VOLUME XXXII

SUMMER 1996

VOLUME 32 NO. 2

T A B L

C O N T E N T S
EDITORIAL
Madagascar: Making Its Mark
Richard T. Liddicoat

FEATURE ARTICLES
Sapphires from the Andranondambo Region, Madagascar
Dietmar Schwarz, Eclzehard I. Petsch, and Tan Kanis
Russian Demantoid, Czar of the Garnet Family
W m . Revel1 Phillips and Anatoly S. Talantsev

Opal from Shewa Province, Ethiopia

Mary L. Johnson, Robert C. Kammerling,
Dino G. DeGhionno, and John I. Koivula

REGULAR FEATURES
Gem Trade Lab Notes
Gem News
Thank you, Donors
Book Reviews
Gemological Abstracts
Guidelines for Authors

ABOUT THE COVER: Perhaps no gem material in modern times carries the mystique of the demantoidgarnet. Found in commercial quantities only in Russia's
Ural Mountains, demantoids were esteemed by scientists and royals alike when
first discovered in the mid-19th century. When Czarist Russia gave way to
Communist Russia, however, demantoids-like other gems-fell out of favor.
With virtually no mining for three-quarters of a century, few of these bright green
garnets entered the marketplace except as part of historical pieces of jewelry like
the fanciful Edwardian "dragonfly" shown here. Today, some efforts are being
made to revive the historical localities and explore for new ones. The second article
in this issue looks at the history of demantoid mining in Russia and reports on the
current situation at the two main demantoid districts, north and south of
Ekaterinburg. The dragonfly pin, platinum over gold with diamonds and rubies,
contains 17 demantoids; the largest are 5.2 m m and 4.7 m m in diameter. The
loose demantoid weighs 8.95 ct. The pin and the stone are courtesy of Michael M.
Scott, Sunnyvale, California.
Photo 0 Harold

a ) Erica

Van Pelt-Photographers, Los Angeles, CA.

Color separations for Gems & Gemology are by Effective Graphics, Compton, CA.
Printing is by Cadmus journal Services, Baltimore, MD.
0 1996 Gemological Institute of America

All rights reserved

ISSN 0016-626X

Editor-in-Chief
Richard T. Liddicoat
Associate Editors
William E. Boyajian
D. Vincent Manson
John Sinkankas
Technical Editor
Carol M. Stockton
Assistant Editor
Irv Dierdorff
e-mail: [email protected]

Editor
Alice S. Keller
1660 Stewart St.
Santa Monica, CA 90404
(310) 829-2991 x 251

e-mail: [email protected]
Subscriptions
Jin Lim Cristina Chavira
(800) 421-7250 x 2 0 1

Fax: (310)453-4478
Contributing Editor
John I. Koivula

Editor, Gem Trade Lab Notes

C. W. Fryer
Editors, Gem News
Mary L.Johnson
John I. Koivula
Editors, Book Reviews
Susan B. Johnson
Jana E. Miyahira
Editor, Gemological Abstracts
C. W. Fryer

PRODUCTION
STAFF

Art Director
Christine Troianello

Production Assistant
Gail Young

EDITORIAL
REVIEW
BOARD

Alan T. Collins
London, United Kingdom
G . Robert Crowningshield
New York, New York

C. S. Hurlbut, Jr.
Cambridge, Massachusetts

P.0.Lebanon, New Jersey

Alan Jobbins
Caterham, United Kingdom

George Rossman
Pasadena, California

John Emmett
Rush Prairie, Washington

Anthony I<. Kampf

Los Angeles, California

Kenneth Scarratt
Bangkok, Thailand

Emmanuel Fritsch
Nantes, France

Robert E. Kane
Helena, Montana

Karl Schmetzer
Petershausen, Germany

C. W. Fryer
Santa Monica, California

John I. Koivula
Santa Monica, California

James E. Sliigley
Santa Monica, California

Henry A. Hanni
Basel, Switzerland

A. A. Levinson
Calgary, Alberta, Canada

Christopher P. Smith
Lucerne, Switzerland

Kurt Nassau

SUBSCRIPTIONS

Subscriptions in the U.S.A. are priced as follows: $59.95 for one year (4 issues), $149.95 for three years (12 issues).
Subscriptions sent elsewhere are $70,00 for one year, $180.00 for three years.
Special annual subscription rates are available for all students actively involved in a G1A program: $49.95, U.S.A.;
$60.00, elsewhere. Your student number must be listed at the time your subscription is entered.
Single issues may be purchased for $15.00 in the U.S.A., $18.00 elsewhere. Discounts are given for bulk orders of 10
or more of any one issue. A limited number of back issues of G d G are also available for purchase.
Please address all inquiries regarding subscriptions and the purchase of single copies or back issues to the
Subscriptions Department,
To obtain a Japanese translation of Gems d Gemology, contact the Association of Japan Gem Tmst, Okachimachi Cy
Bldg., 5-15-14 Ueno, Taito-ku, Tokyo 110, Japan. Our Canadian goods and service registration number is 126142892RT.

MANUSCRIPT
SUBMISSIONS

Gems 6) Gemology welcomes the submission of articles on all aspects of the field. Please see the Guidelines for
Authors in this issue of the journal, or contact the editor for a copy. Letters on articles published in Gems e)
Gemology and other relevant matters are also welcomc.

COPYRIGHT

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 tcchniques (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 editor.
Gems d Gemology is published quarterly by the Gemological Institute of America, a nonprofit educational organization for the jewelry industry, 1660 Stewart Street, Santa Monica, CA 90404.
Postmaster: Return undeliverable copies of Gems ed Gemology to 1660 Stewart Street, Santa Monica, CA 90404.
Any opinions expressed in signed articles are understood to be the opinions of the authors and not of the publishers.

AND REPRINT

PERMISSIONS

MAKING
ITSMARK
Richard T. Liddicoat, Editor-in-Chief
T h e lead article in this issue of Gems e? Gemology describes a new find of gem blue sapphires
on the island of Madagascar. Over the years, we have come to think of Sri Lanlza as the gem
island, but Madagascar certainly vies with it in the variety of fine-quality gem materials that
have been found there. It is interesting to compare the two islands historically as gem sources.
Looking back to 1958, and the 13th edition of G. F. Herbert Smith's famous textbook Gem
Stones, we see that the index to this volume refers to Sri Lanlza 28 times and to Madagascar, 24
times. Both islands are underlain by ancient Precambrian rocks, with an abundance of later pegmatite dikes. Yet Madagascar is nine times the area of Sri Lanlza, and it has not been prospected
as extensively as the smaller island. The potential is enormous.
Over the years, the main gem materials to enter the international market from Madagascar were
magnificent aquamarines and large morganites, plus topaz and tourmaline. (I must confess that
Madagascar is my personal favorite source for the last of these gem materials, because it is the
type locality for liddicoatite, the calcium end-member of the tourmaline group.) Within the last
two decades, important quantities of fine emeralds have been found as well, primarily in the
vicinity of Mananjary, on the island's east coast.
Both ruby and sapphire were among the stones Professor Smith listed for Madagascar in 1958,
but the discoveries up to that date were not of any great significance. Today, however, major
amounts of fine sapphires are coming from the Andranondambo region (including the newer
Antsiermene area). From all reports in the literature and within the trade, this new find is important to the colored stone market, especially in light of decreasing production of blue sapphires at
major localities such as Kanchanaburi (Thailand)and elsewhere. Even more recently, as we report
in a Gem News item this issue, commercial quantities of gem rubies are also now coming from
Madagascar.
Politically, Madagascar was a French territory from 1896 to 1958, when it became a republic of
the French community, under the name Malagasy Republic. It gained independence in 1960.
Since 1975, it has been the Democratic Republic of Madagascar. There was a time during the
heady early days of independence from France when those operating mining ventures were summarily removed from their properties. In recent years, however, the island's government has been
friendlier to mining activities and has even made some efforts to attract former operators back to
the island.
As in many third-world countries, Madagascar's burgeoning population has resulted in the
destruction of much of the forest cover. While ecologists worry, with good cause, about the
future of the fauna and flora of this unique island, the encroachment into the wilderness area has
led to the discovery of new gem deposits and has improved the economic outlook for Madagascar
as an important source of gem minerals. As long as the political situation is encouraging to
would-be prospectors and miners, Madagascar's growing role as a world power in the gem community seems assured. 1-1

Editorial

GEMS & GEMOLOGY

Summer 1996

79

S M P H I ~FROM
S THE
By Dietmar Schwarz, Eckehard J. Petsch, and Jan Kanis

since 1993, large amounts of gemq~1alitysappl~iresfrom the Andranondambo region o f southern Madagascar
have entered the internutionul gem
market. These sapphires, which are
found i n metamorphic sliarn-type
deposits, show a broad range o f gemological and chemical properties, m a n y
of which are similar to those of sapphires from sri Lanka, Myanmar
(Burma), and w e n Kashmir. Most of
the Andranondambo sapphires are
heat treated in Bunglzoli; these can be
separated from their non-heat-treated
counterparts on the basis of i n c l ~ ~ s i o n
features and absorption spectra. Key
factors i n the separation of Andranondambo sapphira jrom synthetic sapphires of different manufacturers are
chemistry (especially the Ga content
of he natural material) and internal
features.

Sapphires from Madagascar

34

adagascar has been a major gem-producing cornn the past) the most importry for many years. I
tant gems have been various pegmatite minerals
such as beryls and tourmalines (see)for example)
Chil~ayama~
1989) and emeralds from schist-type deposits
[Schwarz)1994).Most recently! however! fine sapphires have
emerged onto the international gem marlzet (figure1).
Corundum occurrences i n Madagascar have been
reported occasionally in the literature (for geographic locations! see figure 2). Schmetzer (1986)mentions rubies from
Vatomandry and Gogogogo in the southern region. Chikayama (1989)cites Gogogogo and Ejeda on the southwestern
part of the island (rubies);Amboasary in the southeast! near
Tolanaro (sapphires);and Antanifotsy in the central region,
about 100 lun south of Antananarivo (rubiesand sapphires).
Koivula et al. (1992)describe an unusual type of m~~lti-colored sapphire from a locality called Ianlzarolza~southwest of
Betrolza, in Toliara Province. Non-gem-quality rubies
examined in the Gubelin Laboratory in 1994 reportedly
originated from Ihosyl about 120 lzm north of Betroka (L.
Gentile) pers. comm.l 1994).Whereas none of these localities has had much commercial importance, however! the
Andranondambo (also called Andranondamtso) deposit in
southern Madagascar has turned out to be a significant new
source for fine blue sapphires. This has been of special
interest for the gem trade) because Banglzolz dealers report
that some of the classic sources for high-quality blue sapphires in Kashmir) Myanmar (B~lrma))
and Sri Lanlza are
declining in production.
The much-talked-about Andranondambo sapphire
deposit is not a recent discoveryl although fine-quality sapphires from here first appeared on the world marlzet only a
few years ago (Eliezri and Kremlzow, 1994; Kammerling et
al.! 1995a). As early as 1952/53! French geologist Paul

GEMS & GEMOLOGY

P i p e 1. In only the last

f e w years, Madagascar has
produced nzzmerous fine
sapphires. Although most
of the crystals are small,
some excellent large
stones have also been
recovered. The fine sapphires in this suite are all
reportedly born the
Madagascar deposits. The
16 gradua~edpear-shaped
sapphires in the yellowgold necldace weigh a
total 01 149.87 CL,the sapphires in the earrings
weigh a total of 45.68 ct,
and the ring is set with a
54 ct pear-shaped sapphire. Courtesy of
Mozzawad 1ewelers.

