Proceedings of the 4th International Gemological
Symposium and GIA Gemological Research
Conference
Winter 2006
The Impact of Internal Whitish and Reflective
Graining on the Clarity Grading of D-to-Z
Diamonds at the GIA Laboratory
Identification of Chocolate Pearls Treated by
Ballerina Pearl Co.
Leopard Opal from Mexico
The Cause of Iridescence in Rainbow Andradite
from Japan
Spring 2007
Pink-to-Red Coral: Determining Origin of Color
Serenity Coated Colored Diamonds
Trapiche Tourmaline from Zambia
Summer 2007
Global Rough Diamond Production since 1870
Durability Testing of Filled Diamonds
Chinese Freshwater Pearl Culture
Yellowish Green Diopside and Tremolite from
Tanzania
Polymer-Impregnated Turquoise
Fall 2007
The Transformation of the Cultured Pearl Industry
Nail-head Spicule Inclusions in Natural Gemstones
Copper-Bearing Tourmalines from New Deposits
in Paraba State, Brazil
Type Ia Diamond with Green-Yellow Color Due to Ni
Winter 2007
Latest CVD Synthetic Diamonds from Apollo
Diamond Inc.
Yellow Mn-Rich Tourmaline from Zambia
Fluorescence Spectra of Colored Diamonds
An Examination of the Napoleon Diamond Necklace
Spring 2008
Copper-Bearing (Paraba-type)
Tourmaline from Mozambique
A History of Diamond Treatments
Natural-Color Purple Diamonds
from Siberia
Summer 2008
Emeralds from Byrud (Eidsvoll), Norway
Creating a Model of the Koh-i-Noor
Diamond
Coated Tanzanite
Coloring of Topaz by Coating and
Diffusion Processes
Fall 2008
Identification of Melee-Size Synthetic
Yellow Diamonds
Aquamarine, Maxixe-Type Beryl, and
Hydrothermal Synthetic Blue Beryl
A New Type of Synthetic Fire Opal:
Mexifire
The Color Durability of Chocolate Pearls
Winter 2008
Color Grading D-to-Z Diamonds at the GIA
Laboratory
Rubies and Sapphires from Winza, Tanzania
The Wittelsbach Blue
For a complete list of articles from 1981 forward, visit www.gia.edu/gandg.
Get PDF Issues at

gia.metapress.com
Electronic (PDF) versions of all issues from
Spring 1981 forward are available as part of
Gems & Gemology Online. Full issues are
$20; individual articles and sections are $10.
Order Print Back Issues at

www.gia.edu/gandg
Call Toll Free 800-421-7250 ext. 7142
or 760-603-4000 ext. 7142
Fax 760-603-4070
E-Mail gandg@gia.edu
or mail this form to
Gems & Gemology
PO Box 9022, Carlsbad, CA
92018-9022, USA
U.S. Canada
$ 12
Single Issues
Complete Volumes*
$ 40
19922008
$115
Three-year set
$190
Five-year set
Intl
$ 15
$ 18
$ 48
$ 135
$ 220
$ 60
$ 170
$ 280
10% discount for GIA Alumni and active

GIA students.
*Limited issues from 19851991 are also
available. Please call or visit our website
for details on these and the 2009 issues as
they are published.
THE FRENCH BLUE AND THE HOPE:

NEW DATA FROM THE DISCOVERY OF
A HISTORICAL LEAD CAST
Franois Farges, Scott Sucher, Herbert Horovitz, and Jean-Marc Fourcault
A lead cast of the French Blue diamond, a mythic item in the French Crown Jewels, was recently
found in the mineral collection of the Musum National dHistoire Naturelle (MNHN) in Paris.
The details of this diamondstolen in 1792 during the French Revolutionhave up to now been
known only from a drawing of an insignia of the Golden Fleece belonging to King Louis XV that
was published in 1889 and, more recently, from an unpublished rendering dated as early as
1749. Computer modeling of the French Blue from a laser scan of the lead cast revealed details of
the cut that could not be inferred from these drawings. Models of both the lead cast and the Hope
diamond confirm that the latter could have been recut from the French Blue. The additional discovery of the catalog entry associated with the lead cast at the MNHN suggests that Henry Philip
Hope may have owned the French Blue diamond after its 1792 theft and before it was recut.
n the course of his several visits to India during

the mid-1600s, famed French gem dealer and
adventurer Jean-Baptiste Tavernier (16051689)
obtained many exceptional diamonds (Tavernier,
1676; Morel, 1988). Among these was a large blue
stone weighing 112 316 old carats (115.16 modern
carats; figure 1), later called the Tavernier Blue by
Anglo-American scholars (Balfour, 2000; Kurin,
2006), though the diamond went unnamed at the
time. Based on Taverniers writings, it has been
speculated that the diamond came from the Kollur
mine near Golconda. It was cut to preserve weight
at the expense of symmetry and brilliance, which
was a typical practice in ancient India (Morel, 1986;
1988). In 1668, Tavernier sold the diamond to
Frances King Louis XIV (16381715) for the bargain
price of 220,000 livres (Bapst, 1889). This is roughly
equivalent to $5 million today, and the stone was
probably worth twice that: Tavernier gave his king
a very good deal. In 1671, Louis ordered the diamond recut to improve its brilliance, a responsibili-
THE FRENCH BLUE AND THE HOPE
ty that fell to Jean Pitau (16341676), the court jeweler (Bapst, 1889). In 1673, Pitau delivered a shieldshaped stone weighing around 69 0.03 ct (Morel,
1988; again, see figure 1). Jean-Baptiste Colbert,
King Louiss minister of finance, dubbed the stone
the Diamant Bleu de la Couronne (Blue Diamond
of the Crown). As with Tavernier Blue, French Blue
is a modern anglicism; however, this name will be
used in this article for ease of understanding.
In 1749, Louis XV (17101774) asked Paris jeweler Pierre-Andr Jacquemin (17201773) to mount
the stone in a ceremonial insignia of the Order of
the Golden Fleece (Morel, 2001). Jacquemin produced two color renderings (Farges et al., 2008). The
first (believed to be the final version; figure 2) shows
the French Blue and Bazu diamonds, as well as the
See end of article for About the Authors and Acknowledgments.

GEMS & GEMOLOGY, Vol. 45, No. 1, pp. 419.
GEMS & GEMOLOGY
SPRING 2009
Figure 1. The Tavernier

Blue diamond (bottom;
shown here in a drawing from Tavernier,
1676) was an ~115 ct
flat slab cut primarily to
conserve weight. Louis
XIV ordered it recut in
1671. The resulting ~69
ct stone came to be
known as the Blue
Diamond of the Crown
or the French Blue (computer rendering, top
left), which is believed
to have been recut later
into the 45.5 ct Hope
diamond (top right).
This computer model of
the French Blue was created from a lead cast
recently discovered in
the Musum National
dHistoire Naturelle in
Paris. Photo of the Hope
courtesy of the Smithsonian Institution.
107 ct Cte de Bretagne spinel (originally thought

to be a ruby), which is carved in the shape of a dragon. A second version (Farges et al., 2008, p. 17) bears
two large table-cut blue sapphires (six- and eightsided, respectively); it does not appear that production of this version went any further than a rendering. The finished insignia was a masterpiece of
Rococo jewelry, known as the Toison dOr de la
Parure de Couleur or Golden Fleece of the Colored
Adornment (Bion et al., 1791). At some unknown
time after the French Blue was mounted, Jacquemin
(or another crown jeweler) created a lead cast from
the insignia, which was later recovered by heirs of
the Bapst family, who also served as French crown
jewelers (Bapst, 1889). Germain Bapst later wrote
that it was a tradition in his family to create lead
copies of the crown jewels for documentary purposes. The whereabouts of this lead cast are unfortunately not known; the drawing reproduced in
Bapsts book (figure 3) is the primary record. A lead

cast of the French Blue itself was also prepared at
some point (possibly as late as 1812, see below),
though the party responsiblewhether Pitau,
Jacquemin, or another jeweleris also unknown.
In September 1792, during a wave of revolutionary rioting that swept across Paris, a gang of thieves
broke into the Royal Storehouse, the Garde-Meuble,
and stole most of the French Crown Jewels (including many loose gemstones and pearls) over the
course of five nights (see, e.g., Bapst, 1889; Morel,
1988). Bapst suggested that one of the thieves, Cadet
Guillot Lordonner, left Paris with the Golden Fleece
of the Colored Adornment on the first day of the
theft and unmounted the French Blue and the Cte
de Bretagne spinel from the setting at some point
during his journey between Nantes and le Havre.
He made his way to London, where he tried to sell
the Cte de Bretagne to exiled French monarchists.
GEMS & GEMOLOGY
SPRING 2009
(This rapid exit from the country would have been

necessary because Guillot had taken perhaps the
two most recognizable colored gems in the entire
collection.) Although most of the large diamonds
were later recovered, none of the jewels, such as the
Golden Fleece insignia, ever reappeared. Only the
Cte de Bretagne spinel surfaced in London in 1797
and was reintegrated into the French Crown Jewels
in 1824; it is now housed at the Louvre Museum.
In 1804, Napoleons government issued a law
providing for a 20-year statute of limitations on
crimes committed during the revolution (Winters
and White, 1991). This meant that criminal liability
for the theft would apparently end in 1812 (though
Morel, 2001, suggested otherwise, arguing that this
law did not apply to the crown jewels).
The French Blue was never seen again in its
Figure 3. These drawings by Lucien Hirtz (from

Bapst, 1889, pp. 268269) depict a lead cast of the
Golden Fleece that was made at some unknown
time after the jewel was created. Compare it to the
Jacquemin rendering; note especially the different
diamond mounted at the top.
Figure 2. This color rendering of Louis XVs Order

of the Golden Fleece by jeweler Pierre-Andr
Jacquemin (after 1749) is the only surviving contemporaneous drawing of the insignia, and the only
known color drawing of the French Blue diamond.
Note the presence of a large rounded square brilliant (the Bazu diamond, according to Morel, 1988)
above the red spinel dragon.
GEMS & GEMOLOGY
SPRING 2009
original form, but another large blue diamond did

appearexactly 20 years and two days after the
theft. In 1812, London jeweler John Francillon
(17441816) described a 45.5 ct deep blue diamond
without specks or flaws that he had seen
(Francillon, 1812; figure 4). The owner of the diamond was not named; Francillon simply reported
that he examined it by leave of Mr. Daniel Eliason
(17531824), who was a London diamond merchant
at the time (Patch, 1976; Balfour, 2000). Why Eliason
revealed this stonethe future Hope diamondto
Francillon, and whether he owned it himself or was
acting at the behest of another, is not known.
The first documented (Hertz, 1839) owner of this
stone was Henry Philip Hope (17741839), from
whom it obtained its current name. However, no
clear or reliable evidence exists to document how
and when Hope acquired his blue diamond; he was
known to maintain secrecy about his collection,
presumably for tax reasons (Rivington and Rivington, 1845). Eventually, the blue diamond passed
through the Hope family to America by way of
French jeweler Cartier (Ross, 2005; Kurin, 2006). In
1958, New York jeweler Harry Winston donated the
Hope diamond to the Smithsonian Institution in
Washington, DC (Patch, 1976; Balfour, 2000).
PREVIOUS ATTEMPTS TO
RECONSTRUCT THE FRENCH BLUE
In his 1889 book on the French Crown Jewels, Bapst
claimed that Hirtzs drawing of the lead cast of Louis
XVs Golden Fleece (again, see figure 3) was at the
correct (1:1) scale. However, Morel (1988) determined that Hirtzs drawing of the French Blue, if
indeed to scale, was too narrow to accommodate the
Hope (figure 5, left). In attempting to prove his thesis
that the Hope was recut from the French Blue,
Morel expanded Hirtzs drawing to the dimensions
published a century earlier by Brisson (1787). Morels
converted metric dimensions were 31.00 24.81
12.78 mm. To assess the validity of this approach,
Morel compared the dimensions of the Regent diamond given by Brisson to a more modern measurement reported in 1884 by Jacob (described in Morel,
1988). He found the exact same valuesdown to a
hundredth of a millimeterfor the Regent: 31.58
29.89 20.86 mm.
Unfortunately, in checking Morels work ourselves, we found that his dimensions for the French
Blue were likely underestimated because Jacob had
overestimated those of the Regent. The Louvre
Figure 4. This 1812 sketch and description by London

jeweler John Francillon is the earliest public record of
what would become known as the Hope diamond.
Courtesy of the U.S. Geological Survey Library.
Museum, where the Regent is now housed, reports

the measurements to be 30.5 28.9 20.3 mm
(Diamond, known as the Regent, 2009). Thus,
Brissons measurements are not as accurate as
Morel estimated ( 0.05 mm on average) but rather
0.9 mm, a far more plausible range given the limitations of 18th-century instruments. Assuming similar discrepancies with Brissons measurements for
the French Blue, we calculated a revised estimate
for that stone of 29.99 23.96 12.11 mm. These
differences confirm that we cannot know the
dimensions of the French Blue below the millimeter
level from these records alone.
GEMS & GEMOLOGY
SPRING 2009
In addition to these problems, even after Morel

increased the calculated length of the French Blue
diamond from 28.0 to 31.0 mm, two small but significant inconsistencies remained between the
known dimensions of the Hope and those of his
hypothetical French Blue. Nevertheless, his final
model for the French Blue fully encloses the Hope,
so we believe Morel distorted Hirtzs drawing for
this purpose. In the end, Morels version of the
French Blue diamond more closely resembled a regular heptagon (figure 5, center) than the truncated triangle in Bapst (1889).
Based on Morels information, researchers at the
Smithsonian Institution coordinated a reconstruction of the Tavernier Blue and French Blue diamonds and reexamined their relationship with the
Hope. As part of that effort, Attaway (2005) and
Sucher (2005) created two similar three-dimensional (3D) computer models of the French Blue (e.g.,
figure 5, right), as well as replicas in cubic zirconia
and plastic. By comparing those models to a computer model of the Hope, Attaway and Sucher both
confirmed what had been suspected since at least
the mid-19th century (Barbot, 1858) and which was
analyzed first by Morel (1988): that the Hope was
cut from the stolen French Blue, leaving insufficient material for smaller diamonds.
However, as pointed out by both Morel (1988)
and Kurin (2006), Hirtzs line drawings (and subsequent printings) are likely subject to error and artistic license. This can be clearly demonstrated by analyzing the Cte de Bretagne spinel. Comparing a
contemporary high-resolution photo of the spinel to
the drawings (figure 6, left and inset), Hirtzs distortions of the Cte de Bretagne can be seen to be significanton the millimeter level, which is relative-
ly large given the size of the carving (~45 17.6

mm; these dimensions were estimated from a comparison with the Regent because the actual artifact
was not directly available to us; thus, scaling errors
in Hirtzs drawing cannot be traced precisely).
In addition, because Hirtz depicted both the
front and back of the ornament, the drawings can be
checked for consistency by overlaying a mirror
image of the back side onto the front. Although this
exercise showed remarkably few inconsistencies
(figure 6, right), some small but significant differences do exist (indicated by arrows in figure 6). This
again confirms that any reconstructions from those
drawings cannot be accurate below about the millimeter level.
Last, and perhaps most importantly, the Hirtz
drawings show only the crown and pavilion of the
French Blue, with no side views. Thus, all studies
based on these drawings have had to estimate the
total depth and the girdle details using the French
Blues reported weight (Attaway, 2005; Sucher,
2005). This has a critical influence on the appearance
of the gem, as the precise angles between facets cannot be determined from the Hirtz drawings.
THE LEAD CAST OF THE FRENCH BLUE

Discovery. A recent (2007) update of the inventory
of the mineral and gem collection of the Musum
national dHistoire naturelle (MNHN) in Paris
turned up the lead cast of a large (30.38 25.48
12.88 mm) shield-shaped diamond (figure 7) with
dimensions similar to those previously reported for
the French Blue (Farges et al., 2008; see table 1).
This cast, catalogued in 1850 (its acquisition year is
not known), is also notable because of its entry in
Figure 5. Various line drawings of the French Blue (in red)

are compared to a drawing of
the Hope diamond (in blue,
courtesy of the Smithsonian
Institution): left=based on the
original Hirtz drawing (Bapst,
1889); center=based on the
model calculated by Morel
(1986) after stretching and distorting the drawing by Hirtz;
and right=from the computer
model by Sucher (2005), after
only an iterative stretch of
the drawing by Hirtz.
GEMS & GEMOLOGY
SPRING 2009
Figure 6. At left, the

107 ct Cte de Bretagne
spinel in the Hirtz
drawing, when compared to the actual
shape of the dragon
(inset and red contour
line, which has been
rescaled for the best fit
with the drawing),
shows a number of
errors and distortions.
At right, when the front
and the horizontally
flipped back of the
Hirtz drawings are
overlaid, additional
inconsistencies (gray
arrows) between the
two drawings become
apparent. The inset
shows the actual spinel
(~45 17.6 mm), which
is currently housed in
the Louvre; photo
Runion des Muses
Nationaux/Art
Resource, New York.
the MNHN catalogue (inventory no. 50.165; figure

8), which reads (in French) Mr. Achard, Lapidary
Lead model of a diamond belonging to the Crown of
Portugalcut following the shape of a diamond.
The subsequent entry (50.166), also catalogued in
1850, reads, ibid, ibid [i.e., also Mr. Achard,
Lapidary]Model of a diamond, remarkable for its
claritybelonging to Mr. Hoppe of London
(emphasis added). The piece associated with this
second entry is another lead cast, having the shape
of a table-cut diamond, also known as a mirror
(Morel, 1988). The mirror cut typically resembles
half of a diamond octahedron (Tillander, 1996). The
cast labeled 50.166 is 18 17 11 mm, which puts
the approximate weight of the original diamond at
29.3 ct. Items 50.165 and 50.166 were donated to
the MNHN at the same time by the same person,
Achard, lapidarist, though they could have been
obtained earlier, as they are listed among minerals
donated by Alexandre Wattemarre (17961864), a
practice that he started in 1843, and those in the

collection of MNHN mineralogy curator Ren-Just
Hay (17431822), which was acquired in 1848.
The catalogue for 1850 lists numerous samples dating from the French Revolution or the First Empire;
these were not catalogued for decades afterward
because of the lack of space and a lack of funding
during the post-Napoleonic period.
The 50.165 lead cast is, as evident in figure 7, a
shield-shaped stone, certainly not a cut that follows
the shape of a diamond. Further, there have been
no reports of a shield-shaped diamond of about
6869 ct (the approximate weight of the diamond
from which it was cast) in the Portuguese Crown
Jewels (Twining, 1960; Morel, 2001).
The remark in the 50.166 catalogue entry about
the clarity of the diamond obviously cannot be verified by comparing it to the cast. However, the
50.166 cast does resemble half of a natural diamond
crystal. Further, the mirror cut, in existence since
GEMS & GEMOLOGY
SPRING 2009
Figure 7. These two lead

castsboth table-up
and table-down views
are shown herewere
catalogued in 1850 by
the MNHN, Paris:
MNHN 50.165 (top) and
MNHN 50.166 (bottom).
Item 50.165 is a cast of
the French Blue. Photos
by F. Farges, MNHN.
the late 1500s, was found in many crown jewel collections throughout Europe during the Renaissance
period. In addition, the 50.166 cast is stored inside a
box with a N1 written on the interior, although
it is listed second in the catalogue.
The discrepancies between the entries and their
corresponding casts can be rectified if we assume the
numbers were inadvertently transposed when the
casts were logged in or that the casts were switched
at some point, probably before the donation was
received by Armand Dufrnoy (17921857), who catalogued them in 1850. (Dufrnoy had been tasked
with singlehandedly cataloging tens of thousands of

specimens because of the backlog noted above.)
Thus, it is logical to conclude that the 50.166 entry,
attributing the original to a Mr. Hoppe of London,
actually corresponds to the 50.165 cast resembling
the French Blue. The comment about the originals
remarkable clarity would certainly fit a Golconda
diamond such as the French Blue (Brisson, 1787), as
well as what is known about the Hope from both
Francillons initial report and more modern examinations (e.g., Crowningshield, 1989, which reports a
GIA clarity grade of VS1).
TABLE 1. Dimensions for the French Blue from historical references, compared to those of the lead cast MNHN 50.165.
Reference
Length
Width
Thickness
Weight
Bion et al. (1791)
Brisson (1787)
Brisson (1787), converted by
Morel (1988)
Brisson (1787), converted based
on a modern set of dimensions for
the Regent diamondd
Hirtz (Bapst, 1889), measured from
drawing at 1:1 scale
Average error
Lead cast MNHN 50.165
13 3 4 lignes b
31.00 mm
11 lignes
24.81 mm
52/3 lignes
12.78 mm
29.99 mm
23.96 mm
12.11 mm
2681/8 grains, poids de marca

(68.97 ct)
260 grains, poids de marca
69.03 ct
(69.05 ct)c
28 mm
24 mm
30.38 mm
0.9 mm
25.48 mm
12.88 mm
0.06 ct
68.3 ct
1 ligne = 2.2558 mm (Morel, 1988).

grain, poids de marc 0.0531147 g 3.765 ct (Lionet, 1820).
cMorel underestimated the actual weight of Brisson by 0.02 ct; the accurate value is 69.05 ct.
dAs provided by the Louvre Museum (30.5 28.9 20.3 mm).
b1
10
GEMS & GEMOLOGY
SPRING 2009
Figure 8. The entries in the 1850

catalogue showing the numbers
and descriptions for MNHN
50.165 and 50.166 do not match
the items. Entry 50.165 reads,
M. Achard, LapidaireModle en
plomb dun diamant appartenant
la Couronne de Portugaltaill
suivant la forme du diamant.
(Mr. Achard, LapidaryLead
model of a diamond belonging to
the Crown of Portugalcut following the shape of a diamond.).
Entry 50.166 reads, ibid, ibid
Modle dun diamant remarquable
pour sa limpiditappartenant
M. Hoppe de Londres (ibid,
ibidModel of a diamond,
remarkable for its claritybelonging to Mr. Hoppe of London). The
labels were apparently switched
accidentally at some point.
The Donors: The Achard Family of Parisian Jewelers.

Little information could be found concerning the
identity of the lead casts donor. In 1817, Ren-Just
Hay described a certain Mr. Achard as one of the
most knowledgeable jewelers of this city [Paris] for
everything that deals with the objects [gems] of his
business (1817, p. 235). Babinet (1857, pp. 1415 and
57) wrote that Charles Achard (likely the previous
Achards son, given the date) was known to be
involved more than any other in France concerning
the business of colored gemstones. This author also
mentions Charles Achards father, though not by
name. A third Achard, Edouard (most likely the
grandson of the senior Achard), was the director of
the Parisian Chamber of Commerce for the trading of
gemstones, and was appointed by the Third Republic
to supervise the disastrous sale of what remained of
the French Crown Jewels in 1887. Based on the time
frame, the MNHN donor was likely Charles Achard,
a contemporary of Babinet (17941872) and probably
then a prominent Parisian lapidary. It is possible that
his father was also named Charles, though we found
a David Achard joaillier [jeweler] in some 19thcentury Parisian archives, with no precise dates
(birth, wedding, or death).
Origin of the Cast. We do not know who made the
cast or how the Achards obtained it. Nor do we
know its age. The patina of the lead certainly suggests that the model is quite old or has been extensively used, or both. Pitau, as the original cutter, is
the most logical fabricator, in keeping with his duty
as one of Louis XIVs jewelers (though it appears

that few such casts were actually produced; Bapst,
1889). It is also possible that Jacquemin prepared a
model to better construct his Golden Fleece for
Louis XV. Neither scenario explains how the cast
ended up with the Achards, who were never granted
the right to work directly for the kings of France
prior to the Revolution (Morel, 1988). It is known,
however, that some of Jacquemins possessions
were auctioned after his death in 1774; these could
have included the renderings of the insignia (again,
see figure 2) as well as the cast.
Alternatively, the senior Achard, who would have
been an apprentice during the first years of the
Revolution, could have created a cast of the French
Blue before its theft in 1792. As Brisson was allowed
by the authorities in charge of the Royal Storehouse
to measure the dimensions and density of the French
Bluewhich would have necessitated unmounting
in 1787, it could have been unmounted on other
occasions as well. Achard could conceivably have
obtained the diamond from the Royal Storehouse just
for the purpose of making the cast, but this seems
unlikely: Unlike Brisson, who was a famous scientist
and a member of the French Academy of Sciences,
the Achards did not become prominent until the
Revolution or the First Empire (Hay, 1817; Babinet,
1857) and were not known to be among the jewelers
who served the aristocracy (Bapst, 1889). Thus, there
is very little chance that the Achards could have borrowed the diamond from the Royal Storehouse to
produce the cast during this period.
GEMS & GEMOLOGY
SPRING 2009
11
Finally, based on the label, it is possible that the

cast was created while the diamond was in the possession of Mr. Hoppe of London. This would likely
mean that the cast was made just before the French
Blue was recut, much as similar models were made
for the rough of the Pitt diamond, cast in London
before it was cut into what would become the Regent
(Morel, 1988), and for the Koh-i-Noor prior to its
recutting in 1851 (Sucher and Carriere, 2008). Morel
(1988) proposed that John Francillon (born Jean
Franillon), a French Huguenot lapidary (like the
Achards) who had emigrated to London, could have
been the party tasked with recutting the French Blue
into the Hope; thus, he might have produced a cast of
the French Blue before he began his work.
Whatever its origin, this cast allowed, for the first
time, precise calculation of the shape and dimensions
of the French Blue, including the missing thickness
information (girdle, pavilion and crown) as well as
the angles for each facet. To take advantage of this
unique opportunity, we used laser scanning and computer modeling to recreate an exact model of the diamond, based on the lead cast and the Jacquemin
drawing that shows the diamond in color. We also
Figure 9. The original laser-scan data from the lead cast
of the French Blue (top), when cleaned of scanning artifacts and imported into DiamCalc, produced the facet
diagram at the bottom (table-down, left; table-up, right).
used this simulation to reconstruct Louis XVs colored Golden Fleece insignia based on the information
available. Special care was given to propose a reconstruction that was plausible for a jeweler, as the
drawings of Jacquemin and Hirtz omit important
details, such as the hundreds of smaller diamonds the
jewel was known to contain (specifics of this reconstruction are discussed in Farges et al., 2008). Last, we
examined the implications of this new information
for the history of the Hope and the French Blue.
MODELING THE FRENCH BLUE

FROM THE LEAD CAST
Methodology. The cast was laser-scanned by
Diamond Matrix Technology in Antwerp, Belgium,
using an Octonus Helium 1:4 scanner (Sucher and
Carriere, 2008). The accuracy of the 3D GemCad
model is better than 40 m (analog) or 28 m (digital). The scan data generated a solid consisting of
2,792 planar surfaces (figure 9, top). The resulting 3D
image was then cleaned of scanning artifacts (figure
9, bottom) by importing the GemCad model into
Diamond Calculator 3.0 software (DiamCalc; Sucher
and Carriere, 2008).
DiamCalc can also create a 3D model that
adjusts reflection and refraction based on a programmable refractive index. We employed this
method to get an indication of how well diamond
cutting in 1670 was guided by the laws of optics, and
to compare the brilliance of the lead castderived
model with that created by Sucher (2005) based on
Hirtz. By default, DiamCalc produces simulations of
colorless diamonds, so to simulate the French Blue
we used a color profile of the Hope diamond provided by Dr. Jeffrey Post of the Smithsonian.
Matching the Lead Cast to the French Blue. The
shape and facet patterns of the cast as derived from
the laser scan (again, see figure 9) were somewhat
different from those recorded by Jacquemin (figures
2 and 10a) but were strikingly close to those reported by Hirtz (Bapst, 1889), especially the crown facet
patterns (figures 3 and 10b). They deviated significantly from Morel (1986; figure 10c) but less so from
Sucher (2005; figure 10e). The pavilion facet pattern
in Hirtz was similar to that of the replica, but with
a few notable differences: The size, shape, and placement of the culet matched, as did the first row of
facets; however, the casts pattern is more complex
than the drawing in its upper pavilion facets (figure
11). Still, the patterns in Hirtz and in the cast are
12
GEMS & GEMOLOGY
SPRING 2009
Figure 10. Shown here

are diagrams of the
crown facet patterns of
the French Blue as
derived from various
sources: (a) Jacquemins
1749 drawing; (b) Hirtzs
drawing (Bapst, 1889);
(c) Morel (1986); (d)
Attaway (2005); (e)
Sucher (2005); and (f) the
MNHN 50.165 lead cast
(this study). All drawings are scaled except for
(a), for which no scale
was provided (so it is
scaled to Brisson, 1787).
sufficiently consistent with each other, and with

the diamond-cutting style used in the 17th century
(Tillander, 1996), that we feel confident in identifying the cast as being taken from the French Blue.
Weight. A variety of weights have been reported for
the French Blue, though they are consistent for the
most part. A 1791 inventory of the crown jewels
(Bion et al.) gave the weight as 268 18 gr (gr =
grains, poids de marc); the same weight is also given
on Jacquemins 1749 drawing (again, see figure 2).
This value is equal to that given in a 1691 inventory
(67 ks 18; ks = old carat); both values convert to
68.97 modern carats (Morel, 1988). Similarly, Brisson
(1787) reported a weight of 260 grains, poids de
marc, which Morel (1988) calculated to be equal to
69.03 ct (erroneously; the actual value is 69.05 ct).
From this, Morel (1988) estimated the average
weight to be 69.00 0.03 ct. Brisson (1787) reported
his weight error to be 164 of a grain (0.004 ct), but
his values have since been shown to be closer to 14
grain (0.06 ct, which seems reasonable for 1787).
Using a Tescan scanning electron microscope
operating at 15 kV under low vacuum conditions, we
performed chemical analysis of the cast. The results
showed a composition of 97 wt.% lead, 2 wt.% tin,
and traces of iron and zinc. Based on the density of
this measured composition (11.2 1 g/cm3) and the
density of 3.5254 g/cm3 for the French Blue reported
by Brisson (1787), we estimate that the cast is equivalent to a 68.3 0.2 ct diamond (n.b.: the accepted
density of diamond using modern techniques is

3.513.52 g/cm3; Bari and Fritsch, 2001).
However, the weight represented by the cast in its
current condition is probably not what it was originally because of the rounded edges and worn and weathered surface [partially covered by hydrocerussite,
Pb3(CO3)2(OH)2]. Assuming the original edges of the
cast were reduced by 0.5 mm each, this gives a loss of
~0.3 ct of diamond. Thus, our estimated weight for
the French Blue derived from the cast, 68.3 ct, is 0.4 ct
lower than the low end of Morels (1988) range, which
is only a ~0.5% variance. This is further strong evidence tying the cast to the French Blue diamond.
Figure 11. Overlaying the computer model derived
from the lead cast (blue) on that derived from Hirtz
(red) shows notable differences in the arrangements of
the crown and pavilion facets.
GEMS & GEMOLOGY
SPRING 2009
13
Dimensions. The published dimensions and weights

estimated for the French Blue are shown in table 1.
The dimensions of the cast are 30.38 25.48 12.88
mm (0.01 mm). These match those reported by
Brisson (1787)converted based on the modern set of
dimensions for the Regent diamondwithin 0.9
mm on average. No sink marks (depressions from
contraction of the lead) were observed on the cast,
suggesting that significant lead shrinkage did not
occur. Based on the linear expansion coefficient of
lead (~28 106 mm/K), we calculated that the cast
should have contracted by a maximum of 0.3 mm
along the length and width, and 0.1 mm in thickness,
during the cooling from its molten state (i.e., over
400C). Such shrinkage can be reduced by the use, for
instance, of warm molds, a technology well known
since the end of the 17th century.
If we compensate for this lead shrinkage, the corrections to the French Blues dimensions are still
within Brissons margin of error. However, the
weight would then have to be increased by 4.5 ct,
which is significant even for Brisson. Therefore, we
believe that no significant shrinkage occurred during the making of the cast. If we assume the weight
of the diamond to be 69.0 ct, then only a maximum
of 0.1 mm shrinkage in length and width and <0.01
mm in thickness could have taken place. Therefore,
we estimate that the maximum dimensions were
30.5 0.1 25.6 0.1 12.9 0.1 mm.
Design and Shape. The cast has a thin crown (<3
mm) and a relatively thick pavilion (9.97 mm) as
compared to earlier replicas, and a thin (<0.5 mm)
and very even girdle. Hirtzs drawing (again, see figure 3) shows a central culet with seven culet facets
around it. The culet facets are surrounded by seven
kite-shaped primary main facets, with another row
of seven kite-shaped secondary main facets surrounding these. Fourteen break facets define the girdle. However, the cast has a second row of main
facets that are horizontally split (total 14), plus a tertiary row of 14 additional main facets. In contrast to
the 14 break facets in Hirtzs drawing, there are
only 13 on the cast. The drawing has a total of 36

pavilion facets; 57 are present in the cast (table 2).
However, the break facets along the girdle are vertically split, with break side facets. The presence of
break facets on both the crown and the pavilion
(again, see figure 9) contributes to the artistic symmetry of Pitaus design.
When we compare our model to that of Attaway
(2005), it is evident that the facet patterns are similar,
with the extra facets on the pavilion of the cast model
primarily being the result of the horizontal splitting of
some facets on Attaways model (again, see figure 11).
The angular differences between the two halves of the
split facets range in the vicinity of 35, just enough
to discern them as separate facets rather than one
rounded facet. There is also one extra row of facets on
our model. The crown facet patterns differ only in the
configuration of the break facets: They are vertically
split in Attaways model and not split on ours.
Hirtzs model can also be analyzed and compared
to the other measurements using its length-to-width
ratio. Hirtzs has a ratio of 0.8571, within 1.5% of
Morels model (0.8459). Brissons measurements (as
corrected by Morel) have a ratio of 0.8003, and the
lead cast has a ratio of 0.8387, between Hirtz and
Brisson. This result confirms the distortions in
Hirtzs drawing (Bapst, 1889), and demonstrates again
the relative inaccuracy of Brissons measurements.
The overall shape of the cast (figure 10f) is much closer to that of Hirtz (figure 10b) than that of Morel (figure 10c), although the length-to-width ratio of the
cast is closer to that of Morel than of Hirtz. Attaways model shows a stone with a length-to-width
ratio of 0.8589. Extended over 25 mm (approximately
the length of the diamond), this results in a difference
of about 0.5 mm from our model. This does not
negate the validity of the earlier model, in that the
Hope still fits neatly inside it. Although the depths of
the two models differ, they still encompass the same
volume and, hence, weight.
Computer Simulation of the French Blue. Comparison between our model and that of Sucher (2005)
TABLE 2. Summary of pavilion facet differences between Hirtzs drawing (in Bapst, 1889)
and the lead cast MNHN 50.165.
Source
1749 drawing
MNHN 50.165
14
Culet
Culet
facets
Primary
main facets
Secondary
main facets
Tertiary
main facets
Break
facets
Total
1
1
7
7
7
8
7
14
0
14
14
13
36
57
GEMS & GEMOLOGY
SPRING 2009
addition, in Greek mythology, Apollo, the god of

peace and fine arts, is commonly represented by the
sun. Through its art and architecture, Versailles connected these symbolic relationships, and the palace in
its entirety conveyed the message that Louis was the
Sun King, ordained with the divine right to rule
(e.g., figure 13). The French Bluewith its seven-fold
Figure 12. This computer rendering of the French

Blue diamond, based on the color profile of the
Hope diamond (courtesy of the Smithsonian), represents the model Sucher (2005) composed using
Hirtzs drawing and Brissons measurements.
Compare it to the French Blue model in figure 1,
derived from the lead cast, which has noticeably
greater brilliance and scintillation.
Figure 13. Louis XIV was very fond of symbolism and

mythology. This allegorical scene shows King Louis as
Apollo, the Greek god of the sun. The seven-fold symmetry of the French Blue mimics the radial beams of the sun,
as seen here illuminating the Sun King. Painting by
Joseph Werner II (16371710?), courtesy of Runion des
Muses Nationaux/Art Resource, New York.
using DiamCalc (figure 12) shows that the earlier

model, based on published drawings and dimensions,
results in a gem of lesser scintillation compared to
the one derived from the lead cast (again, see figure
1). We believe the additional facets (57 vs. 36) are
responsible for the superior optical effects of the lead
cast model. This evidence also suggests that the lead
cast is an excellent replica of the original, showing an
accurate set of facets with an exact set of angles that
also enhance brilliance. Thus, this lead cast was probably not copied from a distance or by memory or by
an inexperienced jeweler, but rather was cast directly
from the original diamond.
This simulation also demonstrates that the diamond was a masterpiece of lapidary work due to its
odd number of facets (seven) around the culet (a difficult pattern to cut given the technology existing at
the time) and its greatly increased scintillation and
brilliance over Taverniers original stone. Pitaus peculiar cut design became known as the rose de Paris or
la mode des deux cts (Morel, 1986) and would later
be used for the Hortensia diamond (Morel, 1988).
The number of facets surrounding the culet was
surely not an arbitrary decisioncourt life at
Versailles under Louis XIV was heavily steeped in ceremony and symbolism. Seventeenth-century France
was a deeply Christian kingdom, and in the Bible the
number seven represents spirituality and divinity (e.g.,
the seven days of creation and the seven sacraments
in western Catholicism) and the perfection of the
human form of God (e.g., the number of prophets). In
GEMS & GEMOLOGY
SPRING 2009
15
Figure 14. These views show the model of the Hope diamond (blue) inside the model of the French Blue that was
derived from the lead cast (black). Note the close fit. Models B and C represent views from the respectively marked
arrows on model A.
symmetry, blue color, and brilliancewas clearly

intended to echo this sentiment. Pitaus design was
likely intended to represent a sun with seven radial
beams set in a blue sky.
The importance of this design to King Louis can
be seen in the extravagant sum he paid Pitau to facet
the diamond: the equivalent of half a million dollars,
or about one-tenth the cost of the Tavernier Blue
(Bapst, 1889)this at a time when labor costs were
not normally a meaningful element of a jewels
value. As just one example, inventories of the
French Crown Jewels estimated value only by calculating the worth of the gems and did not include the
metals (gold, silver, bronze, etc.) or manufacturing
costs (e.g., Bion, 1791).
The simulation of the French Blue represents the
rediscovery of a true masterpiece of French Baroque
lapidary art. Created well before the Regent diamond
(cut 17041706 in London), the French Blue is also
one of the earliest examples of the brilliant cut, a
clear departure from the classical octahedral cuts of
the 17th century, such as the mirrors and Mazarins.
Comparisons with the Hope. We created a model for
the Hope diamond by independently reconstructing
Attaways model using photos from Attaway (2005)
and DiamCalc; there were no substantive differences
between the earlier model and ours. Using GemCad,
we inserted the model of the Hope diamond into our
model of the French Blue (figure 14). Model resolution was 43 pixels/mm, and the Hope fit inside the
French Blue with as little as 4 pixels (less than 0.1
mm) distance between the outside edges of the
stones when they were rotated about all three axes.
This is important, as the Attaway study reconstructed the French Blue solely from the available line
16
drawings, without benefit of the proportions and

details of the crown and pavilion provided by the
lead cast. Thus, this study confirms the conclusion
in the earlier works of Morel (1986) and Attaway
(2005), as postulated by Barbot (1858), that the Hope
could have been cut from the French Blue.
Reconstruction of the Golden Fleece. A reconstruction of the colored Golden Fleece was also painted
by artist Pascal Monney of Geneva for this study.
This simulation was then refined based on the 1791
inventory (Bion, 1791) to eliminate inconsistencies
in some of the diamond shapes and settings, either
misdrawn by Hirtz, missing in Jacquemin, or
deemed irrelevant by Morel (1988) and Tillander
(1996). We also added elements of the Rococo style
that dominated during this period by reference to
examples of other Golden Fleece insignia in museums in Lisbon, Munich, Austria, and elsewhere in
Europe. Based on this, Mr. Monneys gouache of the
historic piece (figure 15) is the first reconstruction
that is realistic for a jeweler of the 18th century
Rococo period. More detail, as well as an alternate
version of the insignia, is presented in Farges et al.
(2008) and will be discussed in a future paper.
IMPLICATIONS FOR THE FATE OF THE