Hibon reported the occurrence of small (up to 10

mm) eluvial sapphire crystals that came from an
area 1 lzm northeast and 2 lzm south of Andranondambo village; the sapphires were found together
with a previously unknown mineral that was subsequently named hibonite (Noizet and Delbosl
1955; Curien et a l l 1956).
About four years agol in 19921local miners and
Malagasy traders first offered on the Antananarivo
marlzet parcels of sapphires in various toncs of
blue, including a millzy blue type (geuda),of which
90% were very sinall crystals, When one of the
authors (EJI?) visited Andranondambo in March
1995 with one of the claim ownersl Chabany! he
learned that rumors first circulated in Fort
Dauphin) now called Tolanarol that a new sapphire
deposit had been found near *e city of Beldyl in
the central portion of southern Madagascar. It was
subsequently shownl though) that the stones were
from farther southeastl near Andranondambo. Thai
merchants soon discovered that the pale inillzy
blue sapphires reacted very well to heat treatment!
changing to an attractive blue that in some cases

Sapphires from Madagascar

was comparable to that of Kashmir sapphires. From

then onl Thail Indian) and other traders rushed to
Madagascar to purchase rough material.
As news of the rich sapphire deposit spread
quiclzly in Madagascarl thousands of gem miners
traveled to the Andranondambo areal leaving behind
their aquamarine and tourmaline mines in other
regions of the island. As a resultJproduction of these
latter gem minerals dropped drainatically. Even in
the Mananjary region! a decline in emerald production was noted. It is estimated that as many as
10,000 miners were aggressively working the new
area at different times, leading to numerous fights
and even some murders. According to various
Banglzolz dealers) since the end of 1994 approximately 100 1% of rough Andranondambo sapphires have
been shipped to Banglzolz monthlyl with as much as
80% of the material ultimately usable for jewelry
purposes (usually after heat treatment). Although
most of the crystals are smalll about 10%-15% are
2-7 ct and crystals as large as 50-60 gramsalthough not entirely gem quality-have appeared in
the marlzetplace. The largest Madagascar sapphire

GEMS & GEMOLOGY

LOCATION AND ACCESS

Situated in the Indian Ocean' Madagascar is the
world's fourth largest island! 1'580 lzrn long and
580 lzrn at its widest point. The Andranondambo
deposit is located at 24'26' E and 4637' S. Travel
from Tolanaro to Amboasary! about 70 lzml is on a
good paved road that follows the southern coastline
(again! see figure 2). From Amboasary! however! the
journey continues on a gravel road north through
Behara and Tranomaro to the village of Andranondambo. The conditions of this latter road are so
poorl however! that it talzes five to six hours to
cover the approximately 150 lzrn from Amboasary to
the mining areal through sparsely populated territory. This trip is possible only with a well-equipped
four-wheel-drive vehicle. The road to Tranomaro
crosses arid land covered by huge forests of cactuslilze succulents that are so dense they are virtually
impenetrable. There is very little rainfall in this
F@re 3. Seen from the air, loolzing north, the
Andranondambo sapphire deposit looks like
Swiss cheese, with hundreds of small shaf~s
made by local miners over a distance of about 3
lzm. The buildings in the far north, near the
river, belong to an overseas mining company.
Photo by E. J. Petsch.

Figure 2,The Andranondambo sapphire deposit is

located in southern Madagascar, east oj the village
of Andranondambo, approximately 150 lzrn by gravel road from Amboasary.

reported to date is a 17.9 lzg piece of rough that was

recently described by Gary DuToit! of the Asian
Institute of Gemological Sciences (AIGS)laboratoryl
as "definitely gem-grade sapphire! a fine gem blue
color1' ("The find of a lifetimel'' 1996).
During the March 1995 visit to the sapphire
deposit! Mr. Petsch's reconnaissance flight in a small
twin-engine plane over the Andranondambo mining
area (figure3)was followed by a Landcruiser safari to
the sapphire deposit to collect first-hand information
and samples. The present study is based on information gathered during this visit! and on the examination and analysis of samples obtained from the
deposit at that time as well as from marlzeting channels in Banglzolz and Switzerland.

Sapphires from Madagascar

GEMS & GEMOLOGY

Summer 1996

40

Figure 4. This geologic map of the

Andranondambo region (adapted from Besairie,
1970) shows the location of the Fort Dauphin
and Tranamaro groups of the Androyen system,
in which he sapphires have been found. Also
shown are the Cretuceous volcanics, younger
granites, and 1JrecumbrianAnosyen granites.

region) and the climate is extremely hot. After

Tranomarol the succulent forests become less dense
and the higher) semi-desert plateau presents a more
pleasant) hilly landscape. There are no major rivers
or lakes in this plateau region, which has an average
altitude of 500 m above sea level, with several hills
rising up to 11200m. The dry climate inalzes it possible to reach the mining area year round.
Andranondambo is a very small village) but the
shanty town that has sprung up near the deposit is
extensive; as noted earlier) at times it has housed as
many as 10)OOO peoplel under the most primitive
hygiene conditions. By March 19951 only about
3,000 miners were still searching for sapphires in
Andranondambo proper. One reason for this is that
the local miners have only primitive equipment)
which makes it impossible to sinlz extremely deep
shafts. Therefore) only a limited number of mhers
can worlz effectively at any one shaft (see "Mining1)
below). Another) perhaps more importantl reason is
the discovery of sapphires neqby, about 10-12 lun
north of Andranondamb~~
at Antsiermene. Amrdmg
to Thomas Banker, of GemEssence Ltd. in Bangkok
(pers. comm.) June 1996))there is a new shanty
town of 3~0004~000
diggers in that area) which is
also responsible for many of the fine southern
Madagascar sapphires that are entering the marlzet.

Sapphires from Madagascar

GEOLOGY AND OCCURRENCE

Serious geologic studies of Madagascar began only
after its annexation by France in 1896, The first geol o p map of the entire island was published in 1900)
on the occasion of the International Geological
Congress. The classic three-voluine work on the
mineralogy and petrology of Madagascar was published by Lacrok (1923).
The southern three-quarters of Madagascar is
occupied by the Precambrian basement complex, of
which the oldest system (> 3 billion years old) is
the Androyen. The Androyen system is subdivided
into three groups, two of which are found in the
Andranondambo area (figure 4): the Fort Dauphin
group, which is composed mainly of hornfels with
cordierite; and the Tranomaro groupl which consists mainly of a varied series of originally sedimentary roclzs that were subsequently subjected to
high-grade (granulite)metamorphisin (Rakotondrazafy et al.) 1996).
The sapphire deposits in the Andranondambo
region occur in the high-grade granulite facies)
metamorphic roclzs of the Tranoinaro group; in this
area) the granulite facies consist of crystalline limestone (marblel figure 5) containing some diopsidel
anorthite-rich plagioclase, and wolla~tonite~
as well
as gneisses and pyroxenites. The granulite belt in
the Andranondambo region is approximately 30
km wide and is sandwiched between two younger
Figure 5, Miners dig through the crystalline
Limestones along the eastern contact of he
Andranondambo deposit. In the bacligroz~ndis
the Precambrian Anosyen granite massif.
Photo by E. 1. Petsch.

GEMS & GEMOLOGY

Summer 1996

The sapphires actually occur in thin veins in

the inetamorphic rock. These have been described
as fine strings! meandering through the roclz, that
seem to start and stop without any real geologic
definition, An American mining engineer working
in the area describes the occurrence as a "pea soup1'
mixture of minerals (T.Banlzer1pers. coinm.l 1996).
H. Hanni suggested that the sapphires formed
locally in nests and poclzets in the reaction zones
between pegmatite dikes and pyroxenite (as reported in Kammerling et al.! 1995b).
Numero~~s
explanations for the origin of sapphires and other varieties of gem corundum in calcareous metamorphic rocks are fo~lndi11 the literature. The explanation applicable to each deposit
depends 011 the extent of metamorphism in the
region! the nature and abundance of mineral impurities (e.g./clay minerals)in the origmal rockl whether
or not additional elements have been added during
the metamorphic event (e.g./inetasomati~m)~
as well
as other factors, the discussion of whch is beyond
the scope of this article.

Figwe 6. At the Andronondambo mining area, it

is dangerous to wall<between the deep shafis,
which are sometimes not more than 50 cm aport.
Note in the sl~aftsthe profile of the cataclastic
structure of the calcareous sapphire-bearing rocli.
Photo by E. J. Petsch.

formations: the extensive Cretaceous volcanic

massif (mountainous mass) that outcrops in the
vicinity of Vohitsimbe in the west (again, see figure
41, and the Precambrian Anosyen granite massif in
the east (figure 51. The area also has experienced
episodes of intrusion by younger granites and the
formation of slzarns (Ralzotondrazafy et al.! 1996).
Clearlyl the rocks of the Androyen system have
been repeatedly subjected to various types of geologic processes (e-g.,metamorphisml igneous activity)
throughout their long hstory. The profile exposed
in the various shafts (see, e.g., figure 6) shows a cataclastic structure (i.e.! a rock texture resulting from
tectonic forces)! which indicates that intense
mechanical forces have crushed the roclzs.

Sapphires from Madagascar

MINING
The irregularly shaped Andranondambo sapphire
deposit covers an area that is at least 3 l a long and
varies in width between 500 and 11000m (again!
see figure 3). From the air, one can easily observe
the location and extent of the exposed mining area.
The northern extremity, crossed by a small river
and visible on figure 3! is the only section of the
deposit where mechanical mining was being donel
by an overseas companyl in March of 1995. Today!
there are a number of Thai groups, a Swiss group! a
French group! and more than one Israeli concern
working in the area! usually in partnership contracts with local residents. Mming concessions have
been granted to some of these g r o ~ ~ with
p s ~ each concession 2.5 km x 2.5 km. Howeverl local miners
often do not observe the boundaries of the concessions. They regularly mine on the land illegally but
with the approval of the local chiefs! who malze their
own often-powerful claims on the basis of ancestral
rights (T.Banlzer)pers. c o ~ n mJune
. ~ 1996).
The local miners work independently in small
groups by sinlzing narrow (2 to 2.5 in in diameter)
shafts as much as 20 to 30 m deep (figures 6 and 7))
which is as far as they can safely dig using manual
mining methods. Fortunatelyl the area is very dryl
so there is no rain or groundwater to cause the collapse of these shafts, which would further endanger
the lives of the miners. The workers excavate the

GEMS & GEMOLOGY

Summer 1996

calcareous rock with 2-m-long crowbars and shovels to open their shafts. The material is hoisted in
buckets to the surface with a rope! and after dry
sorting the waste is carried to huge dumps. Every
group of miners has its own small claim area where
shafts are sometimes no more than half a meter
apart! malzing it very dangerous to wallz over some
parts of the deposit (again! see figure 6).
At the time of Mr. Petschls visit) the deposit
was being worlzed very inefficiently) with some of
the waste material being dumped on unexploited
sections of the mining area (again)see figure 5). To
datel however) no serious effort has been made to
bring in heavy machinery to mine the deposit by
open pit and set up a sophisticated processing
plant. This is largely due to the nature of the veins!
which! as noted above! are relatively thin! discon-

Figure 7. Small groups of miners work at each

shaft, descending by ropes, tree trunks, or crude
wooden ladders as deep as 20-30 m to reach the
gem-bearing veins. Pholo by E. /, Zletsch.

tinuousl and do not appear to follow any set pattern

(T. Banker! pers. comm.l 1996). Consequently, the
occurrence of sapphhe crystals is erratic.
PRODUCTION AND DISTRIBUTION
During his visit! Mr. Petsch saw no more than a
few hundred grams of small gem-quality sapphiresl
most of which varied between 0.2 and 0.5 grams
each. This left the impression that prod~~ction
at
the time was very limited! considering the vast
n ~ ~ m bof
e rminers that had been worlzing the area.
Although larger crystals (2+ grams) have been
reported) not a single crystal of more than 1 gram
was seen during the visit (which consisted of two
days at the mine and several days in the general
area].There are always traders in the mining villagel
buying most of the miners1production at the end of
each day. In fact! some Thai dealers were living permanently in the shanty town. Amornpongchai
(1995)p. 6 ) reported that "aro~lnd100 Thai traders
are in the country buying rough at the moment1'
(that is) at the beginning of 1995).
Most of the Andranondambo sapphire production is sent directly to Banglzolz for heat treatment
and cutting. K. Siu of Tai Hang Gems in Bangkolz
(pers. comm.l 1996) estimates that about 90% of
the gem-quality rough material is submitted to
heat treatment before cutting. Part of the material
is also cut in Israel. The finished goods are marlzeted mainly through channels i n Thailand and
Switzerland.
As noted earlier) dealers in Bangkolz report that
on average 100 1% of rough enters that city inonthly
from the Andranondambo regon! 80 lzg of which is
gem quality. Most of the crystals are small) c~itting
stones less than 4 mm. Nevertheless) as much as
15% of the gem-quality crystals yield cut stones
over 2 ct. Very large crystals-50-60 grams-have
been recovered! but these usually must be cut before
heat treatment to remove the potentially damaging
negative crystals. However! a number of 15-20 ct
cut stones have been reported in the trade and seen
at the Gubelin Laboratory. The largest of the
Madagascar sapphires shown in figure 1 is 54 ct)
and (again as noted above) a 17.9 lzg rough blue sapphire from Madagascar recently appeared in
Banglzolz.
MATERIALS AND METHODS
The test sample consisted of more than 800 nonheat-treated crystals of varying (including gem)
quality that ranged from about 0.2 to 4 ct (see)e.g.)

Sapphires from Madagascar

GEMS & GEMOLOGY

Summer 1996

figure 8). All were purchased by Mr. Petsch during

his March 1995 visit to the mining site. From this
collection, we selected a number of pieces for
gemological research. One or two windows were
polished on 200 non-heat-treated crystals to facilitate testing. Sixty additional crystals were subjected to heat treatment! half of these by T. Hager at
Figure 8. As these crystals from the test sample
show, Andranondambo sapphires have a broad
range of variabiLity in their morphology. The
most common type is the dipyramid, shown here
at the top right and cenLer right. Note also the
dissolution features on the surface of the crystal
on ihe bottom rjght in particular. The crystals
range from 0-46ct (bottom left) to 0.98 ct (center
left). Photo 0 GIA and Tino Hammid.

Sapphires from Madagascar

- -

the University of Maim! Germany! and the other

half by K. Siu of Tai Hang Gems Ltd.! Banglzolz.
Windows were also polished on these heat-treated
samples for examination. In additionl several
parcels of faceted hdranondambo sapphires (about
60 total! ranging from 1 to 6 ct] were obtained though
marlzeting channels in Bangkok and Switzerland for
examination and chemical analysis. These stones had
been heat treated in Banglzolz (the treatment conditions are not lznown).
Refractive indices! birefringence! optic character! and pleochroism were recorded for each of 50
non-heat-treated and 50 heat-treated samples.
Specific gravity was determined hydrostatically on
60 faceted stones and 40 crystals. The fluorescence
behavior was checked for the entire test sample!
more than 800 crystals and faceted stones. Color
and fluorescence of 60 of these samples were
checked before and after heat treatment.
One hundred non-heat-treated sapphires! out of
the 200 polished (windowed) samples, and 100
heat-treated sapphires (40 windowed and 60 faceted
stones) were subjected to spectroscopic examination. Polarized ultraviolet-visible-near infrared
spectra (280 to 880 nm) were run on a P e r l h Elmer
Lambda 9 spectrophotometer. We recorded a total
of 100 spectra [both o [ordinary ray] and e [extraordinary ray]; 50 from samples that had not been heat
treated! and 50 from heat-treated stones). Twenty
samples were measured before and after heat treatment. Infrared analyses were performed on about
20 of the stones with a Pye-Unicam FTIR 9624
spectrometer. A total of about 80 polished samples
and faceted stones were analyzed by means of energy-dispersive X-ray fluorescence (EDXRF)spectroscopy. These analyses were performed on a
Tracor Northern Spectrace 5000 system! using a
program specially developed by Prof. W. B. Stern, of
the Institute of Mineralogy and Petrographyl
University of Basel.
The internal features (growth characteristics]
were studied in all 320 windowed or faceted samples. To identify the mineral inclusions, we had
selected samples polished down until the incl~isions to be analyzed were exposed at the surface.
Analyses of about 40 mineral inclusions were carried out with a scanning electron microscope
equipped with an energy-dspersive spectrometer
(SEM-EDS)at the S W A laboratory! Lucerne. The
mineral inclusions of about 20 polished samples
were examined by Raman spectroscopy at the
AIGS laboratory, in Bangkok.

GEMS & GEMOLOGY

Summer 1996

As mentioned above! some stones in our test

sample were heat treated in Banglzolz and some
were treated at the University of Mainz. Mr. Siu
explained that the details of his heat-treatment
process for the Andranondambo. sapphires depend
on the color of the original material. Pale blue sapphires are heated in a charcoal oven in a reducing
atmosphere. For darlzer crystalsl the material is
first heated in an oxidizing environment. The sapphires are then heated in a gas oven at high temperatures (120U-170UC)1 for varying lengths of
time! depending on the nature of the starting material. The oxidation/reduction can be controlled by
changing the 02/H2 gas ratio. Mr. Hagcr used a
slightly oxidizing atmosphere and temperatures of
about l85O0C for approximately five hours. He
heat-treats the stones in alumina crucibles without
adding A1203 powder.
CHARACTERIZATION OF THE
ANDRANONDAMBO SAPPHIRES
Visual Appearance. A large number of t h e
Andranondambo sapphires showed more or less
well-developed crystal habits (againl see figure 8).
These can be classified into four main types: (a)
dipyramidal (with or without a basal face)! (b)prismatic (always with a basal face)( (c) transition or
and (d)
combination type-prismatic + di~yramidal~
distorted plate-lilze or distorted dipyramidal crystals. By farl the most common type is the dipyramid. Most of the Andranondambo sapphires
showed at least a few crystal faces; rarely were they
entirely irregular or fragments. Some of the crystals
&splayed interesting dissolution features on their
surfaces (again! see figure 8). Before any treatment!
the samples were typically weak to saturated light
blue to darlz blue; almost all of the crystals showed
distinct color zoning.
Once facetedl the heat-treated samples are typically blue with tones that range from medium darlz
to very dark [figure 9); rarelyl they appear almost
blaclz in daylight. Eye-visible color zoning is often
present! but eye-visible internal features such as
minerals or fissures are relatively rare. Cut stones
of less than 1 ct can be quite cleanl even when
examined with a gein microscope at moderate
(2Ox40xJmagnification.
Gemological Properties. The standard gemological
properties for the Andranondambo sapphires (see
table 1) were found to be consistent with corundum in general.

Sapphires from Madagascar

Pleochroism. All heat-treated samples exhibited

distinct to strong dichroisin. In paler non-heattreated crystalsl the pleochroism was sometimes
less distinct. Normallyl the colors seen in the
dichroscope are blue (parallel to the c-axis) and
greenish blue (perpendicularto the c-axis).
Fluore.scence. Most of the Andranondambo sapphires (heat treated and non-heat-treated) were
inert to both long- and short-wave W radiation.
Rarely! we observed a bluish white fluorescence in
non-heat-treated crystals exposed to long-wave W.
Some heat-treated stones showed a challzy blue or

GEMS & GEMOLOGY

Summer 1996

Table 1. Gemological characteristics of sapphires from Andranondambo, Madagascar,

No. samples
(natural/
heat treated)a

Propefly

Clarity

800160

Refractive indices

50150

Birefringence
Optlc character
Specific gravityb
Pleochroism

Fluorescent@

(reaction to
UV radiation)
Optical absorption
spectrum (UVNis)

Chemistry
(trace and
minor elements)

Internal feature9
(growth
characteristics)

50150
50/50
60 faceted
40 crystalsb

Heat treated

Natural (non-heat-treated)
Weak to saturated colors ranging from light to dark
blue. Almost all crystals show distinct color zoning.

Medium to highly saturated blue with medium to

very dark tones. The color zoning, in general, Is
less distinct.

Very clean to heavily included. Most faceted material

slightly included to clean.

Same as non-heat-treated.
Same as non-heat-treated.
Same as non-heat-treated.

Uniaxial negative
3.99-4.01
3.89-3.99
Light greenish blue (perpendicularto c-axis).
Saturated blue; rarely, violetish blue (parallel to c
axis), Sometimes less intense in paler crystals.
Usually inert to long- and short-wave; sometimes
faint bluish white to long-wave; very rarely, a faint red
fluorescence to long-wave UV is seen.
Most pronounced absorption minimum, in general,
around 490 nm (e-spectrum).
Absorption edge, in most cases, at relatively high
values (compared to sapphires from Myanmar and
Sri Lanka): o - 320-335 nm, e - 32Ck340 nm.
Very rarely, absorption shoulder around 320 nm.
Fe203 = 0.12-0.61
Ti02 = 0.01-0.10
Ga203 = 0.01-0.04
Cr203 < 0.01
V205 < 0.01
MnO < 0.01
Strong color zoning, mainly parallel to the basal
face c; sometimes very dark blue or brownish bands,
Prominent growth structures (parallel to the basal
pinacoid c, to various dipyramids [normally n,z],to
the 2nd order prism a, and to the rhombohedron r).
Healing fissures with varying textures; often "frosty"
appearance.
Negative crystals showing a large variation in sizes
and shapes.
Stringers of pinpoints.
Fine hollow tubes.
Mineral inclusions: calcite, apatite, feldspar (plagioclase, K-feldspar), phlogopite, Mg-hornblende,
pyroxene rhedenbergite"), rutile (grains and needles), Fe-mineral (exact nature not yet determined,
probably hematite), Ca- and K-silicates [exact nature
not yet determined).
Tentatively identified (based on visual appearance):
"sulfide," "spinel," thorianiteluraninite.

Same as non-heat-treated.
Same as non-heat-treated.
Light greenish blue (perpendicularto c-axis).
Satured blue; rarely, violetish blue (parallel to caxis). Distinct to strong.
By far, most samples are inert to long- and
short-wave U v rarely, chalky blue or green to
long- and short-wave; a weak red fluorescence
to long-wave UV is extremely rare,
Almost always, the absorption minimum shifts
to 360 or 420 nm (0- and e-spectrum).
Absorption shoulder (around 320-330 nm)
becomes more pronounced and more frequent
than in non-heat-treated sapphire.
Same as non-heat-treated.