FRENCH BLUE AND THE HOPE
Ownership. The correct catalog entry for the lead
cast of the French Blue states that the diamonds
owner was Mr. Hoppe of London. Who was Mr.
Hoppe? One of Achards most important customers
was Henry Philip Hope, the first known owner of
the Hope diamond. Hope was also a friend of RenJust Hay, the mineralogy curator at the MNHN
GEMS & GEMOLOGY
SPRING 2009
until 1822. In his treatise on gems, Hay (1817)

thanks only two people for their readings and gem
donations: Henry Philip Hope and Charles Achard.
Hope had previously donated a collection of gems to
Hay, which now resides at the MNHN; this donation could well have included the lead cast, as discussed above. Further, another donation to the
MNHN by the Achards, MNHN 50.167 (note the
numbering immediately after the lead casts; again,
see figure 8), is a glass replica of a large deep blue sapphire that the catalogue states was sold to Mr.
Hoppe of London. Given these connections, it
seems logical to assume that Hoppe was in fact
Henry Philip Hope.
In 1849, Hopes oldest nephew, Henry Thomas
Hope (18081862), formally inherited the Hope diamond (Morel, 1988; Kurin, 2006). The trial over
Hopes will was public and acrimonious (Rivington
and Rivington, 1845), so by 1850 a jeweler as connected as Achard was surely aware of the diamonds
existence, even assuming he had not become aware
of it because of his relationship with Henry Philip
Hope (who had published a catalogue of his gem
collection 11 years earlier; see Hertz, 1839). Yet the
lead cast is clearly that of the French Blue, not the
Hope, and an experienced French jeweler like
Achard could hardly have confused the two. The
label suggests that one of the Achards (most likely
the father) was somehow able to link Henry Philip
Hope to the French Blue.
Two possibilities exist: (1) that Achard simply
assumed the Hope was the recut French Blue, or (2)
that the family had actual knowledge of this fact
from their relationship with Henry Philip Hope. As
the Hope and the French Blue were the largest blue
diamonds of their time, such an assumption would
have been logical enough. There is no written record
of this connection prior to Barbot (1858), though this
does not preclude the idea having been in circulation
earlier. That said, like all gem dealers, the Achards
could have had confidential information on the
gems they came into contact with, and on their
clientssuch as Henry Philip Hope. Had Hope possessed the French Blue, the Achards would have
been among the very few people who might have
known about it.
However, had this information become known
by other jewelers (especially the kings jewelers like
Bapst), the French government or royalists hoping
to please the surviving exiled heirs to the crown
could have claimed or attempted to buy back the
diamond, as they did with the Cte de Bretagne
Figure 15. This gouache of King Louis XVs Golden

Fleece of the Colored Adornment, based on information about the piece and its gems gleaned from this
study, was created in 2008 by Pascal Monney, Geneva,
Switzerland (reprinted by permission of the owner).
spinel and numerous other gems lost in the 1792

robbery. The fact that this did not occur with the
French Blue suggests that it was likely recut soon
after the theft. In the absence of hard evidence,
such as the lead cast (which was not publicly
announced until 2008), it would have been difficult
to proveespecially in courtthat the recut stone
was once the French Blue.
GEMS & GEMOLOGY
SPRING 2009
17
Who Authorized the Recutting? Based on Francillons memo, the recutting of the French Blue into
the Hope could not have taken place any later than
September 1812. Does the Achard label represent
proof that Henry Philip Hope was involved in that
recutting? Hope was certainly one of the few collectors wealthy and passionate enough to quietly purchase what would have been the most prominent
stolen diamond in existence at the time. Perhaps
Hope saw an opportunity to obtain the diamond,
then have it recut to hide its origins.
The intriguing lack of records concerning Hopes
acquisition of his blue diamond has been noted by
many scholars (e.g., Balfour, 2000; Kurin, 2006). A
logical assumption from Francillons 1812 memo is
that the recut diamond was then owned by London
gem merchant Daniel Eliason. But other information suggests that the situation may not have been
so simple.
The banking firm of Hope & Co., established in
Amsterdam in 1726, was well connected with the
crowns of Europe for many years (Kurin, 2006). The
company provided financing to the governments of
Britain, Russia, and Portugal; it also worked with the
U.S. government and the French crown to provide
funding for the Louisiana Purchase in 1804. And one
of the companys private clients was Daniel Eliason,
who used Hope & Co. to fund certain Brazilian mining activities.
Knowledge of the whereabouts of the French
Blue was still a risky proposition in 1812. So why
would a memo surface that year announcing the
Hopes existence? Hope & Co. was acquired by
Barings in 1813according to Balfour (2000), due to
the declining state of the Hope fortune. If Henry
Philip Hope owned the stone at the time, he might
have commissioned Eliason to sell it, perhaps in
part to raise funds to prevent a takeover.
To date, there is no hard evidence to confirm the
validity of either scenario. It is interesting to note,
however, that English jeweler and art expert Bram
Hertz, who would later publish a catalogue of Hopes
collection (Hertz, 1839), became Hopes agent for
purchasing diamonds shortly after Eliasons death in
18
1824 (Rivington and Rivington, 1845). In that job,

Eliason was possibly Hertzs predecessor.
CONCLUSION
The discovery of the lead cast of the French Blue
reveals new details about the appearance of this historic diamond and allows a computer reconstruction
more accurate than those of previous studies. Its
quantitative reconstruction shows that the mythic
diamond was a masterpiece of mid-17th century diamond cutting, a fitting symbol for Louis XIV to support his religious dominance and political authority.
Our work confirms earlier studies (Morel, 1986,
1988; Attaway, 2005) that indicate the Hope diamond could have been recut from the French Blue.
In addition, the MNHN label attributing the
French Blue to Mr. Hoppe of London suggests
that Henry Philip Hope may have owned the
French Blue at some point before its recutting. This
is a possibility that has not been documented before
this research (see also Farges et al., 2008). This new
information is in agreement with the post-theft scenario proposed by Bapst (1889). However, if the label
is correct, then our discovery is not fully in agreement with Kurins (2006) German scenario
involving the Duke of Brunswick.
Acquiring and keeping a stone of the importance
and visibility of the French Blue, in any form, would
have required a confluence of exceptional criteria,
which would only have been possible for an individual in a position of power and great wealth. Hopes
connections to the crowns of Europe would have
provided an insiders view of European politics at the
time, and with his personal connections to Eliason,
Achard, and Hay, Hope would have been in a position to know of the availability of any exceptional
stones. Additionally, he had one of the finest personal gemstone collections in all of Europe. The statement on the MNHN label, Mr. Hoppe of London,
is not conclusive, but Henry Philip Hope certainly
had the method, motive, and opportunity to acquire
the French Blue and have it recut quickly to hide his
possession of a stolen royal diamond.
GEMS & GEMOLOGY
SPRING 2009
ABOUT THE AUTHORS

Dr. Farges (farges@mnhn.fr) is professor of mineralogy and
curator of minerals and gems at the Musum National
dHistoire Naturelle (MNHN) in Paris, and consulting professor
at the Department of Geological and Environmental Sciences,
Stanford University, California. Mr. Sucher is principal of The
Stonecutter in Tijeras, New Mexico. He specializes in the
recreation of famous diamonds, and served as a gem cutter
for the Discovery Channel documentary Unsolved History:
The Hope Diamond. Mr. Horovitz is a jeweler in Geneva,
Switzerland, who specializes in jewelry history, including the
history of Louis XVs Golden Fleece. Mr. Fourcault is the technician in charge of the mineral collection at the MNHN.
ACKNOWLEDGMENTS
This article is dedicated to the late Bernard Morel, who
extensively and exhaustively researched the French Crown
REFERENCES
Attaway N. (2005) The French connection. Lapidary Journal,
Vol. 59, No. 3, 2005, pp. 2428.
Balfour I. (2000) Famous Diamonds. Christies, London.
Babinet J. (1857) Etudes et Lectures sur les Sciences dobservation et Leurs Applications Pratiques [Studies and Lectures on
the Observation Sciences and Their Practical Applications],
Vol. 3. Mallet-Bachelier, Paris.
Bapst C.G. (1889) Histoire des Joyaux de la Couronne de France
[History of the Crown Jewels of France]. Hachette, Paris.
Barbot C. (1858) Trait Complet des Pierres Prcieuses [Comprehensive Treatise of Gemstones]. Morris, Paris.
Bari H., Fritsch E. (2001) The natural history of diamonds. In H.
Bari and V. Sautter, Eds., Diamonds: In the Heart of the
Earth, in the Heart of the Stars, at the Heart of Power, transl.
by Michael Hing, Musum National dHistoire Naturelle,
Paris, pp. 2043.
Bion J.-M., Delattre F.-P., Christin C.-G.-F. (1791) Inventaire des
diamants de la couronne, perles, pierreries, tableaux, pierres
graves, et autres monumens des arts & des sciences existans
au garde-meuble (...) [Inventory of diamonds of the crown,
pearls, precious gems, paintings, engraved stones, and other
monuments of arts & sciences in the royal storehouse].
Imprimerie Nationale, Paris.
Brisson M.J. (1787) Pesanteur Spcifique des Corps [Specific Gravity
of Matter]. Imprimerie Royale, Paris, http://gallica2.bnf.fr/ark:/
12148/bpt6k94851w.f92.
Crowningshield R. (1989) Grading the Hope diamond. G&G,
Vol. 25, No. 2, pp. 9194.
Diamond, known as the Regent (2009) www.louvre.fr [date
accessed Feb. 25, 2009].
Farges F., Sucher S., Horovitz H., Fourcault J.-M.(2008) Deux
dcouvertes majeures autour du diamant bleu de la
Couronne [Two major discoveries about the Blue
Diamond of the Crown]. Revue de Gemmologie, No. 165,
pp. 1824.
Francillon J. (1812) Handwritten description of a blue diamond
weighing 177 grains, Kunz Collection, U.S. Geological Survey
Library, Reston, VA, www.usgs.gov/125/articles/library.html.
Hay R.-J. (1817) Trait des Caractres Physiques des Pierres
Prcieuses [Treatise on the Physical Characteristics of Gems].
Jewels. We thank him for his passion and dedication to the

documentation of this collection, in particular of the French
Blue and Louis XVs Golden Fleece of the Colored
Adornment. Dr. Jeffrey Post of the Smithsonian Institution in
Washington, DC, is thanked for supplying research on the
Hope diamond. Pol van der Steen and Jan de Henau of
Diamond Matrix Technology (Antwerp, Belgium) are thanked
for their laser scanning of the French Blue lead cast. Sergei
Sivovolenko (Octonus Software, Moscow) is thanked for
assistance with DiamCalc and GemAdviser. Noted gemologist Rui Galopim de Carvalho and Teresa Maranhas, from
the Ajuda Palace in Lisbon, are thanked for information
about the Portuguese Crown Jewels. Brenda Graff of the
U.S. Geological Survey Library, Reston, Virginia, is thanked
for providing a scan of the Francillon memo in figure 4. The
authors also thank the reviewers for their thorough reading
of the manuscript.
Courcier, Paris, http://gallica2.bnf.fr/ark:/12148/bpt6k91813m.

Hertz B. (1839) A Catalogue of the Collection of Pearls and
Precious Stones Formed by Henry Philip Hope Esq. William
Clowes and Son, London.
Kurin R. (2006) Hope Diamond: The Legendary History of a
Cursed Gem. HarperCollins, New York.
Lionet P.L. (1820) Manuel du Systme Mtrique. Vanackere,
Lille, France.
Morel B. (1986) La vritable histoire du Diamant bleu [The true
history of the Blue diamond]. Revue de Gemmologie, No.
86, pp. 510.
Morel B. (1988) The French Crown Jewels. Fonds Mercator,
Antwerp.
Morel B. (2001) The diamonds of the European monarchies. In H.
Bari and V. Sautter, Eds., Diamonds: In the Heart of the
Earth, in the Heart of the Stars, at the Heart of Power, transl.
by M. Hing, Musum National dHistoire Naturelle, Paris, pp.
248291.
Patch S.S. (1976) Blue Mystery: The Story of the Hope Diamond.
Smithsonian Institution, Washington, DC.
Rivington F., Rivington J. (1845) The Annual Register, or a View
of the History and Politics of the Year 1844. George Woodfall
and Son, London.
Ross J.F. (2005) Hope sleuth cracks case? Smithsonian, April
2005, p. 42.
Sucher S. (2005) French Blue. http://museumdiamonds.com/
FrenchBlueDiamond.htm.
Sucher S., Carriere D. (2008) The use of laser and X-ray scanning
to create a model of the historic Koh-i-Noor diamond. G&G,
Vol. 44, No. 2, pp. 124141.
Tavernier J.P. (1676) Les six voyages de Jean Baptiste Tavernier,
Ecuyer Baron dAubonne quil a fait en Turquie, en Perse, et
aux Indes [The Six Voyages of Jean Baptiste Tavernier, Baron
d Aubonne, that He Made to Turkey, Persia, and the Indies].
Clouzier, Paris.
Tillander H. (1996) Diamond Cuts in Historic Jewelry 13811910.
Art Books Intl., London.
Twining E.F. (1960) A History of the Crown Jewels of Europe.
B. T. Batsford, London.
Winters M.T., White J.S. (1991) George IVs blue diamond.
Lapidary Journal, Vol. 45, Nos. 910, pp. 3440.
GEMS & GEMOLOGY
SPRING 2009
19
GRAY-TO-BLUE-TO-VIOLET
HYDROGEN-RICH DIAMONDS FROM
THE ARGYLE MINE, AUSTRALIA
Carolyn H. van der Bogert, Christopher P. Smith, Thomas Hainschwang, and Shane F. McClure
The Argyle diamond mine is the only known source of type IaB hydrogen- and nitrogen-rich
diamonds colored gray to blue to violet. Twenty such diamonds were studied to investigate
the relationships between their spectroscopic characteristics and color grades. The unusual
color is the result of broad absorption bands centered at ~520565 nm and 720730 nm.
A pronounced 551 nm band is superposed on the 520565 nm feature. Spectral deconvolution of this feature suggests that it may be a composite, including H-related absorptions at
~545 and 563 nm and bands at 520530 and 551 nm. The near- and mid-IR regions exhibit
strong absorptions, including those related to H and N, many of which become more intense
with increasing color saturation. The PL spectra exhibit peaks associated with nickel-related
defects, which may play an important role in the coloration of the more violet diamonds in
this group. H-rich gray-to-blue-to-violet diamonds, which are not known to be treated, can be
readily separated from similar-hued diamonds that may be HPHT enhanced or synthetic.
lthough a common impurity in type Ia diamonds, hydrogen is thought to influence the

color of only a small proportion of them
(Fritsch and Scarratt, 1992). These relatively H-rich
diamonds fall into four color groups, as described by
Fritsch et al. (1991, 2007a) and Fritsch and Scarratt
(1992, 1993): (1) brown to grayish yellow to green,
(2) white, (3) chameleon, and (4) gray to blue
to violet. While H-rich brown and yellow diamonds are very common, gray-to-blue-to-violet diamonds are quite rare, particularly stones with high
color saturation and those that are larger than half a
carat (figure 1). Because these diamonds are characterized by their high H content and unusual color,
this article will refer to them as HGBV (H-rich
gray to blue to violet) diamonds.
The gemological characteristics of HGBV diamonds were first described by Fritsch and Scarratt
(1989, 1992, 1993) and Fritsch et al. (1991), along
with their dominant infrared (IR) and ultravioletvisible (UV-Vis) spectral characteristics. The IR and
UV-Vis characteristics were also reported in De
Weerdt and Van Royen (2001). A recent study of
20
GRAY-TO-BLUE-TO-VIOLET DIAMONDS
thousands of small stones provided an updated view

of these spectral characteristics for H-rich diamonds
in general, including the HGBV color group (Fritsch
et al., 2007a). All these studies showed that the IR
spectraof all the color groups of H-rich diamondsare dominated by H-related peaks, and suggested that their UV-Vis spectra may be influenced
by the relatively high H content. Photoluminescence (PL) optical characterization (Iakoubovskii
and Adriaenssens, 2002) and electron paramagnetic
resonance (EPR) measurements (Noble et al., 1998)
of HGBV diamonds have been performed as well;
these studies revealed nickel defects that may also
influence their color.
The present article describes gemological observations and a focused spectroscopic (UV-Vis, IR, and
PL) and DiamondView imaging study of 20 HGBV
diamonds. We describe how these data characterize

GEMS & GEMOLOGY
SPRING 2009
Figure 1. Two of the largest

HGBV diamonds faceted to
date are set in the yellow
metal rings shown here
a 2.34 ct Fancy Dark violetgray emerald cut and a 1.06
ct Fancy Dark gray-blue
ovaland were included in
this study. The ring on the
upper right is a striking
combination of Argyle diamonds, with a 1.38 ct Fancy
red center stone, flanked by
gray-violet and near-colorless diamonds. The loose
0.73 ct pear-shape, also part
of this study, is Fancy Deep
grayish violet. The yellow
metal rings were designed
and manufactured by
Jean Mahie; the other items
were provided by L. J. West
Diamonds. Photo by
Robert Weldon.
the material and relate to its GIA color grades. In

particular, we investigated the different UV-Vis-NIR
spectra of violet- as compared to blue-hued samples
through analysis of their two major absorption minima, which affect their apparent color (e.g., Fritsch
et al., 2007a). In addition, we performed Gaussian
modeling of the major 520565 nm absorption feature to explore whether it is composed of several
individual bands (i.e., a composite), as suggested by
Fritsch et al. (1991).
SOURCE OF HGBV DIAMONDS:

THE ARGYLE MINE
The Argyle mine in northwestern Australia (see,
e.g., Shigley et al., 2001) is the source of H-rich diamonds from at least two different color groups:
brown to grayish yellow to green (Fritsch and
Scarratt, 1993; Massi, 2006; Fritsch et al., 2007a),
and gray to blue to violet (Fritsch and Scarratt, 1992,
1993; Noble et al., 1998; Iakoubovskii and Adriaenssens, 2002). Argyle is currently the only known
source of HGBV diamonds, which represent a
minute proportion of the mines output. The rough
typically occurs as irregular fragments, most of
which are fashioned as melee ranging from a few
points to about 0.20 ct. Much of this material has
an unremarkable, unsaturated color.
Some larger pieces of rough with more saturated
color are encountered, but faceted stones over half a
carat are very rare, and only a few weighing more

than one carat have been recovered in the 25+ years
since mining began at Argyle. When available, the
finest stones of this color variety have been sold
through the yearly Argyle pink diamond tender. For
example, two of the 63 stones offered at the 2006 tender were HGBV diamonds, a proportion that approximately reflects the relative production of tender-quality HGBVs to pinks at the mine (Max, 2006).
Between 1993 and 2008, a total of 20 HGBV diamonds, ranging from 0.39 to 2.34 ct, were offered at
the tender (Argyle Diamonds, 19932008; see G&G
Data Depository at www.gia.edu/gemsandgemology).
The 2.34 ct stone was a Fancy Dark violet-gray emerald cut (again, see figure 1). In spring 2009, Rio Tinto
will offer its first exclusive tender of blue and violet
diamonds, representing select material from several
years of production (Rio Tinto, 2009).
MATERIALS AND METHODS

Samples. During the course of this research, more
than 30 HGBV diamonds passed through the GIA
Laboratory in New York for origin-of-color determination and grading services. Of these, 18 representative stones were selected for study (see table 1 and,
e.g., figure 2). In addition, two exceptional stones set
in yellow metal rings were loaned to the laboratory
(again, see figure 1); both had received GIA Colored
Diamond grading reports prior to being set. All 20 of
GEMS & GEMOLOGY
SPRING 2009
21
the samples were reportedly from the Argyle mine;

they had Argyle inscriptions or were submitted by
Rio Tinto NV (or both). Several were also offered at
various Argyle pink diamond tenders.
Some spectroscopic measurements could not be
made on all the diamonds in this group because of
time constraints. In addition, some samples were
not suited to certain spectroscopic measurements
due to interference from their jewelry settings or
the cut geometry of the stone. However, all the
samples were examined using other standard gemological tests (see below).
Grading and Testing Methods. GIA color grades
were determined by experienced graders using the
standard conditions and methodology of GIAs colored diamond grading system (King et al., 1994).
Internal features were observed with a standard
binocular microscope using a variety of lighting
techniques and magnification up to 100. Most of
the stones received Colored Diamond Identification
and Origin Reports, which do not include clarity
TABLE 1. The 20 HGBV diamonds examined for this
study, listed from bluest to grayest to most violet.
GIA color gradea
FDK gray-blue b
FDK gray-blue
FDK gray-blue c
FAN gray-blue
FAN gray-bluec
FAN gray-blue
FAN gray-bluec
FAN grayish bluec
FDK grayc
FAN violet-grayc
FDK violet-grayb
FAN gray-violet
FDK gray-violet
FDK gray-violetc
FAN grayish violet
FAN grayish violetc
FDP grayish violet
FDP grayish violet
FDP grayish violet
FDP grayish violetc
Weight
(ct)
Shape
Spectra
collected
1.06
1.02
0.23
0.51
0.24
0.21
0.23
0.22
1.57
0.15
2.34
0.26
0.56
0.25
0.16
0.09
0.52
0.65
0.73
0.10
Oval
Shield
Marquise
Pear
Cushion
Pear
Round
Modified heart
Cushion
Round
Emerald
Cushion
Modified oval
Pear
Round
Round
Cushion
Oval
Pear
Round
UV-Vis-NIR; IR; PL
UV-Vis-NIR
UV-Vis-NIR; IR; PL
UV-Vis-NIR; IR; PL
UV-Vis-NIR; PL
IR
UV-Vis-NIR
UV-Vis-NIR; IR; PL
UV-Vis-NIR; IR; PL
UV-Vis-NIR; PL
UV-Vis-NIR; IR; PL
UV-Vis-NIR; IR
UV-Vis-NIR; IR; PL
UV-Vis-NIR; IR; PL
UV-Vis-NIR; IR; PL
UV-Vis-NIR; IR
FDK = Fancy Dark, FAN = Fancy, FDP = Fancy Deep

These diamonds were set in rings, which prevented the collection
of some spectroscopic data.
c
Complete data sets for these stones could not be collected due to
time constraints and/or unfavorable conditions caused by cut
geometry.
b
22
grades, so any discussion of clarity characteristics is

unrelated to specific clarity grades. Reaction to UV
radiation was observed in a darkened room with a
conventional 4-watt combination long-wave (365
nm) and short-wave (254 nm) UV lamp. All the
unmounted stones were also examined with a De
Beers Diamond Trading Co. DiamondView deep-UV
(<230 nm) luminescence imaging system (Welbourn
et al., 1996). We used a handheld spectroscope to
view absorption features in the visible range, with
the stones both at room temperature and after
immersion in liquid nitrogen (~77 K). We checked
for electrical conductivity by putting the diamond
on a metal plate and touching it in several places
with an electrical probe.
Most of the samples were analyzed by a variety
of spectroscopic techniques. For 15 diamonds, we
recorded absorption spectra in the UV-Vis-NIR
region (250850 nm) using a Thermo-Spectronic
Unicam UV500 spectrophotometer with a sampling
interval of 0.1 nm and a spectral bandwidth of 0.5
nm. The samples were mounted in a cryogenic cell
and cooled with liquid nitrogen. Because the path
lengths of the light through the faceted stones were
not known, it was impossible to normalize the UVVis-NIR data so the individual samples could be
directly compared. However, we were able to compare ratios of areas and heights calculated for each
individual spectrum (e.g., Gaffey et al., 1993).
Absorption spectra in the mid-IR (6000400 cm1;
1 cm1 resolution) and near-IR (110002700 cm1; 4
cm1 resolution) regions were recorded for 12 diamonds at room temperature with a Thermo-Nicolet
Nexus 6700 Fourier-transform infrared (FTIR) spectrometer, equipped with KBr and quartz beam splitters. A 6 beam condenser focused the incident beam
on the sample, and 1,024 scans per spectrum were
collected to improve the signal-to-noise ratio. Since
true normalization was not possible, we scaled the
data at 2700 cm1 in the three-phonon region to compare the amplitudes of the IR peaks between stones
of different color grades.
Low-temperature PL spectra were collected on 10
stones using a Renishaw 1000 Raman microspectrometer at three different laser excitations. We used
50-milliwatt Ar-ion lasers for two different laser
excitations: 488.0 nm for the 490850 nm range and
514.5 nm for the 517850 nm range. A He/Ne laser
produced a 632.8 nm excitation for the 640850 or
690850 nm range. The samples were cooled by
direct immersion in liquid nitrogen. Three scans
were collected for each measurement to improve the
GEMS & GEMOLOGY
SPRING 2009
Figure 2. These are three of the HGBV diamonds examined for this study: a 0.65 ct Fancy Deep grayish violet
oval, a 0.56 ct Fancy Dark gray-violet modified oval, and a 1.02 ct Fancy Dark gray-blue shield cut. Photos
by C. H. van der Bogert.
signal-to-noise ratio. For the general PL measurements, the laser was slightly defocused to collect a
bulk measurement. In addition, focused measurements of one sample were taken at the 488.0 nm
excitation to investigate differently fluorescing
regions. All the spectra were scaled to the diamond
Raman peaks, so the amplitudes of the absorption
features could be compared.
Gaussian Modeling. Fritsch et al. (1991) suggested
that a broad band observed in HGBV diamonds at
520565 nm was a composite of multiple absorption
bands. To test this theory, we performed basic spectral deconvolutions for 10 of the highest-quality UVVis-NIR spectra using Origin analytical software
(www.originlab.com) to identify the major absorption bands contributing to the composite feature.
Individual absorption bands were generally assumed
to have a Gaussian (bell curve) shape, and composite
absorption bands were considered the sum of multiple individual Gaussian curves, termed Gaussian
components (e.g., Burns, 1993). The Gaussian components were modeled by spectral deconvolution,
which determines the component positions and
widths that best reproduce a composite absorption
band, such as the one seen in HGBV diamonds (figure 3). The modeling technique requires the establishment of a baseline from which the calculations
are made. This study used a straight baseline defined
by the absorption minima that bounded the proposed composite feature.
with lighter toneFancy Light, Very Light, Light, or

Faintor with higher saturation and moderate tone
Fancy Intense or Fancy Vivid. Thus, the samples had
relatively low saturation and dark tone. When saturation is very low, a hue modifier is assigned, because
the stone appears brownish or grayish (King et al.,
1994). All the blue and violet samples had gray modifiersgray-blue, grayish blue, gray-violet, and grayish
violet. Some samples were primarily gray with blue or
violet modifiersbluish gray and violet-gray (table 1).
Figure 3. Spectral deconvolution of the 520565 nm

broad band using Gaussian components suggests that
it is a composite of multiple absorption bands. The
peak positions were modeled using a baseline defined
by the minima on each side of the composite band.
RESULTS AND ANALYSIS

Color Appearance. The HGBV diamonds in this study
were found to have natural color within a limited
range of fancy grades (table 1): Fancy, Fancy Deep, and
Fancy Dark. None of the samples had fancy grades
GEMS & GEMOLOGY
SPRING 2009
23
Figure 4. Many of the diamonds displayed subtle blue fluorescence zoning when examined with the DiamondView
(A, magnified 20). However, half the stones contained discrete zoned areas with alternating intensities of brighter
blue fluorescence, sometimes with very thin yellowish green fluorescent layers (B, magnified 30) and/or isolated
yellowish green fluorescing regions (C, magnified 40). Photomicrographs by C. H. van der Bogert.
One of the samples had such low saturation that it

was impossible to discern a specific hue, so it was
classified as gray (again, see table 1). None of the samples exhibited bluish violet or violetish blue hues.
Ultraviolet Fluorescence and Phosphorescence. All
the diamonds in this study had consistent fluorescence reactions. When exposed to long-wave UV
radiation, each exhibited a slightly chalky yellow
reaction of moderate-to-strong intensity. Exposure to
short-wave UV produced a weak-to-moderate reaction of the same color. However, samples with weaker saturation (i.e., those that appeared grayer) tended
to phosphoresce weak-to-strong yellow to both longand short-wave UV, while those that were relatively
stronger in saturation did not phosphoresce.
DiamondView Imaging. The samples showed a
variety of reactions when examined with the
DiamondView. Typically, they fluoresced medium
blue with subtle alternating brighter and darker
blue planar and angular zones (figure 4A). Ten of the
18 unmounted samples exhibited discrete areas of

brighter and darker blue layers alternating with very
thin yellowish green layers conforming to cubooctahedral growth zones (figure 4B). Occasionally,
yellowish green fluorescence was present in small
isolated zones (figure 4C); though such zones were
found in samples of all color grades, they tended to
be larger and brighter in violet-hued stones. The differences between the blue and yellowish green areas
were further investigated with focused PL analyses
(see below).
Electrical Conductivity. None of the samples in this
study were electrically conductive.
Microscopic Characteristics. Inclusions. In general,
the HGBV diamonds contained only a limited range
of inclusions, compared to the wider range seen in
diamonds overall. Most common were etch features
(including etch channels), tubes, acicular cavities,
and pits. The etch channels occurred along cleavage
directions and were quite deep in some cases (figure
Figure 5. HGBV diamonds exhibited different types of characteristic internal features. These included (left, magnified
52) coarsely textured channels, where chemical etching occurred along preexisting cracks or cleavages; they can sometimes be very deep, as seen in the 0.73 ct stone in the center (magnified 40). Another type of feature (right, magnified
52), either etch or growth tubes, consists of hollow, icicle-like acicular inclusions. Photomicrographs by C. P. Smith.
24
GEMS & GEMOLOGY
SPRING 2009
Figure 6. Also present in

the HGBV diamonds were
shallow cavities that
resembled furrows with
radiating acicular margins
(left) and fields of small pits
(right), which may be either
etch features or remnants of
inclusions removed during
cutting. Photomicrographs
by C. H. van der Bogert;
magnified 40.
5). Long, icicle-like tubes (figure 5, right), cavities

with clusters with radiating acicular margins (figure
6, left), and small pointy pits (figure 6, right) were
also seen in many of the samples. All these features
occurred singly or as tight clusters and all were surface-reaching. The small pits appeared to be the
remnants of larger features that were removed during cutting.
Only two of the 20 samples contained large
crystal inclusions. These were colorless (possibly
negative) crystals and black to dark brown acicular
crystals, all of which were inert in the DiamondView (figure 7) and could not be identified with the
analytical techniques available. We observed graphitized stress fractures in two other samples. Pinpoints
and clouds were seen in most of the samples, sometimes present as wisp-like patterns or stringers.
Internal Growth Structures, Color Zoning, and
Strain. Viewed through an optical microscope, color
zoning typically occurred as subtle zones with higher saturation (darker gray, or more blue or violet)
than the surrounding areas (figures 7, left, and 8).
The zones typically had either straight or slightly
wavy planar boundaries. Occasionally, rectangular
planar brownish zones or sectors were observed. A
few samples exhibited surface grain lines that were
frequently associated with the color zoning.
Obvious internal growth structures were generally
absent, though turbidity along some of the growth
structures was sometimes seen both in transmitted
light and with the DiamondView imaging system.
The areas with stronger coloration had a weaker
blue luminescence than the weakly colored areas
(figure 7). In transmitted light, the green fluorescent
areas observed with the DiamondView were not
noticeably different from the weaker blue fluorescent areas.
The stones showed weak strain following the
cubo-octahedral growth planes when observed using
crossed polarizers with transmitted light (figure 9).
UV-Vis-NIR Spectroscopy. Observations of Spectra.

When we viewed the HGBV diamonds with a standard desk-model spectroscope, typically only a
weak-to-moderate broad band at approximately 550
nm was observed, though occasionally a weak band
at 594 nm was noted. Stones with more dominant
gray color also exhibited a weak line at 415 nm.
The dominant absorptions in the spectrophotometer data were a series of broad bands centered at
520565 nm, 720730 nm, ~760 nm, and ~835 nm
(figure 10). The 520565 nm feature extended from
approximately 460 to 600 nm, with a typical maximum at 530550 nm. It was superposed by a narrower band centered at 551 nm, typically extending
from 545 to 560 nm at the base. While the centers of
all the broad absorption features varied within the
study group, the 551 nm peak did not.
Figure 7. Mineral inclusions were uncommon in the
HGBV diamonds. When present, they were primarily
clusters of radiating dark brown to black acicular crystals (left), with a form similar to those of some etch features. The planar, slightly wavy color zoning adjacent to
the inclusion was also seen with the DiamondView
(right). Areas with stronger violet coloration had weaker
blue fluorescence than areas with less violet coloration.
Major color zoning is marked with arrows. Photomicrographs by C. H. van der Bogert; magnified 40.
GEMS & GEMOLOGY
SPRING 2009
25
Figure 8. Subtle color zoning

in many of the diamonds
typically consisted of moresaturated violet zones adjacent to less-saturated and
occasionally brownish zones,
which followed indistinct
growth structures or occurred
in sectors. Photomicrographs
by T. Hainschwang (left, in
immersion) and C. H. van
der Bogert (right, diffused
light); magnified 20.
Many weaker peaks and broad bands were also

seen in the UV-Vis range (again, see figure 10). There
was a general increase in absorption into the UV,
with a subtle broad feature centered at 350 nm.
Other weak peaks included a 379/380 doublet and
individual peaks at 405, 415 (N3 zero-phonon line),
425, 445, and 448 nm. Commonly, a weak composite
band centered at 594 nm, roughly 10 nm wide, was
present. In addition, the strongly saturated samples
exhibited very weak peaks at 603, 608, 610, 618, and
630 nm.
Spectral Analysis. Variations between the violetand blue-hued samples were evaluated by comparing
different aspects of their spectra (figure 11A). First,
we investigated the center positions of the two broad
absorption minima in the blue (Bmin) and red (Rmin)

regions. These positions were found to vary with
hue. Samples with violet hues had absorption mini-
Figure 10. The UV-Vis-NIR spectra of HGBV diamonds

are characterized by a strong broad band at 520565
nm, with a smaller superposed 551 nm broad band, a
second strong broad band centered at 720730 nm,
and two smaller broad bands at ~760 and ~835 nm.
The position of the 551 nm band did not vary, but the
positions of the other broad bands varied slightly with
color grade. The spectra were collected at ~77 K and
are offset vertically for clarity. FAN = Fancy, FDK =
Fancy Dark, and FDP = Fancy Deep.
Figure 9. A low degree of internal strain was characteristic of the HGBV diamonds, which exhibited only
subtle gray and blue interference colors, with angular
streakiness characteristic of cubo-octahedral growth.
This is in contrast to the banded and tatami-patterned strain characteristic of type IIb blue and type
IIa pink diamonds, and the high degree of mottled
strain typical of type Ia pink-to-purple diamonds.
Photomicrograph by T. Hainschwang; magnified 20.
26
GEMS & GEMOLOGY
SPRING 2009
ANALYSIS OF UV-VIS ABSORPTION SPECTRA
ABSORBANCE
Ba = Blue transmission area
Ra = Red transmission area
Bd = Blue transmission depth
Rd = Red transmission depth
Bmin = Blue absorption minimum
Rmin = Red absorption minimum
Rd
Bd
Ra
Ba
Rmin
Bmin
400
450
500
550
600
650
700
WAVELENGTH (nm)
POSITION OF MINIMA IN RED AND BLUE REGIONS

620
MINIMUM IN RED REGION-Rmin (nm)
Gaussian Modeling. Each of the 10 UV-Vis-NIR

spectra was modeled in two different ways to
explore the possibility that the 520565 nm absorption was a composite of multiple bands. First, we
allowed the software to automatically calculate peak
positions based on three to four starting positions.
For example, the starting positions selected for the
composite feature in figure 3 were 522, 542, 551, and
563 nm, based on the inflection points at the last
three positions and the large width of the feature
toward the UV. Our software allowed the peak positions and widths to vary without constraint in order
to maximize the fit of the Gaussian components to
the composite feature. A model was considered successful when multiple model runs using a variety of
starting positions resulted in reasonable positive
peak heights and areas, as well as similar final peak
positions for each modeling attempt, and when R2 (a
measure of how well the model fits the spectrum)
was 0.99 or greater. The goal was to achieve the best
statistical fit of the composite feature while using as
few peaks as possible. While models with more than
three or four peaks could yield improved R2 values,
the additional peaks would represent a statistically
insignificant proportionless than 1%of the overall composite feature.
Using this approach, we found that the 520565
nm absorption in most of the samples was a
composite feature, best modeled as a combination
of four independent absorption peaks centered at
520530 nm (with an average value of 527 nm),
FDP grayish violet

FAN grayish violet
610
FDK gray-violet
FAN violet-gray
FAN gray-blue
600
FDK gray-blue
590
580
570
560
440
450
460
470
480
490
500
MINIMUM IN BLUE REGION-Bmin (nm)
RATIO OF AREA TO DEPTH

1.8
1.6
Bd/Rd
ma at 455467 nm in the blue region and 601613

nm in the red, whereas samples with blue hues had
minima at higher wavelengths (460490 nm) in the
blue and lower wavelengths (568602 nm) in the red
(figure 11B). On average, the positions of the minima
were 127 nm apart for the violet-hued samples, and
108 nm apart for the blue-hued samples.
Next, we compared the relative strengths (i.e.,
the areas; see figure 11A) and depths (figure 11C) of
the absorption minima in the blue and red regions.
We calculated the ratio of the area of the blue
absorption minimum (Ba) between 380 and 551 nm
to the area of the red absorption minimum (Ra)
between 551 and 720 nm, and the ratio of the depth
of the blue absorption minimum (Bd) to the depth of
the red absorption minimum (Rd). Most of the samples with blue hues had Ba/Ra (1.92.1) and Bd/Rd
(1.51.7)higher than those with violet hues. The
violet-hued samples had Ba/Ra of 1.41.7 and Bd/Rd
of 1.31.5 (figure 11C).
FDP grayish violet
1.4
FAN grayish violet