Changes after heat treatment:

Mineral inclusions become "turbid9'-translucent
or even opaque; supeficially, they appear glassor mirror-like, some porcelain-like.
Stress fissures are formed around crystal inclusions and negative crystals; these show a mirror
effect (mostly, they have a frosted appearance).
In other healing fissures, the original fluid inclusions become rounded and look like highly
reflective spheres. Rarely, textures are similar to
those observed in some synthetic flux corundums.
Most striking are fine-, medium- or, sometimes,
coarse-grained bands that are grayish white;
these bands may have a bluish gleam.
Color zoning (mostly weak to moderate with
light to medium blue tones), parallel to the basal
pinacoid c, to different dipyramids, to the 2nd
order prism a, and to the rhombohedron r.

Fw "color" and "fluorescence," 60 of the 800 non-heat-treated samples were also tested afler heat treatment.
Some of the crystals contained many mineral inclusions,
Most of the minerals were identified by SEM-EDS; calcite, apatite, phlogopite, and rutile were also found by RAMAN spectroscopy.

green fluorescence in short- or long-wave UV; this

was probably induced by heat treatment, as we did
not obseme it in any of the 800 non-heat-treated
stones we examined (see also Themelis, 1992).
Very rarely, we observed a faint red fluorescence to

88

Sapphires from Madagascar

long-wave U V in both heat-treated and non-heattreated samples.

Internal FeaturesIGrowth Characteristics, Growth
Structures. The dominant crystal forms are the

GEMS & GEMOLOGY

Summer 1996

Fi~wre10. These growth structures and color zoning were common in the Andranondambo sapphires examined (viewperpendicular to the caxis): narrow growth planes parallel to the basal
face c and repetitive sequences of the dipyramids
n and z, Color bonds t11at show different shades
of blue are confined to these well-defined
domains. Magnified 40X.

Figure 11. Also seen were growth stmctwes and

color zoning that consisted of sequences parallel
to the clipyramid n and the rhombohedron r.
Immersion; magnified 6 0 ~ .

dipyramids n { 2 m ] and z {2231]and the basal pinacoid c (OOOl]; the second order prism (1 {11%0}and
the rhombohedron r (10il) may also be of importance. Under the gem microscopel with an immersion liquid! most of the Andranondambo sapphires
showed prominent growth characteristics that
reflected most of the morphological properties that
have been observed macroscopically. These consisted mainly of straight and angular sequences of
growth planes/bands ('lzonesll)parallel to the basal
pinacoid cl the dipyramids n and z (often in repetitive sequences)! the prism a! and the rhombohedron r (see figures 10 and 11).

In some cases! the central zones were also accompanied and delineated by growth structures parallel
to the basal c plane and to the dipyramidal faces.
A few crystals showed a darlz blue border zone
(rim)along the dipyramidal faces. These zones varied in thiclzness on the different faces but! in generall were less than 1 inin. The opposite case-that
is, a large! intensely colored blue central zone (core]
and a narrow near-colorless outer zone-also was
observed. Here! the color zoning was parallel to the
faces of the second-orderprism a (figure 14).When
loolzing parallel to the c-axis of the Andranonda~nbosapphires! we often saw different types of
color zoning. Most frequent were very compact
central zones of intense blue color and hexagonal

Color Zoning. In most of the non-heat-treated crystals, color zoning was very pronouncedl with darkl
intense areas. After heat treatment) color zoning
usually was less distinctl and for the most part the
color bands were light to medium blue. The color
bands were! in general! very narrow (figure 12)and
tended to be concentrated in certain areas of the
crystal. Even without magmficationl the color zoning in the non-heat-treated crystals was often seen
as strong! well-defined domains of a darlz blue
(almost blaclz) or darlz brownish blue that were typically delineated by faces parallel to the basal pinain repeticoid cf the dipyramids n and z ji~onnally~
tive sequences; again) see figure 10)andl rarely! the
prism a. We commonly saw darlzer central zones
(tube-!cone-! col~imnar-Ior pipe-lilze in appearance;
seel e.g./ figure 131 with more-or-less well-defined
outlinesl similar to those observed in Mong Hsu
rubies (Smith and Surdez! 1994; Peretti et al.! 19951.

Sapphires from Madagascar

Figure 12, The color bands in Andranonclam bo

sapphires, here shown parallel to he basal pinacoid, were typically veiy narrow. Magnified 8 0 ~ .

GEMS & GEMOLOGY

Summer 1996

Figure 13. Commonly seen in the Andranondambo

sapphire crystals was a central zone with a cone- (as
here) or pipe-like appearance. Such central zones are
delineated b y planes parallel to the basal c face and
by repetitive dipyramidal (n,z) faces. These central
zones are similar in appearance to those observed in
Mong Hsu (Myanmar)rubies. Magnified 60x.

outline, as well as hexagonal "growth rings" that

varied in color intensity and thickness.

Twinning. Pronounced twinning is rare. In most

cases, the presence of twin planes and intersection
lines was only suggested.
Mineral Inclusions. The minerals we observed in
Andranondambo sapphires were randomly disFigure 14. In a few samples, a large, intensely
colored blue core was surrounded b y a narrow
(< 1 mm) near-colorless "rim."Magnified 2 0 ~ .

Sapphires from Madagascar

tributed. The most common mineral appeared to

be calcite (chemical analyses showed an almost
pure Ca-carbonate). The calcite crystals varied
greatly in size [up to 1 mm) and shape. Some were
rounded and some were elongated (figure 15),
whereas still others were plate-like. Frequently, the
calcite inclusions were well-developed, presenting
various morphologies. Although usually transparent and colorless, the calcite crystals themselves
sometimes contained numerous small (usually
fluid) inclusions, which gave them a turbid white
appearance.
Other colorless and transparent crystal inclusions were identified as apatite. They occurred
most often as irregularly rounded prisms (figure 16))
or as elongated hexagonal prismatic crystals. They
also appeared as plate-like crystals with many
faces. Two types of feldspar were identified.
Plagioclase most commonly occurred as colorless,
transparent grains of varying size (normally less
than 0.5 mm) and irregular morphology, but it was
sometimes seen as whitish crystals or as "filling
material" in fissures. Typically, the plagioclase was
of anorthite composition. The I<-feldspar crystals
appeared as transparent, often slightly yellow
grains and as small, irregularly shaped crystals.
Colorless, transparent, needle-like to long-prismatic or stallz-like crystals distributed without any orientation were identified as the amphibole Mghornblende.
The second-most-common included mineral
was biotiteJphlogopite (figure 17). As a rule, it
formed isolated orangy brown transparent platelets
or mineral aggregates. The chemical analyses actually distinguished two types of mica: In addition to
the "normal" biotite/phlogopite composition, some
of the analyzed crystals showed distinct Ti concentrations (a few weight percent TiOi). Sometimes,
the mica crystals were intergrown with other mineral inclusions (most often, with calcite and pyroxene). A Ca/Al-silicate that normally occurs in the
form of greenish brown grains or irregularly shaped
crystals belongs to the pyroxene group (hedenbergite). The chemical analyses proved some crystals
to be Ca- and K-silicates, but we have not yet determined their exact nature. Some grayish to black
grains with metallic luster showed only the element titanium in the chemical analyses. With
Raman spectroscopy, we identified these as rutile.
We also saw rutile in the form of oriented needlelike crystals of varying length and as plate-like
inclusions.

GEMS & GEMOLOGY

Summer 1996

rqpre 13. Colorless cuicite crystals were the

most common mineral inclusions identified in
the Andianondambo sapphires. They occurred
rounded or elongated (as shown here) or even
plate-like. Magnified 80x.

A few mineral inclusions could not be analyzed chemically because they were too small or
were in faceted gems that could not be polished to
bring the minerals to the surface. These inclusions
were identified on the basis of their visual appearance under the gem microscope: (a)opaque, brownish yellow grains with metallic luster (sulfide);(b)
opaque, blaclz, cube-like crystals with metallic luster, sometimes accompanied by stress-fissures,
Figure 16. Also identified in the sapphires from
southern Madagascar were irregularly rounded
prisms of apatite, like the sample shown here at
the top center with some colorless to near-colorless calcite crystals and a few opaque black
grains (probablyrutile or spinel). Magnified 60x.

Sapphires from Madagascar

which are probably uraninite or uran-thorianite

(see also "Discussion" section); and (c) other
opaque grains with a grayish blaclz metallic luster
and more-or-less rounded outlines, which probably
belong to the spinel group.
In one non-heat-treated sample, we observed
fine yellowish brown needles and pinpoints that, in
reflected light, had a strong metallic luster; these
were concentrated in some well-delineated areas of
the host crystal (figure 18).Although we could not
conclusively identify these needles, chemical analyses in areas of the host crystal where the needles
reached the surface revealed high Fe concentrations. This indicates that the needles are an Fe
mineral (possibly hematite, which has been identified-e.g., by Barot et al.,1989-in sapphires from
Kenya). We also observed, but could not identify,
long, fine needles that ran parallel to the basal face
of the host crystal.

Negative Crystals and Fluid Inclusions. Relatively

common in the Andranondambo sapphires were
so-called negative crystals and their fluid fillings.
These inclusions varied greatly, from minute particles and flat, disk-like cavities to elongated, irregularly shaped cavities and large forms delineated by
many faces or showing bizarre shapes. Some of the
larger, elongated negative crystals were accompanied by tails or "seams" of smaller negative crystals (figure 19). Small negative crystals were often
arranged in rows, giving the appearance of strings of
Figure 1 7. Biotite/phlogopite was the second
most common mineral observed i n the southern
Madagascar sapphires, often appearing as ininera1 aggregates. Photomicrograph by E. 1. Giibelin;
magnified 66x.

GEMS & GEMOLOGY

S u m m e r 1996

Figure 18. One crystal had concentrations (in

well-defined areas) of fine needles and "pinpoints" of a yellowish brown color and a strong
metallic luster in reflected light. Chemical analyses of the areas with high concentrations of these
needles cutting the surface had distinctly higher
Fe contents than the rest of the stone. Thus, the
needles probably represent an Fe mineral (probably hematite). Magnified 60X.

The nature of the material(s) filling the negative crystals is still not well known. Most of the
negative crystals looked like single-phase fluid
inclusions. However, they are probably two-phase
inclusions in which one liquid phase occupies
almost the entire cavity. One large three-phase negative crystal contained a colorless liquid, a brownish liquid, and a gas bubble. Rarely, we saw grayish
black platelets with strong luster (graphite?)in the
fluid inclusions; more common, however, were
needle-like inclusions that we have not yet been
able to identify.
The fact that the negative crystals in Andranondambo sapphires are often concentrated on
growth planes can have significant consequences
for heat treatment. As these inclusions commonly
rupture at high temperatures, they may result in
the breakage of the host crystal or cut stone.

Color Bands. The most striking internal feature in

the heat-treated Andranondambo sapphires examined were the grayish white (rarely, grayish brown)
fine-, medium-, or coarse-grained bands (figure 211.
Because we did not see such bands in non-heattreated Andranondambo stones, we believe that
their formation is directly related to the heat-treatment process. In the Andranondambo sapphires,
these HT-bands (the designation HT is given to
emphasize that these structures formed during heat
treatment) were observed in most of the samples
examined; sometimes, especially in small stones,
they were not easily seen. These bands were often
accompanied by color zoning. Fine-grained HTbands often appeared very compact, giving the
impression of three-dimensional "block structures." Others were less compact and appeared
fainter and more delicate. With the microscope,
using oblique fiber-optic illumination, we frequently observed a bluish "gleam" to the HT-bands.
Some very compact bands appeared brown in transmitted light.
-

pearls, or they were dispersed over former growth

planes (figure 20). Sometimes, the negative crystals
on the growth planes were accompanied by small
crystals of different minerals. The faces of larger
negative crystals (especiallythe flat ones) frequently showed a strong reflection (mirror effect) or striation of the cavity walls. All types of negative crystals may be accompanied by relatively flat healing
fissures with random orientation.