FDK gray-violet
FAN violet-gray
FAN gray-blue
1.2
FDK gray-blue
1.2
1.4
1.6
1.8
2.0
2.2
Ba/Ra
Figure 11. The spectral differences between the samples

were investigated by comparing the positions of the
absorption minima of the major transmission windows
in the blue and red regions (A and B), and the ratios of
the areas of these windows to their depths (A and C).
GEMS & GEMOLOGY
SPRING 2009
27
position of the 543549 nm peak varied widely. For

these spectra, reproducible model fits were possible
only by using three peaks: 526530 nm (average
529 nm), 551 nm (without variation), and 556572
nm (average 563 nm).
Then, based on these results, we tested the
hypothesis that the composite feature consists of a
551 nm peak plus two hydrogen-related peaks545
and 563 nmseen in other colors of H-rich diamonds.
For these model runs, we constrained the position of
the 545, 551, and 563 nm peaks to test whether valid
models could be fit to the spectra under these conditions. For all 10 spectra, we were able to fit the composite feature with peaks at 545, 551, and 563 nm if a
fourth peak in the region 520528 nm was allowed to
be present. The position of the fourth peak was not
constrained. For these models, R2 ranged between
0.987 and 0.998, indicating that these peaks are reasonably good fits to the composite feature. It should
be kept in mind, however, that these models do not
represent unique solutions.
Figure 12. A strongly colored (Fancy Deep) grayish

violet diamond had the most intense absorption
peaks across all measured regions as shown in this
composite spectrum (A). The peaks were seen in
other HGBV diamonds, but with lower intensity.
HGBV diamonds are characterized by multiple
sharp, strong, hydrogen-related peaks. These include
the 3107 cm1 peak and its associated peaks at 6068,
5888, 5556, and 4495 cm1, as well as those at 3237
and 2786 cm1. Numerous other small peaks
observed throughout the mid- to near-IR region (B)
were characteristic of all the diamonds in this study
(see also figure 13). It was possible to identify these
diamonds as type IaB and, in some cases, to calculate the concentration of IR-active nitrogen (C).
543549 nm (average 546 nm), 551 nm (without

variation), and 557566 nm (average 563 nm). For
some samples, however, the models were not consistently reproducible using four peaks; the final
28
Infrared Spectroscopy. All the diamonds proved to

be type IaB, with a very high concentration of
hydrogen (Woods and Collins, 1983; e.g., figure 12).
Although total absorption in the one-phonon region
was common, for five samples (e.g., figure 12C) it
was possible to use spectral fitting techniques
(Kiflawi et al., 1994; Boyd et al., 1995) to determine
that the concentration of IR-active nitrogen ranged
from 1350 to 2700 ppm. We observed a platelet
peak extending from 1376 to 1368 cm1, seen in
another study of Argyle diamonds (Noble et al.,
1998), in only a few samples. We typically recorded
bands at ~1550, 1515, and 1499 cm1 (figure 12A); in
some samples, peaks at 1435, 1405 cm1 were also
present, along with broad features at ~1010, 780,
and 550 cm1 (figure 12C).
After the IR absorptions associated with nitrogen, the most significant were those related to
hydrogen (e.g., Woods and Collins, 1983; Fritsch and
Scarratt, 1989; Fritsch et al., 1991). A series of sharp
peaks were recorded between 3300 and 2700 cm1,
with two dominant peaks at 3237 and 3107 cm1
(figures 12B and 13A,B). Several other weaker bands
were recorded between 4705 and 4168 cm1 (figures
12A and 13C). Very weak bands at 4678, 4412,
4236, and 4224 cm1 were sometimes present.
Weak, sharp bands were observed at 6068, 5888, and
5556 cm 1 (figure 12). Another series of broad
absorption bands were positioned at 10570 cm1
(946 nm), 10300 cm1 (970 nm), 10200 cm1 (980
GEMS & GEMOLOGY
SPRING 2009
Figure 13. Across the IR spectrum, the strength of the IR peaks and broad bands correlated with color. The violet-hued stones had stronger IR absorptions than the blue-hued stones for peaks between 4800 and 2700 cm 1
that have been attributed to hydrogen (A, B, C). Broad peaks of unknown origin between 10570 and 8250 cm 1
were also stronger in violet-hued stones, but the strengths of those at 7850 and 7490 cm 1 did not seem to
depend on color (D). The spectra were scaled at 2700 cm 1 and are offset vertically for clarity.
nm), 10090 cm1 (991 nm), and between 8630 and

7490 cm1 (figures 12 and 13D).
When we compared the IR peaks of stones with
different color grades, we found that the strengths of
these peaks generally depended on diamond hue.
The 3107 cm1, associated 2786 cm1, and 3237 cm1
peaks were more intense in violet- than in bluehued diamonds (figure 13A). The same relationship
between peak intensity and hue also was seen for
the weaker peaks in the regions 33003000 cm1
(figure 13B) and 48004000 cm1 (figure 13C), as
well as for the 6068 and 5888 cm1 peaks. The
strongest broad peaks between 11000 and 8000 cm1
were recorded in Fancy Deep grayish violet stones
(figure 13D). However, except for one such grayish

violet diamond with very strong absorption, the
Fancy gray-blue stones exhibited stronger absorptions at 7850 and 7490 cm1 than the diamonds
with other color grades.
Photoluminescence Spectroscopy. As noted above,
we used three different laser excitations to investigate the PL properties of the samples. Many PL centers were activated by different excitations to varying degrees, while others were only active for specific excitations (figure 14). All the spectra were scaled
to the Raman peaks, so we could compare the
amplitudes of the absorption features.
GEMS & GEMOLOGY
SPRING 2009
29
Figure 14. These spectra, typical for HGBV diamonds, were collected for a Fancy Deep grayish violet diamond cooled to ~77 K at three excitation
wavelengths (488, 514, and 633 nm). The 488 nm
excitation caused strong luminescence in the
650750 nm region, with dominant PL bands at 669,
689, and 701 nm, and weaker bands at 659 and 717
nm. Weak broad peaks were present at 503.2 (H3),
604, 613, 626, and 645 nm. The 514 nm excitation
exhibited strong peaks at 669 and 701 nm, but it also
showed distinct peaks at 694 and 717 nm. A 794 nm
(NE8) peak was strongest under 633 nm excitation.
The spectra were scaled to the Raman peaks (R) and
offset vertically for clarity.
488 nm Excitation. With 488 nm excitation, the

dominant luminescence consisted of a series of
strong broad bands between ~650 and 750 nm (figures
14 and 15A). These included strong peaks at 669, 689,
and 701 nm, with weaker bands at 659 and 717 nm.
The asymmetries of some of these features suggest
they are composite bands. In particular, the 689 nm
peak likely includes a 694 nm luminescence, as
observed with the 514 nm excitation (see below).
A weak 604 nm peak was often present with its
associated vibronic system at 613 and 626 nm, as
described by Iakoubovskii and Adriaenssens (2002).
Some stones had weak broad bands positioned at 496,
504, and 512 nmthe higher-energy phonon replicas
of the S2 (NE1) zero-phonon line at 489.2 nm, which
itself could not be observed because of its proximity
to the laser line. Seen in natural type IaB diamonds,
the NE1 is associated with nickel-vacancy complex
defects (Nadolinny and Yelisseyev, 1994).
Two stones exhibited a 496.7 (S3; Zaitsev, 2001)
nm band, which is associated with mixed-habit
cubo-octahedral growth. Some of the spectra also
exhibited peaks at 503.2 nm (figure 14) or at 502.8
and 503.2 nm (figure 15A, inset). The 503.2 nm
peak is seen in nitrogen-containing diamonds (H3;
30
Figure 15. PL spectra were collected for different colors

of HGBV diamonds cooled to ~77 K using three excitation wavelengths: (A) 488 nm, (B) 514 nm, and (C) 633
nm. The spectra were scaled to the Raman peaks (R),
and the insets and (C) are offset vertically for clarity.
Zaitsev, 2001). Most of the spectra had a small

broad peak centered at 524 nm. The violet-hued
stones exhibited a 550.6 nm peak, which corresponds to the 551 nm peak seen in the UV-Vis-NIR
spectra (figure 15A). A 746.8 nm peak, observed in
the Fancy Dark gray-violet stone, is thought to be
GEMS & GEMOLOGY
SPRING 2009
associated with Ni (Yelisseyev and Kanda, 2007).

The major luminescence peaks (669, 689, and
701 nm) were stronger for violet-hued diamonds, and
their intensity increased with color saturation (figure
15A). The 604 nm system was active only in the
stones with the most violet color, those graded grayish violet. The 604, 659, 669, and 701 nm centers
have been attributed to Ni defects (Bokii et al., 1986;
Iakoubovskii and Adriaenssens, 2002), while the
645, 689, and 694 nm peaks are tentatively attributed to Ni defects (Iakoubovskii and Adriaenssens,
2002). The strengths of the peaks between 496 and
524 nm do not seem to have a direct relationship to
hue or saturation.
514 nm Excitation. Many of the same features seen
with 488 nm excitation were also present under 514
nm excitation, but they had slightly different characteristics (figures 14 and 15B). The dominant peaks
again included those at 669 and 701 nm. The composite structure of the 689 nm peak became apparent as the 689 and 694 nm peaks resolved into a pair
of peaks. In addition to the 659 and 717 nm bands
on the flanks of the strongest luminescence peaks,
there was a weak peak at 727 nm. The 717 and 727
nm peaks are phonon sidebands associated with the
701 nm vibronic system (Iakoubovskii and
Adriaenssens, 2002).
The 604 nm system was also present at this
excitation, in addition to a weak peak at 645 nm,
both of which were stronger in the violet-hued
stones. Some of the samples exhibited small peaks
at 524.1 nm and weak bands at 521.8, 528.4, 532,
537.8, and 542.7 nm (figure 15B, inset). The most
violet of these stones also showed a 529.8 nm peak.
Two of the samples only exhibited small peaks at
528.4 and 542.7 nm. All had a 584.4 nm peak, the
strength of which was greater in the violet-hued
stones. As in the 488 nm excitation, the overall
strength of the dominant peaks was strongest in the
stones with the most saturated violet color.
633 nm Excitation. With this excitation, the 794 nm
(NE8) peak was dominant (figures 14 and 15C).
Additional peaks were present at 694, 701, 717, and
727 nm. The 669 nm peak appeared below the
Raman peak in some cases. Many of the stones
exhibited a GR1 zero-phonon line (741/744 doublet),
which indicates radiation damagein this case, natural radiation damage, because the stones were not
treated in any way.
For this excitation, the relationship to color is
less clear (figure 15C). However, the two stones with

the strongest 794 nm peaks were both grayish violet.
The gray-blue stones, however, had stronger luminescence peaks than the Fancy Dark gray-violet
stone. Some stones with more saturated violet color
also exhibited broad weak peaks at 788, 813, and 830
nm. The 813 and 830 nm features are associated
with the zero-phonon line at 794 nm (Iakoubovskii
and Adriaenssens, 2002). The 788 and 794 nm centers have been attributed to Ni defects (Bokii et al.,
1986; Iakoubovskii and Adriaenssens, 2002).
488 nm Excitation Spot Analyses. We also investigated the PL differences between the yellowish
green and blue fluorescing regions seen with the
DiamondView (figure 4C) in a Fancy Deep grayish
violet stone using the 488 nm excitation. We collected PL spectra from three focused areas: a yellowish green spot, a weak-to-moderate blue spot, and a
stronger bright blue spot (figure 16). The bright blue
spot corresponded to a region of less saturated color
in the sample, while the yellowish green and weakto-moderate blue fluorescing areas correlated with
more saturated violet color. The yellowish green
spot exhibited the strongest photoluminescence.
Figure 16. PL spectra for three areas with differently

colored ultra short-wave UV fluorescence were collected from a Fancy Deep grayish violet diamond
using an excitation wavelength of 488 nm. Areas
with yellowish green fluorescence exhibited stronger
Ni-related peaks than those with blue fluorescence.
The spectra were scaled to the Raman peak (R).
3
2
1
GEMS & GEMOLOGY
SPRING 2009
31
It contained the typical strong 669, 689, and 701 nm

peaks, with accompanying 659 and 717 nm peaks.
The 604 nm series was also present. The spectra
also showed peaks at 496.7, 504.1, 513, and 524 nm
(figure 16, inset).
The spectra from the blue-fluorescing regions
had much lower overall luminescence in the 669,
689, and 701 nm peaks. Yet they had greater luminescence in the 495575 nm region, where the spectra exhibited a 503.2 nm (H3) peak and its vibronic
bands. No 524 nm peak was observed, but a 535.9
nm peak associated with (natural) radiation damage
was present (again, see figure 16, inset).
DISCUSSION
Gemological Characteristics. The gemological characteristics of the HGBV diamonds in this study
etch features, color zoning, inclusions, and fluorescence reactionswere consistent with the diamonds described in Fritsch and Scarratt (1992, 1993)
and more than 100 other small (0.030.10 ct) HGBV
stones that were studied by Fritsch et al. (2007a).
Microscopic Characteristics. The internal features
seen in HGBV diamonds are similar to those seen in
pink diamonds from Argyle (e.g., Hofer, 1985;
Fritsch et al., 2006). Some may have formed via geologic processes, such as dislocation- or cleavage-controlled etching (Lu et al., 2001; Fritsch et al., 2006),
while others may be the result of the dissolution of
crystal inclusions during processing and cleaning
(Chapman, 1992). Indeed, the icicle-like acicular features and the cavities with radiating acicular points
were quite similar in form to the inclusions with
radiating acicular crystals (figure 7, left), so it is possible that mineral inclusions exposed at the surface
of the diamond dissolved during cleaning.
Fluorescence. The yellow fluorescence of the samples
to both long- and short-wave UV radiation may be
related to their high H content, because it also occurs
in H-rich diamonds of other colorschameleon,
gray-green, and orange (Eaton-Magaa et al., 2007).
When observed at shorter wavelengths using
the DiamondView, the samples generally exhibited blue fluorescence. Caused by the N3 defect
groups of three nitrogen atoms and a vacancy
along a {111} plane (e.g., Collins, 1999)this is
the most common fluorescence color in diamonds (Moses et al., 1997). The presence of N3
defects is not surprising, since the IR-active N
32
contents of the samples ranged up to 2700 ppm.

Some samples also exhibited yellowish green
layers following the growth structures. Green fluorescence has typically been ascribed to the H3 center (Dyer and Matthews, 1958)a vacancy plus an
A-aggregate of nitrogen (Davies, 1972). However,
Noble et al. (1998) and Iakoubovskii and Adriaenssens (2002) suggested that green fluorescence in
HGBV diamonds might be related to nickel impurities. Spot PL analyses of one sample (figure 16) support this idea (see below).
Color Appearance and UV-Vis Characteristics.
Because HGBV diamonds typically have very low
color saturation and dark tone, no single hue or
color description succinctly captures the color
grades exhibited by this group. This has led both the
trade and scientific literature to use a range of
names to describe this materialfor example,
hydrogen blue or Argyle blue (Max, 2006), violet (Federman, 2003, 2007), and blue-gray
(Iakoubovskii and Adriaenssens, 2002), gray-blue
(Noble et al., 1998), and blue-gray-violet (De
Weerdt and Van Royen, 2001).
One of the goals of this article was to refine the
color terminology for HGBV diamonds. However, the
range of color grades exhibited by the study samples
further complicates the description of these diamonds based solely on color. We believe that calling
them gray to blue to violet, as in Fritsch et al.
(2007a), or HGBV diamonds, remains the simplest
solution, with the caveats that these color descriptions typically do not include unmodified blue or violet hues, and that none of the study samples (nor any
of the diamonds offered at tender from 1993 to 2008)
had bluish violet or violetish blue hues.
The apparent color of these diamonds results
from the additive effect of the two broad absorption
minima or transmission maximaalso known as
transmission windowsin the blue and red regions
(figure 10; see also Fritsch et al., 2007a). We found
that the center positions of the transmission windows changed with hue: Samples with minima in the
blue and red regions were violet-hued, whereas those
with minima in the orange-to-yellow and blue-togreen regions were blue-hued (again, see figure 11B).
The apparent colors, however, depend not only
on the positions of the transmission windows, but
also on the windows relative strengths (Fritsch et
al., 2007a), which we investigated through a comparison of the ratios of the areas and depths of the features (figure 11C). As the ratio Ba/Ra increases, the
GEMS & GEMOLOGY
SPRING 2009
UV-VIS-NIR ABSORPTION SPECTRA
Comparison with Type IIb Diamonds. Some HGBV

color grades overlap those of electrically conductive
type IIb diamonds (King et al., 1998), but there are
usually subtle differences. This was demonstrated
by Darley and King (2007) for another H-rich bluegray diamond that was observed in GIAs New York
laboratory. Even some HGBV diamonds that fall
within the same color space as type IIb diamonds
can be distinguished on the basis of these subtle
color differences. For example, grayish blue and
gray-blue HGBV diamonds tend to be slightly more
violet than their type IIb equivalents, though they
occupy the same color space and thus receive the
Type
Type
Type
Type
IaAB pink
IaA purple
IaB HGBV
IIb blue
H3
N3
725730 band
350
band
551
ABSORBANCE
area (strength) of the blue window increases relative

to the red window. For example, a value of 1 means
that the absorptions in the blue and red regions were
of equal strength, whereas a value of 2 means that
the absorptions in the blue region were twice as
strong as those in the red. Our data showed that
HGBV diamonds with blue hues had higher Ba/Ra
and Bd/Rd values than those with violet hues (figure
11C). The blue stones absorption in the blue region
was about twice as strong as in the red region. Violet
stones had Ba/Ra = 1.41.7 and Bd/Rd = 1.31.5, so
the area and depth of the absorption in the blue
region still exceeded that of the absorption in the red
region, but less than they did for the blue stones. It is
unclear whether stones with B a /R a 1.4 and
Bd/Rd 1.3 would be predominantly gray, blue, or
violet, because there were samples with three different color grades in this area of the plot (figure 11C).
In summary, when the absorption band in the blue
region is broad, flat, and has an area twice as large as
that of the red band, the stone appears blue; when it
is relatively deep, with steep flanks, and an area ~1.5
times that of the red band, the stone appears violet.
The presence and relative strengths of the two
transmission windows also explain why HGBV diamonds, particularly those with violet hues, have
slightly different colors under different lighting conditions. Those with strong color saturation tend to
appear slightly bluer under fluorescent lighting
(stronger illumination in the blue region) and slightly
more violet under incandescent lighting (stronger
illumination in the red region). The violet-hued
stones have transmission windows that are positioned wider apart than the blue-hued stones, so they
are more sensitive to lighting conditions, especially
those with different intensities in the blue and red
wavelengths. The difference, however, is not great
enough to consider them color-change diamonds.
280 nm
minimum
300
400
500
600
700
800
WAVELENGTH (nm)
Figure 17. The UV-Vis-NIR spectra of HGBV diamonds

differ greatly from the spectrum of a typical type IIb
blue diamond; they are more similar to pink and purple diamonds. However, the 520565 nm composite
feature in HGBV diamonds is different from the
550560 nm feature seen in some pink and purple diamonds, as shown by Gaussian spectral deconvolution
(figure 3). The spectra were collected at ~77 K.
same color grades. In all cases, their UV-Vis-NIR

spectra are very different, with most type IIb diamonds characterized by a simple spectrum consisting of an absorption minimum at 280 nm and a
smooth, almost linear increase in absorption into
the near-infrared region (figure 17). HGBV diamonds
are also readily distinguished from type IIb diamonds by their nonconductivity and differences in
UV fluorescence (e.g., Fritsch and Scarratt, 1992).
The quick separation of HGBV and type IIb diamonds is useful because HGBV diamonds are not
known to be treated to enhance their color, whereas
the color of type IIb diamonds may be improved via
high-pressure, high-temperature (HPHT) treatment.
In addition, synthetic blue diamonds typically fall
into two categoriesmixed types IIb+IIa or IIb+Ib
that also have different gemological properties from
HGBVs (e.g., Shigley et al., 2004). The identification
of a diamond as HGBV supports a natural origin and
precludes special testing to determine whether
HPHT treatment has occurred.
One unusual natural mixed type IIa+IIb diamond
that was color graded light violet-gray has been documented (Reinitz and Moses, 1993). However, it
was concluded that the color of this diamond was
caused by the additive effect of separate pink type
IIa and gray type IIb zones.
GEMS & GEMOLOGY
SPRING 2009
33
Comparison with Pink and Purple Diamonds.

Previous work on HGBV diamonds (Noble et al.,
1998) suggested that the 530550 nm band had the
same origin as the 550560 nm broad band seen in
some pink and purple diamonds (Collins, 1982;
Moses et al., 2002; figure 17). In type IIa pink diamonds, the 550560 nm broad band may result
from nitrogen-vacancy defects (N-V centers), which
also cause two strong absorptions at 575 and 637
nm (e.g., Scarratt, 1987; Fritsch et al., 2007b). The
520565 nm composite band in HGBV diamonds
clearly is not related to the N-V center, because
neither the Gaussian decomposition nor the PL
spectra showed major absorption features centered
at 575 or 637 nm.
The defect responsible for the 550560 nm broad
band in type Ia pink and purple diamonds has not
been established. It may be caused by plastic deformation, which has been observed in many pink and
purple diamonds as anomalous double refraction
(ADR) or high strain, and is typically, but not exclusively, confined to thin strain-related lamellae
(Orlov 1977; De Weerdt and Van Royen, 2001; King
et al., 2002; Moses et al., 2002; Titkov et al., 2008).
HGBV diamonds do not show a high degree of ADR
or strain, and their color is related to growth structures rather than deformation lamellae (figure 7); if
the 550560 nm absorption feature in pink and purple diamonds is the result of strong deformation,
then the 520565 nm absorption feature in HGBV
diamonds does not have the same origin.
Roles of Hydrogen, Nitrogen, and Nickel.
Hydrogen. The relatively high hydrogen concentration in HGBV diamonds has been documented
previously (e.g., Fritsch and Scarratt, 1989; Fritsch
et al., 1991) and was recently reviewed and updated (Fritsch et al., 2007a). The H-related features
usually seen in IR spectra include the 3107 cm1
band and associated peaks at 6068, 5556, 4495,
4168, 2786, and 1405 cm1, which are likely the
result of C-H vibration interacting with aggregated
nitrogen. The peaks between 4705 and 3237 cm1
may correspond to N-H vibration (Fritsch et al.,
2007a). The origins of all the remaining IR peaks
have not been determined thus far, but many have
been observed in H-rich diamonds of various colors (Fritsch et al., 2007a).
The possible role of hydrogen in the coloration
of H-rich diamonds, including HGBV diamonds,
was first suggested by Fritsch and Scarratt (1989).
Indeed, as illustrated by our Gaussian modeling of
34
the UV-Vis spectra of HGBV diamonds, they may

exhibit some of the same optical absorption bands
as other H-rich diamonds. Peaks at 545 and 563 nm,
possibly related to high hydrogen content, have
been seen in brownish yellow or Cape yellow (e.g.,
De Weerdt and Van Royen, 2001) and chameleon
(e.g., Hainschwang et al., 2005) diamonds. To date,
however, it has not been shown definitively that IRactive hydrogen causes optically active absorption
features, including the 545 and 563 nm bands
(Fritsch et al., 2007a). Regardless, the presence of
hydrogen probably is not the only factor involved in
the coloration of H-rich diamonds, which exhibit a
range of colors. It seems clear, though, that it may
be an important prerequisite for the formation of
the defects responsible for color in H-rich diamonds,
as suggested by Iakoubovskii and Adriaenssens
(2002) for HGBV diamonds.
Nitrogen. Five of the samples examined here exhibited high concentrations of IR-active nitrogen, present
primarily as B aggregates (four-nitrogen defects),
from 1350 to 2700 ppm. This confirms previous estimates that HGBV diamonds contain at least 500
ppm N (Fritsch and Scarratt, 1989). Nitrogen was
found in HGBV diamonds not only as B aggregates,
but also as low concentrations of N3 and N2 complexes (two types of three-nitrogen complexes;
Iakoubovskii and Adriaenssens, 2002). Nitrogen and
hydrogen concentrations are linked, such that they
increase commensurately (e.g., Iakoubovskii and
Adriaenssens, 2002; Rondeau et al., 2004), so it is not
surprising that HGBV diamonds also have relatively
high N contents.
Nickel. This element is an important factor in the
PL spectra of HGBV diamonds. Numerous peaks
attributed and tentatively attributed to Ni-N defects
were seen (604, 645, 659, 669, 689, 694, 701, 788,
and 794 [NE8] nm centers). In fact, in the present
study the dominant luminescence features were
stronger for the violet stones and also depended on
the color saturation (figure 15). This indicates that
the violet diamonds had higher concentrations of
certain Ni-related defects than the blue diamonds.
While the PL spectra for the blue fluorescing
areas did exhibit a peak at 503.2 nm (H3), the PL
spectra for the yellowish green fluorescing area did
not (figure 16). In addition, the yellowish green
area had much stronger Ni-related peaks than the
blue areas. This evidence supports the idea that
yellowish green fluorescence in HGBV diamonds
GEMS & GEMOLOGY
SPRING 2009
is related to Ni-defects rather than the H3, as suggested by Iakoubovskii and Adriaenssens (2002)
and Noble et al. (1998).
It is unknown what defects are responsible for
most of the IR peaks from 10570 to 7490 cm1, but a
peak at 10300 cm1 has been seen in some synthetic
Ni- and N-doped diamonds (Zaitsev, 2001). It
remains to be tested whether these peaks, also seen
in some other H-rich diamonds (Fritsch et al.,
2007a), are related to H-, Ni-, and/or other defects.
One outstanding question is whether the Nirelated PL peaks in HGBV diamonds are optically
active, as are all the Ni-related centers discovered
thus far during EPR studies of synthetic diamonds
(e.g., Yelisseyev and Kanda, 2007). One study of
HPHT synthetic diamond grown in the Fe-Ni-C
system (Yelisseyev et al., 2002) observed optical
absorptions at 515.5, 518.0, 520.0, and 527.3 nm
resulting from Ni-N complex defects that are possibly related to the 727 nm PL center (Lawson and
Kanda, 1993; Yelisseyev and Nadolinny, 1995).
These defects may contribute to the 520530 nm
component of the composite band seen in HGBV
diamonds, especially since a 727 nm PL center was
seen at the 514 and 633 nm excitation wavelengths.
Alternatively, the 520530 nm component could be
related to the 524 nm band observed in the PL spectra. It is not clear whether this band might be related to the NE3, a Ni-related center at 523.3 nm.
A 732 nm optical center is also related to Ni-N
complex defects (Lawson and Kanda, 1993), as
observed in HPHT synthetic diamonds (Yelisseyev
et al., 2002). Previous observations of HGBV diamonds (Noble et al., 1998) noted that it is impossible
to speculate on the origin of the broad band centered
at 720730 nm because of the absence of sharp lines
or structures. However, our data indicate that this
absorption is stronger for violet-hued HGBV diamonds with more saturated color (figure 10). Further
EPR studies of HGBV diamonds with a range of
color grades are necessary to investigate the origins
of possible Ni-related optical features more fully.
Higher Ni concentrations in diamond may also
correspond to higher N and H concentrations (e.g.,
Lang et al., 2004). It follows that when both Ni and
N are present more Ni-N defects could form, for
example, in HGBV diamonds. Indeed, Noble et al.
(1998) suggested that Ni-N interactions might
influence the color of these diamonds. Yet Ni is
also present, for example, in chameleon diamonds
(Hainschwang et al., 2005) and some natural-color
green-yellow diamonds (Wang et al., 2007); thus,
Figure 18. This rare 1.41 ct Fancy Dark gray-violet

octagonal-cut HGBV diamond, named Ocean
Seer, was sold at the 2008 Argyle pink diamond
tender. Courtesy of Argyle Diamonds.
the colors of these diamonds do not depend simply

on the presence of Ni. Perhaps the variety of colors
could result from different Ni defects. For example,
one of the EPR centers observed by Noble et al.
(1998), proposed as Nis with N + in a fourth nearestneighbor position, was seen only in HGBV diamonds in their study and not in other diamond colors from the Argyle mine. High H and N concentrations may be required for different Ni and Ni-N
defects to form in HGBV and other H-rich diamonds that ultimately give rise to different apparent colors. A slight variation in the relative concentrations of specific defects could be responsible for
the color variations in HGBV diamonds.
CONCLUSIONS
One of the more unusual varieties of fancy-color diamonds is that represented by type IaB hydrogen-rich
gray to blue to violet, or HGBV, diamonds (see,
e.g., figure 18), which to date are only known from
the Argyle mine in Australia. This study of HGBV
diamonds revealed systematic relationships between
their spectroscopic properties and color grades. It
also confirmed previous work documenting numerous H-related IR peaks in HGBV diamonds. Many
H-related peaks were stronger in the violet- than the
blue-hued samples, indicating that violet-hued
stones may have higher H concentrations. We also
confirmed and expanded on work showing that nick-
GEMS & GEMOLOGY
SPRING 2009
35
el-related defects are observed in the PL spectra of

HGBV diamonds. Indeed, the luminescence of the
Ni-related peaks was stronger in the violet- than the
blue-hued samples, suggesting a higher concentration of these Ni-defects in the violet-hued stones. In
addition, we showed that yellowish green fluorescence in HGBV diamonds is correlated with PLactive Ni-defects.
The color of HGBV diamonds is the additive
effect of two transmission windows in the blue and
red areas of the visible region. The positions and
ABOUT THE AUTHORS

Dr. van der Bogert is a senior research scientist at Westfaelische
Wilhelms University in Muenster, Germany. At the time this article was accepted, Mr. Smith was vice president and chief
gemologist at the American Gemological Laboratories in New
York. Mr. Hainschwang is director of the Gemlab Laboratory for
Gemstone Analysis and Reports in Liechtenstein. Mr. McClure is
director of Identification Services at the GIA Laboratory in
Carlsbad, California. Both Dr. van der Bogert and Mr. Smith performed research for this article while employed at the GIA
Laboratory in New York, and Mr. Hainschwang did so with the
former GIA GemTech Laboratory in Geneva, Switzerland.
ACKNOWLEDGMENTS
The authors thank John King, Matthew Hall, Dr. Wuyi Wang, Dr.
REFERENCES
Argyle Diamonds (19932008) Pink Diamond Tender [annual
catalog]. West Perth, Australia.
Bokii G.B., Bezrukov G.N., Klyuev Yu.A., Naletov A.M., Nepsha
V.I. (1986) Natural and Synthetic Diamonds. Nauka, Moscow
(in Russian).
Boyd S.R., Kiflawi I., Woods G.S. (1995) Infrared absorption by the
B nitrogen aggregation in diamond. Philosophical Magazine B,
Vol. 72, No. 3, pp. 351361.
Burns R.G. (1993) Mineralogical Applications of Crystal Field
Theory. Cambridge University Press, Cambridge, UK, 575 pp.
Chapman J. (1992) Letters: Hollow hexagonal columns in diamond not etch pits. G&G, Vol. 28, No. 1, p. 73.
Collins A.T. (1982) Colour centers in diamond. Journal of
Gemmology, Vol. 18, No. 1, pp. 3775.
Collins A.T. (1999) Things we still dont know about optical centers in diamond. Diamond and Related Materials, Vol. 8, No.
89, pp. 14551462.
Darley J., King J.M. (2007) Lab Notes: Natural color hydrogenrich blue-gray diamond. G&G, Vol. 43, No. 2, pp. 155156.
Davies G. (1972) The effect of nitrogen impurity on the annealing
of radiation damage in diamond. Journal of Physics C: Solid
State Physics, Vol. 5, No. 17, pp. 25342542.
De Weerdt F., Van Royen J. (2001) Defects in coloured natural
diamonds. Diamond and Related Materials, Vol. 10, No. 37,
pp. 474479.
Dyer H.B., Matthews I.G. (1958) The fluorescence of diamond.
Proceedings of the Royal Society of London A, Vol. 243, No.
1234, pp. 320335.
36
strengths of these windows directly affect apparent

color. Spectral deconvolution showed that the composite absorption feature centered at 520565 nm
could be a combination of peaks observed in other
H-rich diamonds (545 and 563 nm) and 551 and
~520530 nm peaks. This finding underscores their
close relationship to other H-rich diamonds. Overall,
the gemological and spectroscopic characteristics of
HGBV diamonds separate them from other types of
diamonds of similar color that may be treated or
even synthetic.
Andy H. Shen, and Dr. Christopher M. Breeding of the GIA

Laboratory, Thomas Gelb (formerly of the GIA Laboratory),
Franck Notari of GemTechLab in Geneva, and the manuscript
reviewers for their many useful comments and suggestions.
David Kondo and Paul Johnson of the GIA Laboratory, Kyaw
Soe Moe (formerly of the GIA Laboratory), Dr. Wang, and Dr.
Breeding aided in the collection of spectra. Special thanks are
extended to Joseph Casella, Robyn Ellison, Anthea Lema, and
Gavin Pearce of Argyle Diamonds and Rio Tinto Diamonds, West
Perth, for images and information about the production and sale
of these diamonds. The following companies and individuals
kindly supplied the samples and jewelry items described in this
study: Argyle Diamonds; Rio Tinto Diamonds NV, Antwerp;
Gemcut SA, Geneva; Jean Mahie, New York; Alan Bronstein,
Aurora Gems, New York; and L. J. West Diamonds, New York.
Eaton-Magaa S., Post J.E., Heaney P.J., Walters R.A., Breeding

C.M., Butler J.E. (2007) Fluorescence spectra of colored diamonds using a rapid, mobile spectrometer. G&G, Vol. 43, No.
4, pp. 332351.
Federman D. (2003) Gem ProfilePurple and violet diamonds:
Kissing cousins. Modern Jeweler, Vol. 102, No. 9, pp. 4748.
Federman D. (2007) Gem ProfileViolet diamond. Modern
Jeweler, www.modernjeweler.com/web/online/DiamondGem-Profiles/Violet-Diamond/2$284 [accessed Oct. 27, 2008].
Fritsch E., Scarratt K. (1989) Optical properties of some natural
diamonds with high hydrogen content. In A. Feldman and S.
Holly, Eds., Diamond Optics II, Society for Photo-optical
Instrumentation Engineers, Bellingham, WA, pp. 201206.
Fritsch E., Scarratt K. (1992) Natural-color nonconductive grayto-blue diamonds. G&G, Vol. 28, No. 1, pp. 3542.
Fritsch E., Scarratt K. (1993) Gemmological properties of type Ia
diamonds with an unusually high hydrogen content. Journal
of Gemmology, Vol. 23, No. 8, pp. 451460.
Fritsch E., Scarratt K., Collins A.T. (1991) Optical properties of
diamonds with an unusually high hydrogen content. In R.
Messier, J.T. Glass, J.E. Butler, and R. Roy, Eds., Materials
Research Society International Conference Proceedings,
Second International Conference on New Diamond Science
and Technology, Washington, DC, Sept. 2327, Materials
Research Society, Pittsburgh, PA, pp. 671676.
Fritsch E., Rondeau B., Notari F. (2006) Cleavage resistance of
plastically deformed natural diamonds revealed by dissolved
planar features. Diamond and Related Materials, Vol. 15, No.
9, pp. 13101313.
GEMS & GEMOLOGY
SPRING 2009
Fritsch E., Hainschwang T., Massi L., Rondeau B. (2007a)