Figure 19. Negative crystals were common in the

Andranondu~nbosapphires. Here, large, elongated negative crystals are accompanied by a
"seam" ( ' nailer ones. Magnified E O X ,

Healing Fissures. Healing fissures were also quite

common in the Andranondambo sapphires. They
were typically flat, rarely wavy, with net-like,
(rarely) grain-like, or tube-like "textures." They
might consist of isolated rounded or elongated fluid
inclusions, of larger negative crystals, or sometimes
of "stringers" of small rounded or slightly elongated particles. Even in non-heat-treated stones, many
healing fissures had a distinct frosty appearance. In
the healing fissures of some heat-treated stones, the

Sapphires from Madagascar

GEMS & GEMOLOGY

Summer 1996

sapphires showed grayish white to white needlelike inclusions. Because the samples in which we
observed these needles were seen only after heat
treatment, we could not establish whether the needles were formed during treatment or were present
before but changed their appearance during heating.

Figure 20. In some of the southern Madagascar

sapphires, negative crystals were dispersed over a
former growth plane. Magnified 60x.

original fluid inclusions were rounded and looked

like highly reflective spheres, or they had textures
similar to those seen in some synthetic flux corundum~
(figure22).
Other Internal Features. In the Andranondainbo
sapphires, we also observed: (a)stringers of pinpoint-like inclusions in different arrangements
(rarely, a sheaf-like appearance), some of which
were very delicate (almost cobweb-like];and (b)fine
hollow tubes (sometimes needle-like), which have
been reported to contain polycrystalline material
(by H. Hanni in Kammerling et al., 1995b).Any of
the inclusion minerals described in the preceding
sections may be accompanied by (c)wing-like healing fissures, unhealed stress fissures, or long, thin,
tube- or canal-likeinclusions (figure23). Sometimes,
we also observed swarms of delicate, divergent tubes
and stringers, which were almost identical to the
"comet tails" seen in many Kashmir sapphires.
After heat treatment, many of the originally
transparent crystal inclusions turned turbid or even
opaque. In addition, the surface of a mineral inclusion sometimes changed dramatically: Many crystal faces developed a porcelain-,or glass-like appearance, whereas others became so reflective as to display a mirror effect. Sometimes, a surface acquired
a frosted or even crust-like appearance. In some
mineral inclusions, we observed the formation of
small tension fissures at the contact with the host
crystal. Several of the heat-treated Andranondainbo

Sapphires from Madagascar

Absorption Spectra. Non-Heat-Treated Sapphires.

A typical absorption spectrum for most of the nonheated Andranondambo sapphires is shown in figure 24a. The spectrum in figure 24b was seen in
non-heat-treated material as well as in heat-treated
samples. The broad absorption bands around 570
nm in the o-spectrum and around 700 nrn in the espectrum are responsible for the blue color (see,
e.g., Ferguson and Fielding, 1971 and 1972;
Schmetzer and Bank, 1980; Fritsch and Rossman,
1987, 1988a and b; Moon and Phillips, 1994).The
main difference between the two spectral types is
the presence/absence of an absorption shoulder in
the 320-330 nm range. The absorption minima for
both spectral types lie around 360,420, and 490 nm
in both the o- and e-spectra. As a rule, in non-heattreated Andranondambo sapphires the deepest
absorption minimum for the e-spectrum was
observed at 490 nm. In both heat-treated and nonheat-treated samples with the 320-330 nm absorption shoulder, the absorption minimum almost
always shifted to 360 or 420 nrn.

Figure 21. The most striking internal feature of

the heat-treated Andranondambo sapphires
examined were bands of fine-, medium; or
coarse-grained textures and grayish white (rarely
grayish brown) color. The formation of these HTbands is directly related to the heat-treatment
process. Magnified 80x.

GEMS & GEMOLOGY

Summer 1996

absorption trend, beginning at about 700 nm and

extending into the near-infrared region as a result
of Fez+ <-> Fe3+ charge transfers (Krebs and
Maisch, 1971; Ferguson and Fielding, 1971 and
1972; Schmetzer and Banlz, 1980), may become
more intense (figures 24 c and d). Very strong Fe3+
absorption features at 375 and 387 nm, which are
considered typical for sapphires associated with
basaltic rocks [see figure 24e, the spectrum of a sapphire from Antanifotsy, in central Madagascar),
were not observed in the Andranondambo sapphires. For comparison, the absorption spectra considered typical for sapphires from Kashmir, Burma
(Myanmar),and Sri Lanka are presented in figure 25.

1
Figwe 22. Healing fissures i n some heat-tieated Andranondambo sapphires were reminiscent of the textures observed in some fluxgrown synthetic wrundums. Magnified 60x.

Heat-Treated Sapphires. Most of the heat-treated

samples had absorption spectra similar to that
shown i n figure 24b. The absorption shoulder
around 320-330 nm is more or less pronounced,
and in general the absorption edge lies between 290
and 300 nm. In non-heat-treated material, the
absorption edge was usually located between 320
and 340 nm. It must be emphasized, however, that
a strong 320-330 nm absorption shoulder was not
always present in heat-treated Andranondambo
sapphires. The shoulder may be very weak or even
completely absent. As illustrated in figure 24 a and
b, for the non-heat-treated and for most of the heattreated Andranondambo sapphires, the absorption
bands around 570 and 700 nm were dominant.
These bands are ascribed to charge-transfer transitions between Fez+ and Ti4+ (see, e.g., Krebs and
Maisch, 1971; Ferguson and Fielding, 1971 and
1972; Schmetzer and Banlz, 1980; Moon and
Phillips, 1994).In rare cases, however, an additional

Sapphires from Madagascar

Chemical Analysis. Table 2 gives the (semi-quantitative) EDXRF results for trace and minor elements
in 80 sapphires from the Andranondambo deposit
and in four sapphires of basaltic origin from Antanifotsy, central Madagascar.
Compared to the sapphires from Kashmir, which
are associated with pegmatite intrusions (Levinson
and Cook, 1994), and those from Myanmar and Sri
Lanka, which are of metamorphic origin, the
Andranondambo "skarn sapphires" have similar
Fe203 + Ti02 and V205 + Cr203 + MnO + Ga203
contents (although the latter may have a slightly
broader range of variability; these figures are based
on preliminary data from an ongoing Gubelin laboratory research project).The main difference between
Figure 23. It was not uncommon in the
Madagascar sapphires to see minerals like this calcite crystal accompanied by a tube-like inclusion.
Magnified 40x.

GEMS & GEMOLOGY

Summer 1996

..

Mil

400

50C
600
700
WAVELENGTH [nm)

SOU

9<Kl

HO

4W

500
600
7M
WAVELENGTH (nm)

800

SOU

Figure 24. These ore the most common

absorption spectra recorded in the Andranondambo sapphires: (a)non-heuttreated;(b)heat treated and some nonheat-treatedmaterial; and (c)and (d)
heat treated with dominant F& <->
F&- charge-transferabsorption in the
near-infrared. Spectrum (e)was recordedfzom a "basaltic" sapphire from
Antanifotsy, central Madagascar; it
shows a typical strong Fe3+absorption
in the ultraviolet at 375 and 387nm.

w S 4 M 5 W 6 0 0 ~ 8 0 0 W
WAVELENGTH (mi)

WAVELENGTH (om)

these and the basalt-associated Antanifotsy sapphires

lies in the iron concentration, which is higher for the
sapphires found in basaltic deposits. We did not see
any definite correlations (for example, higher Ti
contents where Fe concentrations are higher) in the
Andranondambo sapphires. Also, there was no evidence of a simple correlation between iron content
and color intensity.
DISCUSSION
Geology and Occurrence. The Andranondambo
deposit-like the occurrences in Kashrnir, Myanmar,
and probably Sri Lanka-results from metamorphic
conditions. The sapphires of the Andranondambo
mining region in southeast Madagascar were

formed in U-Th skarns that belong to the Pan

African granulitic formations. Ralzotondrazafy et
al. (1996)defined two main stages of crystallization
in the skarns: The minerals aluminous diopside,
C03-scapolite, titanite or spinel, and thorianiteluraninite-as well as corundum-are characteristic of stage 1; one of the main mineral reactions
of stage 2 is the crystallization of hibonite at the
expense of corundum and spinel. Many of the
Andranondambo sapphires are well-developed
crystals (figure 81, which often show interesting
dissolution features at their surfaces. The dissolution phenomena are the result of changes in the
mineralogic environment during stage 2 of slzarn
metasomatism.

Figure 25. The three spectra shown here are typical for (a) Kashmir, (b)Bzrma (Myanmar), and (c) Sri
Lankan blue sapphires.
I

I'^\

M O w e w o e w 7 M w e s o o

Sapphires from Madagascar

s 4 0 0 i o c w o 7 0 1 n o s a

WAWN61H (null

WA-~

GEMS & GEMOLOGY

Imn)

Summer 1996

The mineral inclusions in the Andranondambo

sapphires-predominantly Ca minerals, especially
calcite-reflect the nature of the surrounding metamorphic carbonate and calc-sihcate host rocks and
the paragenesis of these sapphires. Other minerals
found in sapphires from this deposit include: apatite,
feldspar (plagioclase and K-feldspar),phlogopite, Mghornblende, pyroxene (hedenbergite),rutile, Ca- and
K-sihcates, spinel, thorianite/uraninite, and sulfide.
The following minerals related to sapphire genesis
were identified in sapphire-bearing rock samples
obtained at Andranondambo for this study: calcite,
anorthite, wollastonite, phlogopite, hibonite, scapolite, pyroxene, K-feldspar, and amphibole. Behier
(1960)also described spinel as occurring with sapphire in the Andranondambo area.
The lack of a continuous increase or decrease
in the color intensity of the growth bands evident
in most of the Andranondambo sapphires (figures
10-12) means that there was no continuous
increase or decrease in the coloring agents present
in the nutrient fluid during crystal growth. Rather,
the sequences of alternating color zones indicate a
multi-stage growth with the several growth phases
characterized by changes in the genetic environment (e.g., variations in the composition of the
nutrient).
Comparison to Sapphires from Other Localities.
Most of the Andranondambo sapphires that reach
the gem market are heat treated. The fine-grained
bands induced by heat treatment in the Andranondambo sapphires are not exclusive to this locali-

TABLE 2. EDXRF analyses of trace and minor elements

in sapphires from the skam-associated Andranondambo
deposits and the basalt-associatedAntanifotsy deposits.
Oxide
Fez03
Ti02
Ga203c
v2

Cr203

MnO

Content (wt.%)
Andranondamboa
Antanifotsyb
1.80-2.09
0.04-0.08
0.01-0.02
< 0.01
< 0.01
< 0.01

0.12-0.61
0.01-0.1 0
0.01-0.04
< 0.01
< 0.01
< 0.01

a Concentration ranges in 80 samples from the Andrandambo deposits.

Concentration ranges in 4 samples from the Antanifotsy region.

In 10 synthetic sapphires from Kyocera, examined by the senior author,
the Ga concentration was <0.01 wt.% (GayOi); in eight Seiko synthetic
sapphires and in 36 Chatham flux-grown blue and orange synthetic sapphires, the Ga contents were below the detection limit of about 0.005
wt.% (Ga2O3)(see also Schrader and Henn, 1985).
f~

96

Sapphires from Madagascar

ty. Clouds of fine-grained particles following internal growth planes have been described in non-heattreated sapphires from southern Vietnam (Smith et
al., 1995).They are also laown in sapphires from
other basaltic sources (Kiefert and Schmetzer, 1987)
and in sapphires from Kashmir (Schwieger, 1990)
and Sri Lanlza (based on Gubelin laboratory data).
Because of their very small size, we could not identify the particles forming the HT-bands in the
Andranondambo samples to determine if they are
different from those seen in non-heat-treated sapphires from other deposits.
T h e appearance during heat treatment of
unhealed tension fissures, or of healing fissures
that extend into the host crystal, relates to differences in the expansion coefficients of the minerals
involved. The newly formed fissures show, in general, two modes of appearance: (a)flat and circular
with a "seam"; and (b)healing fissures with frosted
textures. At the Gubelin laboratory, we have also
seen both types in heat-treated sapphires from
other sources. The absence of such fissures, as well
as the presence of negative and/or mineral crystals
that do not show evidence of rupture or other damage, provide good indication that a cut stone has
not been subjected to heat treatment (figure26).
Non-heat-treated samples, in general, show
absorption spectra (figures 24 a and b, respectively)
that are similar to those of Kashmir sapphires or,
more rarely, to sapphires from Myanmar (figure 25).
In the first case, the absorption edge is normally
positioned at higher values (o - 320-335 nm; e 320-340 nm), compared to sapphires from
Myanmar or Sri Lanka (again, see figure 25).
The heat-treatment process influences the
absorption behavior of Andranondambo sapphires.
However, the appearance of the absorption shoulder at about 320-330 nm cannot automatically be
related to heat treatment, both because the absorption shoulder was not seen after treatment in many
samples and because the shoulder was observed in
some unheated stones (figure 24b). This variable
absorption behavior may be due to different heating
conditions or to differences in the original materials. In most cases, the absorption edge in heat-treated stones was still high (>315 nm); only in a few
samples was it lower (< 305 nm). Consequently,
the position of the absorption edge and the presence (or absence) of an absorption shoulder around
320-330 nm is of only limited use in establishing
whether an Andranondambo sapphire has been
heat treated. In comparison to the original material,

GEMS & GEMOLOGY

Summer 1996

the broad absorption bands in the 500-800 n m

range do increase as a result of heat treatment.
Empirical experience at the Gubelin Laboratory
indicates that, in general, the absorption spectra of
heat-treated Andranondambo -sapphires are more
similar to those of most Sri Lanlzan sapphires
(when no absorption shoulder is present; see figure
25) or of Burmese sapphires (when a pronounced
shoulder is developed; see figure 25). Whereas the
absorption spectra of some Andranondambo sapphires (figures 24c-d) in the 500-800 n m range
resemble those of sapphires originating from basaltassociated deposits, the Fe3+ absorption i n the
ultraviolet, at 375 and 387 nm, is never as intense
as in typical basaltic sapphires (figure 24e).
The concentrations of Fe, Ti, and Ga in the
Andranondambo sapphires sometimes reach values
higher than those that are considered typical of sapphires from the metamorphic deposits in Myanmar
and Sri Lanlza (up to 0.61 wt.% Fefi, 0.10 wt.%
TiOo, and 0.04 wt.% Ga203).The iron contents of
the Andranondambo sapphires, however, are lower
t h a n those measured in t h e four (basaltic)
Antamfotsy sapphires, but they may overlap those
of sapphires from other basaltic deposits (compare,
e.g., Guo et al., 1992; Smith et al., 1995). T h e
chemical behavior of the Andranondambo sapphires is probably related to the special genetic conditions that exist during slzarn metasomatism.
The range of mineralogical and gemological
properties shown by Andranondambo sapphires is
relatively broad, but within this range are specimens with spectral and chemical characteristics as
well as internal features similar to those of sapphires
from Sri Lanka, Myanmar, and even Kashmir.
Internal features of non-heat-treated Andranondambo sapphires that resemble those of Sri
Lanlzan or Kashmir sapphires include prismatic
amphibole crystals; black, cube-like grains; colorless,
irregularly rounded crystals; and short rutile needles.
However, in our sample, we did not see the long,
fine rutile needles that are typical for Sri Lankan sapphires (compare Giibelin, 1973; Webster, 1983;
Anderson, 1990), or the nestslbands of short, dull
rutile needles that are more common in Burmese
sapphires (Gubelin, 1973). Nor did we see those
internal features that are most typical of Kashmir
sapphires (the velvety appearance that is caused
mainly by clouds, lines, strings, and flake-like
arrangements of dust-like inclusions; or the association of zircon, tourmaline, pargasite, plagioclase,
allanite, uraninite [compare Schwieger, 19901).

Sapphires from Madagascar

Figure 26. This 0.24 ct sapphire from southern

Madagascar shows no evidence of rupture of the
fluid inclusion or stress cracking around the small
solid inclusions. This suggests that the stone was
not subjected to high-temperature heat treatment.
Stone courtesy of I. Z. Eliezri; photomicrograph by
fohn I. Koivula, magnified 3 0 ~ .

After heat treatment, many Andranondambo

sapphires lack specific inclusion features. Small
stones ( < 1 ct) may be quite clean, even when
examined with the microscope. For such stones,
growth characteristics are of little or no diagnostic
value, and identification should be based on absorption spectra and chemical data. It appears that the
reaction to UV radiation is a useful additional test
to separate the Andranondambo sapphires from
their Sri Lankan counterparts. Sri Lanlzan sapphires
very often show quite strong red or orange-red fluorescence to long-wave UV-radiation.
T h e properties for t h e few samples from
Antanifotsy appear to be consistent with sapphires
from other basaltic deposits (see, e.g., fiefert and
Schmetzer, 1987). The distinction between these
basaltic sapphires and the Andranondambo sapphires is relatively easy based on the inclusion features, absorption spectra (figures 24a-e), and chemical properties (table 2).

Separation from Synthetic Sapphires. Andranondambo sapphires can be fairly easily separated from
synthetic blue sapphires grown by different methods. The synthetics (Verneuil, Chatham, Kyocera,
Seiko) normally have very little or no Ga (see the

GEMS & GEMOLOGY

Summer 1996

footnote to table 2). In addition, most of the internal features seen thus far in Andranondambo sapphires are quite different from those observed in
laboratory-grown sapphires (compare, e.g., Kane,
1982; Gubelin, 19831.
CONCLUSIONS
The appearance and properties of the sapphires from
the Andranondambo deposit are related to the
metamorphic geologic (slzarn) environment in
which these sapphires formed. Some Andranondambo sapphires, at least in some aspects, resemble
sapphires from Sri Lanlza, Myanmar, or Kashmir.
On the one hand, non-heat-treated stones revealed
some inclusions that are similar or almost identical
to those seen in some Sri Lanlzan or Kashrnir sapphires. On the other hand, many (heated and notheated) Andranondambo sapphires showed "Burmatype" absorption spectra. In most cases, however,
these Madagascar stones could be separated from
sapphires of other localities by means of, in addition
to inclusion features, absorption spectra in combination with chemical data.
The separation of heat-treated from non-heattreated Andranondambo sapphires is easy when the
so-called HT-bands are present. Additional features
that indicate that the stone has been heat treated
are changes observed in the appearance of many
inclusions and, to some extent, the absorption
spectra.
Although some internal features of Andranondarnbo sapphires may be similar to those observed
in certain synthetic sapphires (especially the healing fissures in heat-treated Andranondambo stones
and those seen in flux-grown synthetics), the overall inclusion scenes should make confusion unlilzely. For sapphires that lack (typical) inclusions, a
quite reliable separation is possible based on Ga
content: The known synthetic sapphires have very
little or no Gal whereas the Madagascar stones can
have quite high Ga values.
Precise production figures for the sapphire
deposits of Andranondambo i n southern

Madagascar are not available; nevertheless, we

believe that thousands of kilograms of these sapphires have reached the gem market since 1992193.
For the trade, this has been important because
these sapphires could, at least in part, compensate
for the inconsistency of production from the traditional localities. In addition, a certain percentage of
the Andranondambo material represents an alternative to "Burma-type" sapphires, and some stones
have even been compared to Kashmir sapphires. As
the relatively recent discovery of sapphires at the
Antsiermene perimeter, less than 12 km north of
Andranondambo, indicates, the regional geologic
conditions in this part of the island favor the occurrence of other slzarn-associatedsapphire deposits.

Acknowledgments: The authors are grateful to Prof.

W. Hofmeister and to Tobias Hiiger, Institute of
Gemstone Research, University of Mainz, Germany,
for help with the discussion of genesis and occurrence, as well as for some mineral identification i n
samples from the Andranondambo deposit. Kenneth
Siu, Tai Hang Gems Ltd., Bangkok, together with
Tobias Hager, heat treated the Andranondambo sapphires. Dr. A. Perretti, Adligenswil, and Dr. H. D. von
Scholz, SUVA, Lucerne, Switzerland, did SEM-EDS
analyses on mineral inclusions. Roman spectroscopy
of mineral inclusions was performed by Gary D~zToit
of the Asian Institute for Gemological Sciences,
Bangkok. Prof. W. B. Stern, o f the Institute of
Mineralogy and Petrography, University of Basel,
developed the program used for trace-element analyses. George Bosshart, of the Giibelin Gemmological
Laboratory helped edit the original manuscript.
Dr. Karl Schmetzer, Petershausen, Germany, and
Christopher P. Smith, of the Gubelin Gemmological
Laboratory, provided discussion of the growth structures of the Andranondambo sapphires. The largest
part of the test sample was provided b y the firm
/ulius Petsch Jr., Idar-Oberstein, Germany.
Additional samples came from Kenneth Siu, W.
Spaltenstein, Golden M. Co., and the AIGS laboratory, all of Bangkok. Unless otherwise noted, the photomicrographs are b y Dietmar Schwarz.

REFERENCES
Amompongchai A. (1995)Mining in Madagascar. Asian Jewellery,
April 1995, p. 6.
Anderson B.W. (1990) Gem Testing, 10th ed. Rev. by E. A.
Jobbins, Butterworths, London.
Barot N.R., Flamini A,, Graziani G., Gubelin E J . (1989)Star sap-

98

Sapphires from Madagascar

phire from Kenya, Journal of Gemmology, Vol. 21, No. 8, pp.