Hydrogen-related optical centers in natural diamond: An
update. New Diamond and Frontier Carbon Technology, Vol.
17, No. 2, pp. 6389.
Fritsch E., Rondeau B., Hainschwang T., Quellier M.-H. (2007b)
A contribution to the understanding of pink color in diamond:
The unique, historical Grand Cond. Diamond and
Related Materials, Vol. 16, No. 8, pp. 14711474.
Gaffey M.J., Bell J.F., Brown R.H., Burbine T.H., Piatek J.L., Reed
K.L., Chaky D.A. (1993) Mineralogical variations within the
S-type asteroid class. Icarus, Vol. 106, No. 2, pp. 573602.
Hainschwang T., Simic D., Fritsch E., Deljanin B., Woodring S.,
Del Re N. (2005) A gemological study of a collection of
chameleon diamonds. G&G, Vol. 41, No. 1, pp. 2035.
Hofer S.C. (1985) Pink diamonds from Australia. G&G, Vol. 21,
No. 3, pp. 147155.
Iakoubovskii K., Adriaenssens G.J. (2002) Optical characterization of Argyle diamonds. Diamond and Related Materials,
Vol. 11, No. 1, pp. 125131.
Kiflawi I., Mayer A.E., Spear P.M., Van Wyck J.A., Woods G.S.
(1994) Infrared absorption by the single nitrogen and A defect
centres in diamond. Philosophical Magazine B, Vol. 69, No. 6,
pp. 11411147.
King J.M., Moses T.M., Shigley J.E., Liu Y. (1994) Color grading of
colored diamonds in the GIA Gem Trade Laboratory. G&G,
Vol. 30, No. 4, pp. 220242.
King J.M., Moses T.M., Shigley J.E., Welbourn C.M., Lawson
S.C., Cooper M. (1998) Characterizing natural-color type IIb
blue diamonds. G&G, Vol. 34, No. 4, pp. 246268.
King J.M., Shigley J.E., Guhin S.S., Gelb T.H., Hall M. (2002)
Characterization and grading of natural-color pink diamonds.
G&G, Vol. 38, No. 2, pp. 128147.
Lang A.R., Yelisseyev A.P., Pokhilenko N.P., Steeds J.W.,
Wotherspoon A. (2004) Is dispersed nickel in natural diamonds associated with cuboid growth sectors in diamonds
that exhibit a history of mixed-habit growth? Journal of
Crystal Growth, Vol. 263, pp. 575589.
Lawson S.C., Kanda H. (1993) An annealing study of nickel point
defects in high-pressure synthetic diamond. Journal of
Applied Physics, Vol. 73, No. 8, pp. 39673973.
Lu T., Shigley J.E., Koivula J.I., Reinitz I.M. (2001) Observation of
etch channels in several natural diamonds. Diamond and
Related Materials, Vol. 10, No. 1, pp. 6875.
Massi L. (2006) Etudes des Dfants dans les Diamants Bruns et
les Diamants Riches en Hydrogne. PhD thesis, University of
Nantes, France, 372 pp. (in French).
Max D. (2006) One in a millionThe Rio Tinto Diamonds Argyle
pink diamond tender report. Idex Magazine, No. 199, November
13, pp. 7579, www.idexonline.com/portal_FullMazalUbracha.
asp?id=26439 [accessed Oct. 27, 2008].
Moses T.M., Reinitz I.M., Johnson M.L., King J.M., Shigley J.E.
(1997) A contribution to understanding the effect of blue fluorescence on the appearance of diamonds. G&G, Vol. 33, No.
4, pp. 244259.
Moses T.M., King J.M., Wang W., Shigley J.E. (2002) A highly
unusual 7.34 carat vivid purple diamond. Journal of Gemmology, Vol. 28, No. 1, pp. 712.
Nadolinny V., Yelisseyev A. (1994) New paramagnetic centers
containing nickel ions in diamond. Diamond and Related
Materials, Vol. 3, No. 12, pp. 1721.
Noble C.J., Pawlik Th., Spaeth J.-M. (1998) Electron paramagnetic resonance investigations of nickel defects in natural diamonds. Journal of Physics: Condensed Matter, Vol. 10, pp.
1178111793.
Orlov Y.L. (1977) The Mineralogy of Diamond. J. Wiley & Sons,
New York.
Reinitz I.M., Moses T.M. (1993) Gem Trade Lab Notes: Light violet-gray diamond. G&G, Vol. 29, No. 3, p. 199.
Rio Tinto (2009) Rio Tinto tenders a rare offering of blue diamonds.
www.riotintodiamonds.com/ENG/media/media_releases_
1144.asp [accessed Mar. 9, 2009]
Rondeau B., Fritsch E., Guiraud M., Chalain J.-P., Notari F.
(2004) Three historically asteriated hydrogen-rich diamonds: Growth history and sector-dependent impurity incorporation. Diamond and Related Materials, Vol. 13, pp.
16581673.
Scarratt K. (1987) Notes from the laboratory 10. Journal of
Shigley J.E., Chapman J., Ellison R.K. (2001) Discovery and mining of the Argyle diamond deposit, Australia. G&G, Vol. 37,
No. 1, pp. 2641.
Shigley J.E., McClure S.F., Breeding C.M., Shen A.H.,
Muhlmeister S.M. (2004) Lab-grown colored diamonds from
Chatham Created Gems. G&G, Vol. 40, No. 2, pp. 128145.
Titkov S.V., Shigley J.E., Breeding C.M., Mineeva R.M., Zudin
N.G., Sergeev A.M. (2008) Natural-color purple diamonds
from Siberia. G&G, Vol. 44, No. 1, pp. 5664.
Wang W., Hall M., Breeding C.M. (2007) Natural type Ia diamond
with green-yellow color due to Ni-related defects. Gems &
Gemology, Vol. 33, No. 3, pp. 240243.
Welbourn C.M., Cooper M., Spear P.M. (1996) De Beers natural
versus synthetic diamond verification instruments. G&G,
Vol. 32, No. 3, pp. 156169.
Woods G.S., Collins A.T. (1983) Infrared absorption spectra of
hydrogen complexes in type I diamond. Journal of Physics
and Chemistry of Solids, Vol. 44, No. 5, pp. 471475.
Yelisseyev A., Babich Yu., Nadolinny V., Fisher D., Feigelson B.
(2002) Spectroscopic study of HPHT diamonds, as grown at
1500C. Diamond and Related Materials, Vol. 11, No. 1, pp.
2237.
Yelisseyev A., Kanda H. (2007) Optical centers related to 3d transition metals in diamond. New Diamond and Frontier Carbon
Technology, Vol. 17, No. 3, pp. 127178.
Yelisseyev A., Nadolinny V. (1995) Photoinduced absorption lines
related to nickel impurity in annealed synthetic diamonds.
Diamond and Related Materials, Vol. 4, No. 3, pp. 177185.
Zaitsev A.M. (2001) Optical Properties of Diamond: A Data
Handbook. Springer-Verlag, Berlin, 502 pp.
For online access to all issues of GEMS & GEMOLOGY from 1981 to the present, visit:
gia.metapress.com
GEMS & GEMOLOGY
SPRING 2009
37
NOTES & NEW TECHNIQUES
HACKMANITE/SODALITE FROM
MYANMAR AND AFGHANISTAN
David Kondo and Donna Beaton
In recent years, significant amounts of gemquality sodalite/hackmanitesome unusually

transparenthave been produced from the
Mogok region of Myanmar and Badakhshan
Province in Afghanistan. Samples from both
countries varied in color, transparency, UV
fluorescence/phosphorescence, tenebrescence,
and UV-Vis-NIR spectra. The Burmese material was generally more included and showed
weaker fluorescence and phosphorescence
than the Afghan samples. EDXRF spectroscopy
revealed traces of sulfur in all samples. The
tenebrescence of many of the stones from
both localities was strong enough for classification as hackmanite.
ackmanite [Na8Al6Si6O24(Cl2,S)] is a sulfur-bearing variety of sodalite [Na8Al6Si6O24Cl2] that

gemologists typically distinguish according to its
tenebrescencethat is, its ability to change color in
response to the application or absence of certain
wavelengths of light (e.g., white light or UV radiation;
Japan Germany Gemmological Laboratory, 2008).
Interestingly, a survey of the literature showed no
Editors note: Consistent with its mission, GIA has a vital role in conducting research, characterizing gemstones, and gaining knowledge
that leads to the determination of gemstone origins. The gemstones
studied in this report are not subject to the Tom Lantos Block Burmese
JADE Act of 2008, and their import was in accordance with U.S. law.
38
NOTES AND NEW TECHNIQUES
consistent definition for the term hackmanite. Some

sources say it is sulfur-bearing sodalite that shows fluorescence (e.g., Jackson, 1997), possibly with the
valence state of sulfur being a critical factor. Others
stipulate tenebrescence as the defining criterion (e.g.,
Simpson and Weiner, 1989). Various sources indicate
that hackmanite can either fade or deepen in color
when left in darkness for days to months, depending
on the origin of the material (e.g., Webster, 1994;
Hainschwang, 2007; Tunzi and Pearson, 2008). In
most cases, the material turns pink to purple/violet
with exposure to ultraviolet radiation, and the color
fades on exposure to sunlight or artificial white
light sources. This behavior is reversible unless the
stone is exposed to heat. Heating to more than 500C
destroys the UV sensitivity of hackmanite, leaving it
in its bleached state (Medved, 1954; Kirk, 1955).
Hackmanite is usually found as translucent-toopaque crystalline aggregates, often intergrown with
other minerals, especially nontenebrescent sodalite.
Until recently, examples of transparent faceted hackmanite were reported only rarely (see, e.g., Koivula and
Kammerling, 1989a,b). Hackmanite is known from
Canada (Mont Saint-Hilaire, Quebec; and Bancroft,
Ontario); Magnet Cove, Arkansas; Libertyville, New
Jersey; Minas Gerais, Brazil; the Kola Peninsula,
Russia; and Greenland (Miser and Glass, 1941;
Webster, 1994; Bernard and Hyrsl,
2004). More recently, it was reported from the Mogok area of Myanmar
and the Badakhshan Province of Afghanistan (e.g.,
Johnson and Koivula, 1998; Moore, 2001, 2002; Liu et
al., 2004; Japan Germany Gemmological Laboratory,
2008; Tunzi and Pearson, 2008).
In April 2007, gem dealer Hussain Rezayee
informed the authors about additional production of
hackmanite/sodalite from Myanmar, in an area 11 km
east of Mogok at Pyang Gyi, near Pein Pyit. Production
GEMS & GEMOLOGY
SPRING 2009
Color after fading
Color after UV exposure
Long-wave UV
Sample 1
Myanmar
Samples
6, 4, & 2
Myanmar
Samples 16, 15,

14, 11, 12, & 13
Afghanistan
Samples
7, 8, 9, & 10
Afghanistan
Figure 1. These samples of Burmese and Afghan hackmanite/sodalite, which were studied for this report, are
shown in their desaturated color state, after exposure to short-wave UV radiation, and during exposure to longwave UV. See table 1 for sample weights. Photos by Robert Weldon.
started in mid-2003, but originally was of low quality.

Beginning in 2007, about 2,000 carats per month of
various sizes were being cut into cabochons and
faceted stones, according to Mr. Rezayee. He loaned
several of the Burmese cabochons to us for examination, and Bangkok-based gem dealer G. Scott Davies
donated a faceted Burmese hackmanite to GIA.
Mr. Rezayee also loaned us samples of hackmanite/sodalite that were produced in Badakhshan
since 2002. These samples generally appeared similar to the material from Myanmar, though some
were unusually transparent, and Mr. Rezayee reported that clean stones up to 18 ct have been faceted
from the Afghan material. He has cut approximately
1,000 carats of faceted stones and 10,000 carats of
cabochons (ranging up to ~40 ct) of the Afghan hackmanite/sodalite.
Rough material from both localities is commonly
oiled to enhance its transparency. According to Mr.
Davies and F. Hashmi (pers. comm., 2008), some is

oiled after being cut and polished; since any recutting will cause the cracks to reappear, such stones
are re-oiled to improve their apparent clarity.
MATERIALS AND METHODS

Six samples from Myanmar (a 0.64 ct cushion cut
and five cabochons weighing 20.2356.20 ct) and 10
from Afghanistan (six faceted stones weighing
0.444.83 ct and four cabochons of 5.9229.94 ct)
represented as hackmanite were examined for this
study (e.g., figure 1). We evaluated all samples for
color in a Gretag Macbeth Judge II light box, using a
D65 daylight-equivalent fluorescent lamp. To assess
tenebrescence, we examined the stones after they
were faded by exposure to a standard 100-watt household incandescent bulb or a 4.5-watt daylight-equivalent fluorescent lamp, and immediately after their
GEMS & GEMOLOGY
SPRING 2009
39
color was intensified by exposure to short-wave UV

radiation (using a 4-watt bulb). Generally it was necessary to expose the samples to white light for a period of several hours to more than one day to attain the
maximum amount of fading, whereas only several
minutes were needed to deepen the color with the
short-wave UV lamp. We also examined the color
(starting in both color states) of the most tenebrescent samples (nos. 1, 2, 4, and 15) after they were held
in the dark for at least three weeks.
Refractive indices were measured with a standard
refractometer, and specific gravity was determined
hydrostatically. We examined all samples with a
gemological microscope using various lighting techniques (darkfield, diffused light, reflected light, fiberoptic illumination, etc.). We observed fluorescence
and phosphorescence in a darkened room using a
standard long- and short-wave UV lamp. All samples
were characterized with Raman, Fourier-transform

infrared (FTIR), and energy-dispersive X-ray fluorescence (EDXRF) spectroscopy. For comparison, we
also characterized three transparent blue sodalites of
unknown locality from the reference collection at
the GIA Laboratory in New York by FTIR and
EDXRF spectroscopy. UVvisiblenear infrared (UVVis-NIR) spectroscopy was performed on 15
Burmese and Afghan samples that showed sufficient
diaphaneity using a double-beam spectrophotometer
scanning from 900 to 200 nm; spectra were acquired
for 13 of these samples in both their faded and UVexcited color states.
RESULTS AND DISCUSSION

Standard gemological properties and tenebrescence of
all the Burmese and Afghan hackmanites/sodalites
TABLE 1. Properties, including tenebrescence, of 16 hackmanite/sodalite samples from Myanmar and Afghanistan.a
Source
Myanmar
Sample
Diaphaneity
Weight
(ct)
RI
SG
Color after fading

(white light)
1
2
Faceted cushion
Oval cabochon
Transparent
Translucent
0.64
24.83
1.479
1.47
2.26
2.29
Light grayish violet

Very light purple
Oval cabochon
Translucent to
semi-opaque
56.20
1.47
2.29
Oval cabochon
Translucent to
semi-opaque
20.23
1.47
2.44
Marquise
cabochon
Triangular
cabochon
Pear cabochon
Translucent
23.77
1.47
2.29
Translucent
22.26
1.47
2.30
Blue unchanged;
purple got paler and
more violet
Light-to-medium
pinkish purple; nearcolorless areas
unchanged
Medium purple with
violet
Medium violet
Semi-opaque
29.94
1.45
2.28
Translucent
Translucent
Transparent
16.97
12.39
5.92
1.45
1.45
1.46
2.30
2.30
2.30
11
12
Pear cabochon
Oval cabochon
Triangular
cabochon
Faceted oval
Faceted cushion
Transparent
Transparent
4.83
4.25
1.480
1.480
2.31
2.31
13
Faceted oval
Transparent
3.34
1.480
2.31
14
15
Faceted oval
Faceted oval
Transparent
Transparent
2.59
1.45
1.480
1.480
2.31
2.31
16
Faceted oval
Transparent
0.44
1.480
2.31
6
Afghanistan
Cut
7
8
9
10
Medium purple with

very dark purple area
Light purple
Light violet
Near colorless
Very light blue
Very light grayish
greenish blue
Very light violetish
blue
Very light violet
Medium pinkish
purple
Very light pink
Color after UV
exposure
Tenebrescence
Name
Medium violet
Medium-to-dark
purple
Medium-to-dark
purple ranging to
blue
Medium purple and
near colorless
Strong
Strong
Hackmanite
Hackmanite
Weak
Hackmanite
Strong
Hackmanite
Dark purple to
violet
Dark violet
Weak
Hackmanite
Moderate
Hackmanite
Dark purple
Very weak
Sodalite
Medium purple
Medium violet
Very light violet
Weak
Weak
Very weak
Hackmanite
Hackmanite
Sodalite
Very light blue

Very light grayish
greenish blue
Light bluish violet
None
Very weak
Sodalite
Sodalite
Moderate
Hackmanite
Light violet
Medium-intense
purple
Light purple
Moderate
Strong
Hackmanite
Hackmanite
Moderate
Hackmanite
The color states described above were assessed after fading with a 100-watt incandescent light for several hours or longer at a distance of
approximately 15 cm (6 in.), and after inducing color with several minutes of exposure to a standard short-wave UV lamp.
40
GEMS & GEMOLOGY
SPRING 2009
Figure 2. Crystalline
inclusions were present
in some of the Burmese
samples of hackmanite/sodalite studied
(leftsample 4, right
sample 1). Photomicrographs by D. Kondo;
fields of view ~4.8 and
3.3 mm, respectively.
are summarized in table 1. Exposure to short-wave

UV for several minutes perceptibly changed the color
of 15 of the 16 stones: Eight showed strong-to-moderate tenebrescence, four showed a weak change, and
four had a very weak or no change. For the purposes
of this study, we defined samples with weak to
strong tenebrescence as hackmanite, and samples
with no or very weak tenebrescence as sodalite. The
tenebrescent effect can be seen in figure 1, which
shows samples in their faded, UV-excited, and fluorescent states. The stones showed a much faster reaction to UV radiation and to white light than from
being kept in the dark. The daylight-equivalent fluorescent lamp induced fading as rapidly and effectively
as the incandescent bulb.
Since these samples were obviously sensitive to
some wavelengths of radiation, we carefully examined their condition before and after all spectroscopy, as the various lasers (514 and 830 nm), Xrays, and UV, visible, and infrared wavelengths used
in the equipment might affect the color state. We
found that no perceptible changes in color occurred
during the testing process, although Hainschwang
(2007) noted a fading of the color of Burmese hackmanite after exposure to a green laser (514 nm) for
two minutes.
Samples 1, 2, and 4 (from Myanmar) and 15 (from
Afghanistan) generally showed a rapid and distinct
change from light purple or violet to medium-dark
purple or violet within a few seconds of exposure to
UV radiation. Several Afghan samples showed moderate tenebrescence by deepening to a medium purple or violet. Burmese sample 4 had an uneven
change of color: A broad white patch across part of
the stone (not visible in figure 1) showed no change,
although areas at the stones periphery had strong
tenebrescence. The area that did not change color
was identified as nepheline by Raman analysis. The
significant nepheline component of this stone
explains its abnormally high SG (2.44) compared to

the published range for sodalite/hackmanite
(2.152.35; for nepheline, the values are 2.552.65
according to Webster, 1994).
The four samples that were kept in the dark for
at least three weeks showed no change in color
from their faded state (induced by exposure to the
incandescent bulb for one day), with one exception:
Sample 15 became slightly less saturated, indicating
that our starting color state had not been completely faded. After the color of these samples was deepened by exposure to short-wave UV radiation for
several minutes, storage in the dark for at least
three weeks caused their color to become slightly
less saturated.
For all stones except sample 4, the physical properties were consistent with published values for
sodalite and hackmanite (see Webster, 1994; Johnson
and Koivula, 1998). In general, the Burmese samples
showed fractures, irregular white masses, and inclusions of transparent crystals (e.g., figure 2). Raman
analysis identified the transparent crystals as pyroxene in sample 3 and mica in sample 5, but we were
unable to identify the transparent crystals in other
stones. The Afghan samples examined for this study
were for the most part much less saturated, less
included, and more transparent than the Burmese
material. They often contained fingerprints (figure
3, left), transparent to whitish included crystals (figure 3, center), wispy cross-hatched inclusions (figure
3, right), and fractures containing a whitish foreign
material.
Both a whitish appearance in Afghan samples and
fluorescence seen in fractures in Burmese stones
suggested clarity enhancement. This was confirmed
in all eight of the Afghan stones by the presence of
absorption bands between 3050 and 2830 cm1 and
at ~3412 cm1 in the FTIR spectra. The 0.64 ct
faceted Burmese sample did not show any visual or
GEMS & GEMOLOGY
SPRING 2009
41
Figure 3. These images show some typical internal features observed in Afghan hackmanite/sodalite, including a
large fingerprint (left; sample 13), numerous transparent crystals (center; sample 11), and some wispy crosshatched clouds (right; sample 12). Fields of view are approximately 6.8 mm, 3.1 mm and 4.9 mm, respectively.
Photomicrographs by D. Kondo.
spectral evidence of clarity enhancement, while the

four Burmese cabochon samples tested had saturated
absorption in the region of interest, so no conclusions could be drawn.
All the samples fluoresced to long-wave UV radiation. The Afghan stones generally gave a strong yellow to orange reaction, and most of the Burmese
material fluoresced weak orange (samples 1 and 2
showed a moderate-to-strong orange). When exposed
to short-wave UV, the Afghan samples fluoresced a
weak-to-moderate red or orange that was quickly
obscured by a stronger moderate-to-strong white to
yellowish white fluorescence, whereas the Burmese
material showed strong greenish yellow fluorescence
to short-wave UV in fissures; the Burmese host material itself did not react to short-wave UV, except for
sample 1, which had a weak red reaction and lacked
fissures.
Phosphorescence to UV radiation was also present
in varying amounts. After exposure to short-wave
UV, the Afghan material showed a moderate-tostrong yellowish white phosphorescence lasting for
several minutes; the long-wave reaction had weaker
intensity but similar duration. Burmese samples 1
and 2 showed very weak to weak white phosphorescence to short-wave UV that lasted for perhaps one
minute. Similar phosphorescent reactions for
Burmese and Afghan samples were described by
Tunzi and Pearson (2008).
As expected, the Raman spectra of all the samples were indistinguishable from the spectrum of
sodalite. In the mid-infrared region, the FTIR spectra
showed some differences, most notably between
2750 and 2250 cm1. In addition, the sodalite reference samples showed absorption peaks in the
infrared spectrum at 4874, 4690, 4110, 3971, 3033,
42
2655, and 2272 cm1 that were not seen in any of the
samples submitted for this study. Qualitative chemical analysis by EDXRF spectroscopy of all the
Burmese and Afghan samples showed a weak sulfur
peak, which the sodalite references lacked.
In the literature, tenebrescence and fluorescence
of hackmanite are generally attributed to S2 (e.g.,
Liu et al., 2004; Sidike et al., 2007; Japan Germany
Figure 4. These UV-Vis-NIR spectra for representative

hackmanites from Myanmar and Afghanistan show
an increase in absorption at ~550 nm for the UVexcited state (in blue) compared to the desaturated
color state (in red). These samples were run with the
beam entering the table and exiting the culet. Path
lengths for the 0.64 ct Burmese and the 2.59 ct
Afghan samples are ~3.45 and 5.60 mm, respectively.
GEMS & GEMOLOGY
SPRING 2009
Gemmological Laboratory, 2008). To better understand the absorption features that lead to the perceived color, figure 4 provides UV-Vis-NIR absorption spectra of representative Burmese and Afghan
hackmanites (showing both strong and weak
tenebrescence) in their desaturated and UV-excited
color states. In general, all the spectra had a broad
band in the mid-500 nm range, although in some
cases the band peaked closer to 590 nm. The lesstransparent Burmese samples (all six tested) only
showed this ~550 nm band and an absorption edge
ranging from ~310 to 350 nm. The Afghan samples
(eight out of 10 tested) showed this mid-500 nm feature plus other bands in many cases. For example, a
minor band in the 410412 nm range was present in
six of the Afghan samples, but not in any of the
Burmese samples. We also saw peaks with typical
positions of 277 and 313 nm in Afghan samples;
however, we cannot say if these bands are present in
the Burmese samples as well since this region was
saturated in those spectra.
For tenebrescent samples, the band in the mid500 nm region grew in absorption after the stone was
excited with UV radiation: This is the band responsible for the color and phenomenon. Many samples in
the desaturated color state showed weak peaks superimposed on the main band, with the most prominent
secondary band centered at 672 nm. Similar results
were recorded by Hainschwang (2007).
CONCLUSION
REFERENCES
Association of Hong Kong, Vol. 25, pp. 8590.

Medved D.B. (1954) Hackmanite and its tenebrescent properties.
American Mineralogist, Vol. 39, pp. 617629.
Miser H.D., Glass J.J. (1941) Fluorescent sodalite and hackmanite
from Magnet Cove, Arkansas. American Mineralogist, Vol. 26,
pp. 437445.
Moore T. (2001) Whats new in mineralsTucson show 2001.
Mineralogical Record, Vol. 32, No. 3, pp. 245257.
Moore T. (2002) Whats new in mineralsDenver show 2001.
Mineralogical Record, Vol. 33, No. 1, pp. 8399.
Sidike A., Sawuti A., Wang X.-M., Zhu H.-J., Kobayashi S., Kusachi
I., Yamashita N. (2007) Fine structure in photoluminescence
spectrum of S2 center in sodalite. Physics and Chemistry of
Minerals, Vol. 34, pp. 477484.
Simpson J.A., Weiner E.S.C. (1989) The Oxford English Dictionary,
2nd ed., Vol. VI. Clarendon Press, Oxford, pp. 10011002.
Tunzi J., Pearson G. (2008) Hackmanite, tugtupite and afghaniteTenebrescence and fluorescence of some sodalite related minerals. Australian Gemmologist, Vol. 23, No. 8, pp.
349355.
Webster R. (1994) Gems: Their Sources, Descriptions and Identification, 5th ed. Rev. by P. G. Read, Butterworth-Heinemann,
Oxford, UK, p. 375.
Bernard J.H., Hyr`sl J. (2004) Minerals and their Localities. Granit,

Prague, Czech Republic.
Hainschwang T. (2007) A study of an unusual hackmanite from
Myanmar. www.gemlab.net/website/gemlab/index.php?id=187,
posted November 30.
Jackson J. (1997) Glossary of Geology, 4th ed. American Geological Institute, Alexandria, VA, p. 288.
Japan Germany Gemmological Laboratory (2008) Strange jewel
hackmanite. Gem Information, Vol. 3738, July 23, pp.
2632 (in Japanese).
Johnson M.L., Koivula J.I., Eds. (1998) Gem News: Hackmanite
from Myanmar. G&G, Vol. 34, No. 3, pp. 223224.
Kirk R.D. (1955) The luminescence and tenebrescence of natural
and synthetic sodalite. American Mineralogist, Vol. 40, No.
22, pp. 2231.
Koivula J.I., Kammerling R.C., Eds. (1989a) Gem News:
Hackmanitea remarkable variety of sodalite. G&G, Vol. 25,
No. 2, pp. 112113.
Koivula J.I., Kammerling R.C., Eds. (1989b) Gem News: Update
on hackmanite. G&G, Vol. 25, No. 4, pp. 245246.
Liu S.I., Peng M.S., Tse E.Y.L. (2004) The tenebrescence of hackmanite from Afghanistan. Journal of the Gemmological
Clearly, many of the Burmese and about half the

Afghan samples examined in this study showed the
distinct tenebrescence that is characteristic for hackmanite. However, a few saturated blue-to-purple
Burmese stones and some very desaturated Afghan
stones showed little or no change in color with exposure to UV radiation or bright white light sources, or
to placement in the dark for extended periods. Stones
with no or very weak tenebrescence may best be
referred to as sodalite, despite containing traces of
sulfur. Although hackmanite is commonly described
in the gemological literature as a sulfur-bearing variety of sodalite that is distinguished by its tenebrescence, there are no guidelines clearly separating hackmanite from sodalite. We suggest that only sodalite
with noticeable tenebrescence be called hackmanite.
ABOUT THE AUTHORS
Mr. Kondo (david.kondo@gia.edu) is gemological
research associate, and Ms. Beaton is manager of
Identification Services, at the GIA Laboratory in New York.
ACKNOWLEDGMENTS
The authors thank Hussain Rezayee (Pearl Gem Co.,
Beverly Hills, California) and G. Scott Davies (AmericanThai Trading, Bangkok) for supplying information and
samples for this report. Farooq Hashmi (Intimate Gems,
Jamaica, New York) is thanked for helpful discussions.
GEMS & GEMOLOGY
SPRING 2009
43
SOLUTION-GENERATED PINK COLOR SURROUNDING

GROWTH TUBES AND CRACKS IN BLUE TO BLUE-GREEN
COPPER-BEARING TOURMALINES FROM MOZAMBIQUE
John I. Koivula, Kevin Nagle, Andy Hsi-Tien Shen, and Philip Owens
Several transparent, faceted, blue to blue-green

copper-bearing tourmalines containing growth tubes
and cracks surrounded by sleeves of pink color were
examined for this report. On the basis of microobservation, it is theorized that a radioactive solution
was the probable cause of the pink color. The presence of the pink zones also supplied visual evidence
that the host tourmalines had not been heat treated.
ver the past year, we examined nine blue to bluegreen Cu-bearing tourmalines from Mozambique
that contained surface-reaching growth tubes and cracks
that were outlined or sleeved with obvious pink color
zones (e.g., figure 1). These gems came from four different
gem dealers over the course of the year.
The first gem, from Simon Watt, was a 14.12 ct blue
heart-shaped mixed cut measuring 15.82 13.68 10.82
mm that was purportedly from Mozambique. As shown in
figure 2, this tourmaline contained a surface-reaching
growth tube sleeved by a pink zone of moderate intensity.
Soon thereafter, Bill Vance and David Freeland Jr. sent us
the 27.63 ct cushion mixed cut shown in figure 1. This gem,
also said to be from Mozambique, measured 17.82 17.13
12.95 mm. It contained a surface-reaching macroscopic
growth tube under its table that was enveloped along its
length by an intense zone of pink (almost red) color, which
created a clear contrast against the blue bodycolor of its host.
The third and largest of the gems came into the GIA
Laboratory for identification and origin determination. This
tourmaline was a blue-green pear-shaped modified brilliant
cut that weighed 33.26 ct and measured 24.34 19.88
12.89 mm. It contained several thin surface-reaching growth

44
RAPID COMMUNICATIONS
tubes that were all sleeved by narrow pink color zones.

While these features were too small to be seen with the
unaided eye, they were clearly evident with magnification.
Chemical analysis by laser ablationinductively coupled
plasmamass spectrometry (LA-ICP-MS) revealed that the
country of origin for this gem was also Mozambique (for
more on this technique, see Abduriyim et al., 2006).
The remaining six tourmalines, all pear shapes
believed to originate from Mozambique, came from Mark
H. Smith. They ranged from 1.07 to 2.66 ct, and were all
light blue-green in color. Pink color zoning was associated
with cracks extending from the pink-zoned growth tubes in
these stones.
Although detailed gemological investigations have
been performed relatively recently on Cu-bearing tourmalines from Mozambique (Abduriyim et al., 2006; Laurs et
al., 2008), no inclusions of this nature were mentioned or
illustrated in these published works. This suggests such
inclusionsand their formation mechanismare relatively rare in tourmaline.
Proposed Coloration Mechanism. It is well known and

scientifically established that radiation can produce pinkto-red color in tourmaline (Nassau 1984), so it is logical to
surmise that radiation is responsible for the pink color surrounding the surface-reaching growth tubes and related
cracks in these blue to blue-green Cu-bearing tourmalines.
The most likely mechanism for the formation of the
pink color in these tourmalines can be hypothesized from
previous research by one of the authors (JIK) on smoky
quartz and green diamond. The first of these articles
(Koivula, 1986) described the coloration of smoky quartz
crystals by naturally occurring radioactive hydrothermal
fluids. The brown color was confined to a surface layer
only a few millimeters thick; the color was darkest around
surface features and surface-reaching cavities, as well as
surrounding near-surface fluid inclusion chambers that
were still intact, containing both liquid and gas phases. In
fact, the presence of brown clouds of color surrounding
these fluid inclusion chambers proved that the fluid they
contained was radioactive at one time, since the smoky
GEMS & GEMOLOGY
SPRING 2009
causing radioactive solutions can invade the crystal by capillarity, thereby imparting color to the inner surfaces of
those features. If at some later point the outer surface of
such a crystal is removedeither by natural causes such as
etching or water abrasion, or through lapidary processes
then only the surfaces that have evaded removal (e.g., within growth tubes, etc.) will show the color imparted by the
radioactive solutions. Since it is well known that exposure
to radioactivity causes a pink-to-red color in tourmalines
(see, e.g., Reinitz and Rossman, 1988), that is the probable
explanation for the presence of the pink sleeves in these Cubearing tourmalines.
Micro-Examination. First, we used standard gemological
Figure 1. Of the Cu-bearing tourmalines examined for

this report, this rich blue 27.63 ct gem contained the
largest and most obvious rubellite-colored growth
tube, which is clearly visible under the table facet.
Photo by Robert Weldon.
color in quartz is the result of radiation (Nassau, 1984).
The second study (Koivula, 1988) documented a section of a natural diamond crystal with a negative-crystal
cavity that was open to the surface through a thin neck.
Trapped in the cavity was a small loose diamond crystal
that was too large to escape through the neck. Brownish
green radiation stains covered the inner walls of this cavity, its neck, and the surface of the trapped diamond crystal. However, no such stains were observed anywhere on
the host diamonds outer surface.
These two studies show that radioactive solutions can
impart a post-growth shallow layer of color in gem materials. If they have surface-reaching features such as etch pits
and cavities, channels, cracks, or growth tubes, then color-
Figure 2. Discovered in a 14.12 ct Cu-bearing tourmaline said to be from Mozambique, this pink-zoned
growth tube was the first one of these inclusions
examined for this report. Photomicrograph by J. I.
Koivula; field of view 4.9 mm.
techniques to confirm that the gems were tourmaline. Then

we turned to energy-dispersive X-ray fluorescence (EDXRF)
spectroscopy to determine that they were Cu-bearing, and
we followed that with LA-ICP-MS chemical analysis on the
four largest stones to confirm that their geographic origin
was Mozambique.
The next step in the documentation process, and the
purpose of this article, was to determine the nature of these
unusual pink-sleeved growth tubes and cracks, and to document them photographically. This was accomplished
using a gemological photomicroscope and various lighting
techniques.
Like all growth tubes in tourmaline, these were oriented
parallel to the optic axis (c-axis) of their hosts. In all instances
where pink coloration was observed around a growth tube,
that growth tube reached the surface of the host gem. Where
the growth tubes were completely confined in the tourmaline, we did not see any associated pink color.
When looking down the length of a pink surface-reaching needle (figure 3), we observed bleeding of the pink
Figure 3. When viewed down the length of one of the

pink needles, the radiation-induced color can be
seen to bleed out into the surrounding tourmaline
host, becoming weaker until it gradually fades away.
The diameter of the growth tube is 0.06 mm.
Photomicrograph by J. I. Koivula.
GEMS & GEMOLOGY
SPRING 2009
45
Figure 4. Viewed down the length of the

growth tube, a dark red-to-pink zone can
be observed immediately surrounding
the tube, which itself is lined on the
inner walls with a rusty-looking epigenetic coating (left). Wherever pink-zoned
growth tubes reached the surface, their
edges looked ragged, apparently due to
the cutting process (right, reflected light).
For a growth tube to show a pink sleeve
of color, it must reach the surface of the
host tourmaline. This tube is 0.07 mm
in diameter at its broadest point.
Photomicrographs by J. I. Koivula.
color into the surrounding tourmaline host, which became

weaker until it gradually faded away. This is exactly what
would be expected if a radioactive solution had entered by
capillary action into open growth tubes on the surface of a
tourmaline crystal. A thin zone of much darker pink-to-red
color immediately surrounded all the colorized tubes (figure 4, left), while a rusty-looking epigenetic coating lined
the inner walls of the growth tubes. Exposure to the
radioactive solutions could have taken place within the
gem pockets that originally hosted the tourmalines, or
within a specific part of the alluvial deposit after the tourmalines weathered from their host pegmatite. The radioactive solutions may have originated from interaction with
radioactive minerals; such minerals are common in some
Mozambique pegmatites (e.g., Dias and Wilson, 2000).
With magnification and surface-reflected light (figure 4,
right), the edges of all the surface-reaching growth tubes
looked ragged and rough. This is apparently due to damage
Figure 5. Pink-colored growth tubes in Cu-bearing

tourmalines have a wide range of diameters; only the
largest will be clearly visible without magnification.
Photomicrograph by J. I. Koivula; field of view 2.9 mm.
46
along the rims of the growth tubes that occurred during

the faceting process.
As shown in figure 5, we observed a wide range in the
diameters of the pink-colored growth tubes. Only the
largest of these, such as that shown in figure 1, were visible without magnification.
We also noticed that when epigenetic matter completely blocked a growth tube, the color-inducing radioactive solution only penetrated to the point of blockage, so
the coloration stopped there as well. As a result, some
growth tubes had partial sleeves of pink color (figure 6).
During this examination, we also saw pink color zoning associated with cracks extending from or between the
pink-zoned growth tubes, as well as from some surfacereaching cracks (figure 7, left). Just as we observed with the
growth tubes, higher magnification and immersion clearly
showed that the pink coloration emanated from the inner
walls of these cracks and gradually dissipated into the host
tourmaline (figure 7, right).
Figure 6. As shown here, if epigenetic matter blocks a

growth tube, then the color-inducing radioactive solution can only penetrate to the point of blockage, effectively stopping the pink coloration. Photomicrograph
by J. I. Koivula; field of view 2.2 mm.
GEMS & GEMOLOGY
SPRING 2009
Figure 7. The inner walls of any fissures or cracks that reached the surface (or surface-reaching growth tubes) in
these Mozambique tourmalines were also colored pink to red by natural irradiation (left; field of view 2.9 mm).
At higher magnification, the penetration of the pink-to-red color into the surrounding tourmaline is clearly visible
(right; field of view 0.8 mm). Photomicrographs by J. I. Koivula.
Abduriyim et al. (2006) indicated that heat treatment
at around 500C is used to produce a desirable neon blue
color in Cu-bearing tourmaline. This article also stated
that purplish pink and pink colors faded when exposed to
temperatures between 400 and 500C. And in their 2008
report, Laurs et al. indicated that a temperature of 530C
was used to drive off pink-to-purple color and produce
vivid blues and greens. The fading of radiation-caused
pink-to-red color in tourmaline through heat treatment is
well documented in the literature (Nassau, 1984). The
temperatures mentioned in such fading experiments range
from 260C to 400C, with no red or pink color possible at
all above 750C. Together with Dr. Emmanuel Fritsch of
the University of Nantes, one of the authors (JIK) did a
number of fading experiments on a variety of gem materials in the early 1990s using both heat and light. In these
experiments, the color faded completely between 450C
and 500C for all the pink-to-red tourmalines tested.
In view of this, we believe that the blue to blue-green
bodycolor shown by these Cu-bearing tourmalines must

be of natural origin and not the result of heat treatment. If
these gems had been heat treated, then the pink zones surrounding the growth tubes would have faded, and would
not show such an intensity of color.
Conclusion. To the authors knowledge, the coloration of

tourmaline surrounding surface-reaching growth tubes and
cracks by invading radioactive solutions is not mentioned
anywhere else in the literature. The fact that all of the examples described in this report came from Mozambique suggests that this type of inclusion feature may be characteristic
of that locality, although granitic pegmatites worldwide are
known to host radioactive solutions. The presence of the
pink zones surrounding surface-reaching growth tubes in
these otherwise blue to blue-green gems also provides clear
proof that the host tourmalines were not heat treated, since
the temperature required to artificially produce such colors
in Cu-bearing tourmalines would fade the pink color.
REFERENCES
ABOUT THE AUTHORS
Mr. Koivula (jkoivula@gia.edu) is chief gemologist, Dr. Shen
is research scientist, and Mr. Owens is staff gemologist at
the GIA Laboratory, Carlsbad. At the time this article was
prepared, Mr. Nagle was staff gemologist at the GIA
Laboratory, Carlsbad.
ACKNOWLEDGMENTS
The authors thank Simon Watt (Mayer & Watt, Maysville,
Kentucky), Bill Vance (Vance Gems, Newark, Delaware),
David Freeland Jr. (Tucson, Arizona), and Mark H. Smith
(Thai Lanka Trading, Bangkok) for bringing these most
interesting tourmalines to our attention, and for allowing
us to document their stones for this study.
Abduriyim A., Kitawaki H., Furuya M., Schwarz D. (2006)

Paraba-type copper-bearing tourmaline from Brazil, Nigeria,
and Mozambique: Chemical fingerprinting by LA-ICP-MS.
Gems & Gemology, Vol. 42, No. 1, pp. 420.
Dias M.B., Wilson W.E. (2000) Famous mineral localities: The
Alto Ligonha pegmatites, Mozambique. Mineralogical Record,
Vol. 31, No. 6, pp. 459497.
Koivula J.I. (1986) Solution coloration of smoky quartz. Journal of
Koivula J.I. (1988) Remarkable dissolution in diamond. Zeitschrift
der Deutschen Gemmologischen Gesellschaft, Vol. 36, No.
3/4, pp. 149152.
Laurs B.M., Zwaan J.C., Breeding C.M., Simmons W.B., Beaton
D., Rijsdijk K.F., Befi R., Falster A.U. (2008) Copper-bearing
(Paraba-type) tourmaline from Mozambique. Gems &
Gemology, Vol. 44, No. 1, pp. 430.
Nassau K. (1984) Gemstone Enhancement. Butterworths, London.
Reinitz I.M., Rossman G.R. (1988) Role of natural radiation in tourmaline coloration. American Mineralogist, Vol. 73, pp. 822825.
GEMS & GEMOLOGY
SPRING 2009
47
IDENTIFICATION OF THE ENDANGERED PINK-TO-RED

STYLASTER CORALS BY RAMAN SPECTROSCOPY
Stefanos Karampelas, Emmanuel Fritsch, Benjamin Rondeau, Aude Andouche, and Bernard Mtivier
All corals within the Stylasteridae family (including

the Stylaster genus) are listed in Appendix II of CITES;
this means they are protected and their trade requires
an export permit, unlike corals from the Corallium
genus, which include most pink-to-red corals used in
jewelry. Raman scattering demonstrates that corals
from the Stylaster genus contain carotenoid pigments
(polyenic pigments substituted with methyl groups),
whereas those from the Corallium genus are colored
by unmethylated polyenic pigments. Additionally,
Stylaster corals are made of aragonite, whereas those
from Corallium are composed of calcite. Through
Raman scattering analysis, the fully protected
Stylaster pink-to-red corals may be distinguished
from this other type of gem coral.
n June 2007, delegates from 171 countries convened at

The Hague to decide which species to include under the
CITES (Convention on International Trade in Endangered
Species of Wild Fauna and Flora) agreement. The aim of
CITES is to ensure that international trade in plant and animal specimens does not threaten their survival. The species
covered by the convention are listed in three appendices,
according to the degree of protection they need. Appendix I
includes species threatened with extinction, where trade is
permitted only in exceptional circumstances. Species in
Appendix II are not necessarily threatened with extinction,
but their trade must be controlled to avoid use that would
threaten their survival. Appendix III contains species that
are protected in at least one country that has asked other
CITES parties for assistance in controlling the trade.
The gemological significance of this triennial meeting
is that corals from the Corallium genus, the most impor-
tant of all gem coral species, were being considered for protection under Appendix II (CITES, 2008a). Ultimately, it
was decided not to include them. More recently, on April
8, 2008, China, which now has domestic laws to protect
these species, requested that CITES include four
Corallium species (C. elatius, C. japonicum, C. konjoi,
and C. secundum) under Appendix III (Fish and Wildlife
Service, 2008). Meanwhile, the Stylasteridae family, which
includes all Stylaster gem corals (e.g., figure 1), remained
listed under Appendix II of CITES (as of January 18, 1990),
which means a certificate issued by the management
authority from the country (or state) of export is required
(CITES, 2008b).
Pink-to-red corals have been used for ornamental purposes for about 10,000 years (Liverino, 1989). According
to Rolandi et al. (2005), there are two classes, Hydrozoa
and Anthozoa, within the Cnidaria phylum (i.e., cnidarians) that have skeletons durable enough for use in gem
materials and carvings. These two classes each contain a
family (Stylasteridae and Coralliidae, respectively) that
together yield the majority of pink-to-red coral species
used for ornamentation (Pienaar, 1981; Rolandi et
al., 2005; Smith et al., 2007). Most corals found in the
Figure 1. These orangy pink cabochons (12 12 5 mm)

were fashioned from Stylaster coral. No evidence of dye
or impregnation was detected in these specimens. Photo
by B. Rondeau.

48
GEMS & GEMOLOGY
SPRING 2009
Corals from the Corallium genus are found in waters

throughout the worldnotably the Mediterranean Sea
and the Atlantic, Pacific, and Indian Oceansas are
corals from the Stylaster genus (e.g., Pienaar, 1981;
Rolandi, 1981; Rolandi et al., 2005; Smith et al., 2007;
CITES, 2008a). Stylaster corals (sometimes referred to as
lace corals; figure 4) are typically impregnated and/or
dyed and may be used as a substitute for Corallium
mostly as beads and cabs. In such forms, it is sometimes
difficult to precisely identify any coral by routine gemological examination. Stylaster corals are also found with
natural purple and violet coloration (e.g., S. californicus
and S. subviolaceus).
Careful observation of the structural features found in
the Stylaster and Corallium genera may be useful in identifying the particular species of a coral jewel. Stylaster
corals typically contain surface pores (again, see figure 4)
that are arranged in circular patterns called cyclo-systems
(Pienaar, 1981; Rolandi, 1981; Rolandi et al., 2005). Corals
from the Corallium genus typically have striated or scalloped structures (Smith et al., 2007).
The specific gravity of coral ranges from 2.37 to 2.75,
and it is strongly dependent on the porosity of the individual piece. The refractive indices of both calcitic and aragonitic corals are a function of their calcium carbonate configuration. RI is also influenced by the magnesium content
of calcitic corals and the strontium content of aragonitic
corals (Rolandi et al., 2005). Because corals from these two
genera can be difficult to separate using only classical
Figure 3. The pink Corallium coral in this 6-cm-tall

clip is set with emerald, amethyst, and diamond.
Courtesy of Van Cleef & Arpels.
Figure 2. This early 15th century Portuguese reliquary

holding a fragment of the Holy Cross features red
Corallium coral. The piece measures 53.2 cm tall and
20.0 cm wide. Courtesy of Museu Nacional de
Machado de Castro, Coimbra, Portugal; collection no.
MNMC 6036, DDF/IMC.
market today are from the Corallium genus, including
the prized ox-blood (dark red; figure 2) and angelskin (light pink; figure 3) colors.
GEMS & GEMOLOGY
SPRING 2009
49
detailed taxonomy, color description, and geographic origin. All seven specimens were represented as being of natural color, and this was verified by standard gemological
examination (i.e., microscopic observation and fluorescence to UV radiation). Note that we could not obtain a
group photo of the samples because we did not have access
to all of them at the same time.
We recorded Raman scattering spectra with a Jobin
Yvon T64000 spectrometer coupled with an Olympus
microscope at the University of Nantes. The study used
an excitation laser emitting at 514 nm (argon ion laser),
with a power of 2 mW and a resolution of 1 cm1 at room
temperature. A low-power laser was selected to avoid
destroying the corals fragile organic matter. For samples
with varying color distribution, spectra were taken in different-colored areas as identified by the spectrometers
microscope (50 magnification). The analyzed areas measured about 2 2 m. Measurements were repeated at
least twice in the same area to ensure reproducibility.
Exposure time was 240 seconds.
Figure 4. Corals of the Stylasteridae family, such as

this pink specimen of the species Stylaster roseus
(sample no. 1; other portions of this same specimen
were red) are protected by CITES. Note the distinctive surface pores of this lace coral. Courtesy of
MNHN, Paris (Collection no. MNHN-Hyd.00000001); photo by S. Karampelas, image width 9 cm.
Results and Discussion. Raman spectra of two light
gemological methods, the present study explores their

identification based on Raman scattering analysis.
Materials and Methods. This study was carried out on

seven pink-to-red coral specimens from the collections of
the Musum National dHistoire Naturelle (MNHN;
National Museum of Natural History) in Paris and the
Centre de Recherche Gemmologique (CRG; Center for
Gemmological Research) in Nantes. Two of the samples
(nos. 1 and 2) belong to the Stylaster genus, and the other
five (nos. 37) to the Corallium genus. See table 1 for their
pink corals of different species, Corallium secundum and

Stylaster sanguineus, in the 1600600 cm1 range are presented in figure 5. In the C. secundum spectrum, two
bands at ~1088 and 714 cm1 correspond to the 1 symmetric and 4 in-plane bending of carbonate ions (CO2
3 ) in calcite, respectively (Urmos et al., 1991). The two bands at
~1520 and 1130 cm1 are characteristic of unmethylated
polyenic pigments; they are assigned, respectively, to C-C
(1) and C=C (2) stretching (Merlin and Del-Dubois,
1986; Karampelas et al., 2007).
Polyenes (or polyacetylenes, cited here as unmethylated polyenes) are organic compounds that contain several
sequences of alternating double and single carbon-carbon
bonds (i.e., a polyenic chain). Polyenic molecules can have
TABLE 1. Taxonomya, color range, and geographic origin of the seven coral samples.
Sample
no.
Class
Order
Family
Genus
and species
Color range
Geographic
origin
Collectionb
1
2
Hydrozoa
Hydrozoa
Anthoathecatae
Anthoathecatae
Stylasteridae
Stylasteridae
Stylaster roseus
Stylaster sanguineus
Light pink to red

Light to dark pink
MNHN
MNHN
Anthozoa
Alcyonacea
Coralliidae
Corallium rubrum
Light to dark red
Anthozoa
Alcyonacea
Coralliidae
Corallium rubrum
Light to dark red
5
6
Anthozoa
Anthozoa
Alcyonacea
Alcyonacea
Coralliidae
Coralliidae
Corallium rubrum
Corallium rubrum
Anthozoa
Alcyonacea
Coralliidae
Corallium secundum
Light to dark pink

Light to dark red
to orange
Light to dark pink
Unknown
Pacific Ocean
(Hawaiian Islands)
Mediterranean Sea
(southern France)
Mediterranean Sea
(southern France)
Unknown
Atlantic Ocean
(Republic of Senegal)
Pacific Ocean
(Hawaiian Islands)
CRG
CRG
MNHN
MNHN
MNHN
Taxonomy follows the Integrated Taxonomic Information System, www.itis.gov (Phylum: Cnidaria).
= Musum National d'Histoire Naturelle (Paris); CRG = Centre de Recherche Gemmologique (Nantes, France).
b MNHN
50
GEMS & GEMOLOGY
SPRING 2009
various substitutions at their terminations. The general

chemical formula of these compounds is R-(-CH=CH-)n-R
(where n = number of double bonds, and R and R = end
groups; Merlin and Del-Dubois, 1986; Karampelas et al.,
2007). We obtained similar results for different areas in all
the samples within the Corallium genus (1: 1130 5 cm1;
2: 1520 10 cm1). Raman peaks at the same positions
have been documented in the spectra of other Corallium
species (see Merlin and Del-Dubois, 1986; Urmos et al.,
1991; Kaczorowska et al., 2003; Rolandi et al., 2005; Smith
et al., 2007; and Fan and Yang, 2008).
In the spectrum of S. sanguineus, we observed a band
at about 1085 cm1 and a doublet at 706 and 702 cm1 (see
inset to figure 5); the band and doublet correspond to the
1 symmetric and 4 in-plane bending of CO32 ions in
aragonite, respectively (Urmos et al., 1991). Two additional peaks at ~1511 and 1155 cm1 are characteristic of carotenoid pigments (Merlin and Del-Dubois, 1986). Carotenoid pigments are also polyenic molecules, with various
substitutions on their terminal ends and an additional
Figure 5. In the Raman spectrum of Corallium secundum (red line; sample no. 7), bands due to calcite are
observed at ~1088 cm1 and 714 cm1. The two bands
at ~1520 and 1130 cm1 are due to unmethylated
polyenic pigments. In the spectrum of Stylaster sanguineus (orange line; sample no. 2), aragonite bands are
observed at ~1085, 706, and 702 cm1 (see inset).
Two sharp bands at ~1511 and 1155 cm1 are due to
carotenoids. Among corals of gemological interest, only
the Stylaster genus appears to contain aragonite and
carotenoid pigments. Note that the spectra are normalized to the major Raman peak of carbonate (~1088 cm1
for calcite and 1085 cm1 for aragonite). The spectra
have been stacked and shifted vertically for clarity.
four methyl groups attached to their polyenic chain. In

carotenoids, the 1 vibration is modified by the presence
of additional methyl (CH3) groups in the polyenic chain
(methyl in-plane bending modes); hence, the Raman peak
is shifted 25 (10) cm1 (Okamoto et al., 1984). This difference (1130 vs. 1155 cm1 in figure 5) is substantial enough
to separate unmethylated polyenes (corals of the
Corallium genus) from carotenoids (corals from Stylaster)
unambiguously.
The position of the other peak related to polyenic
chains, at 1520 10 cm1 (under 514 nm excitation), is
dependent on the number of carbon double bonds in the
chain. Thus, for a given number of double bonds, this particular Raman peak is in the same position for a carotenoid
and an unmethylated polyenic molecule (Okamoto et al.,
1984). In different areas of the two Stylaster corals, we
obtained similar results for each of the two peaks (1155 5
cm1 and 1520 10 cm1). Rolandi et al. (2005), again, measured peaks at the same positions for other species within
the Stylaster (stated as Allopora) genus. Destructive studies on extracted pigments from Stylaster corals also
demonstrated that they contain carotenoids (Ronneberg
et al., 1979).
Moreover, a recent study has shown that the different
colors seen in some corals are actually due to mixtures of
several types of carotenoids (e.g., orange Stylaster) or
unmethylated polyenes (e.g., pink Corallium; Karampelas
et al., 2007). Thus, the natural colors of pink-to-red
Stylaster and Corallium corals result from the nature and
relative proportions of several carotenoids and unmethylated polyenes, respectively.
Conclusion. Gem corals from the Stylaster genus are protected by Appendix II of CITES, and a certificate issued by
the management authority of the country (or state) of
export is required for any new material released
(import/export of a Stylaster coral from old stock is permitted; for more information about the legal framework of
importing/exporting species protected by Appendix II, consult CITES, 2008c). Currently, there are no such restrictions on corals of the Corallium genus.
This study showed that pink-to-red Stylaster corals
contain carotenoid pigments and are aragonitic, while
those from the Corallium genus contain unmethylated
polyenic pigments and are calcitic, so they can be separated on this basis using Raman spectroscopy. A review
of the literature indicates that Stylaster is the only pinkto-red gem coral that contains carotenoid pigments.
Thus, if a coral from the Stylaster genus cannot be distinguished by its surface features, it can be identified nondestructively using Raman scattering. Raman spectra of
additional natural-color coral specimens from these and
other genera need to be collected to refine this criterion.
Raman spectroscopy may also prove useful to gemologists in detecting other materials protected by CITES,
such as ivory and pearls.
GEMS & GEMOLOGY
SPRING 2009
51
ABOUT THE AUTHORS

At the time this article was prepared, Dr. Karampelas
(s.karampelas@gubelingemlab.ch) was a Ph.D. student in the
Department of Mineralogy-Petrology-Economic Geology at
Aristotle University of Thessaloniki in Greece and at the Institut
des Matriaux Jean Rouxel at the University of Nantes (IMNCNRS), France. He is now a postdoctoral researcher at the
Gbelin Gem Lab in Lucerne, Switzerland. Dr. Fritsch is professor of physics at IMN-CNRS. Dr. Rondeau is assistant professor at the University of Nantes. Mrs. Andouche is curator of
the Cnidaria collection, and Dr. Mtivier is assistant professor,
at the Musum National dHistoire Naturelle in Paris.
ACKNOWLEDGMENTS
The authors thank the Musum National dHistoire Naturelle in
Paris, which loaned five of the samples for this study. Rui Galopim
de Carvalho (Sintra, Portugal) and Pedro Miguel Ferro (Museu
Nacional de Machado de Castro, Coimbra, Portugal) kindly
arranged for the photo in figure 2, and Dominique Dufermont (Van
Cleef & Arpels, Paris) generously provided the photo in figure 3.
REFERENCES
Convention on International Trade in Endangered Species (2008a)
Consideration of proposals of amendment of Appendices I and
II. www.cites.org/eng/cop/14/prop/E14-P21.pdf [date accessed:
Oct. 25, 2008].
Convention on International Trade in Endangered Species (2008b)
Appendices I, II and III. www.cites.org/eng/app/appendices.shtml
[date accessed: Oct. 25, 2008].
Convention on International Trade in Endangered Species (2008c)
Article IV: Regulation of trade in specimens of species included in Appendix II. www.cites.org/eng/disc/text.shtml#IV [date
accessed: Oct. 25, 2008].
Fan L., Yang M. (2008) In situ resonance Raman spectra of organic
pigments in Momo coral. Journal of China University of Geosciences, Vol. 19, No. 2, pp. 146151.
52
Fish and Wildlife Service (2008) New CITES Appendix-III listing

by China for four species of Corallium coral. www.fws.gov/le/
PubBulletins/PB040808CITESAppendixIIIListingChina.pdf
[date accessed: Sept. 30, 2008].
Kaczorowska B., Hacura A., Kupka T., Wrzalik R., Talik E.,
Pasterny G., Matuszewska A. (2003) Spectroscopic characterization of natural corals. Analytical and Bioanalytical Chemistry,
Vol. 377, No. 6, pp. 10321037.
Karampelas S., Fritsch E., Sklavounos S., Soldatos T. (2007) Polyacetylenic pigments found in pearls and corals. Proceedings of
the 30th International Gemmological Conference, Institute of
Problems of Chemical Physics, Russian Academy of Sciences,
Moscow, pp. 4951.
Liverino B. (1989) Red Coral, Jewel of the Sea. Transl. by J. H.
Johnson, Analisi Editions, Bologna, Italy, 208 pp.
Merlin J.C., Del-Dubois M.L. (1986) Resonance Raman characterization of polyacetylenic pigments in the calcareous skeleton.
Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, Vol. 84B, No. 1, pp. 97103.
Okamoto H., Saito S., Hamaguchi H., Tasumi M., Eugster C.
(1984) Resonance Raman spectra and excitation profiles of
tetradesmethyl--carotene. Journal of Raman Spectroscopy,
Vol. 15, No. 5, pp. 331335.
Pienaar H.S. (1981) African star coral, a new precious stylasterine
coral from the Aghulas bank, South Africa. Journal of Gemmology, Vol. 17, No. 8, pp. 589601.
Rolandi V. (1981) Les gemmes du rgne animal: tude gemmologique des scrtions des Cnidaires. Revue de Gemmologie
a.f.g., Vol. 66, No. 1, pp. 39.
Rolandi V., Brajkovic A., Adamo I., Bocchio R., Landonio M.
(2005) Gem corals: Classification and spectroscopic features.
Australian Gemmologist, Vol. 22, No. 7, pp. 285297.
Ronneberg H., Fox D., Liaaen-Jensen S. (1979) Animal carotenoidscarotenoproteins from hydrocorals. Comparative Biochemistry
and Physiology Part B: Biochemistry and Molecular Biology,
Vol. 64B, No. 4, pp. 407408.
Smith C.P., McClure S., Eaton-Magaa S., Kondo D. (2007) Pinkto-red coral: A guide to determining origin of color. G&G, Vol.
43, No. 1, pp. 415.
Urmos J., Sharma S.K., Mackenzie F.T. (1991) Characterization of
some biogenic carbonates with Raman spectroscopy. American
Mineralogist, Vol. 76, No. 34, pp. 641647.
GEMS & GEMOLOGY
SPRING 2009
Editors
Thomas M. Moses and
Shane F. McClure
GIA Laboratory
CUBIC ZIRCONIA Reportedly

Coated with Nanocrystalline
Synthetic Diamond
It has been two decades since G&G
first reported on cubic zirconia with a
thick diamond-like coating (Spring
1987 Gem News, p. 52) and E. Fritsch
et al. commented on the remote possibility of growing a thinner monocrystalline film on cubic zirconia (CZ)
that would give the thermal conductivity of diamond (A preliminary gemological study of synthetic diamond thin
films, Summer 1989 G&G, pp.
8490). Recently, the GIA Laboratory
had the opportunity to study some
new, commercially available samples
of cubic zirconia reported to be coated
with nanocrystalline synthetic diamond (typically defined as having grain
sizes less than 500 nm). Serenity Technologies (Temecula, California) and
Zirconmania (Los Angeles) supplied
the lab with material they market as
EternityCZ and Diamond-Veneer,
respectively.
The Serenity Technologies website
claims it is virtually impossible to
visually identify EternityCZ as anything but a diamond. The only way to
Editors note: All items are written by staff

members of the GIA Laboratory.
LAB NOTES
Figure 1. This 0.31 ct cubic zirconia from Serenity Technologies is

reportedly coated with nanocrystalline synthetic diamond.
positively identify EternityCZ is by its

weight, hardness and chemical component (www.serenitytechnology.com).
It also states that the RI and dispersion
change due to the nanocrystalline diamond coating. Zirconmania makes
similar claims about Diamond-Veneer
(http://diamondveneer.net).
We examined 14 round brilliant
samples from Serenity (0.290.32 ct;
e.g., figure 1) and four from Zirconmania (0.132.36 ct). Seventeen of the
specimens corresponded to the D range
on the GIA diamond color grading
scale; the last was equivalent to an E.
All the analyses we performed successfully identified the samples as diamond simulants. Microscopic examination with darkfield illumination
revealed the orange pavilion flash typi-
Figure 2. In reflected light, the

coating on the surface of this 2.36
ct cubic zirconia is clearly visible.
Also, the chips look conchoidal,
in contrast to the typical step-like
appearance of diamond. Field of
view 1.7 1.3 mm.
cal of CZ. All also tested as not diamond with a thermal conductivity
diamond tester and the DiamondSure
instrument. All the specimens showed
chips, in various sizes, which appeared
conchoidal, not step-like as one might
expect for diamond. Additionally, all
the samples revealed the presence of a
coating on the crown and pavilion
when viewed in reflected light (figure
2). The coatings appearance varied
within facets and particularly at the
GEMS & GEMOLOGY
SPRING 2009
53
facet junctions. Finally, the coating

could be scratched with a corundum
(Mohs 9) hardness pointtherefore, it
did not seem to add significantly to the
CZs durability.
SG values ranged from 5.91 to
5.96, as calculated by the DiaVision
noncontact measuring device. This
range coincides with the reported SG
of 5.95 for yttrium-stabilized CZ (M.
ODonoghue, Ed., Gems, 6th ed.,
Butterworth-Heinemann, Oxford, UK,
2006). We were unable to measure the
RI of the coated CZs using a standard
gemological refractometer, but readthrough observations, which provide
a relative approximation of RI, yielded
results more consistent with CZ than
diamond.
Raman, photoluminescence (PL; at
325, 488, 514, and 830 nm laser excitations), and Fourier-transform infrared
(FTIR) analyses using standard techniques revealed no peaks associated
with diamond. The Raman and FTIR
spectra matched those of CZ. Although we have not yet had an opportunity to determine the thickness of
the coating, it appears to be too thin to
contribute significantly to the spectra
dominated by the underlying material.
The results of our tests establish
that these EternityCZ and DiamondVeneer samples are easily separated
from diamond. Characterization of the
coating material is the focus of ongoing
research. If the coating material is
nanocrystalline synthetic diamond, it
does not take much imagination to predict that natural diamond, instead of
CZ, might be used as a future substrate
material to improve the appearance or
the color of the stone (see, e.g., Summer
1991 Gem News, pp. 118119). Should
such a treatment become commercially available, it could be far more difficult for gemologists to identify.
Sally Eaton-Magaa and
Karen M. Chadwick
DIAMOND
Assemblages of K-Feldspar,
Hematite-Magnetite, and Quartz
in Etch Channels
Most minerals seen in diamond occur
54
LAB NOTES
which made up most of this assemblage, as a mixture of hematite and

magnetite. The bottom part of this
mixture was totally enclosed in the
diamond and showed well-formed
stepped surfaces. The transparent gray
inclusions were identified as Kfeldspar, and the white inclusions
were quartz.
A 1.67 ct cut-cornered rectangular
step-cut Fancy Dark brown diamond
contained an assemblage almost as
large (~0.9 0.6 0.5 mm), which
broke the surface of the pavilion near
a corner. Raman spectroscopy identified this assemblage as a mixture of
hematite-magnetite and quartz.
The presence of dark sectorial
clouds in all five stones suggested that
these diamonds could have formed in
a similar environment. Since quartz
and K-feldspar would not be stable in
the high-temperature and high-pres-
Figure 3. This 2.46 ct Fancy black

diamond contained dark sectorial clouds as well as assemblages
of mineral inclusions, which
formed in etch channels.
as single crystals. Rarely have we

encountered inclusions of mineral
assemblages formed at conditions
outside the diamond stability field.
Recently, the New York laboratory
examined a group of five diamonds
(1.673.70 ct) submitted together by a
single client. These stones contained
dark sectorial clouds, as well as
numerous etch pits and etch channels. Three of the diamonds were
color graded Fancy black, and the
other two were graded Fancy Dark
brown. Infrared spectroscopy showed
a relatively high concentration of
hydrogen in all the stones, which is
the likely cause of the sectorial clouds
that produced the dark colors.
Of greatest interest were the
assemblages of mineral inclusions
seen in the etch pits and channels of
all the diamonds. Microscopic examination revealed these as opaque dark
brown, transparent gray, and transparent near-colorless to white materials.
The largest assemblage (~1.1 0.5
1.0 mm)observed in a 2.46 ct oval
modified brilliant-cut Fancy black diamond (figure 3)consisted of four
minerals (figure 4) in an etch channel
that broke the surface at the crown
shoulder. Raman spectroscopy identified the opaque dark brown portion,
Figure 4. This assemblage of

hematite-magnetite, K-feldspar,
and quartz crystallized in an
etch channel of the diamond in
figure 3 as secondary inclusions.
The orange color is likely due to
iron staining from weathering of
the iron-bearing hematite-magnetite. Field of view 0.8 mm.
Hematitemagnetite
Quartz
GEMS & GEMOLOGY
K-feldspar
SPRING 2009
sure stability field of diamond, these

assemblages could only have formed
at conditions outside the diamond stability field, such as within the continental crust, where these minerals are
stable. Furthermore, the well-developed crystalline quality and morphology of these included minerals suggest
formation and growth after the diamond was brought to a relatively shallow depth in the earth.
Most of the mineral inclusions we
observe in the laboratory are protogenetic or syngenetic inclusions formed
within the diamond stability field.
These assemblages provide excellent
examples of epigenetic mineral inclusions that formed outside the diamond stability field and are typically
associated with crustal processes.
Wai L. Win and Ren Lu
Clarity Grading Radiation Stains

Radiation stains can appear as green or
brown patches in diamond. They are
typically associated with naturals,
indented naturals, feathers, or etch
features, and are thought to be caused
by exposure to radioactive elements in
a near-surface, low-temperature environment. Radiation stains are green
when they form and can turn brown if
the diamond is subjected to relatively
high temperatures, such as those that
occur during the polishing process.
Their impact on a diamonds clarity
grade depends on whether or not they
penetrate the surface of the stone.
Recently, the New York laboratory
examined a 1.01 ct round brilliant cut
submitted for grading. An etch channel
extended into the diamond from a
bezel surface; it was identifiable by its
distinct elongated form and angular
outline, as well as the growth markings along its edges (figure 5). Spherical
brown zones were visible reaching
beyond the etch channel in two areas.
Their unusual appearance, color, and
relationship to the etch channel immediately identified them as radiation
stains. They probably formed when
radioactive particles lodged in the etch
channel, affecting only those areas.
To assess the effect of a radiation
LAB NOTES
considered inclusions if the penetration is visible at 10 magnification, as

was the case with the stains illustrated here. Otherwise, the staining is
treated as a blemish, which has only a
minor effect on the clarity grade.
Vincent Cracco and
Alyssa Grodotzke
Rare Mixed Type (Ia/IIb) Diamond

with Nitrogen and Boron Centers
Figure 5. The two brown patches

of color in this etch channel,
shown here at 100 magnification, are radiation stains. The
etch channel was visible at 10
magnification; as a result, these
stains were considered inclusions
for the purpose of clarity grading.
stain on a diamonds clarity grade, the

grader must first determine if the stain
is an inclusion or a blemish. While all
radiation stains penetrate into the diamond to a certain extent, they are only
Type IIb diamonds are among the

rarest and most valued of all natural
diamonds. Their characteristic blue
color originates from a very low concentration of boron impurities, which
is also responsible for their distinctive
properties (see, e.g., J. M. King et al.,
Characterizing natural-color type IIb
blue diamonds, Winter 1998 G&G,
pp. 246268). In general, type IIb diamonds do not have the quantity and
variety of inclusions often observed in
type I and some type IIa diamonds,
which are differentiated by the presence (type I) and relative absence (type
IIa) of nitrogen impurities.
The New York laboratory recently
examined a very rare natural type IIb
diamond with a noticeable type Ia
component. This 0.17 ct round brilliant cut was graded Fancy Light grayish blue (figure 6, left). Numerous
Figure 6. This 0.17 ct Fancy Light grayish blue type IIb diamond (left) contained unusual amounts of nitrogen and hydrogen, characteristic of a type
Ia diamond. The DiamondView fluorescence image (right) shows the heterogeneous distribution of boron (darker blue areas represented by spots 1
and 4) and nitrogen (lighter blue areas represented by spots 2 and 3).
SP1
SP4
SP3
GEMS & GEMOLOGY
SP2
SPRING 2009
55
graphite particles were the only inclusions observed. Electrical conductivity,

as measured with a gemological conductometer, was consistent with that
of a typical type IIb diamond. The
stone showed very weak blue fluorescence to long-wave ultraviolet (UV)
radiation and was inert to short-wave
UV. It showed both blue and red phosphorescence, as is typical of natural IIb
stones. DiamondView images revealed
zones with various hues of blue fluorescence (figure 6, right), which suggested a heterogeneous distribution of
defects and impurities.
The mid-infrared spectrum (figure
7) had a dominant IIb character, with
a boron component indicated by a
band at ~2801 cm1. However, evidence of a nitrogen component with
both A and B aggregates was clearly
present in the 12801170 cm1
region, indicating a type Ia nature as
well. A noticeable amount of hydrogen was also observed at 3107 and
1405 cm1, whichto the best of this
contributors knowledgeis the first
time hydrogen has been directly
observed in a natural type IIb diamond. The relative intensities of the
boron, nitrogen, and hydrogen bands
varied noticeably among the regions
sampled. Unfortunately, the nature
of our infrared system and the shape
of the stone did not allow us to correlate the different IR features to specific regions that were indicated by the
DiamondView image.
However, we were able to correlate low-temperature PL spectra at
325, 488, and 514 nm excitations to
those specific regions (figure 8). This
technique allowed us to probe pointby-point for the presence or absence of
nitrogen-related features, effectively
mapping the stones type IIb and Ia
regions. Specifically, PL spectra taken
from the lighter blue region (spots 2
and 3 in figure 6, right) showed nitrogen features typical of type Ia stones,
such as the N3, H3, H4, and NV0 centers; spectra from the darker blue
regions (spots 1 and 4) exhibited virtually none of these features, correlating
to type IIb. This PL mapping allowed a
rare direct observation of the N3
56
LAB NOTES
Figure 7. The mid-IR spectrum of the 0.17 ct diamond shows a dominantly

type IIb nature, with a boron-related band near 2801 cm1, but alsosee
insetsevidence of a type Ia nature, with nitrogen (e.g., the 1173 cm1 band,
correlating to B-aggregates) and hydrogen (3107 and 1405 cm1) impurities.
defect at 415 nm (along with H3, H4,

and NV0) in a mostly type IIb stone,
which is quite noteworthy because
the N3 defect is considered the key

feature in the distinction of type I
from type II diamonds.
Figure 8. Taken at 325 nm UV wavelength, low-temperature PL spectra collected from the lighter blue areas (spots 2 and 3) in figure 6 showed optical
centers (e.g., N3, H3, H4, and NV0) consistent with type Ia diamonds,
whereas spectra from the darker blue areas (spots 1 and 4) were free of
these nitrogen features, which is consistent with a type IIb diamond.
GEMS & GEMOLOGY
SPRING 2009
Figure 9. The parcel of rough spinel on the left, reportedly from Tajikistan, contains pieces weighing up to
48.5 g. The seven faceted spinels on the right (9.0428.16 ct) were fashioned from some of this rough.
Mixed-type diamonds with a type

IIb component have been examined
previously at the GIA Laboratory (e.g.,
Lab Notes: Summer 2000, pp. 156157;
Summer 2005, pp. 167168; and Winter
2008, pp. 364365). However, type IIb
diamonds with an extensive type Ia
component have been observed only
rarely. The mixed-type nature indicates
a substantial change in the geochemical environment during the diamonds
crystallization. Further analysis of the
available data may shed light on the
interaction between boron, carbon,
nitrogen, and hydrogen, and their
impact on the spectral and physical
properties.
Ren Lu
Purplish Pink SPINEL from

TajikistanBefore and
After Cutting
In December 2007, Pakistan-based
client Syed Iftikhar Hussain submitted a parcel of spinel rough reportedly
from Tajikistan (figure 9, left). These
84 samples, the largest weighing 48.5
g, exhibited varying saturations of purplish pink color. Little has been written on the properties of Tajik spinel
(see, e.g., J. I. Koivula and R. C. Kammerling, Examination of a gem
spinel crystal from the Pamir Mountains, Zeitschrift der Deutschen
Gemmologischen Gesellschaft, Vol.
38, 1989, pp. 8588), so in November
2008 the Bangkok laboratory was for-
Figure 10. Among the inclusions observed in the rough spinels were a fine
euhedral crystal (left, magnified 75) and crystals with white particulate
trails forming comet tails (right, magnified 35).
LAB NOTES
tunate to have an opportunity to

briefly examine seven stones that the
client had faceted from this parcel (figure 9, right).
Most of the original rough consisted of broken pieces, and only a few
showed the octahedral crystal forms
typical of spinel. We could not perform
accurate RI measurements because of
the lack of flat surfaces, so we had to
rely on other tests. The hydrostatic SG
measurements, spectra seen with a
handheld spectroscope, polariscope
reactions, and UV fluorescence were
consistent with spinel. These observations were further substantiated by PL
spectroscopy (514 nm laser excitation
at room temperature) on the largest
piece, which proved it was natural
spinel. The most prominent inclusions
seen in the samples were euhedral
crystals, needles, and crystals with
white particulate trails forming
comet tails (figure 10).
After the rough was cut, we
obtained standard gemological properties for the seven faceted stones. The
results were fairly consistent: RI
1.7121.713, SG3.593.62, strong
red fluorescence to long-wave UV
radiation and weak red to weak-tomoderate orange (some with a chalky
greenish cast) fluorescence to shortwave UV, and a characteristic organ
pipe spectrum (with some general
absorption in the orange/yellow and
part of the green region) seen with the
GEMS & GEMOLOGY
SPRING 2009
57
11.96 ct pear shape hosted the most

internal features, which consisted of a
plane of octahedral negative crystals
and some euhedral crystals (figure 11).
Tiny negative crystals were only
faintly visible in one other stone.
Unfortunately, there was no time to
identify inclusions in either the rough
or cut spinels with Raman spectroscopy. The PL spectrum of the
pear-shaped stone closely matched
that of the rough sample.
Nicholas Sturman
Figure 11. One of the faceted spinels contained a plane of negative crystals
(left, magnified 50) and a group of euhedral crystals (right, magnified 20).
spectroscope. While the refractive
indices were almost identical to that
of the crystal detailed by Koivula and
Kammerling (and a Tajik spinel reported in the Spring 1989 Lab Notes, pp.
58
LAB NOTES
3940), the SGs varied slightly.

Since the cleanest pieces of rough
were likely selected for faceting, it
was no surprise that six of the cut
stones showed few inclusions. The
PHOTO CREDITS
Robison McMurtry1; Karen M.
Chadwick2; Jian Xin (Jae) Liao3, 6 left;
Wai L. Win4; Jason Darley5; Ren Lu
6 right; Ken Scarratt9 left, 10; Suchada
Kittayachaiwattana9 right; Nicholas
Sturman11.
GEMS & GEMOLOGY
SPRING 2009
Editor
Brendan M. Laurs (blaurs@gia.edu)
Contributing Editors
Emmanuel Fritsch, CNRS, Institut des Matriaux
Jean Rouxel (IMN), University of Nantes, France
(fritsch@cnrs-imn.fr)
Michael Krzemnicki, SSEF Swiss
Gemmological Institute
Basel, Switzerland
(gemlab@ssef.ch)
Franck Notari, GemTechLab,
Geneva, Switzerland
(franck.notari@gemtechlab.ch)
Kenneth V. G. Scarratt, GIA Laboratory,
Bangkok, Thailand
(ken.scarratt@gia.edu)
Despite the global economic downturn, the annual Tucson

gem and mineral shows again offered a wide variety of
materials, including a spectacular pendant set with an
untreated 3 ct Colombian emerald and a 5 ct D-Flawless
diamond (figure 1). In addition, the shows saw the debut of
some interesting new gem materials and localities, many of
which will be described in future issues of G&G.
Overall, dealers at the show had low sales expectations, but many were pleasantly surprised. Although show
attendance was light compared to previous years, those
dealers with unusual and attractive merchandise at good
prices typically did okay. Likewise, Gems & Gemologys
sales of subscriptions, back issues, In Review books, and
charts surpassed expectations, with special interest in the
new G&G flash drives preloaded with back issue PDFs.
This years theme for the Tucson Gem and Mineral
Society show was Mineral Oddities. Next years Tucson
Gem and Mineral Show will take place February 1114,
and the theme will be Gems & Gem Minerals, which
should be of particular interest to G&G readers.
G&G appreciates the assistance of the many friends
who shared material and information with us this year,
and also thanks the American Gem Trade Association for
providing space to photograph these items during the
AGTA show.
New play-of-color opal from Welo, Ethiopia. A new

source of high-quality play-of-color opal was discovered in
early 2008 in Welo Province, Ethiopia, about 500 km
north of Addis Ababa. This deposit is geographically
Figure 1. One of many pieces of fine jewelry seen at the

Tucson shows, this Van Cleef & Arpels diamond necklace features a detachable 3.03 ct untreated Colombian
emerald and a 5.09 ct D-Flawless diamond. Courtesy of
Robert E. Kane and Fine Gems International, Helena,
Montana; photo by Tino Hammid.
Editors note: Interested contributors should send information and illustrations to Brendan Laurs at blaurs@gia.edu or
GIA, The Robert Mouawad Campus, 5345 Armada Drive,
Carlsbad, CA 92008. Original photos can be returned after
consideration or publication.
GEMS & GEMOLOGY, Vol. 45, No. 1, pp. 5975
GEM NEWS INTERNATIONAL
GEMS & GEMOLOGY
SPRING 2009
59
Figure 2. These opals (7.5523.48 ct) originate from a

new deposit in Welo Province, Ethiopia. This material
typically has a lighter bodycolor than opals from
Shewa Province. Photo by Robert Weldon.
Figure 3. Several Welo samples, such as this 8.19 ct

piece, showed a columnar structure of play-of-color
opal within common opal, which is characteristic of
opals from Ethiopia. Photo by B. Rondeau.
distinct from the Mezezo deposit in Shewa Province,

which was discovered in the early 1990s (see, e.g., Spring
1994 Gem News, pp. 5253).
These contributors examined a parcel of about five
rough and 30 cut Welo opals supplied by Opalinda and
Eyaopal, the main distributors of this material. The cabochons showed good play-of-color (figure 2); the vast majority were white and transparent, but some had a bodycolor
varying from light yellow to dark chocolate brown.
Compared to Mezezo opals (e.g., J.-P. Gauthier et al.,
Lopale dEthiopie: Gemmologie ordinaire et caractristiques exceptionnelles, Revue de Gemmologie a.f.g., No.
149, 2004, pp. 1523), those from the new deposit generally appear much whiter. We noted all spectral colors in the
play-of-color in our samples. Most of the cabochons were
similar in appearance to opals from Australia or Brazil.
However, many samples displayed a columnar structure of
play-of-color opal within common opal (figure 3), as first
described in material from Mezezo (again, see Gauthier et
al., 2004). This feature is only very rarely observed in opals
from sources outside Ethiopia.
The hydrostatic SG of the opals ranged from 1.80 to
2.10. This broad range is in part due to the high porosity
of some samples, as revealed by a significant weight
increase after immersion in water (up to 8%). Fluorescence varied from inert to moderate yellowish white to
both long- and short-wave ultraviolet (UV) radiation.
Samples that were inert displayed an unexpected greenish
phosphorescence of moderate intensity. No luminescence
was observed in the opals with a yellow-to-brown bodycolor, even the light ones; these darker bodycolors are
probably due to the presence of iron, which quenches
luminescence. The yellow-to-green luminescence is likely due to the presence of uranium (E. Gaillou et al., The
geochemistry of gem opals as evidence of their origin,
Ore Geology Reviews, Vol. 34, 2008, pp. 113126).