46 7-473.
Behier J. (1960) Contribution A la mineralogie de Madagascar.
Annales gdologiqzzes de Madagascar, No. 29, Tananarive.
Bcsairie H. (1970) Carte geologique de Madagascar: Feuille

GEMS & GEMOLOGY

Summer 1996

Ainpanihy No. 8. Service gcologique de Madagascar,

Tananarive.
Chikayama A. (1989) Gemstones from Madagascar. In 22nd
International Gemmological Conference, Italia, 1989,
Proceedings ATTI, CISGEM, Milan, Italy, p. 29.
Curien H., Guillemin C., Orcel J., Sternderg M. (1956) La
hibonite, nouvelle espfece mineral;. Convtes Rendus, Vol.
242, pp. 2845-2847.
Eliezri I.Z., Kremkow C. (1994)The 1995 ICA world gemstone
mining report. ICA Gazette, December, pp. 1-19.
Ferguson J., Fielding P.E. (1971)The origins of the colors of natural yellow, green, and blue sapphires. Chemical Physics
Letters, Vol. 10, No. 3, pp. 262-265.
Ferguson J., Fielding P.E. (1972)The origin of the colors of natural yellow, blue, and green sapphires. Australian Journal of
Chemistry, Vol. 25, pp. 1371-1385,
The find of a lifetime (1996)JewelSiam, June-July,pp. 86-87.
Fritsch E., Rossman G.R. (1987)An update on color in gems.
Part 1: Introduction and colors caused by dispersed metal
ions. Gems a) Gemology, Vol. 13,No. 3, pp. 126-139.
Fritsch E., Rossman G.R. (1988a)An update on color in gems.
Part 2: Colors involving multiple atoms and color centers.
Gems el Gemology, Vol. 24, No. 1, pp. 3-15.
Fritsch E., Rossman G.R. (1988b) An update on color in gems.
Part 3: Colors caused by band gap and physical phenomena.
Gems a ) Gemology, Vol. 24, No. 2, pp. 8 1-102.
Giibelin E. (1973) Innenwelt der E d e l s t e i ~ ~ABC
e . Verlag,
Zurich.
Giibelin E. (1983)Identification of the new synthetic and treated
sapphires. Journal of Gemmology, Vol. 18, No. 8, pp.
677-706.
Guo J., Wang F., Yakoumelos G. (1992)Sapphires from Changle
in Shandong Province, China. Gems a) Gemology, Vol. 28,
No, 4, pp. 255-260,
Kammerling R.C., Koivula J.I., Fritsch E. (1995a) Gem news:
Sapphires from Madagascar. Gems a) Gemology, Vol. 31,
No. 2, pp. 132-133,
Kammerling R.C., Koivula J.I., Johnson M.L. (1995b)Gem news:
The sapphire deposit in southern Madagascar. Gems d
Gemology, Vol. 31, No. 4, pp. 283-284,
Kane R.E. (1982) The gemological properties of Chatham fluxgrown synthetic orange sapphires and synthetic blue sapphires. Gems a ) Gemology, Vol. 28, No. 4, pp. 255-260.
Kiefert L., Schmetzer K. (1987) Blue and yellow sapphire from
Kaduna Province, Nigeria. Journal of Gemmology, Vol. 20,
No. 718, pp. 427442.
Koivula J.I., Kammerling R.C., Fritsch E. (1992) Gem news:
Sapphires from Madagascar. Cems el Gemology, Vol. 28,
NO.3, pp. 203-204,
ICrebs J.J., Maisch W.G. (1971) Exchange effects in the optical

Sapphires from Madagascar

absorption spectrum of ~ e in~A1203.

+
Physical Review B,
Vol. 4, No. 3, pp. 757-769,
Lacroix A. (1923)Minhlogie de Madagascar. Challamel Ed.,
Paris.
Levinson A.A., Cook F.A. (1994) Gem corundum in alkali
basalt: Origin and occurrence. Gems es) Gemology, Vol. 30,
No. 4, pp. 253-262.
Moon A.R., Phillips M.R. (1994) Defect clustering and color in
Fe,Ti:oi-%03. Journal of the American Ceramics Society,
Vol. 77, No. 2, pp. 356-367.
Noizet G., Delbos L.A. (1955) Etude geologique des schisles
cristallins de LiAndroy Mandrareen. Service Geologique,
Tananarive, Madagascar.
Peretti A., Schmetzer I<., Bernhardt H.J., Mouawad F. (1995)
Rubies from Mong Hsu. Gems es) Gemology, Vol. 31, No 1,
pp. 2-26.
Rakotondrazafy M.A.F., Moine B., Cuney M. (1996). Mode of
formation of hibonite within the U-Th skams from the granulites of S-E Madagascar. Contributions to Mineralogy and
Petrology, Vol. 123, No, 2, pp. 190-201.
Schmetzer K., Bank H. (1980)Explanations of the absorption
spectra of natural and synthetic Fe- and Ti-containing corundums. Neues lahrbuch fur Mineralogie, A b h d l u n g e n , Vol.
139, pp. 216-225.
Schmetzer I<. (1986) Natiirliche und synthetische Rnbine. E.
Schweizerbarttsche Verlagsbuchhandlung Nagele und
Obermiller, Stuttgart, Germany, p. 76.
Schrader H.W., Henn U. (1985) Uber die Problematik der
Galliumgehalte als Hilfsmittel zur Unterscheidung von
natiirlichen Edelsteinen und synthetischen Steinen.
Zeitscl~riftder Deutschen Gemmologischen Gesellschaft,
Vol. 34, No. 314, pp. 152-159.
Schwarz D. (1994) Emeralds from the Mananjary Region,
Madagascar: Internal features. Gems es) Gemology, Vol. 30,
No. 2,pp. 88-101.
Schwieger R. (1990) Diagnostic features and heat treatment of
Kashmir sa~vhires.Gems a ) Gemology, Vol. 26, No, 4, pp.
267-280.
Smith C.P., Surdez N. (1994)The Mong Hsu ruby: A new type of
Burmese rubv.
, ,IewelSiam. Vol. 4, No. 6, December-January,
pp. 82-98.
Smith C.P., Kammerling R.C., Keller A., Peretti A., Scarratt
K.V., Khoa N.G., Repetto S. (1995).Sapphires from southern
Vietnam. Gems ft)Gemology, Vol. 31, No. 3, pp. 168-186.
Themelis T. (1992)The Heat Treatment of Ruby and Sapphire.
Gemlab Inc. [No city or country given]
Webster R. (1983) Gems, Their Sources, Descriptions, a n d
Identification, 4th ed. Rev. by B. W. Anderson, Butterworths,
London.

..

GEMS & GEMOLOGY

Summer 1996

99

CZAR OF THE

GARNETFMILY

By Wm. Revel1 Phillips and Anatoly S.Ta1antsev

Demantoid, green andradite garnet, was

discovered i n the Central Ural Mozzntains
of Russia i n the mid-19th century. A
favorite of the Czar's court, demantoid
was ano~hervictim of the 191 7 Bolshevik
Revolution, when mining of this and other
Russian gems was halted. Today, however, independent miners are recovering
notable quantities of stream-worn demantoid pebbles from two major districts:
Nizhniy Tagil, about 115 k m north, and
Sisser~slz,about 75 lzm s o u ~ h of
,
Ekaterinburg. As a resz~lt,these distinctive bright "golden" green to dark green
garnets are re-emerging i n the gem marl ~Although
.
cut stones continue t o be
small for the most part, a number of fine
demantoids over 1 ct have been seen.

(T

he popular garnet family is one of the most prolific

mineral groupsJ with a family tree rooted primarily in
five common end-member garnets and branching to a
spectrum of colors and gem varieties. HistoricallyJ garnet
was loved for its deep) rich red colorl and the word pyrope is
derived from the Greelz word meaning 'lfire-lilzell (Dana,
1958).
The green garnets are of a younger generation and the
lldarhgsJJof the family. Demantoid emerged first! in the 19th
century! when it entered the royal courts of Czarist Russia
from the valleys of the Ural Mountains. Pwged in the early
20th century by the Bolshevilz revolutionl demantoid all but
disappeared from the international gem scene. In recent
decades) it was largely replaced by its younger sib& tsavorite! from the East African bush cotlntry. In a new Russia)
howeverl the demantoid sites are again becoming active! with
a trickle of green fire reentering the market (figure1).
Demantoid (Ca3Fe2[SiO4I3)is the chromium-green gem
variety of andradite! but it usually exhibits a yellow overtone due to intrinsic ferric iron. With a hardness slightly
less than 7! demantoid is not a good ring stone. However!
its high index of refraction [1.89)and dispersion [0.057; diamond is 0.044) malze it a gem of great brilliance and fire for
other jewelry purposes.
The following account looks at the early history of
demantoid as a gem material and its present status with
regard to localitiesl geologyl productionJ and marketing! as
demantoid reemerges from virtual obscurity into a promising future. Locality information is based largely on the
authors1own experience and knowledge of the Urals region.
A review of demantoidJs place in the garnet family helps
explain its distinctive propertiesl especially in contrast to
tsavoritel with which it has been compared in the txade.

HISTORY
About 1853) children from the settlement of Elizavetinslzoye [also spelled l'ElezavetinslzayaJ1)lsouthwest of the
larger village of Nizhniy Tagil (also known as Nazhniy-

100

Russian Demantoid Garnets

GEMS & GEMOLOGY

Summer 1996

Figure 1. A number of fine

demantoids from Russia
have appeared in the gem
market recently. These
unusually large demantoids
(l.64-5.40 ct) range from
"golden" green to a deep
"emerald" green. Stones
courtesy of Gebr. Henn, IdarOberstein, Germany; photo
GIA and Tina Hammid.

Tagilsk) in the Central Urals [figure 2), found unusual "grass"-green pebbles (probably similar to those
shown in figure 3) in the heavy spring runoff of the
Bobrovka River (Eichmann, 1870; Samsonov and
Turingue, 1985). Local jewelers identified the green
gem as chrysolite (olivine, [(Mg, FehSi041-generally
known by gemologists as peridot). Mineral collecting was very stylish in late-19th-century Czarist
Russia, especially among the nobility, and mineral
collectors soon converged on Nizhniy Tagil in
search of specimens of this gem. This "Ural chrysolite" (which also was called "Bobrovsk emerald,"
"Uralian emerald," and "Siberian chrysolite") soon
appeared in the jewelry shops of Moscow and St.
Petersburg, mostly as small, fiery calibrated stones
framing enamel work or the larger gems of the Urals,
such as pink topaz or beryl.
Nils von Nordensheld [also spelled "Nordenski6Id"), a Finnish mineralogist who first gained
fame in the Urals with his identification of the new

Russian Demantoid Garnets

mineral phenakite [Be2,Si0 4), arrived at Nizhniy

Tagil a year after the initial demantoid discovery.
He remained for a year and a half studying copper
deposits in the area. P.V. Eremeyer (also spelled
"Eremeev"), his friend and biographer, suggests
that von Nordensheld never visited the actual site
where the green stones were found, but rather
examined those in several collections. On the basis
of these examinations, von Nordensheld declared
the green gem a new mineral and certainly not
chrysolite.
On February 20, 1864, von Nordensheld described the beautiful green gem before the St.
Petersburg Mineralogical Society as green andradite
garnet colored by a small chromium content
(Eremeyer, 1871). The fact that andradite has the
highest R.I. and dispersion of any of the garnets
explained the unusual brilliance and fire, and he
proposed the name demantoid [i.e., "diamondlike") for the little "green diamonds" (Clark, 1993,

GEMS & GEMOLOGY

Summer 1996

101

Figure 2. The main two

demantoid localities
described in this article
are reached by paved
and unsurfaced roads
fiom Ekaterinburg, the
largest city of the Central
Urals in Russia. The
northern-Nizhniy Tagil
(A)-district is near
Elizavetinskoye, about
35 lzm by road southsouthwest of N W y
Tagil. T%esouthernSissertslz (B)-district lies
between Poldneva ya and
Verlzhniy Ufaley, about
35 lim by road south of
Polevslz'oy. The numbers
on the roads represent
the number of kilometers
between the distance
marlzers (stars).