Fourier-transform Raman spectra were obtained for several samples using a Bruker RFS 100 spectrometer. All spectra were consistent with opal-CT, with Raman bands at
about 1070, 780, 670, and 345 cm 1, and water-related
bands at about 3200 and 2950 cm 1.
Welo opal is found in volcanic rock, possibly a rhyolite.
The rough samples we examined consisted of opal (either
common or play-of-color) cementing fragments of the host
rock. By contrast, opal from Mezezo fills cavities in rhyolite, forming nodules. Despite these differences, the fact
that columnar structures are seen in opals from both
deposits (but very rarely from elsewhere) seems to indicate
similarities in the conditions of their formation.
Benjamin Rondeau (benjamin.rondeau@univ-nantes.fr)
CNRS, Team 6112
Laboratoire de Plantologie et Godynamique
University of Nantes, France
60
Francesco Mazzero
Opalinda, Paris, France
Eyassu Bekele
Eyaopal, Addis Ababa, Ethiopia
Jean-Pierre Gauthier
Centre de Recherches Gemmologiques
Nantes, France
Emmanuel Fritsch
Gem-quality rhodochrosite from China. The Wudong
mine in China has long been thought to be a source of
fine rhodochrosite, but mining operations have been sporadic and largely undocumented until recently. The mine
is located in the Wuzhou area of Guangxi Zhuang
Autonomous Region, approximately 480 km (300 miles)
northwest of Hong Kong. In early 2007, a group of
GEMS & GEMOLOGY
SPRING 2009
Figure 5. Some fine stones (here, 3.6129.56 ct)

have been cut from the Chinese rhodochrosite.
Courtesy of Collectors Edge Minerals Inc.; photo
by Robert Weldon.
Figure 4. This well-formed rhodochrosite crystal (5
4 cm) from Chinas Wudong mine is attached to a
matrix of quartz, galena, pyrite, and very minor fluorite. Photo by Jeff Scovil.
investors purchased the mine and shifted its emphasis

from base metals (lead-zinc-silver) to specimen- and gemgrade rhodochrosite. In 2008, Collectors Edge Minerals
Inc. made an arrangement with Wudongs owners to
excavate the rhodochrosite pockets. The mine reaches a
depth of over 150 m in a maze of tunnels, stopes, and
shafts that have been excavated using jack-leg drills,
explosives, and ore cars. Rhodochrosite and associated
minerals from Wudong were recently described by B.
Ottens (Rhodochrosit aus dem Blei/Zink-Bergwerk
Wudong bei Liubao, Guangxi, China, Lapis, Vol. 33, No.
10, 2008, pp. 5356).
This contributor and colleagues mapped the site in
August 2006. The mineralized veins vary from several
centimeters to more than 2 m wide, and dip steeply to
nearly vertical. Rhodochrosite occurs as solid fillings,
making the veins appear like red streaks. Where the veins
widen and the structure allows, an open pocket will contain fine crystals. There are several similarities to the
mineralization patterns seen at the now-closed Sweet
Home mine in Colorado (see K. Knox and B. K. Lees,
Gem rhodochrosite from the Sweet Home Mine,
Colorado, Summer 1997 G&G, pp. 122133). Although
Wudongs host rocks are sedimentary and Sweet Homes
are granitic, rhodochrosite crystals from both mines may
be large and sometimes gemmy, occurring with fluorite,
galena, wolframite, chalcopyrite, apatite, quartz, barite,
and sphalerite. A major difference between the two is that
the Wudong mine does not appear to contain tetrahedrite,

which was abundant at the Sweet Home mine.
Most of the rhodochrosite crystals from Wudong are
slightly to heavily etched, making them appear pink and
opaque. A few lack this etching, however, and are quite
attractive with good luster (e.g., figure 4). They range up to
nearly 13 cm in maximum dimension and are rhombohedral, like those from Sweet Home, but often show thin,
bladed habits. The internal characteristics of the Wudong
rhodochrosites indicate a turbulent growth history. Many
of the crystals have significant inclusions and banding,
which makes faceting a challenge and severely limits the
potential for large gems.
Mining so far has produced nearly 100 kg of lapidarygrade material, in addition to mineral specimens. The
lapidary material is being cut into faceted stones (e.g., figure 5) and cabochons, as well as beads, eggs/spheres, and
carvings. As of February 2009, processing of less than 10
kg of material by this contributor had yielded ~150
faceted stones weighing from <1 ct to 3 ct, ~20 faceted
stones of 310 ct, and four stones weighing 10+ ct. In
addition, there were ~50 cabochons up to 20 ct, ~20 polished rhombs up to 30 ct, and ~15 eggs of <50 ct each.
These polished goods are being distributed by Paul Cory
(Iteco Inc., Powell, Ohio), and it will probably take three
years to cut the inventory on hand. The Wudong mine is
producing enough rhodochrosite rough to continue supplying the market created by the Sweet Home mine,
which has been closed since mid-2004 (Spring 2007 GNI,
pp. 6162).
Bryan K. Lees (bryan@collectorsedge.com)
Collectors Edge Minerals Inc.
Golden, Colorado
GEMS & GEMOLOGY
SPRING 2009
61
COLORED STONES AND

ORGANIC MATERIALS
Gem-quality amethyst from Tata, Morocco. A new source
of amethyst reportedly has been discovered in the AntiAtlas Mountains. According to Jack Lowell (Colorado
Gem & Mineral Co., Tempe, Arizona) and mine owner Ait
Ouzrou Mohamed (Agadir, Morocco), production started
in late 2007. Well-formed crystals of amethyst are recovered from soil on Bouodi Mountain (figure 6), located near
the city of Tata. The deposit is mined by a small number
of workers on an occasional basis, depending on the orders
received for the amethyst. Each worker typically gathers 2
kg of material daily using simple hand tools, and about
30% of the production can be polished into cabochons or
faceted stones. The material is mostly suitable for crystal
specimens or cabochons, although some high-quality facet
rough has been produced.
Mr. Lowell loaned one faceted pear-shaped modified
step cut (13.74 ct) and five crystals (3.9436.6 g) to GIA for
examination (e.g., figure 7). The rough exhibited wellformed pyramidal terminations, with typical horizontal
striations on the prism faces. The samples showed a characteristic deep purple triangular color zone within the crystal terminations that was surrounded by near-colorless
quartz (figure 8). In the faceted stone, this color zone was
carefully oriented to present a uniformly deep face-up color.
The following properties were obtained from the
faceted stone and two of the crystals: diaphaneitytransparent to translucent (rough); pleochroismweak, ranging
from pinkish purple to bluish violet; RI1.5421.552;
Figure 6. The new Moroccan amethyst deposit is

located in weathered rock on a remote mountainside.
Photo by Ait Ouzrou Mohamed.
62
Figure 7. Although the Moroccan amethyst is commonly color zoned, it can be fashioned to show an
even face-up purple color, as seen in this 13.74 ct
pear-shaped modified step cut (faceted by Alan
Morgan of Mesa, Arizona). The Moroccan crystal
weighs 11.1 g. Photo by Kevin Schumacher.
hydrostatic SGaverage of 2.65 (measurements varied

0.01); Chelsea filter reactionnone; fluorescenceprimarily inert to both long- and short-wave UV radiation,
though near-colorless areas did exhibit a faint white reac-
Figure 8. Two distinctive features of the Moroccan

amethyst are near-colorless to deep purple color zoning
and inclusions of reddish brown dendritic hematite.
Photomicrograph by D. Beaton; magnified 10.
GEMS & GEMOLOGY
SPRING 2009
tion to short-wave UV. No distinct absorption bands were

observed with a desk-model spectroscope. The properties of
the Moroccan amethyst are generally consistent with those
listed for amethyst in R. Webster (Gems, 5th ed., revised by
P. G. Read, Butterworth-Heinemann, Oxford, UK, 1994, pp.
225229).
Viewed with crossed polarizers, the cut amethyst
revealed bulls-eye and Airy spiral optic figures, as well as
small areas of Brazil-law twinning. Microscopic examination revealed fingerprints composed of fluid remnants
and two-phase fluid-gas inclusions, which are common in
quartz. Distinctive inclusions of reddish brown dendritic
hematite (again, see figure 8) and an unusually large primary two-phase inclusion were also observed. As noted in
amethyst from Sri Lanka by E. J. Gbelin and J. I. Koivula
(Photoatlas of Inclusions in Gemstones, Vol. 2, Opinio
Publishers, Basel, Switzerland, 2005, p. 561), the hematite
inclusions occurred in a colorless growth zone where they
consumed the locally available iron, so the surrounding
quartz was deficient in this chromophore.
Amethyst from Morocco is typically seen as drusy
crystals in geodes that are mined from lava flows (W.
Lieber, Amethyst: Geschichte, Eigenschaften, Fundorte,
Christian Weise Verlag, Munich, Germany, 1994). This is
the first occurrence of well-formed gem-quality amethyst
crystals in Morocco.
Donna Beaton (donna.beaton@gia.edu)
GIA Laboratory, New York
Anahs new ametrine. The Anah mine in southeastern
Bolivia, near the border with Brazil, has enjoyed consistent
production since coming under private control in 1990.
The mine is best known for its ametrine (amethyst-citrine),
a quartz variety that exhibits two principal colors, purple
and yellow (see P. M. Vasconcelos et al., The Anah
ametrine mine, Bolivia, Spring 1994 G&G, pp. 423).
Bolivia is the worlds only commercial source of this gem.
Mine owner Ramiro Rivero has transformed operations in
recent years to better control production, cutting, jewelry
manufacturing, and retail activities, establishing a clear
mine-to-market chain of custody for his product (see
Winter 2001 GNI, pp. 334335).
In August 2008, GIA staff members visited Anah to
videotape and report on the mine-to-market operations.
The Anah mine presently employs 74 workers and is
active in five tunnels (e.g., figures 9 and 10). The company
has also started to process the tailings piles to recover
material that is now popular in the market, such as pale
amethyst (called anahta in Bolivia). Taking advantage of
abundant groundwater supplies, workers wash the ore on
location and then presort it before transport to Minerales y
Metales del Oriente, Mr. Riveros manufacturing arm in
the Bolivian city of Santa Cruz.
Production at the Anah mine is lower than it was a
decade ago, but Mr. Rivero maintains that is because greater
efficiencies in mining, sorting, and cutting have reduced the
amount of rough needed for the value-added operations.
Figure 9. At Bolivias Anah mine, a worker removes an

ore car loaded with quartz-bearing material that will
then be washed and presorted. Photo by R. Weldon.
Nevertheless, the mine still produces some 2,5003,500 kg

of gem-quality material, from a total 120 tonnes of quartz
mined each year. Amethyst averages the highest production
(44%), followed by ametrine (33%), and citrine (23%);
ametrine remains the most lucrative product.
The proportion of ametrine has actually increased (from
20% a decade ago) because of Mr. Riveros success in marketing gems that do not necessarily show the traditional
split in amethyst and citrine colors. In the early years of the
Anah mine, ametrines were often faceted as emerald cuts
showing a distinct color demarcation to mimic the appeal
of bicolored tourmalines. But cutting ametrine for an even
color split wastes much of the quartz. While fantasy cut
gems have long ignored this purple-yellow color demarca-
Figure 10. This tunnel at the Anah mine exhibits the

richness of the quartz deposit. Photo by R. Weldon.
GEMS & GEMOLOGY
SPRING 2009
63
Figure 11. Anah mine

ametrine is cut to maximize yield while also
blending the yellow and
purple colors. The 33.55
ct gem on the left was
cut by Dalan Hargrave;
the 44.23 ct concave-cut
stone on the right was
faceted in Bolivia. Both
gems courtesy of
Minerales y Metales del
Oriente; photos by
R. Weldon.
tion in favor of free-form shapes, Minerales y Metales del

Oriente today focuses on a blend of colors in more traditional shapes (i.e., round, oval, or pear). This strategy maximizes yield. In addition to purple, the resulting mix of colors may exhibit peach or deep orangy red hues when
viewed face-up (e.g., figure 11).
For many years, Minerales y Metales del Oriente developed its own cutting styles and jewelry design prototypes,
but contracted with overseas manufacturers for large-scale
production. Until recently, it worked with a gem cutting
and jewelry manufacturing plant in China. With the global
economic downturn, today the company is producing all
of its cut gems and finished jewelry in Santa Cruz.
Robert Weldon (robert.weldon@gia.edu)
GIA, Carlsbad
Azurite/malachite from Sonora, Mexico. The 2008 Tucson
gem shows saw the notable availability of large quantities
of well-crystallized specimens of azuriteand malachite
pseudomorphs after azuritefrom a new source, the
Figure 12. These attractive cabochons consist of intergrowths of blue azurite and green malachite from a new
source, the Milpillas mine in Sonora, Mexico. The larger
stone measures 4.9 3.5 cm. Photo by Robert Weldon.
64
Milpillas copper mine in the Cananea District of northern

Sonora State, Mexico (see T. P. Moore, Whats New in
MineralsTucson Show 2008, Mineralogical Record,
Vol. 39, No. 3, 2008, pp. 236237). In addition to these
specimens, there was some massive material available
consisting of intergrowths of azurite and malachite that
were well suited for cutting cabochons. Bill Larson
(Palagems.com, Fallbrook, California) obtained approximately 5 kg of this material, half of which has now been
worked by lapidarist Bud Standley (Standley Collections,
San Diego) into several hundred carats of cabochons and
many free-form carvings, some ranging up to 20 cm in
maximum dimension.
Mr. Larson loaned four of the cabochons to GIA for
examination (e.g., figure 12). They consisted of curvilinear
domains of granular blue azurite that were intergrown with
felty aggregates of light- and dark-green malachite (figure 13;
both minerals confirmed by Raman analysis of one sample).
The malachite sprays typically radiated into the azurite, giving the appearance that the malachite partially replaced the
azurite. This is consistent with the pseudomorphous nature
of the malachite in the mineral specimens mentioned
Figure 13. A closer view of the Milpillas material shows

granular azurite intergrown with fibrous malachite.
Photo by Robert Weldon; field of view 3.2 cm.
GEMS & GEMOLOGY
SPRING 2009
above. It is hoped that more of the colorful crystals and massive azurite/malachite intergrowths will be preserved from
the crusher at the Milpillas copper mine.
Because of the relatively low hardness of azurite and
malachite (Mohs 3124), care is required when cutting/polishing and wearing this material, as it may scratch easily
and become dull. Azurite/malachite is therefore appropriate for jewelry such as necklaces and brooches, but not for
rings or items exposed to daily wear.
Brendan M. Laurs
Light yellow-green grossular from Kenya. Deep green
grossular (tsavorite) was discovered in East Africas
Mozambique Belt in the late 1960s (Tanzania) and 1970
(Kenya), as reported by C. R. Bridges (Green grossular garnets [Tsavorites] in East Africa, Summer 1974 G&G, pp.
290295). According to Bridges (p. 293), the material ranged
from colorless to pale yellowish green to rich grass or
emerald green. Examination of the early finds indicated the
bright green was due to vanadium with contributions from
chromium. Manganese was also found in brightly colored
stones, and iron in stones of more yellowish hue (Bridges,
1974). Later work correlated increasing concentrations of V
and Cr to green coloration and increasing Fe to yellow coloration (D. V. Manson and C. M. Stockton, Gem-quality
grossular garnets, Winter 1982 G&G, pp. 204213).
In November 2008, Dudley Blauwet (Dudley Blauwet
Gems, Louisville, Colorado) loaned GIA four samples of
light yellow-green grossular (figure 14) that were reportedly
from the village of Kabanga in the Voi area of Kenya. Voi
has traditionally produced much darker or more saturated
green material (tsavorite), so we took this opportunity to
characterize these lighter samples. The following properties
were obtained from the four stones: colorlight yellowgreen to light yellowish green; RI1.738 (three stones) or
1.736 (one stone); hydrostatic SG3.583.60; fluorescenceweak to moderate red to moderate orange to longwave UV radiation, and moderate orange-yellow to shortwave UV; and no features seen with the desk-model spectroscope. These properties are consistent with grossular (R.
Webster, Gems, 5th ed., revised by P. G. Read, ButterworthHeinemann, Oxford, UK, 1994, pp. 201202). Between
crossed polarizers, very weak to moderate anomalous double refraction was observed. Microscopic examination
revealed short-to-long needles (apparently etch tubes; some
fibrous) at ~70/110 orientation, clusters of small transparent crystals, small thin films, and/or solid particles (figure
15); these resembled inclusions often seen in tsavorite (M.
ODonoghue, Ed., Gems, 6th ed., Butterworth-Heinemann,
Oxford, UK, 2006, pp. 215216). In contrast, a similarly colored large greenish yellow grossular from an unspecified
location in East Africa reported in the Winter 2005 GNI section (pp. 352353) had the strong roiled growth features
characteristic of the hessonite variety of grossular.
All four stones were chemically analyzed by laser ablationinductively coupled plasmamass spectrometry (LAICP-MS). The results confirmed they were grossular, with
Figure 14. These samples of grossular (2.969.96 ct) are

reportedly from Voi, Kenya. Photo by Robert Weldon.
minute amounts of Mn, Mg, and Ti (<0.7 wt.%), and traces

of Fe, K, V, and Cr (<0.1 wt.%). These are common impurities in grossular, and the low concentrations of chromophores are consistent with the pale color of these samples. UV-Vis-NIR absorption spectra showed very weak
bands at 409, 419, and 430 nm, attributed to Mn2+, which
correlates to yellow coloration (Winter 1991 Gem News, p.
258). Although these samples did contain the tsavorite
chromophores Cr and V, their colors were not saturated
enough for them to be classified as such gemologically.
Mr. Blauwet subsequently obtained some additional
Kabanga material from the same supplier, consisting of
several parcels totaling nearly 100 g (in pieces typically
Figure 15. The Kenyan garnets in figure 14 contained

inclusions such as needles, thin films, and transparent
crystals. Photomicrograph by D. Beaton; magnified 30.
GEMS & GEMOLOGY
SPRING 2009
65
Figure 16. These attractive green kornerupines

(0.250.84 ct), identified as the mineral prismatine, are
reportedly from Tanzania. Photo by Robert Weldon.
weighing 12 g, and rarely up to 5 g). The rough consisted

of broken fragments, with the exception of one parcel that
also contained waterworn pieces. Similar to the material
characterized for this report, the additional Kabanga rough
ranged from light yellowish green to mint green.
HyeJin Jang-Green (hjanggre@gia.edu) and
Donna Beaton
Kornerupine (prismatine) from Tanzania. East Africa produces a wide variety of unusual gem minerals, some of
which are quite attractive as well as scientifically interesting. For example, the Kwale area in southern Kenya is
known to produce a V-bearing kornerupine that is bright
apple-green (M. ODonoghue, Ed., Gems, 6th ed.,
Butterworth-Heinemann, Oxford, UK, 2006, p. 421). In
September 2008, GIA received three similarly colored
kornerupines (0.250.84 ct; figure 16) that were reportedly
from Tanzania, loaned by Dudley Blauwet. According to
his supplier, the material came from a single find in late
2007 in the Usambara Mountains, near Tanga. He obtained
Figure 17. The Raman spectrum of the 0.84 ct sample

shows both of the key boron-related bands (~880 and
~800 cm1), thus identifying the stone as prismatine.
20.6 g of rough that yielded 63 faceted stones, cut in calibrated sizes, totaling 13.25 carats.
Gemological testing of the three faceted samples produced the following results: colorintense green; pleochroismlight brownish yellow, strong green, and light bluish
green; RI = 1.6601.662, = 1.6731.674, and =
1.6751.678; birefringence0.0160.017; SG3.283.32;
fluorescencestrong chalky yellow to long-wave, and faint
yellow to short-wave, UV radiation; spectrumno absorption lines seen with the desk-model spectroscope. These
properties are consistent with those reported by
ODonoghue (2006). Parallel growth tubules, partially
healed fissures, and short needles were observed with magnification, as were included crystals (possibly apatite or zircon) and negative crystals. All of these have been reported in
kornerupine by E. J. Gbelin and J. I. Koivula (Photoatlas of
Inclusions in Gemstones, Vol. 1, ABC Edition, Zurich,
1986; Vol. 2, Opinio Verlag, Basel, Switzerland, 2005).
Minerals of the kornerupine group are ferromagnesian
boron-bearing aluminosilicates that can be represented by the
generic formula ( ,Fe,Mg)(Mg,Fe,Al)9(Si,Al,B)5O21(OH,F). The
group includes two minerals differentiated by their boron (B)
content: kornerupine sensu stricto (B<0.5 per formula unit
[pfu]) and prismatine (B>0.5 pfu; E. S. Grew et al.
Prismatine: Revalidation for boron-rich compositions in the
kornerupine group, Mineralogical Magazine, Vol. 60, 1996,
pp. 483491).
EDXRF spectroscopy of our samples detected the major
elements expected for kornerupine, along with traces of V.
Boron cannot be detected by EDXRF, so we could not use
this technique to determine whether our samples were
kornerupine sensu stricto or prismatine. However, two
Raman bands at ~884 and ~803 cm1 in kornerupine group
minerals are sensitive to the presence of B, and their relative intensities can be used to estimate B content. Prismatine shows both the 884 and 803 cm 1 bands, while
kornerupine sensu stricto shows only the 884 cm1 band (B.
Wopenka et al., Raman spectroscopic identification of Bfree and B-rich kornerupine [prismatine], American
Mineralogist, Vol. 84, 1999, pp. 550554).
Raman spectra of all three samples showed bands at
~880 and ~800 cm1, indicating the presence of significant
boron (figure 17). LA-ICP-MS analysis performed on all
three samples confirmed the presence of V and that B content was >0.5 pfu (see table 1). Consequently, these three
kornerupines were B-rich and classified as prismatine.
Pamela Cevallos (pamela.cevallos@gia.edu)
TABLE 1. LA-ICP-MS data for three prismatines from
Tanzania.
Composition
V average (ppm)
B average (ppm)
B average (pfu)
66
0.25 ct
0.30 ct
0.84 ct
937
8740
0.720
1053
6690
0.941
1069
7679
0.828
GEMS & GEMOLOGY
SPRING 2009
Figure 18. These labradorites (1.10 and 0.75 ct) were

cut from material that was sourced from beach gravels located near Ketchikan, Alaska. Photo by Robert
Weldon; the 0.75 ct stone is GIA Collection no. 37787.
Transparent labradorite from Ketchikan, Alaska. In March

2008, GIA received some transparent very light yellow to
very light brownish yellow samples that were represented as
albite-oligoclase (sodic plagioclase) from Ketchikan, Alaska.
Some were donated and others loaned by Dudley Blauwet,
who received the material from a supplier who collected it
by hand from beach gravels in an area about two hours by
boat from the coastal town of Ketchikan. The supplier has
visited the collecting area once a year for the past 10 years, in
trips lasting 34 days, and typically recovers about 300600 g
of rough on each trip. The pebbles are mostly small, and he
only collects material that will cut stones of 3 mm or larger;
rarely, he has found pebbles up to nearly 5 g. Approximately
40% of the collected rough is facetable. More than 1,000
carats have been cut, typically weighing <12 ct, although
some larger stones (up to 12 ct) have been cut. Most of the
material has been sold to cruise ship tourists visiting
Ketchikan, either loose or set into jewelry.
The samples supplied by Mr. Blauwet consisted of two
oval brilliants (0.75 and 1.10 ct; figure 18) and four pieces of
rough (0.62 g). The following properties were recorded on
all samples (except that RI and birefringence were determined on the two faceted stones only): colorvery light
yellow to brownish yellow; RI1.5611.570; birefringence0.009; hydrostatic SG2.692.72; fluorescence
inert to long-wave and weak red to short-wave UV radiation; and no features seen with the desk-model spectroscope. These properties are consistent with those reported
for labradorite (see, e.g., M. ODonoghue, Ed., Gems, 6th
ed., Butterworth-Heinemann, Oxford, UK, 2006, pp.
263267; Winter 2006 GNI, pp. 274275). Microscopic
observation revealed a few small, dark brown to black,
opaque crystals, as well as numerous needles in one plane.
All six samples were chemically analyzed by LA-ICPMS, using the same procedure as for the labradorite from
Mexico reported in the Winter 2006 GNI entry. As expected
from the RI values listed above, all samples had a composition corresponding to labradorite: ~Ab3242Or2-3An5566
referring to the end-members albite (Na-rich), orthoclase (Krich), and anorthite (Ca-rich). All contained traces of iron,
and we know that Fe3+ in the tetrahedral site of plagioclase
produces a pale yellow color (http://minerals.caltech.edu/
color_causes/metal_ion/index.htm). UV-Vis-NIR spectra

showed a 380 nm peak and a very weak absorption at 420
nm. ODonoghue (2006, p. 267) noted that absorptions at
380 and 420 nm in plagioclase are due to Fe 3+.
Labradorite is a calcium-dominant plagioclase (An5070),
which can be separated from sodic plagioclase (albite-oligoclase, An030) by its higher RI values. Gem-quality colorless
to light yellow labradorite is known from various localities
in western North America, such as Oregon, Utah, New
Mexico, and Mexico (e.g., Winter 2006 GNI, pp. 274275).
This is the first time we have encountered such material
from Alaska.
HyeJin Jang-Green
Gem news from Myanmar. From April to November 2008,
this contributor received information on several new gem
occurrences in Myanmar, as described below.
Blue kyanite comes from Mohnyin Township in Kachin
State, associated with garnet, tourmaline, and quartz.
Cabochons (15 ct; see figure 19) and some faceted gems
have been cut from this material, and such stones have
been sold as sapphires in Yangon, Mandalay, and
Taunggyi.
Gem-quality blue sodalite has been found in Ohnbinyehtuat (~22570 N, 96313 E), which is located 6
km southwest of the previously known sodalite/hackmanite deposits in the Mogok area near Pein Pyit.
Granitic pegmatites in the Sakangyi area (15 km west of
Mogok, at ~225400 N, 962030 E) continue to produce
large crystals of topaz (some exceeding 50 kg), as well as
aquamarine, rock crystal quartz, and green fluorite.
Figure 19. Gem-quality blue kyanite (here, from 0.78

0.65 cm to 1.20 0.73 cm) is being produced from
Mohnyin Township in northern Myanmar. Photo by
U Tin Hlaing.
GEMS & GEMOLOGY
SPRING 2009
67
Figure 20. These spinels

(left, 7.91 ct; right, 4.73
ct) proved to be natural
and synthetic, respectively. In the natural
spinel, note the large
feature under the table
that contains planar
inclusions in two
directions. Photos by
G. Choudhary.
Gem-quality yellow scheelite and large brownish purple

zircon have been found ~55 km west of Mogok, near
Thabeikkyin (~225200 N, 955811 E).
Reddish brown garnet crystals (average 4 g) have been
gathered from weathered gneiss in the Mogok region.
The deposit is located about 13 km east of Momeik
(~2306 N, 9648 E). Bright stones up to 15 ct have been
cut from this material.
Ruby mining at Mong Hsu is taking place in underground workings to a depth of 30 m beneath the original
mining site (~214830 N, 972950 E). The present
ruby production is estimated to be about one-tenth of the
amount produced during the boom time in 19931994.
Ruby cabochons from the John Saul mine in Kenya are
popular in Yangon and Taunggyi, where they are marketed as African ruby. Some of them are glass filled.
Most of the synthetic rubies seen recently by this contributor in Taunggyi show only very faint curved lines
and evidence of heat treatment.
There are several active gem markets in Mogok. The
Garden Gem Market is the largest, with 3,0005,000
merchants. Other markets include Le-U, Mogok
Cinema, Mintado, and Kyatpyin Cinema.
Total sales in the gem section of the Union of
Myanmar Economic Holdings Ltd. (UMEHL) Gems &
Jade Sales in January 2008 was ~$467,000.
U Tin Hlaing
Dept. of Geology (retired)
Panglong University, Myanmar
Interesting natural and synthetic spinels. Microscopic

study is an important part of gem identification, and inclusions may add to the beauty of gems. Recently, the Gem
Testing Laboratory of Jaipur, India, encountered two specimens that displayed interesting inclusion scenes. A purplish pink 7.91 ct oval (figure 20, left) was submitted for
identification, while a 4.73 ct pale grayish yellow oval (fig-
Figure 21. At higher magnification, the planar structures in the natural spinel in figure 20 appeared to be
composed of intermittent liquid films. Photomicrograph by G. Choudhary; magnified 35.
Figure 22. Milky zones

formed lath-like structures
in this spinel (left, magnified
35). At higher magnification, the milky zones were
composed of kite-shaped
domains of fine iridescent
films (right, magnified 80).
Note how the individual
blades are all oriented in the
same direction. Photomicrographs by G. Choudhary.
68
GEMS & GEMOLOGY
SPRING 2009
Figure 23. Lath-like inclusions were also concentrated in

areas between the zones in figure 22. These inclusions
were oriented in two directions intersecting at ~90.
Photomicrograph by G. Choudhary; magnified 60.
ure 20, right) was represented as natural sapphire to the

buyer, who requested an origin determination.
The 7.91 ct specimen had an RI of 1.720, gave a hydrostatic SG of 3.58, and exhibited red fluorescence to longand short-wave UV radiation (with a stronger reaction to
long-wave UV). It displayed a weak strain pattern in the
polariscope. Fine lines in the red region and an absorption
band in the yellow-orange region were visible with the
desk-model spectroscope. These features identified the
stone as natural spinel.
The inclusion features in this spinel were notable. A
large, flat feature under the table displayed fine planar
structures running in two directions. This inclusion was
visible with the unaided eye (again, see figure 20, left). At
higher magnification, the planar structures appeared as
intermittent liquid films (figure 21) that gave the impression of partially healed fractures.
Additional features became apparent when the specimen was rotated and observed using a fiber-optic light
source. Numerous parallel zones of fine whitish films or
platelets created a lath-like effect (figure 22, left). These
zones appeared to follow two different directions that
were inclined to one another. At higher magnification,
these lath-like zones formed kite-shaped domains composed of fine iridescent films in parallel orientation (figure 22, right). In addition, there were fine whitish inclusions oriented in two directions that intersected one
another at ~90 (figure 23). Similar-appearing inclusions
in spinel (identified as hgbomite) were illustrated by E.
J. Gbelin and J. I. Koivula (Photoatlas of Inclusions in
Gemstones, Vol. 2, Opinio Publishers, Basel,
Switzerland, 2005, pp. 693, 714).
The color of the pale grayish yellow specimen (again,
see figure 20, right) was similar to that observed in many
natural sapphires from Sri Lanka, which seemed to support the sellers claim. However, basic gemological testing
revealed its true identity. It had a refractive index of 1.735,
a hydrostatic SG of 3.61, and strong chalky blue fluorescence to short-wave UV, but was inert to long-wave UV.
In the polariscope, it exhibited a strong strain pattern
(ADR effect). The desk-model spectroscope displayed faint
bands in the green, yellow, and orange-red regions, in a
pattern that corresponded to cobalt. These properties indicated a synthetic spinel.
Circular to subhexagonal zones or bands composed of
clouds of fine dotted inclusions were visible with magnification (figure 24, left). The subhexagonal features were
very similar to inclusion patterns in natural corundum. A
profile view showed that these zones were composed of
parallel planes in a layered pattern (figure 24, right). When
the sample was viewed with diffused illumination in
immersion, subtle color zoning was observed. The center
appeared pale blue, and the outer regions were pale yellow
(figure 25).
Other features visible in the synthetic spinel included irregular thread-like inclusions and large spherical
gas bubbles. The subhexagonal inclusion features could
have caused this synthetic spinel to be mistaken for a
natural stone. Although hexagonal patterns have been
reported in other materials, such as synthetic star sapphire (see Summer 2007 GNI, pp. 177178), it is quite
unusual to see them in a gem belonging to the cubic
system.
Gagan Choudhary (gtl@gjepcindia.com)
Gem Testing Laboratory, Jaipur, India
Figure 24. Clouds of fine

dotted inclusions observed in
the synthetic spinel formed
subhexagonal zones or bands
(left) that also appeared
circular in some viewing
directions. In profile view
(right), these zones were
arranged in parallel
planes. Photomicrographs
by G. Choudhary;
magnified 65.
GEMS & GEMOLOGY
SPRING 2009
69
Figure 25. Viewed in immersion, the synthetic spinel

displayed subtle color zones, with a blue central region
and yellow outer portions. Photo by G. Choudhary.
Zoisite from Afghanistan. In mid-2008, gem dealer Mark

Kaufman (Kaufman Enterprises, San Diego) loaned GIA
some light purple samples that were sold to him as zoisite
from the Shinwari tribal area, Nangarhar Province,
Afghanistan. These consisted of four pieces of rough weigh-
Figure 26. These two zoisites, a 1.1 g crystal and a

1.68 ct modified cushion cut, are reportedly from
Afghanistan. Photo by Robert Weldon.
ing up to 1.1 g and a 1.68 ct modified brilliant-cut cushion

that he had faceted from this material (e.g., figure 26). He
first encountered this zoisite in early 2001 at a gem show in
France. From 1.2 kg of mixed-quality rough, Mr. Kaufman
selected 100 g of material. Several years later (in 2006) he
purchased another 850 g (of which ~300 g was cuttable)
from the same parcel. So far, he has faceted four stones
weighing up to 3.71 ct. Although he has pieces of rough
that should cut larger stones, he reported that the material
tends to crack during faceting due to internal stress.
The following properties were obtained from an examination of all five samples: colorvery light brownish purple, with moderate light yellow, grayish blue, and grayish
purple pleochroism; hydrostatic SG3.35 0.02; RI (measured on the faceted sample only) = 1.694, = 1.695,
and = 1.702; fluorescenceinert to long- and short-wave
UV radiation; Chelsea filter reactionnone; and absorption lines at 427 and 452 nm visible with a desk-model
spectroscope. These properties are consistent with those
reported for zoisite (e.g., J. E. Arem, Color Encyclopedia of
Gemstones, 2nd ed., Van Nostrand Reinhold Co., New
York, 1987). The material was confirmed as zoisite using
Raman spectroscopy, which can distinguish between
zoisite and closely related clinozoisite. EDXRF analysis
showed that the nonintrinsic trace elements were dominated by Fe, with minor amounts of V, Cr, Zn, and Sr.
One of the rough pieces consisted of a well-formed crystal (again, see figure 26) that had a flat tabular prismatic
morphology with pronounced striations along its length.
One cleavage direction was obvious in the other rough
pieces. All the samples were characterized by abundant, randomly oriented, prismatic inclusions (e.g., figure 27), which
were identified as actinolite with Raman spectroscopy.
Donna Beaton and Ren Lu
Figure 27. Prismatic inclusions of actinolite were

present in the zoisite from Afghanistan. Photomicrograph by D. Beaton; field of view 1.0 mm.
70
GEMS & GEMOLOGY
SPRING 2009
Figure 28. These freeform cabochons of Australian

chrysoprase weigh 36.85 and 41.67 ct. Photo by
Robert Weldon.
Figure 29. Interesting dendritic patterns are rarely seen

in fashioned chrysoprase. In these samples, they were
formed by inclusions of manganese oxides, which here
appear red-brown in reflected light. Photomicrograph
by R. Befi; image width 5.9 mm.
INCLUSIONS IN GEMS
Australian chrysoprase with dendritic inclusions. Two
freeform cabochons (36.85 and 41.67 ct; figure 28) of
Australian chrysoprase with unusual inclusions were loaned
to GIA by Steve Perry (Steve Perry Gems, Davis, California).
The rough material from which they were cut originally had
been designated for use as a carving material, but was discarded because it contained dark-colored blemishes.
Both samples displayed a yellowish green color typical
of chrysoprase, as well as a few near-colorless areas.
Standard gemological testing confirmed the identification:
spot RI1.54; hydrostatic SG2.57; fluorescenceyellowish green, moderate to short-wave and weaker to longwave UV radiation; Chelsea color filterno reaction. No
mosaic-like veining, as seen in dyed chalcedony, was
observed. EDXRF chemical analysis detected Si and traces
of Ni, but no Cr, which is consistent with the chemical
formula of chrysoprase.
Chrysoprase can range from pale green through apple
green to a deep rich green; the intensity of the color is
directly related to Ni content (J. H. Brooks, Marlborough
Creek chrysoprase deposits, Fall 1965 G&G, pp. 323330).
LA-ICP-MS analysis of both cabochons showed Ni levels
of 0.020.18 wt.% in near-colorless areas, and 0.530.90
wt.% in green areas.
Viewed with magnification, the inclusions in both
stones showed interesting dendritic patterns (figure 29).
Dendrites form along surface-reaching cracks and along
flow layers in chalcedony and other minerals as a result
of epigenetic fluids. These solutions typically deposit
manganese oxides in characteristic branching shapes (in
this case, with a more three-dimensional form than is
typically seen in dendrites). The visual appearance of the
inclusions was highly suggestive of manganese oxides,
and this was supported by Raman spectroscopy of some
dendrites that reached the surface. In addition, LA-ICPMS analyses of the surface-reaching inclusions showed
enriched Mn.
Most fashioned chrysoprase has uniform color without any dendritic inclusions. The presence of such a delicate dendritic pattern makes these cabochons rather special, by transforming a plain interior into exotic and
vibrant material.
Riccardo Befi (riccardo.befi@gia.edu)
Ankangite and celsian inclusions in quartz from Brazil. In
January 2008, gem dealer Sergio Pereira de Almeida purchased (in Tefilo Otoni, Minas Gerais) a 20 kg parcel of
colorless quartz crystals that contained radiating black
needles and lesser quantities of euhedral white and colorless inclusions. The quartz crystals, which ranged from
one to 10 cm long, were prismatic and most were terminated. The entire parcel was cut into cabochons and
faceted gems, yielding approximately 20,000 carats total.
Some of the stones (e.g., figure 30) were donated to the
mineralogy museum at the University of Rome La
Sapienza by Mr. Pereira de Almeida and examined for
this report.
Four samples were characterized using standard gemological techniques, scanning electron microscopy with
energy-dispersive spectroscopy (SEM-EDS), and electronmicroprobe analysis. The gemological properties were
consistent with quartz, except that the SG value of 2.67
was slightly high. The conspicuous black needles in these
stones were 0.11 cm long and 510 m in diameter (as
indicated by SEM). Semiquantitative SEM-EDS analyses
of these needles revealed the presence of Ti, Cr, and V.
The white-to-colorless euhedral mineral inclusions varied
from 10 m to 1 mm in exceptional cases.
Using a Cameca SX-50 electron microprobe at the
Italian National Research Councils Institute of Environmental Geology and Geoengineering (IGAG-CNR) in
Rome, we identified the black needles as ankangite,
GEMS & GEMOLOGY
SPRING 2009
71
quartz, and as such it further enriches the catalog of

known quartz inclusions.
Michele Macr (michele@minerali.it)
and Adriana Maras
Department of Earth Sciences
University of Rome La Sapienza, Italy
Marcello Serracino and Pier Francesco Moretti
IGAG-CNR, Rome
Figure 30. This 27.01 ct quartz cabochon contains

black, needle-shaped sprays of ankangite and colorless-to-white crystals of celsian. Photo by M. Pant.
Ba(Ti,V3+,Cr3+)8O16. This extremely rare oxide mineral is

named after the discovery locality, Ankang County in
Shaanxi Province, China. The white-to-colorless inclusions
were celsian (BaAl2Si2O8), a Ba-rich mineral of the feldspar
group. Chemical analyses of both inclusion types are reported in table 1. (No Ba was detected by the SEM-EDS analyses
described above because the instrument had not been calibrated for this element.)
The presence of both ankangite and celsian inclusions
in this quartz suggests that it originated from a barium-rich
deposit. The quartzs discoverer, a Mr. Nilsinho from
Curvelo, Minas Gerais, is keeping the location within
Brazil confidential.
To the best of our knowledge, this is the first reported
occurrence of the minerals ankangite and celsian in
TABLE 1. Electron-microprobe analyses of inclusions in

two quartz cabochons from Brazil.a
Paraba quartz with copper inclusions from Brazil.

Medusa quartz from the Brazilian state of Paraba is
known to contain two types of inclusions (Fall 2005 GNI,
pp. 271272). The blue or green jellyfish-shaped inclusions near the colorless core of the quartz crystals were
ascribed to gilalite, a rare Cu-silicate. In addition, tiny
blue-to-green acicular crystals of an as-yet-unidentified
Cu-silicate mineral were documented near the surface of
the quartz. Recently, this author bought several polished
pieces of this type of quartz in Brazil (where it is sold as
Paraba quartz) that contained a third type of inclusion.
Overall, this quartz appeared mottled red-brown (e.g.,
figure 31), but when viewed in profile with magnification,
the material was seen to consist of a bottom layer of colorless-to-purplish quartz in sharp contact with an upper
(genetically younger) layer containing abundant fine fibers
that formed thin bundles up to about 1 mm long. The fibers
typically had a copper color that was easily seen in reflected
light (figure 32). A simple electrical conductivity test performed on some exposed fibers proved their identity as copper. Their shape is very unusual for native copper, and they
almost certainly formed as pseudomorphs (replacements) of
the original fibrous Cu-silicate. Additional evidence for this
origin is provided by blue remnants present among the copper fibers. The conversion of a Cu-silicate to native copper
is indicative of a strongly reducing environment.
Jaroslav Hyr`s l (hyrsl@kuryr.cz)
Prague, Czech Republic
Ankangiteb
Celsian
SYNTHETICS AND SIMULANTS
SiO2
TiO2
Al2O3
V2O3
Cr2O3
CaO
FeO
BaO
Na2O
K2O
0.39
56.84
0.17
14.27
6.02
0.02
0.05
20.76
0.07
0.03
41.84
bdl
23.84
bdl
bdl
bdl
0.01
29.19
0.23
4.58
Total
98.62
99.69
Artificial glass showing color change when exposed to

light. The SSEF Swiss Gemmological Institute recently
analyzed a faceted oval that was sold in India as an
extraterrestrial gemstone. The 9.57 ct specimen appeared
light yellow when it was removed from the stone paper and
viewed with incandescent light (figure 33, left). However,
when exposed to a strong fiber-optic halogen lamp for a few
seconds, it turned dark bluish gray (figure 33, center); daylight had a similar, if weaker, effect. Exposure to long-wave
UV resulted in a less-pronounced color change; under
short-wave UV, only a slight color shift was observed.
When the sample was heated in water to about 80C, the
color change quickly reversed, but with an intermediate
step in which the stone turned brown (figure 33, right). The
same reversible color change sequence was also observed
without heating when the sample was kept in the dark for
Oxides (wt.%)
aThe
values given represent the average composition of two inclu

sions of each mineral, analyzed in two different samples. Mg and
Mn were not detected. Abbreviation: bdl = below detection limit.
b
The chemical composition of the ankangites was recalculated analytically
because V interfered with both Cr and Ti in the electron microprobe.
72
GEMS & GEMOLOGY
SPRING 2009
Figure 31. This 11.68 ct quartz cabochon from Brazil represents a new type of Paraba quartz. Photo by J. Hyr`sl.
several hours. In optics, this reversible color change on

exposure to light is called a photochromic effect, with the
reversal rate being temperature dependent (i.e., increasing
the temperature speeds up the reversal).
Standard gemological testing established the following
properties: RI1.524; hydrostatic SG2.395; polariscope
reactionisotropic; UV fluorescenceslight green to longwave and chalky yellow to short-wave. No radioactivity
was detected with a Geiger counter, and microscopic
examination revealed no inclusions.
EDXRF spectroscopy showed major amounts of Si and
Al, along with K, Ca, and Ti as minor elements, and traces of
Pb, Zr, Ag, and Br. Based on the gemological properties,
chemical data, and the lack of inclusions, we identified the
material as a photochromic artificial glass. The Raman spectrum showed no characteristic peaks, highlighting the specimens amorphous state. Interestingly, the intense green laser
beam (514 nm) of the Raman unit produced a reversible dark
gray spot in the stone (again, see figure 33, right), similar to
the color produced by exposure to the fiber-optic lamp.
Figure 32. The new Paraba quartz contains fibrous

inclusions of native copper (here, 1 mm long).
Photomicrograph by J. Hyr`s l.
Using SSEFs portable UV-Vis spectrometer (see

www.ssef.ch/en/news/pdf/UV-Vis.pdf), we examined the
characteristic absorption features for each color state: yellow, bluish gray, and brown (figure 34). With this instrument, absorption and luminescence can be measured
simultaneously in a few seconds over the whole spectral
range, which is especially important for photosensitive
materials. Common sequential spectrophotometers would
only show a mixture of the absorption spectra from the
three color states.
The spectrum of the light yellow state showed only a
very slight increase in absorption toward the blue region.
The spectrum of the dark bluish gray state revealed a
broad absorption at ~600 nm and two smaller bands at
~500 and 650 nm, respectively; a transmission window at
~450 nm was responsible for the bluish gray color. Finally,
the brownish state spectrum showed no particular absorption at 600 nm but a general increase in absorption toward
the shorter wavelengths, resulting in the brown color.
Figure 33. This 9.57 ct photochromic artificial glass appeared light yellow before (left) and dark bluish gray after
(center) exposure to a fiber-optic lamp. This reversible color change had an intermediate brown stage (right). The
dark spot in the right photo is due to the Raman laser. Photo by Luc Phan, SSEF.
GEMS & GEMOLOGY
SPRING 2009
73
Such an effect is well known for industrial photochromic

glass (e.g., for sunglasses) that is doped with halogenides such
as silver bromide (AgBr). Exposure to lightparticularly in
wavelengths ranging from blue to long-wave UVtransfers
an electron from the Br ion to the Ag+ ion, which becomes
light-absorbing metallic silver. When shielded from light, the
glass slowly returns to its original state.
Chiara Parenzan and Michael S. Krzemnicki
(gemlab@ssef.ch)
SSEF Swiss Gemmological Institute
Basel, Switzerland
Some unusual dyed imitations. Researchers at the University of Nantes recently examined two types of imitations fashioned from dyed materials that are rarely used as
simulants. The first type consisted of two faceted ovals
(5.52 and 7.12 ct; e.g., figure 35, left) sold as ruby in Jaipur,
India. With magnification, it appeared that the red color
was concentrated in fractures; the stones were otherwise
virtually colorless (e.g., figure 35, right). The samples were
singly refractive (RI = 1.738) and had a hydrostatic SG of
3.62. These values are consistent with grossular.
Fourier-transform Raman spectra obtained with a
Bruker RFS 100 spectrometer showed peaks at 877, 828,
550, 419, and 374 cm1, also consistent with grossular. The
chemical composition was measured with a JEOL 5800
SEM equipped with a high-resolution Princeton Gamma
Tech IMIX-PTS germanium energy-dispersive detector
operating with an accelerating voltage of 20 kV, a current of
1 nA, and a 37 take-off angle. We measured a composition
(at.%) of ~14.8% Si, 15.0% Ca, 9.5% Al, 0.1% Mn, 0.7%
Fe, and 59.7% O; analyses of both samples showed only
slight variations of <0.2%. This indicated nearly pure endmember grossular, a material seen only rarely.
The second type was a red pierced disk, or pi (50 mm in
diameter, 11 mm thick; figure 36), that was sold as red
chalcedony. It was purchased intact and then accidentally
broken. As the interior of the sample was largely colorless
(figure 36, right), the owner correctly deduced that the
stone had been dyed and she submitted it to our lab for
confirmation. Again, the red color appeared to be concentrated in fractures. The sample was singly refractive (RI =
Figure 34. These UV-Vis absorption spectra illustrate the

differences in the light yellow, bluish gray, and brown
color states of the photochromic artificial glass sample.
1.559) with a hydrostatic SG of 2.60, values consistent with

serpentine. The Raman spectra showed peaks at 1048, 685,
644, 532, 461, 378, and 231 cm1, and the chemical composition (at.%) was 16.96% Si, 24.05% Mg, 0.42% Al, and
58.58% O. Both sets of data also indicated serpentine.
Powder X-ray diffraction analysis, with an INEL CPS
120 X-ray diffractometer, gave a pattern for antigorite. In
particular, the peak at 59.06 2 is unique to antigorite
among the serpentine group minerals.
Both the garnets and the serpentine fluoresced vivid
orange to both long- and short-wave UV radiation. Orange
luminescence is uncommon for serpentine as well as for
garnet species that typically occur in reddish hues. Cleaning
the samples with acetone removed some of the red dye.
It is rare for garnet to be dyed to imitate ruby, or for serpentine to be dyed to imitate chalcedony. The use of unusual gem materials as starting materialsnearly pure grossular and antigoritemade these samples especially notable.
Benjamin Rondeau
Emmanuel Fritsch
Blanca Mocquet and Yves Lulzac
Centre de Recherches Gemmologiques, Nantes
Figure 35. The faceted

oval on the left (7.12 ct)
was sold as ruby, but
proved to be dyed
grossular. Further
examination revealed
concentrations of red
color in fractures
(right). Photos by
B. Rondeau (left) and
B. Mocquet (right).
74
GEMS & GEMOLOGY
SPRING 2009
Figure 36. This pierced

disc (left, 50 11 mm),
sold as red chalcedony,
is actually dyed antigorite. The sample was
accidentally broken,
revealing concentrations
of red color in fractures
(right). Photos by
B. Mocquet.
CONFERENCE REPORTS
International Kimberlite Conference. The ninth IKC was
held August 1015, 2008, at Johann Wolfgang Goethe
University in Frankfurt, Germany. This typically quadrennial meeting, the most important scientific conference on diamond geology, brought together nearly 500
geologists and other researchers to share the latest information on the conditions of diamond formation and current efforts to locate new diamond deposits. Several of
the oral and poster presentations were of gemological
interest. Abstracts of all presentations are available at
www.9ikc.uni-frankfurt.de/scientific_program.html.
George Read (Shore Gold Inc., Saskatoon, Canada)
opened the conference with a review of the current state of
diamond exploration, mining, and marketing, while noting
developments in the understanding of diamond geology
since the 2003 IKC. Dr. David Phillips (University of
Melbourne, Australia) discussed evidence that suggests the
rich alluvial diamond deposits along the west coast of
southern Africa were produced not only by the Orange
River drainage system, but also by paleo-drainage in the
general area of the present-day Karoo River, farther to the
south. Dr. Tania Marshall (Exploration Unlimited,
Johannesburg) presented a genetic model for the formation
of the Ventersdorp alluvial deposit in South Africa, which
has produced an estimated 14 million carats of rough diamonds over the past century. Debbie Bowen (Let`s eng
Diamonds Ltd., Maseru, Lesotho) described the characteristics of diamonds from the two Let`s eng-la-Terae kimberlites, where some 75% of the mine production is gem-quality material and approximately 19% of the diamonds from
the main kimberlite pipe are type IIa. Since 2003, the mine
has produced 25 crystals larger than 100 ct; a 215 ct colorless type IIa diamond was found there in January 2007 and
a 478 ct piece of rough was found in September 2008.
Dr. Sonal Rege (Macquarie University, Sydney,
Australia) presented chemical data for more than 40 elements from the LA-ICP-MS analysis of approximately 500
diamonds. These trace elements are present in microscopic
and submicroscopic inclusions that are believed to represent the fluid from which the diamonds crystallized; there
was too much overlap for the data to be useful for geographic origin determination. From a study of diamonds
from Yakutia, I. Bogush (ALROSA, Mirny, Russia) present-
ed an analysis of IR spectroscopic data suggesting that diamonds from particular geographic sources might be distinguished by their unique spectral signatures and the relative
proportions of various optical defects.
Dr. David Fisher (DTC Research Centre, Maidenhead,
United Kingdom) reviewed the current understanding of the
effect of HPHT treatment on brown diamonds. The brown
color is thought to be due to absorptions related to vacancy
clusters in the diamond lattice. According to this theory,
HPHT annealing breaks up these clusters to release individual vacancies, removing the cause of the brown color orif
the vacancies interact with nitrogen impuritiescreating
other colors. Dr. Victor Vins (New Diamonds of Siberia,
Novosibirsk, Russia) reported on the HPHT color alteration
of type IaA and IaB brown diamonds that exhibited evidence
of strong plastic deformation. Heating above 1800C at 7
GPa for 10 minutes transformed their color to yellow-green,
with accompanying changes in spectral features. Dr.
Lioudmila Tretiakova (GCAL, New York) reviewed the use
of IR and photoluminescence spectroscopy to characterize
the distinctive optical defects of both natural brown and
HPHT-treated type IaAB and IIa diamonds.
The 10th IKC will take place in 2012 in Bangalore, India.
James E. Shigley
GIA Research, Carlsbad
Wuyi Wang
ERRATA
In figure 34 (p. 339) of the Winter 2008 article by D. Schwarz
et al. titled Rubies and Sapphires from Winza, Central
Tanzania, the polarized UV-Vis-NIR spectra should have
been labeled to indicate that the red line is for E c and the
black line is for E c. In addition, the colors of the red and
black lines in spectra A should have been reversed. See the
G&G Data Depository (www.gia.edu/gemsandgemology)
for the corrected version.
The Winter 2008 article on the Wittelsbach Blue (pp.
348363) contained two minor errors: Laurence Graffs
name was misspelled, and the final auction price including
the buyers premium for the diamond was incorrectly
described as the hammer price.
Gems & Gemology regrets the errors.
GEMS & GEMOLOGY
SPRING 2009
75
4. Coloring agents in emerald

include chromium and
A. beryllium.
B. vanadium.
C. titanium.
D. nitrogen.
5. While ________ is typically the
chromophore in natural aquamarine, the color of Maxixe and
Maxixe-type beryl is due to a
radiation-induced color center.
A. cobalt
B. titanium
C. iron
D. copper
The following 25
questions are based on
information from the Spring,
Summer, Fall, and Winter 2008
issues of GEMS & GEMOLOGY.
Refer to the feature articles,
Notes and New Techniques,
and Rapid Communications
in those issues to find the single
best answer for each question.
6. The Koh-i-Noor is a ________

diamond.
A. D-color, type IIa
B. Fancy Deep blue, type IIb
C. Fancy Light pink, type IIa
D. G-color, type Ia
Mark your choice on the response

card provided in this issue or visit www.gia.edu/gandg to take the Challenge
online. Entries must be received no later than Monday, August 3, 2009.
All entries will be acknowledged with an email, so please remember to
include your name and email address (and write clearly!).
Score 75% or better, and you will receive a GIA CONTINUING EDUCATION
CERTIFICATE (PDF format). If you are a member of the GIA Alumni
Association, you will earn 10 Carat Points. (Be sure to include your
GIA Alumni membership number on your answer card and submit your
Carat card for credit.) Earn a perfect score, and your name also will be listed
in the Fall 2009 issue of GEMS & GEMOLOGY. Good luck!
1. Distinct ________ color zoning is

common among Winza rubies
and pink sapphires.
A. bluish violet
B. brown
C. black
D. colorless
2.
76
When vivid blue-to-green tourmaline was discovered in

Mozambique in 2001, the country joined Brazil and ________ as
the only known sources of Cubearing Paraba-type material.
GEMS & GEMOLOGY CHALLENGE
A.
B.
C.
D.
Madagascar
Nigeria
Vietnam
Zambia
3. The 35.56 ct Wittelsbach Blue,

like all other diamonds that
arrived in Europe prior to 1725,
is widely believed to have originated in
A. Africa.
B. Brazil.
C. India.
D. Russia.
7. For orange-to-yellow johachidolite samples, the Be concentration ________________ as the color

desaturates, while rare-earthelement concentrations remain
relatively constant and typically
________ than for green samples.
A. increases/higher
B. increases/lower
C. decreases/higher
D. decreases/lower
8. When color grading D-to-Z diamonds, GIA methodology places
the light source above the tray
at approximately _______ degrees,
and the diamond is observed at
an angle of about _______ degrees
from the stone.
A. 0/45
B. 0/90
C. 45/45
D. 45/90
9. The most significant recent treatment of diamonds involves
________ to either lighten off-color
stones or create various fancy colors.
A. annealing at high pressure
and high temperature
B. filling with a lead-based glass
C. applying synthetic diamond
thin films
D. irradiation
GEMS & GEMOLOGY
SPRING 2009
10. Which of the following past claims

about the Koh-i-Noor was shown
to be false by computer analysis of
a recently rediscovered cast of the
original 186.1 old-carat stone?
A. The stone was intentionally
cleaved to create damage
prior to its surrender to the
British or some other forced
change in ownership.
B. The stone was cut from the
Great Mogul.
C. One of the larger faces
could have been an unpolished cleavage plane.
D. Both A and B
11. Which of the following indications
of treatment were not seen with
recent Co-bearing coatings on
tanzanite?
A. Tiny pink-to-orange flashes
of color related to minute
areas of damage
B. Weak surface iridescence
C. Abraded facet junctions
visible with a microscope
and immersion
D. A difference in luster
between the coated pavilion
and the uncoated crown
12. Of the several hundred kilograms
of tourmaline rough mined to
date in the Mavuco area of
Mozambique, about ________%
show blue-to-green Paraba
colors without heat treatment.
A. 5
B. 10
C. 15
D. 25
15. Purplish pink to purple diamonds

from Siberia are known for
exhibiting unusually strong
A. strain patterns.
B. twinning.
C. UV fluorescence.
D. dichroism.
16. Which of the following has not
been observed in the multiphase
inclusion assemblages in emeralds from Byrud, Norway?
A. halite
B. calcite
C. zircon
D. pyrrhotite
17. XPS concentration profiles of surface-colored topaz indicated that
the color of the treated pink
topaz can be attributed to
A. a diffusion process involving
Mn.
B. a coating of gold-bearing SiO2.
C. a diffusion process involving
Fe and/or Cr.
D. a coating of Co.
18. One important way to distinguish
natural, untreated diamonds
from synthetic diamonds is to
look for ________ in their UV-Vis
spectrum.
A. the 415 nm absorption band
due to the N3 center
B. differences in the relative
size of absorption bands
due to the N3 and N2 centers
C. a broad absorption band at
550 nm
D. absorption due to the H3
center
13. The most distinctive feature of

rubies and sapphires from the
Winza deposit is their ________ ,
which can be considered localityspecific for some stones.
A. chemical composition
B. inclusion scene
C. morphology
D. refractive index
19. Like the Hope, the Wittelsbach

Blue is a Fancy Deep grayish blue
diamond that displays unusual
________ phosphorescence.
A. blue
B. pink
C. red
D. yellow
14. An SG value of ________ is strong

evidence that a fire opal is ______ .
A. greater than 1.77/synthetic
B. less than 1.77/synthetic
C. less than 1.77/crazed
D. greater than 1.77/phenomenal
20. Chocolate Pearls showed the

most damage from exposure to
A. acetone.
B. chlorine bleach.
C. sunlight.
D. hairspray.
GEMS & GEMOLOGY CHALLENGE
21. HPHT-grown yellow synthetic

diamonds often display color zoning ________ because impurities are
incorporated in different concentra
tions along internal growth sectors.
A. that has a mottled, patchy
appearance
B. that is pink or brown and
oriented parallel to octahedral faces
C. that has a curved, banded
appearance
D. that has a combination
cross and square pattern
22. The gota de aceite effect in
Colombian emeralds was once
attributed to a microscopic
dusting of calcite grains in the
interior of the stone, but it is
actually due to
A. parallel hollow tube-like
structures.
B. partially healed fracture planes.
C. transparent angular or
hexagonal growth structures.
D. three-phase inclusions.
23. One drawback of filling cracks in
diamond with glass is that the
process may
A. reduce carat weight.
B. reduce clarity.
C. result in a lower color grade.
D. result in lower dispersion.
24. While natural aquamarine and
blue beryl may be identified by a
wide range of multiphase fluid or
mineral inclusions, with magnification hydrothermal synthetic
blue beryl is distinguishable by its
A. strongly undulating, stepand chevron-like growth
structures.
B. seed plate residues.
C. flux inclusions.
D. Both A and B
25. While yellowish diamonds in
GIAs D-to-Z color grading system
receive a letter grade only, brown
diamonds below _______ are noted
with the letter grade and a word
description.
A. J
B. K
C. L
D. M
GEMS & GEMOLOGY
SPRING 2009
77
BOOK REVIEWS/GEMOLOGICAL ABSTRACTS

The Gems & Gemology Book Reviews and Gemological Abstracts sections are available only in electronic (PDF) format. These sections are available free of charge both on the G&G web site (www.gia.edu/gemsandgemology) and as part of
G&G Online (gia.metapress.com), and are paginated separately from the rest of the issue.
These sections are also included in this full-issue PDF. Accordingly, the Table of Contents included in this file lists
these additional sections, and thus differs from the Table of Contents in the print version. For these reasons, this PDF is
not the official version of this issuethe journal of record for this issue is the combined print/online version that was
released to subscribers. This file is created for archival purposes in order to maintain continuity with previous issues.
Si
BOOK REVIEWS/GEMOLOGICAL ABSTRACTS
GEMS & GEMOLOGY
SPRING 2009
Book
0 Reviews
9
Editors
Susan B. Johnson
Jana E. Miyahira-Smith
Thomas W. Overton
Cartier
By Hans Nadelhoffer, 352 pp.,
illus., publ. by Chronicle Books
[www.chroniclebooks.com],
San Francisco, 2007. US$75.00
While many books have been written
on the House of Cartier, this is by far
the most extensive review to date.
Author Hans Nadelhoffer, Christies
longtime jewelry expert in Geneva,
spent three years compiling it after
being granted full access to the
Cartier archives in Paris, New York,
and London. The result is a beautiful
hardcover volume that includes fullcolor photographs, sketches, and
archival images, along with narratives
and a historical review of the famed
house, while it covers the major periods of influence and design.
Nadelhoffer takes the time to
detail periods of style and how they
were influenced by world events.
Alfred Cartier and his three sons
Louis (head of Cartier in Paris),
Jacques (London), and Pierre (New
York)handled the most influential
clients, including royalty, diplomats,
movie stars, and captains of industry.
Fashion, war, and economic fluctuations influenced the decisions in hiring designers. Readers will understand the close links with arts, crafts,
theatre, literature, and film.
One of my favorite sections, on the
evolution of period and style, covers
Indian, Chinese, and Japanese motifs
and how they found their place in
Cartiers work. These new variations
used circles and ellipses to conceal
purely constructional elements in the
jewelry and create fluent ornamental
forms. Another chapter describes how
BOOK REVIEWS
the designers for the Ballets Russes of

the early 20th century integrated floral
elements from 18th-century France
into their theater designs Also interesting is Nadelhoffers account of the evolution from tiaras in the late 19th century to headbands in the 1920s and
garland-style breast ornaments.
Nadelhoffer has a clever way of
explaining how these shifts in style
took place.
For jewelry designers, this book
provides an important look at how
events of the past 150 years influenced design in Europe and America.
Many of these styles are popular
today, and Nadelhoffer helps the reader understand their foundation.
Sections on the designers explain the
distinct differences between Cartier
and peers such as Boucheron, Faberg,
and Vever, as well as their influences
on each other.
Over 500 beautiful images, lavish
full-page photographs, drawings, and
original sketches of jewelry, clocks,
objets dart, and accessories fill the
pages of this book. Most of these
pieces are housed in collections and
museums; for the true collector, however, these images are important and
accurate.
Hans Nadelhoffer brings to life, in
a charming and compelling way, the
history of the legend that is Cartier.
This book will satisfy the appetite of
even the toughest critics. It is a must
for anyone interested or involved in
estate and period jewelry or jewelry
design.
MELINDA ADDUCCI
Joseph DuMouchelle Auctioneers
Grosse Pointe Farms, Michigan
7000 Years of Jewelry

By Hugh Tait, Ed., 256 pp., illus.,
publ. by Firefly Books, Buffalo, NY,
2008. US$29.95
This is an important book for all jewelersone that many would see as a
concise history of the evolution of
jewelry. It features some 500 pieces of
jewelry from the collections of the
British Museum, founded in 1753. It
is the successor to the original catalogue for the 1976 British Museum
exhibition Jewellery Through 7000
Years, which was primarily illustrated in black and white. In this revised
version, the pieces have been superbly
photographed and printed in color. (In
fact, many illustrations look better
than the actual objects despite the
fact that the museum cases are well
lit.) The book designers have created
an attractive layout, and the figure
captions are informative.
There is a very useful nine-page
introduction with some historical
portraits of bejeweled notables and
contemporary paintings of goldsmiths workshops. The main body of
the text is divided into 18 chapters
arranged by regions and in chronological order. It begins with the Middle
East, 50002000 BC , and covers
almost every region of the world
through various eras, concluding with
Europe, 17001950. Additional chapters focus on amulets, cameos, and
rings.
These are followed by a list of
British Museum accession numbers
for the pieces and a bibliography. Most
of the objects were included in the
1976 catalogue, where fuller descriptions and bibliographies may be found.
GEMS & GEMOLOGY
SPRING 2009
S1
Some revisions have been made in

this version to account for new
research discoveries, and the section
on Europe from 1700 to 1950 has
been completely rewritten with new
illustrations that include the museums recent acquisitions. There is a
useful glossary of technical terms, a
brief general bibliography on historical jewelry, and a list of titles for further reading.
Most of the jewelry displays in the
museum are shown with more utilitarian objects found during archeological excavations in the respective
areas. The museums regional jewelry
displays are often in adjoining rooms,
but others are on different floors, so if
you are fortunate enough to visit the
British Museum you should obtain a
map or guide before you start: You
will save time and energy.
The book weighs nearly three
pounds (1.36 kg), which is substantial
but still light enough to carry on a
visit to the museum, and is an incredibly good value. Jewelers and gemologists will find it an interesting read
and an attractive reference volume for
their shelves.
ALAN JOBBINS
Caterham, United Kingdom
Pegmatites
By David London, 368 pp, illus.,
publ. by the Mineralogical Association of Canada [www.mineralogicalassociation.ca], 2008. US$125.00
This is a clearly written, informative,
and richly illustrated book about this
fascinating rock typewhich is an
important source of gem minerals. For
more than a century, the origin of pegmatites has been fertile ground for
research, discussion, and debate; and
like the research, that debate is far
from over. Pegmatites is a strong
defense of the model proposed by the
author and his colleagues. Indeed, this
is a valuable and well-presented volume. It should be noted, though, that
to date there is no mainstream model
(as illustrated by the amount of materi-
S2
BOOK REVIEWS
al published on this topic), and the present reviewer and colleagues have proposed and supported a contrary model
of pegmatite formation, one of several
currently being pursued by researchers.
None of these models has yet been
universally accepted as definitive.
Dr. Londons approach is based on
his research on granites; he uses the
experimental model of Jahns and
Burnham (1969) as his starting point.
However, this model must be put
into context, since it is capable of
modifications that invalidate many of
Dr. Londons conclusions. The situation with melt and fluid inclusions is
a case in point, as a very superficial
treatment of these features can easily
result in skepticisms, and one feels
this attitude is a problem for the
entire book.
Chapter 17, Internal Evolution of
Pegmatites, is the highlight of Dr.
Londons pegmatite narrative. This
reviewer suggests that the error of
assuming low H2O concentration or
high density, made throughout the
book, culminates in this chapter. The
author dismisses the composition of
individual melt and fluid inclusions
as unrepresentative of the bulk melt
composition in pegmatites, but fails
to note that individual melt and fluid
inclusions are simply snapshots of an
evolving process.
In the epilogue, the author (to his
credit) briefly mentions some doubts
that have been expressed regarding his
own model. Much more research will
be required in our attempts to understand the many mysteries posed by
pegmatites, and we should avoid
attaching ourselves to any single
model, lest in the process we ignore
relevant facts.
It is a testament to the extent to
which pegmatites have been studied
that approximately 800 references are
listed, but many important papers are
still missing. The lack of an index is
also a serious omission. Without it,
many important facts distributed over
the whole volume are difficult to
retrieve.
The book does contain a CDROM with all the illustrations in the
volume as well as the PDF files of the

granitic pegmatite chapters in the
MAC Short Course Handbook on
Granitic Pegmatites, edited by Petr
` erny` and published in 1982.
C
With respect to the form, mineralogy, characteristics, and distribution
of pegmatites, this book is an
admirable discussion of a most fascinating subject, and as such it is recommended to the G&G reader. But
with regard to the origin of pegmatites, a large and important component of the book, Pegmatites should
be seen as the presentation of one particular model of pegmatite genesis,
one that continues to be questioned
by other researchers.
RAINER THOMAS
German Research Centre for
Geosciences GFZ
Potsdam, Germany
Brooches: Timeless Adornment

By Lori Ettlinger Gross, 192 pp.,
illus., publ. by Rizzoli International
Publications, New York, 2008.
US$45.00
Perhaps more than any other piece of
jewelry in a womans wardrobe, a
brooch sends a distinctive message
about the wearers personal style.
While a stunning necklace or flashy
earrings might trigger stares and whispers, the brooch is a sociable jewel
that invites conversation, projects
energy outward, and almost always
has a story to tell.
In Brooches: Timeless Adornment, jewelry historian Lori Ettlinger
Gross shares her passion for these
individual works of art. The style
writer and editor divides her book
into six chapters, each illustrated by
David Behls often whimsical color
photography as well as archived
images. Chapters on history and craft
outline the broochs evolution from
article of necessity to jeweled adornment. Next, chapters devoted to collecting and style explore major motifs
and stylistic periods launched by
trendsetters and technological inno-
GEMS & GEMOLOGY
SPRING 2009
vations. Readers will enjoy historical

accounts of an English prince whose
paramour inspires the creation of a
mysterious lovers eye pin during
the Georgian period (17141830).
More than a century later, Elizabeth
Taylor reveals how she and Richard
Burton discovered their Night of the
Iguana Schlumberger brooch, marking a chapter of their storied romance.
In a chapter she calls Pin-ology,
Ettlinger Gross offers a primer on
pin-pairing and styling inspired by
aquatic, floral, color, and other
themes using traditional and unconventional materials. A closing chapter on pin care offers excellent suggestions for preserving, cleaning, and
repairing costume and precious
brooch jewelry.
The clearly written and engaging
text makes for a breezy read, while
the quality production, layout, and
stylish photos heighten the books
entertainment value. Costume pieces
are lovingly portrayed and hold their
own alongside examples of exquisite
diamond and gemstone jewelry set in
precious metals. Outstanding pieces
include Juliette Moutards ruby and
amethyst starfish brooch for the Paris
house of Ren Boivin (1938); a garnet,
tourmaline, and diamond brooch by
James de Givenchy for Taffin (2004);
and a private collectors ballerina pins
by various artists.
This enjoyable book will delight
brooch fans and is sure to win over a
few converts. Jewelers and casual
readers alike will discover the charm,
inspired artistry, and world of new
looks befitting this classic adornment.
MATILDE PARENTE
Libertine
Indian Wells, California
Collectors Guide to the

Epidote Group
By Robert J. Lauf, 96 pp., illus., publ.
by Schiffer Publishing, Arglen, PA,
2008. US$19.99
There are so many books that
attempt to cover as many different
BOOK REVIEWS
minerals or gems as possible that it is

refreshing to see one devoted to a relatively small group of mineral species.
This work is particularly good news
for the collector who, though able to
find many books on well-known
species or groups like tourmaline, has
difficulty finding anything on a collectable but lesser-known group such
as epidote.
The Collectors Guide to the
Epidote Group is a good effort to provide mineral collectors with the kind
of information suited to their needs.
For the lapidary, gemologist, or jeweler, the book includes the gem varieties
of the epidote group (quartz with epidote inclusions, transparent green to
brown epidote and clinozoisite, catseye epidote, and unakite). The closely
related species zoisite (including the
gem variety tanzanite) is discussed in
this book because historically it was
considered part of the group. (Note:
The International Mineralogical
Association [IMA] recently assigned
new species nomenclature to this
group, which is included in the text,
and ruled that the epidote group is
comprised of only monoclinic crystal
system members. Since zoisite is the
orthorhombic polymorph of clinozoisite, it was removed from the
group.)
There are four main chapters. The
first covers the history and the gem and
lapidary uses of the epidote group.
Taxonomy of the Epidote Group
covers general formula and crystal
structure, with tables of the species and
their formulas, while Formation and
Geochemistry covers epidotes in
igneous and metamorphic rocks. The
final chapter, The Minerals, includes
the clinozoisite, allanite, and dollaseite
subgroups and the related species
zoisite.
For collectors, the meat of the
book is the minerals chapter, devoted
to the actual materials coveted for
their beauty, rarity, or scientific interest. Epidote and clinozoisite occur in
beautiful crystal forms from many
famous localities, such as Knappenwand, Austria, and Prince of Wales
Island, Alaska. The star of the group
from a collectors point of viewif we

leave out tanzaniteis epidote itself.
The book contains numerous
color photographs of epidote-group
and zoisite specimens and fashioned
gems. Although the images and specimens themselves are not as fine as
those seen in some other books and
journals currently in publication, they
do show typical appearances and
include samples from numerous
localities of interestone of the best
features of this book!
Although of great importance, the
information on formation and chemical formulas is of less interest to the
average collector than the locations,
forms, associated species, and desirability of the different species, particularly those that occur as beautiful
specimens. The minerals chapter is
good but would have been of greater
value had it been expanded; better discussion of gem materials would have
been another improvement, especially given the limited attention paid to
tanzanite. That being said, I congratulate the author for creating a book on
a single group of minerals and hope
that others will do the same. For
those who covet epidote and its close
relatives, as I do, this book is a must
have!
MICHAEL EVANS
Gemological Institute of America
Carlsbad, California
OTHER BOOKS RECEIVED

World of Gems Conference. Edited
by Richard Drucker, 90 pp., illus.,
publ. by Gemworld International
[www.gemguide.com], Glenview,
IL, 2008, US$29.95. This handsome
work compiles abstracts and speaker
biographies from the inaugural
World of Gems Conference held
September 1315, 2008, in Rosemont, Illinois. Also included are photos of the conference and an introduction by Mr. Drucker and Dr. Lore
Kiefert of the AGTA Gemological
Testing Center.
TWO
GEMS & GEMOLOGY
SPRING 2009
S3
gemological
0 Abstracts
9
EDITORS
Brendan M. Laurs
Thomas W. Overton
GIA, Carlsbad
REVIEW BOARD
Edward R. Blomgren
Owls Head, New York
Jo Ellen Cole
Vista, California
Sally Eaton-Magaa
GIA, Carlsbad
Eric A. Fritz
Denver, Colorado
R. A. Howie
Royal Holloway, University of London
Edward Johnson
GIA, London
Paul Johnson
Guy Lalous
Academy for Mineralogy, Antwerp, Belgium
Kyaw Soe Moe
West Melbourne, Florida
Keith A. Mychaluk
Calgary, Alberta
Francine Payette
East Victoria Park, Western Australia
James E. Shigley
GIA Research, Carlsbad
Boris M. Shmakin
Russian Academy of Sciences, Irkutsk, Russia
Russell Shor
GIA, Carlsbad
Elise Skalwold
Ithaca, New York
Jennifer Stone-Sundberg
Portland, Oregon
Rolf Tatje
Duisburg, Germany
Dennis A. Zwigart
State College, Pennsylvania
S4
GEMOLOGICAL ABSTRACTS
COLORED STONES AND

ORGANIC MATERIALS
A classification of gem corundum deposits aimed towards gem
exploration. C. Simonet, E. Fritsch [emmanuel.fritsch@
cnrs-imn.fr], and B. Lasnier, Ore Geology Reviews, Vol.
34, 2008, pp. 127133.
This study proposes a classification of ruby and sapphire
deposits based on petrographic data and mode of genesis. The
authors point out the relationships between different gem
deposits and help link them to particular geologic environments, an important aid in prospecting. This also lends support
to determining geographic origin by complementing efforts to
understand inclusion suites seen in gem corundum. Although
mostly based on a review of the available literature, the study
also draws on the lead authors fieldwork, both published and
unpublished.
The authorswhile not concentrating on exact geologic
processes of corundum growth, but rather on favorable conditions of formation and distributionpropose a grouping of
gem deposits according to common characteristics and general
mechanisms of formation. Deposits are classified as primary
or secondary, then further subdivided and extensively
described.
A caveat is added against drawing conclusions about a
given deposit based on similar deposits, as each is unique.
Also, geologically different types of deposits may coexist in
proximity. Many unknowns remain, such as the influence of
metasomatism on ruby-bearing marbles and the influence of
geology on gem quality (transparency). This classification system seeks to provide a basis for future work on the geology of
gem corundum deposits.
ES
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
editors and their abstractors, 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 abstractor of each article is identified by his or her initials at
the end of each abstract. Guest abstractors are identified by their
full names. Opinions expressed in an abstract belong to the abstractor and in no way reflect the position of Gems & Gemology or GIA.
GEMS & GEMOLOGY
SPRING 2009
Optical absorption spectra of Fe 2+ and Fe 3+ in beryl crystals. G. Spinolo [giorgio.spinolo@mater.unimib.it], I.

Fontana, and A. Galli, Physica Status Solidi B, Vol.
244, No. 12, 2007, pp. 46604668.
The authors studied seven beryl samples of varying color
(aquamarine, heliodor, and goshenite), geographic origin,
and iron content to better understand the role of iron in
producing spectral absorption features and coloration. Each
sample was sliced into several optically oriented flat plates
of different thicknesses (0.16 mm), and spectra of these
plates were recorded over the range of 500003000 cm1 at
both ambient and low (16 K) temperatures. Iron in octahedral coordination is principally responsible for blue, green,
and yellow colors in beryl. The authors found no spectral
evidence that iron occurs either in tetrahedral coordination
or within the open channels of the beryl structure. While
Fe2+ concentration in beryl is higher than Fe3+ (by a factor of
3 to 5), Fe3+ plays a more important role in producing the
absorption features related to coloration.
JES
Putting a value on rare organics. M. C. Pedersen
[info@maggiecp.com], Organic Gems, May 2008,
www.maggiecp.co.uk.
Organic materials can be difficult to value because of their
rarity (as with hornbill ivory) or the unavailability of material by which to judge value due to trade bans or other
legal restrictions (as with elephant ivory). In the United
Kingdom, for example, it is unlawful to trade in elephant
ivory, tortoise shell, or rhino horn for commercial gain
even if the piece was owned before trade bans came into
effect. Legally, these materials can only be given away.
The author uses as an example a bead necklace of
root amber (a type of burmite; i.e., amber from
Myanmar that is ~100 million years old) that was offered
for sale. Most gemologists, valuers, and auctioneers have
never seenmuch less handled or valuedthis material.
Root amber accounts for only about 2% of all burmite
mined; the general public is not familiar with this material, and it is not easy to sell. One of two cutters familiar
with working root amber was unable to give any indication of value, but stated that the material is only worth
what someone is willing to pay for it. The age/history of
the necklace had no bearing on the value because it did
not belong to anyone famous and it could be replaced. The
author concluded that the cost approachto purchase and
fashion the material, and then make it into a necklace
should be used to establish a value.
JEC
Quelques varits dambre animal et dambre vgtal
[Some varieties of animal and vegetal amber]. E.
Gonthier and A. Zivkovic, Revue de Gemmologie,
No. 163, 2008, pp. 1117 [in French].
Animal amber (often called ambergris or gray amber) is a
substance produced in the digestive system of sperm
whales and was once used in perfumery. Vegetal amber is

a group of more or less fossilized plant-derived resins of
varying ages. Young, semi-fossil varieties are called copal;
the older, fully fossilized resins are simply amber. The
authors review ambers gemological properties, its uses
throughout history, and its imitations. An extensive glossary of terms related to amber is included.
RT
Structural characterization and chemical composition of
aragonite and vaterite in freshwater cultured pearls.
A. L. Soldati [soldatia@uni-mainz.de], D. E. Jacob,
U. Wehrmeister, and W. Hofmeister, Mineralogical
Magazine, Vol. 72, No. 2, 2008, pp. 579592.
Vaterite and aragonite in freshwater cultured pearls from
mussels of the genus Hyriopsis were structurally and compositionally characterized by Raman spectroscopy, micro
computer tomography, high-resolution field-emission scanning electron microscopy (SEM), and laser ablationinductively couple plasmamass spectroscopy (LA-ICP-MS).
Vaterite is related to the initial stages of pearl biomineralization, but it cannot be a precursor to aragonite. Because
it is not related to a particular crystal habit, vaterite does
not have a structural function in the cultured pearls.
Greater contents of elements such as P and S in vaterite,
as well as larger total organic contents in vaterite compared to aragonite, in conjunction with higher amounts of
Mn 2+ and Mg 2+, imply a stabilizing role for organic
molecules and X2+ ions in biological vaterite.
RAH
DIAMONDS
Characterization of pink diamonds of different origin:
Natural (Argyle, non-Argyle), irradiated and
annealed, treated with multi-process, coated and synthetic. B. Deljanin [brankod@eglcanada.ca], D. Simic,
A. Zaitsev, J. Chapman, I. Dobrinets, A. E Widemann, N. Del Re, T. Middleton, E. Deljanin, and A.
De Stefano, Diamond and Related Materials, Vol.
17, 2008, pp. 11691178.
The use of treatments to alter the color of gem-quality diamonds has become increasingly popular. Radiation treatment, in which diamonds are exposed to high-energy particles, has been used commercially for over 50 years. Since
1999, the use of high-pressure, high-temperature (HPHT)
treatment has received considerable attention. Red and
pink diamonds are the rarest and most expensive naturalcolor diamonds, so artificial enhancement to achieve these
hues is an attractive option. Both natural and synthetic
diamonds can be treated to produce a pink color.
This study mainly focused on using fluorescence techniques to characterize pink diamonds and to compile a
reference library of emission spectra. Long- and shortwave UV fluorescence (365, 254, and 220 nm wave-
GEMS & GEMOLOGY
SPRING 2009
S5
lengths) was used in a custom-built microscope with a

fluorescence camera to record images, and a fluorescence
spectrometer was used to establish a correlation with
color origin. Several additional techniques (UV-Vis-NIR,
Fourier-transform infrared [FTIR], and photoluminescence
[PL] spectroscopy, as well as cathodoluminescence imaging and electrical conductance) were used to establish further criteria for distinguishing natural, treated, and synthetic pink diamonds, and to find correlations with the
fluorescence data. The authors conclude that it is possible
to separate natural-color pink diamonds from all categories of treated pink and laboratory-grown pink diamonds, loose or mounted, by using a combination of standard and advanced gemological instruments.
DAZ
Diamondiferous xenoliths from crystal subduction:
Garnet oxygen isotopes from the Nyurbinskaya
pipe, Yakutia. Z. V. Spetsius, L. A. Taylor [lataylor@
utk.edu], J. W. Valley, M. T. Deangelis, M. Spicuzza,
A. S. Ivanov, and V. I. Banzeruk, European Journal of
Mineralogy, Vol. 20, No. 3, 2008, pp. 375385.
The newly developed Nyurbinskaya kimberlite pipe in
Yakutia, Russia, has yielded an unprecedented array of
xenoliths, all containing diamonds. Garnets were characterized from 121 of these xenoliths, most of them eclogites
but also some pyroxenites and peridotites. The 18O ratios
of most of the peridotitic garnet samples fell within the
range of the average mantle, except for one with a 18O
value of 6.57. Garnets from pyroxenites (websterites)
generally had 18O values above 6.0, with two samples
as high as 7.3 and 8.59, and only two samples as low
as 5.9 and 6.0. Eclogitic garnets had a 18O range of
4.79.7, with more than 80% of them above 6.0.
These garnet oxygen-isotope ratios offer evidence for the
subduction of oceanic crust, as well as major involvement
with the low-temperature, metasomatized upper portion
of the earths crust, in forming the diamond-bearing rocks.
[Abstracters note: This special issue of European
Journal of Mineralogy is in honor of V. S. Sobolev, who
first suggested that diamonds might be found in the rocks
of the northern Siberian platform. Dr. Sobolev followed
up by using indicator minerals such as Cr-rich pyrope to
locate kimberlite, leading to the discovery of the famous
Mir pipe and several hundred others in Yakutia. The issue
contains nine other diamond-related papers in addition to
the one abstracted here.]
RAH
Olivine inclusions in Siberian diamonds: High-precision
approach to minor elements. N. V. Sobolev
[sobolev@uiggm.nsc.ru], A. M. Logvinova, D. A.
Zedgenizov, N. P. Pokhilenko, D. V. Kuzmin, and
A. V. Sobolev, European Journal of Mineralogy, Vol.
20, No. 3, 2008, pp. 305315.
At depths in the continental lithospheric mantle exceed-
S6
ing 120150 km, there are two common geologic environments that support diamond formation. Ultramafic or
peridotitic (U-type) and eclogitic (E-type) environments, as
indicated by mineral inclusions in diamonds, are responsible for diamondiferous xenoliths in kimberlites and lamproites. The ratio of diamonds derived from these two geologic environments varies widely between localities. Utype diamonds predominate, however, in the overwhelming majority of diamond occurrences worldwide. Olivine is
the most typical inclusion in U-type diamonds, along with
enstatite, pyrope, and chromite in those from harzburgitic
or dunitic U-type assemblages.
More than 260 olivine inclusions in diamonds from
major Siberian mines were studied and compared to those
from the Snap Lake deposit in Canada. The olivine composition of eight xenoliths from diamondiferous peridotites in the Udachnaya pipe, representing the rarest
mantle samples, was also reexamined. The inclusions
were analyzed by electron microprobe, and minor-element abundances in most of the olivines varied in the following ranges (wt.%): NiO0.3200.408, CaO
0.0050.045, MnO0.0790.131, Cr2O30.0130.115,
Co0.0090.022, and Al2O30.0070.039. About 70% of
the olivines were very low in CaO, reflecting a relatively
low equilibration temperature for the lherzolitic paragenesis, or the lack of clinopyroxene associated with olivine.
RAH
GEM LOCALITIES
The blue colouring of beryls from Licungo, Mozambique:
An X-ray absorption spectroscopy study at the iron
K-edge. M. O. Figueiredo [ondina.figueiredo@ineti.pt],
T. Pereira da Silva, J. P. Veiga, C. Leal Gomez, and V.
De Andrade, Mineralogical Magazine, Vol. 72, No. 1,
2008, pp. 175178.
This study was undertaken on aquamarine crystals and
fragments with various shades of blue color collected from
a pegmatite field along the Licungo River near Mocuba in
central Mozambique. X-ray absorption near-edge spectroscopy (XANES) was performed to understand the
valence state of iron in this material. This transition metal
was present as predominantly Fe2+ replacing Al3+ in octahedral sites in the beryl structure. No color changes were
noted in the material during X-ray irradiation.
JES
A colour-changing titanite from Afghanistan. T.
Hainschwang [thomas.hainschwang@gemlab.net],
Gems & Jewellery, Vol. 17, No. 4, 2008, pp. 67.
A 3.95 ct gemstone reportedly mined from Badakhshan
Province in northeastern Afghanistan was submitted to
the Gemlab Gemological Laboratory in Liechtenstein and
subsequently identified by FTIR spectroscopy as titanite
(sphene). Although titanite is not typically associated with
GEMS & GEMOLOGY
SPRING 2009
a color-change effect, the stone appeared orangy yellow in

incandescent light and brownish yellowish green in natural daylight. Its UV-Vis-NIR absorption spectra had a broad
absorption band centered near 600 nm; compared to common titanite from Madagascar, this band was more
intense and was shifted ~15 nm toward shorter wavelengths. The increase in absorption intensity and shifted
band position were responsible for the observed color
change, and provide a good example of how sensitive color
change is to small variations in these factors.
ES
Mineralogical and geochemical characterization of the
bituminous agates from Nowy Koscil (Lower
Silesia, Poland). M. Dumanska-Slowik [dumanska@uclagh.edu.pl], L. Natkaniec-Nowak, M. J.
Kotarba, M. Sikorska, J. A. Rzymelka, A. Loboda,
and A. Gawel, Neues Jahrbuch fr Mineralogie
Abhandlungen, Vol. 184, No. 3, 2008, pp. 255268.
Agates from Nowy Koscil in Poland exhibit horizontally
stratified structures often developed as multiradial stars.
These agates are mainly spherical with diameters ranging
from 2 cm up to 4070 cm; most are brownish red, red, or
honey-black. The dark color of their banding is mainly
caused by Fe compounds, rare earth elementbearing minerals, Zn sulfides, and organic matter. The content of
organic matter is relatively low (0.15 wt.%), but it is dispersed within the silica matrix. It forms thin laminae or
irregularly shaped drops or lenses. Asphaltenes are the
dominant bitumen (56%), with the remainder varying
between saturated hydrocarbons (18%), resins (16%), and
aromatic hydrocarbons (10%). Carbon isotope analysis of
the organic matter revealed its algal or mixed algal/humic
origin (13O of 25.928.9).
RAH
Nanostructure of noble opal from the Raduzhnoe deposit,
northern Primorye, Russia. S. V. Vysotskiy
[svys@mail.ru], V. G. Kudryavyi, and A. A.
Karabtsov, Doklady Earth Sciences, Vol. 420, No. 4,
2008, pp. 690692.
The structure of play-of-color opal from Russias Raduzhnoe deposit differs significantly from that of synthetic opal
or Australian opal. The Raduzhnoe material is unusual in
that it comes from a hydrothermal deposit related to volcanic activity, whereas most play-of-color opal is mined
from ancient weathered crusts. During its formation, this
opal also underwent pneumatolytic annealing (the impact
of vapor at elevated pressure-temperature conditions).
X-ray diffraction (XRD) analysis of Raduzhnoe opal
showed that it consists of -cristobalite. The degree of
crystallinity and amount of amorphous silica varied
among samples. The nanoparticles were devoid of the
ordered structure that is characteristic of typical play-ofcolor opal. The partial polymerization of globules due
to their intergrowth led to the increased hardness and the
formation of monolayers and layered packets with relatively even surfaces. These packets and layers, together
with -cristobalite blocks, probably play the role of structured domains and thin films that are responsible for the
gems play-of-color.
A.
M.
Clark
A preliminary stable isotope study on Mogok ruby,
Myanmar. T.-F. Yui [tfyui@earth.sinica.edu.tw], K.
Zaw, and C.-M. Wu, Ore Geology Reviews, Vol. 34,
No. 1/2, 2008, pp. 192199.
Studies of stable isotopes in minerals can provide information on their formation conditions and subsequent geologic history. This study of carbon and oxygen isotopes in calcite marbles hosting gem ruby examined how corundum
formed in these marbles, which are low in Al and Cr, and
the nature of the mineralizing fluids involved. In the
Mogok area, spinel-forsteritebearing marbles, phlogopitegraphitebearing marbles, and ruby-bearing marbles are
enclosed in banded metamorphic gneisses. Isotopic data
for these marbles indicate formation by the interaction of
preexisting rocks with CO2-poor fluids (for the spinelforsterite marble) or CO2-rich fluids (for the phlogopitegraphite and the ruby-bearing marbles) that originated
from unknown external igneous and/or metamorphic
sources. Ruby formation appears to have occurred at temperatures of approximately 600C. Local variations in the
preexisting rocks and the evolving chemical composition
of the infiltrating fluids during cooling account for spatial
differences in the distribution of ruby and other minerals
in the area.
JES
INSTRUMENTS AND TECHNIQUES

Hanneman-Hodgkinson Green Stained Jadeite Filter. A.
Hodgkinson [alan-hodgkinson@talktalk.net],
Australian Gemmologist, Vol. 23, No. 5, 2008, pp.
232237.
This filter was developed to provide a clear and positive
distinction between the various tones and saturations of
natural and dyed green jadeite, as well as a means of distinguishing between natural and synthetic green jadeite.
With an incandescent light source, natural green jadeite
remains green when viewed through the filter, while dyed
material appears pale brownish pink to reddish orange. As
such, this new filter provides a means of separating A-jade
(untreated) from C-jade (bleached and dyed). For B-jade
(bleached and polymer impregnated) and B+C jade (polymer + dye), long- and short-wave UV fluorescence can be
used as further means of separation. Synthetic jadeite can
be identified by its pinkish orange reaction to the filter.
DAZ
Use of software to enhance depth of field and improve
GEMS & GEMOLOGY
SPRING 2009
S7
focus in photomicrography. J. Piper [webmaster@

prof-piper.com], Microscopy and Analysis, No. 90,
2008, pp. 1519.
In photomicrography and light microscopy, two fundamental limitations are depth of field and sharpness of
focus. At high magnifications or when a specimen is relatively thick, it is usually impossible to obtain a sharp
focus over the full depth of the piece. This is because the
depth of field of an objective lens decreases proportionally
as its magnifying power increases. However, by using readily available software, the focal depth and sharpness in
photomicrographic images can be enhanced dramatically.
The author discusses eight free/shareware tools that can
be used for this purpose. A discussion of special techniques applicable for three-dimensional reconstructions is
also provided. According to the authors evaluations, at
least three software solutions will lead to excellent results:
Combine Z 5 (freeware), Picolay (freeware), and Helicon
Focus (shareware).
DAZ
PRECIOUS METALS
Laser surface colouring of titanium for contemporary jewellery. S. OHana [sarah.ohana@manchester.ac.uk],
A. J. Pinkerton, K. Shoba, A. W. Gale, and L. Li,
Surface Engineering, Vol. 24, No. 2, 2008, pp.
147153.
A moderate-power (60 W) pulsed CO2 laser, directed via an
X-Y positioning system and interfaced with commercial
graphics software, has been efficiently utilized to create
controlled, even areas of color, patterns, and drawings on
the surface of commercial-purity Ti-alloy plate for jewelry
design purposes. The laser alters the surface topography of
the metal and results in the deposition of thin layers of
titanium oxides (200 nm or less in total thickness). The
main coloration mechanism is interference, which is produced by the presence of a graded surface layer consisting
of an outer zone of TiO2 (rutile) and underlying zones that
are more Ti-rich (TiO or Ti2O). A wide range of colors and
a variety of delicate designs can be created by this method.
JES
SYNTHETICS AND SIMULANTS

Dependence of crystal quality and value on synthesis
temperature in growing gem diamond crystals. H.Y. Xiao, X.-P. Jia, C.-Yi. Zang, S.-S. Li, Y. Tian, Y.-F.
Zhang, G.-F. Huang, L.-Q. Ma, and H.-A. Ma
[maha@jlu.edu.cn], Chinese Physics Letters, Vol. 25,
No. 4, 2008, pp. 14691471.
The authors report on using temperature to control the
height-to-diameter ratios (-values) of type Ib synthetic
diamond single crystals grown under HPHT conditions.
S8
The synthetic diamonds were grown at 5.4 GPa and

12001300C in a cubic anvil apparatus. The starting
materials included high-purity graphite (the carbon
source), a Ni70Mn25Co5 catalyst, and seeds of high-purity
diamond grit with 0.5 0.5 mm {100} crystal faces. At synthesis temperatures below 1230C, -values of less than
0.40 were realized. These crystals tended to be gray and
opaque, with a wavy surface. They were referred to as
skeleton crystals, and their gray color was attributed to
the inability of carbon to diffuse regularly at such low temperatures, with some diffusing as graphite. By raising the
synthesis temperature to 12301255C, sheet-shaped crystals were formed with -values of 0.400.45. Towershaped crystals with -values of 0.450.60 were observed
at growth temperatures ranging from 1255 to 1280C.
Above this temperature range, tower-shaped crystals
formed with >0.60, but these crystals had abundant
inclusions and surface cavities.
JSS
Early plastics as imitations of gem materials: An introduction to the early plastics often used to copy gem
materials. M. C. Pedersen [info@maggiecp.com],
Organic Gems, No. 7, 2008, www.maggiecp.co.uk.
Plastics have long been employed as gem imitations, particularly of organic materials with structures that can be
copied very closely (e.g., amber, tortoise shell, and horn).
All plastics are polymers, which are very large molecules
made up of many smaller units joined together to form a
long chain. Early semisynthetic plastics are described in
detail, including vulcanite (the first, patented in 1843),
shellac, bois durci, cellulose nitrate (known by the trade
name Celluloid), cellulose acetate, and casein (known as
Galalith or Erinoid). For each one, the development process and the material it imitates are noted. The first completely synthetic plastics, such as phenolic (known as
Bakelite, invented in 1907) and urea formaldehyde (known
as Bandalasta or Beetleware) are similarly described. Many
more plastics were subsequently developed, in the 1930s,
including polystyrene, polyethylene, acrylic, and nylon.
Early plastics can degrade suddenly or crack, craze,
warp, and discolor with time. Once degradation has started
there is no cure, and a degraded plastic can contaminate
other early plastics. Protective measures include avoiding
bright light and keeping the material in a dry, ventilated
place. Also, avoid exposing them to household cleaners,
hairspray, perfumes, and cosmetics.
JEC
Gemmological investigation of a synthetic blue beryl: A
multi-methodological study. I. Adamo, G. D. Gatta
[diego.gatta@unimi.it], N. Rotiroti, V. Diella, and A.
Pavese, Mineralogical Magazine, Vol. 72, No. 3,
2008, pp. 799808.
The authors studied a synthetic Cu/Fe-bearing blue beryl
2+
[(Be2.86Cu0.14)(Al1.83Fe3+
0.14Mn 0.03Mg 0.03)(Si 5.97 Al 0.03)O 18
GEMS & GEMOLOGY
SPRING 2009
(Li0.12Na0.04 0.40H2O)] by electron microprobe, LA-ICPMS, IR spectroscopy, and single-crystal XRD and thermogravimetric analyses. This study aimed to accurately
locate chromophores in the crystal structure, in view of
the increasing production of marketable hydrothermal
synthetic beryls with exotic colors. Fourier maps of the
electron density suggested that Cu was located at the
tetrahedral site along with Be, whereas Fe shared the octahedral site with Al. No evidence was found of extra-framework Cu/Fe sites (i.e., channel sites). IR spectra showed
that the H2O molecules had two configurations, with the
H-H vector oriented both parallel and perpendicular to
[0001].
RAH
High-rate growth and nitrogen distribution in homoepitaxial chemical vapour deposited single-crystal diamond. H.-D. Li [hdli@jlu.edu.cn], G.-T. Zou, Q.-L.
Want, S.-H. Cheng, B. Li, J.-N. Lu, X.-Y. Lu, and
Z.-S. Jin, Chinese Physics Letters, Vol. 25, No. 5,
2008, pp. 18031806.
The authors report on a potential technique for differentiating between natural diamond and CVD-grown synthetic
diamond. In this study, single-crystal synthetic diamond
was grown via microwave plasma chemical vapor deposition using a seed holder made of a CVD polycrystalline
synthetic diamond film. The choice of diamond instead of
molybdenum for the seed holder was intended to reduce
contamination and allow growth at higher temperatures.
Synthetic diamond was grown at a rate of more than 50
m/hour, and Raman and photoluminescence spectroscopy were used to characterize crystallinity and nitrogen
distribution.
The authors note that nitrogen was concentrated on
the central surfaces of the crystals, which were cooler
during growth than the corners and edges. Vertically,
nitrogen was more abundant at a depth up to ~300 m,
and it was homogeneous at greater depths. The surface
was the hottest region during growth, and the authors
propose that nitrogen diffused away from the growing surface toward the cooler as-grown layers below. Diffusion
tapered off below ~600 m depth, where nitrogen saturation was reached. PL spectroscopy of nitrogen vacancies
(NV centers) illustrates the preferential distribution of
nitrogen in the CVD diamond. This distribution is characteristic of high-growth-rate CVD synthetic diamonds,
so the authors recommend using PL spectroscopy to
inspect the nitrogen distribution along the surface and in
cross section to help distinguish between natural and
CVD-grown synthetic diamond.
JSS
Lab alert: Black moissanite. H. Kitawaki [ahmadjan@gaajzenhokyo.co.jp], Gemmology, September 2008, pp.
34 [in Japanese with English supplement].
The author reports on black diamond jewelry contain-
ing synthetic moissanite. This material is difficult to identify in jewelry set with many small stones. Techniques
include magnification (black diamonds typically contain
black graphite inclusions, whereas black synthetic
moissanite is solidly opaque); X-ray inspection (synthetic
moissanites transparency to X-rays distinguishes it from
diamond); use of a moissanite tester (this identifies synthetic moissanite but cannot make a positive distinction
between high temperaturetreated black diamond and synthetic moissanite); X-ray fluorescence (to detect Si, which
is present in synthetic moissanite but absent in diamond);
and Raman microanalysis (which positively identifies diamond, synthetic moissanite, and cubic zirconia). The Xray transparency test is especially useful for jewelry set
with numerous gemstones. For individual stones, techniques such as X-ray fluorescence or Raman microanalysis
are useful.
GL
Monitoring diamond crystal growth, a combined experimental and SIMS study. V. N. Reutsky [reutsky@
uniggm.nsc.ru], B. Harte, Y. M. Borzdov, and Y. N.
Palyanov, European Journal of Mineralogy, Vol. 20,
No. 3, 2008, pp. 365374.
Detailed ion microprobe measurements were performed
on two synthetic diamond crystals grown by the metal
catalyst technique under identical conditions of 1450C
and 5.5 GPa, but with different source nitrogen abundances. The C and N isotope compositions and nitrogen
abundances were measured in traverses across the crystals,
which included cubic and octahedral sectors of both relatively rapid and relatively slow growth. In both crystals, an
early growth phase characterized by falling 13C and rising
Nppm was followed by an extensive growth phase with
fairly constant 13C and gradually decreasing Nppm. The
change in 13C was modeled numerically; stabilization
was achieved once a steady state was attained, and the
synthetic diamond grew with the same 13C composition
as the graphite source. The decreasing Nppm values
appear to be a result of Rayleigh fractionation. The N isotope compositions show major differences of ~30
between octahedral and cubic sectors, possibly representing a consistent difference in N isotope adhesion between
the two faces.
RAH
Trade alert: Flux grown synthetic red spinels again on the
market. M. Krzemnicki [gemlab@ssef.ch], Gem
Market News, Vol. 27, No. 6, 2008, pp. 79.
The author encountered several red flux-grown synthetic
spinels at the September Hong Kong Jewellery & Gem
Fair; these also have been offered for sale in Bangkok.
While not new, they are appearing more often as red spinel
becomes increasingly popular. The material resembles
fine-quality natural spinel and is difficult-to-impossible to
separate using standard gemological tests. Flux-grown
GEMS & GEMOLOGY
SPRING 2009
S9
crystals resemble natural crystals, even containing natural-appearing growth marks, and can fool buyers if mixed
with natural rough in parcels. The author warns that
while blue flux-grown synthetic spinels (colored by cobalt)
have not been seen recently, they may reappear.
Microscopic observation of the synthetics revealed
only a few small jagged or tubular cavities filled with
black to orange-brown flux residues containing gas bubbles; one had a single metallic flake and tiny parallel hollow channels. Natural red spinels can contain brownish
iron hydroxide in cavities that could be confused with
flux residues. Advanced testing indicated a low concentration of Zn, which distinguishes flux synthetics from
natural spinel; the latter contain much greater Zn (by a
factor of 10 or more). Raman spectra showed a distinctly
broader peak shape for the flux-grown synthetics, similar
to the peak broadening observed in Verneuil synthetics.
Strong photoluminescence produced by green laser excitation indicated emission peaks due to chromium in both
natural and flux-grown samples, but in the latter the peak
was less structured, offering another possible separation
technique.
ES
TREATMENTS
Determination of the C defect concentration in HPHT
annealed type IaA diamonds from UV-Vis absorption
spectra. F. De Weerdt [filip.deweerdt@diamondlab.org]
and A. T. Collins, Diamond and Related Materials,
Vol. 17, No. 2, 2008, pp. 171173.
Type Ib diamond is characterized by the presence of single
substitutional nitrogen (i.e., the C defect). This defect is
more often seen in synthetic than in natural diamonds; it
can also be observed in type Ia diamonds. If the concentration of the C defect is at least a few parts per million, it
can be calculated from peak intensities of the 1344 and
1130 cm1 features in the infrared spectrum. If the concentration is between 0.1 and a few parts per million, then the
broad absorption band centered at 270 nm may be used for
this calculation. In type IaA diamonds, however, the
strong absorption of radiation by the A defects makes it
difficult to measure these absorptions. This situation can
occur in HPHT-treated type IaA diamonds, in which A
defects dissociate into C defects.
The authors propose an alternative method of measuring the C-defect concentration by examining the absorption coefficient at 400 nm. To establish the correlation,
they plotted the concentration of C defects against the
absorption coefficient at 400 nm (measured in cm1). Since
N3 centers (ZPL = 415 nm) are one of the most common
defects in diamond, this prompted the researchers to measure the absorption coefficient (as opposed to absorbance)
and deconvolute the absorption spectrum to resolve the
N3 and C defects. The resulting plot of absorption coeffi-
S10
cient against concentration of C defects showed good

agreement, and the calculated proportionality coefficient
was 2.00 0.04 ppm/cm1.
S E M
HPHT treatment of CO2 containing and CO2-related brown
diamonds. T. Hainschwang [thomas.hainschwang@
gemlab.net], F. Notari, E. Fritsch, L. Massi, B.
Rondeau, C. M. Breeding, and H. Vollstaedt,
Diamond and Related Materials, Vol. 17, 2008, pp.
340351.
Ten type I natural diamonds containing CO2 and three
related, non-CO2-bearing diamonds (Pseudo CO2 diamonds) were treated by HPHT processing. Changes in
color, color distribution, inclusions, luminescence, and
spectral features in the visible to mid-infrared regions were
observed. All samples were predominantly brown before
treatment and had inhomogeneous color distribution not
attributed to strain. After treatment, most samples
appeared more yellow, some with a greenish modifier,
although brown remained the dominant hue. Changes in
color were subtler than those induced in classic deformation-related brown type I diamonds, and were not related to any known centers such as H3. These color modifications after HPHT treatment were due to destruction of
broad absorption bands and a general increase in transmission from 400 to 700 nm in the Vis-NIR spectra; however,
typical HPHT-related color centers such as H2 and H3
were not observed. Infrared spectra of the CO2-containing
diamonds showed unusual modification by HPHT treatment. CO2-related bands increased in intensity in diamonds with low to medium CO2 content and were created
in Pseudo CO2diamonds, while nitrogen-related absorptions decreased in both cases. PL spectra exhibited several
unpublished emission peaks that were essentially
unchanged by HPHT treatment.
This work appears to be the first indication of the likely presence of structurally bound oxygen in diamond. The
authors propose that CO2-related absorptions are due to
CO2 bound in the diamond lattice rather than being present as solid CO2 inclusions as was previously thought.
Furthermore, the authors note that their results indicate
that HPHT is a non-identifiable treatment in the case
of CO2-containing and CO2-related brown diamonds, as it
does not produce the color centers indicative of HPHT
annealing in classic type Ia brown diamonds.
Emily V. Dubinsky
MISCELLANEOUS
Are diamonds forever? Using the permanent income
hypothesis to analyze Botswanas reliance on diamond revenue. O. Basdevant [obasdevant@imf.org],
IMF Working Paper 08/80, March 2008, pp. 113,
www.imf.org/external/pubs/ft/wp/2008/wp0880.pdf.
GEMS & GEMOLOGY
SPRING 2009
Botswana has benefited greatly from diamond mining,

which accounts for more than one-third of the countrys
GDP and three-fourths of its exports. While the countrys
government has managed this resource well, it is likely
that diamond production will begin to wind down by
2021, and be depleted by 2029, which would create a precipitous drop in Botswanas development efforts. The
author offers a formula detailing how much revenue the
government must begin saving now in order to forestall
such a problem later.
RS
Certification and artisanal and small-scale miningAn
emerging opportunity for small-scale development.
E. Levin, Communities and Small-Mining Report
Series, June 2008, 24 pp., www.artisanalmining.org/
userfiles/file/CASM-FairTrade.pdf.
This report examines the organizations and certification
procedures involved in developing improved working conditions and greater returns for small-scale miners of colored gems and diamonds in various countries. The article
offers detailed definitions of terms such as fair trade, sustainable, development, and green as applied to mining
practices. It also reviews the various Fair Trade initiatives
that have been created during the past five years, and offers
a resource and contact list for such organizations.
RS
Diamonds not just the proverbial girls best friend
Theyre driving an economic boom in Canadas
north. J. Cooper, CMA Management, Vol. 82, No. 2,
2008, pp. 5254, www.managementmag.com/
index.cfm/ci_id/10798/la_id/1.
In the decade since diamonds have been mined in the
Northwest Territories, the provincial gross domestic product (GDP) has increased from $1.6 billion to $4.3 billion,
while unemployment has fallen from 13.7% to 5.3%.
Mining accounts for 40% of the provincial GDP and has
helped generate more than 4,000 jobs, 25% of which have
gone to aboriginal workers. The report also notes that
Canadian mining operations remain committed to strict
codes of professional conduct, care for the environment,
and involvement of the aboriginal communities to insure
diamond mining continues to be a sustainable benefit to
northern Canada. In 2007, Canada produced an estimated
16 million carats of diamonds valued at $1.7 billion.
RS
Diamonds or development? A structural assessment of
Botswanas forty years of success. E. Hillbom
[ellen.hillbom@ekh.lu.se], Journal of Modern African
Studies, Vol. 46, No. 2, 2008, pp. 191214.
The diamond-rich nation of Botswana is often held up as a
model for successful development through good governance and prudent management of natural resources.
However, the author argues that economic growth and
good governance have not translated into broad-based

development. She points out that much of Botswana
remains technologically backward, and that the country
today is much less self-sufficient in food production than
it was 20 years ago, now importing nearly all of its food.
The nation, she argues, has not used diamond income to
transform its agricultural sector, nor has it promoted
efforts to bring prosperity and employment to all segments
of society.
RS
Evaluation of DNA damage in jewellery workers occupationally exposed to nitric oxide. R. Jayakumar
[jayakumar7979@gmail.com] and K. Sasikala,
Environmental Toxicology and Pharmacology, Vol.
26, 2008, pp. 259261.
Aqua regia, a mixture of nitric and hydrochloric acid, is
commonly used in gold processing and jewelry manufacturing. The chemical reactions that take place when gold
is in contact with nitric acid produce nitric oxide (NO),
which has been shown to cause chromosomal damage and
thus increase cancer risk. The authors studied a group of
Indian jewelry workers who were exposed to aqua regia as
part of their jobs; the workers did not use protective masks
and worked in a small room with poor ventilation.
Compared to a control group, the test subjects showed significantly more chromosomal damage. Cigarette smoking
was shown to increase overall risk, but the length of exposure (i.e., years in the industry) did not. The authors conclude that exposure to aqua regia can cause chromosomal
damage in a jewelry industry setting.
TWO
Repatriation of the Koh-i-Noor diamond: Expanding the
legal paradigm for cultural heritage. S. Ghoshray
[sabyghoshray@sbcglobal.net], Fordham International Law Journal, Vol. 31, 2008, pp. 741-780.
The Koh-i-Noor is one of the worlds great historic diamonds. It currently resides in the Queen Mothers Crown
as part of the English Crown Jewels, but the list of its previous ownersboth known and legendaryis a long one
indeed. At various times in history, it resided in India,
Persia (now Iran), and Afghanistan before being taken to
England following the British conquest of the Punjab. The
author of this essay argues that, because the diamond
changed hands only through wars and violence, never
being honestly bought or sold, it remains the rightful property of India.
The last Indian owner of the diamond was Maharaja
Duleep Singh, the 12-year-old ruler of the Punjab. After
the British conquest, the young maharaja was forced to
sign the Treaty of Lahore, which (among other concessions) specifically transferred the diamond to Britain. The
author argues that this was part of the British attempt to
destroy the political power of the Sikhs by depriving them
of their historical wealth.
GEMS & GEMOLOGY
SPRING 2009
S11
OUR EDUCATION.
YOUR WORLD OF OPPORTUNITY.
Because Public Education

Happens at the Counter.
GIA launches Retailer Support Kit and website
SEOUL 8 : 0 0 P M
LONDON
NOON
GIA alumni network at cultured pearl seminar.
GIA-trained jeweler
advises client on 5 carat solitaire.
NEW YORK
7:00 AM
Diamonds Graduate negotiates

purchase of rough parcel.
TOKYO 8 : 0 0 P M
CARLSBAD
4:00 AM
Student completes gem ID project.
MUMBAI 4 : 3 0 P M
Core gem curriculum updated

to reect new research.
Sales associate explains 4Cs to customer.
HONG KONG 7 : 0 0 P M
Manufacturing exec expands business skills online.
BANGKOK 6 : 0 0 P M
Graduate Gemologist spots treated emeralds in bulk order.
A $97.00 value, shipping

and handling extra.
GIAs Retailer Support Kit has been developed to help

Almost anywhere you go, someone is using education acquired from GIA. Our international campuses, traveling classes,
sales associates educate the public about diamonds,
corporate seminars and online courses help individuals dene and rene vital skills.
the 4Cs, and thoroughly explain a GIA grading report.
And GIA supports that learning with credentials valued throughout the gem and jewelry world.
Take full advantage of all that GIA has to offer by visiting
W W W. G I A . E D U
www.retailer.gia.edu
To order your FREE kit, log on to www.retailer.gia.edu
CARLSBAD
MOSCOW
NEW YORK
MUMBAI
LONDON
BANGKOK
ANTWERP
HONG KONG
FLORENCE
BEIJING
GABORONE
TAIPEI
SEOUL
JOHANNESBURG
OSAKA
TOKYO
GMSGG
RTGG09

SP09

Uploaded by

Copyright:

Available Formats

SP09

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

SP09

Uploaded by

Copyright:

Available Formats

GEMS & GEMOLOGY

The French Blue and the Hope

THE QUARTERLY JOURNAL OF THE GEMOLOGICAL INSTITUTE OF A MERICA

Because Public Education

GIA alumni network at cultured pearl seminar.

Diamonds Graduate negotiates

Student completes gem ID project.

Core gem curriculum updated

Sales associate explains 4Cs to customer.

Manufacturing exec expands business skills online.

Graduate Gemologist spots treated emeralds in bulk order.

A $97.00 value, shipping

GIAs Retailer Support Kit has been developed to help

sales associates educate the public about diamonds,

the 4Cs, and thoroughly explain a GIA grading report.

Take full advantage of all that GIA has to offer by visiting

To order your FREE kit, log on to www.retailer.gia.edu

Volume 45, No. 1

The Dr. Edward J. Gbelin Most Valuable Article Award

FEATURE ARTICLES _____________

Gray-to-Blue-to-Violet Hydrogen-Rich Diamonds from

NOTES AND NEW TECHNIQUES ________

Hackmanite/Sodalite from Myanmar and Afghanistan

RAPID COMMUNICATIONS _____________

Solution-Generated Pink Color Surrounding Growth Tubes and Cracks in

Identification of the Endangered Pink-to-Red Stylaster Corals

REGULAR FEATURES ________________________________________

Gem News International

2009 Gems & Gemology Challenge

Editors, Gemological Abstracts

Editor, Gem News International

Editors, Lab Notes

All rights reserved.

As we celebrate Gems & Gemologys

COPPER-BEARING (PARABA-TYPE) TOURMALINE

MOST VALUABLE ARTICLE AWARD

GEMS & GEMOLOGY

William B. Skip Simmons

of research experience in mineralogy and petrology. Donna Beaton is manager of

COLOR GRADING D-TO-Z DIAMONDS

John M. King, Ron H. Geurts, Al M. Gilbertson, and James E. Shigley

A HISTORY OF DIAMOND TREATMENTS

Congratulations to Rui Galopim de Carvalho of Sintra, Portugal,

MOST VALUABLE ARTICLE AWARD

GEMS & GEMOLOGY

For a complete list of articles from 1981 forward, visit www.gia.edu/gandg.

Get PDF Issues at

Order Print Back Issues at

10% discount for GIA Alumni and active

THE FRENCH BLUE AND THE HOPE:

n the course of his several visits to India during

THE FRENCH BLUE AND THE HOPE

See end of article for About the Authors and Acknowledgments.

GEMS & GEMOLOGY

Figure 1. The Tavernier

107 ct Cte de Bretagne spinel (originally thought

THE FRENCH BLUE AND THE HOPE

Bapsts book (figure 3) is the primary record. A lead

GEMS & GEMOLOGY

(This rapid exit from the country would have been