102

Itussian Demantoid Garnets

p. 176/ gives 1878 as the year the term was introduced and first published). The name was accepted
subsequently by Russian mineralogists and jewelers! but it was initially rejected by the local peoplel
who continued to call it "Ural chrys01ite~~~
as
demontoid so~~nded
too much like a word that was
vulgar in the local dialect.
A second discovery was made about 75 lun (90
lzm by road) south of Elzaterinburg in the Sissertsk
(also spelled " S y ~ e r tand
~ ~ " S y ~ s e r t District
~~)
(also
referred to as the Polevslzloy District or Poldnevaya
District) on the Chusovaya and Chrisolitlza Rivers
of the west slopes of the Ural Mountains. This area
produced gems equall or superiorl in quality to
those at the first locality (Church) 1879).
Demantoid was very popular in Russia from
about 1875 to 1920 (figure 4). It was even incorporated into some of the fabulous creations of Peter
Carl Faberg6 and other court jewelers. Although
most of these gems were used by the Russians/
who preferred the brownish or yellow-green stones
(R. Schafer! pers. comm.! 1996j1a few-including
the less popular (and less brilliant) intense green
gems-were exported to the European marlzet at
exorbitant prices. Edward VII of England favored
green gemsl and demantoid entered the "Belle
Epoque1I(see cover and figure 5).
Serious efforts have been made to find demantoid deposits in similar geologic environments elsewhere in Russia. In 1980, one was reported at a site in
the Kamchatlza Peninsula of eastern Siberia. The
stones are usually small ( 2 4 mmj but of good quality. Several deposits of yellow andradite (sometimes
referred to as lltopazolite~ll
although this term is in

Figure 3. These pebbles, in the form of abraded dodecuhedra, are typical of the alluvial demantoids found
in the Central Urals. The stone on the left clearly
shows the spray of radiating "horsetail" fibers that is
characteristic of demantoid garnets, From left to
right, they weigh 6.63, 4.81, and 4.62 ct. Photo by
David W. Hawkinson, BYU Museum of Art.

GEMS & GEMOLOGY

disfavor with modern gemologists)and uvarovite (the

chrome analog of andradite]also came to W t . Minor
occurrences of demantoid also have been reported
from Zaire! Korea! Sri Lanlza (gem gavels]! California
(San Benito County; Payne! 1981)! and south-central
Mexico (Wilson! 1985). Even some stones from the
Ala Valley of Italy are green enough to be considered
demantoid. Nevertheless! demantoid has largely
remained a gem of the Ural Mountains (Samsonov
and T u n g e l 1985).Crystals from all of these other
localities are rare! seldom inore than a few millimeters! usually very pale! with cut stones over one-quarter carat rare indeed.
With the onset of the Bolshevilz Revolution in
1917! gems went out of vogue in Russia! along with
other symbols of wealth and royalty! and Soviet
resources turned to mineral production more in
demand by industry. Localities of emerald and other
beryls were mined for their berylliuml diamond was
mined for industrial applicationsl and the tailings
piles of chrysoberyl mines were hand-picked for

Figure 4, Popular in Russian jewelry from 1875 to

1920, demantoids provide a field of green for this
antique star brooch manufactured i n Russia.
Courtesy of A La Vielle Russie, New Yorlz City;
photo b y Nicholas DelRe.

molybdenite. In the early stages of World War II!

Joseph Stalin moved the Soviet heavy industry to the
Central Urals! away from immediate Nazi invasion.
The Urals produced planes! tanlzs! and guns for a
desperate nation; demantoid and other gems for personal adornment seemed unimportant.

Russian Demantoid Garnets

Figure 5. Demontoids were particularly popular

i n late 19th- and early 20th-century Edwardian
jewelry. This yellow-gold and silver 19th-century
dragonfly pin is set with two relatively large
demantoids-1.30 and 1.03 ct-in the main body,
as well as assorted smaller demantoids (2,80ct
tola1 weight) throughoz~tthe balance of the piece,
together with diamonds and two rubies, Courtesy
of the Susan Clarlz Gallery of Gem k t ,
Vancouver, BC, Canada.

LOCATION AND ACCESS

Two major districts in the Central Ural Mountains
of R ~ ~ s shave
i a historically yielded demantoid garnet: Nizhniy Tagil (along the Bobrovlza River) and
Sissertslz (near Polevslzloy and Poldnevaya). These
regions are alternating dense forest (sprucel pine!
aspen and birch) and open meadows! with much of
the area marshland (figure 6).
The Nizhniy Tagil district is near the tiny village of Elizavetinskoye! about 115 lzrn north-northwest of Elzaterinburg (figure 7). It is best accessed
from there via Nev'yansk and Nizhniy Tagil over
paved roads. This district contains two deposits: (1)
the placer on the Bobrovlza River (Bobrovslzaya
Placer)! which runs for about 2 lzrn through
Elizavetinslzoye; and (2)a primary! in situ! deposit
at the head of the Bobrovlza River (Tochilny Kluch)!
which is the source of the alluvial demantoid.
The Sissertslz district lies between Poldnevaya
and Verkhniy-Ufaley near the Korlzodin railway station! which is about 75 lzrn south-southwest of
Elzaterinburg (figure8).All but 4 lun of the road from
Elcaterinburg to the site (90 lzm]! via Polevsk'oy, is

GEMS & GEMOLOGY

Summer 1996

paved. In this district there are two primary! in situ,

deposits: at Kladovka (point I in figure 8) and at
Korlzodin (point II in figure 8).In addition! there are
five river placers: Bobrovka (area l l figure 8J1
Zyachiy Log ("Hare Creelzll' area 2)! Ufaleylza [area
3)! Chrisolitka (area 4j1 and Kamenuschka (area 51.
(Note that the Bobrovka area in the Sissertslz district should not be confused with t h e larger
Bobrovka River that is in the Nizhniy Tagil district. J
In 1985)a small demantoid deposit was discovered in the southern part of the Arctic Urals on the
Hulga River. This new deposit contains both in
situ demantoid and a small placer. Crystals are gem
quality) relatively large (6-8 min)! and good green
color. The site is difficult to reach and has not been
studied.
In 1995! another demantoid deposit was reported in the Arctic Urals on the Hadata River. The primary deposit is on the Saum-Kevpyroxenite massif
and is accompanied by a placer deposit about 1 lzm
long. It is scheduled for exploration and study in
the summer of 1996.

Fiszre 6. There is no formal mining site at liizhniy

Tagil. Rather, alluvial pebbles of demantoid are
recovered from stream beds in a wmplex of low,
mature hills covered by dense forest-largely
birch, pine, aspen, and spruce-and, as shown
here, open meadows.

Figure 7. Demantoid was first discovered in

gravels of the Bobrovlia River near the small
settlement of ELizaveti~~slioye,
in what is now
called the Nizhniy Tagil district. Primary crystals are also recovered from the peridotite
intrusion ~021th
of the village, at Tochilny
Kluch, where the demantoid forms in thin
chysotile veins in serpentinized pyroxenjtes.

GEOLOGY AND OCCURRENCE

In the Central Urals) primary deposits of demantoid crystals occur in both major districtsl as shown
in figures 7 and 8. At Nizhniy Tagill serpentine
lenses about 1.5 km long by 200-300 m wide occur
within ultrainafic (pyroxenite-peridotite]intrusions, and are cut by veins of coarse-grained olivine
[chrysolite) and minor dolomite. At Tochilny
Kluch (="Creek'l)lthe demantoid crystals appear to
form around tiny grains of chromite in the highly
fractured contact zones between chrysolite veins
and serpentine, Ultramafic rocks-such as peridotite, pyroxenitel and their alteration product! serpentine-usually contain high concentrations of
chromium and are the source both of chromite
(FeCr204Jand of the chromium for a plethora of

GEMS & GEMOLOGY

Summer 1996

Cr-bearing minerals, including emerald, alexandrite, and demantoid.

Because demantoid crystals are relatively soft
and brittle, they will not withstand the rigors of
extensive stream transportation. Alluvial pebbles
of demantoid (again, see figure 3) are found largely
in the sandy gravels of the shallow headwaters of
small streams fed by large springs (Sainsonov and
Turinge, 1985). Demantoid in the Nizhniy Tagil
district is recovered from Pleistocene river gravels
cut by the active Bobrovlza River; the principal
deposit measures about 500 m along the river valley and is 20 to 100 m wide (again, see figure 7).
The productive sandy, red gravels (sand-size to
gravel-size grains and pebbles of the available rock
types, stained red by iron oxides from the weathering of the iron-rich rocks) are as much as 2.5 m
thick. They lie on an eroded surface of Paleozoic
volcano-sediments, and are covered by several
meters of detritus and soil. The best horizons are
near the base of ancient terraces; these may contain
100 g of demantoid per cubic meter, 80% of which
are pebbles of 4 mm or less. A second deposit on
the Bobrovlza (alsowithin the placer area marked in
figure 7) is even larger (2.5 l a by 50-60 m); much
of the Pleistocene sand and gravel lies below the
present stream cut and may reach a total thickness
of 6 m. Only the modem river bed has been exploited for demantoid crystals, which are poorly formed
dodecahedra (110) and may reach 5-6 cm, although
such large crystals are very rare.
The geology and deposits of the Sissertsk
(Poldnevaya)district are similar to those of Nizhniy
Tagil (again, see figure 8).At both primary deposits,
demantoid occurs in thin (1.5-2 cm) chrysotile
veins in serpentinized pyroxenites. Thus far, these
primary sites have produced only mineral specimens. In all of the five placer deposits, demantoid
is recovered from the lowest gravel bed (the basal
bed), with minor amounts from sand bars.
MINING AND PRODUCTION
Mining specifically for demantoid in Russia has
been very erratic. Before 1915, the stones were usually obtained as a by-product of platinum mining of
the placers at Elizavetinskoye .and gold mining of
the placers in the Sissertslz district. Since then, the
demantoid deposits have been worked mostly by
private miners (figure 9), operating illegally, who
search river gravels or dig pits up to 5 m deep and
then wash and screen the pit gravels. Only the contract companies have licenses to dig, and even

Russian Demantoid Garnets

f Placer deposits
I.
BobfOvka
2. Zayochiy log
3. Ufaleyka
4. Chrlsoinka
5. Kamenuschka

-0.

IWtd

&"&

wg

R o a d
I

Schial-Qnotos

ftaihplw

I
Wollle
Pymxenlte
Ssrpenllne

Figure 8. In the Sissertsk demantoid district, the garnets occur as primary crystals in serpentinite pods in
uhamafic pyroxenites and as detrital pebbles in
stream gravels a few 'kilometers west of the Kladovka
railway station or north of the Korlzodin station.

those may be for exploration only. Illegal digging is

a dangerous profession, as the authorities are
always alert and local competition is keen. Miners
have been subject to arrests, threats, and even
shootings.
In the 1970s, government geologists studied
the Nizhniy Tagil-Elizavetinslzoye deposits for
commercial development, and for three to five
years thereafter, the A 0 Uralquartzsamotsvety
("Ural-quartz-colored stones" Company) attempted
hydraulic mining. Recovery was poor (40%-60%),
and much high-quality rough was lost in the tail-

GEMS & GEMOLOGY

Summer 1996

105

ings. However, most of the demantoid was ultimately recovered by the local people, who work
over every dump and tailings pile.
A very crude estimate would suggest that, prior
to 1990, a total of perhaps 200 kg of demantoid
rough had been mined; the total remaining commercial reserve of the Elizavetinslzoye placers has
been estimated at 2,000-3,000 kg. In 1993, the
deposit was licensed to a metallurgical company
( A 0 NTMK [Nizhniy Tagil Metallurgical Kombinat]) for prospecting, but no commercial production
has resulted. The license expires in 1997, and the
company is desperately seeking foreign financing.
The Korkodin-Chrisolitlza deposit was explored
in the late 1980s, and a license was purchased by a
private conglomerate (TOO "Grani") from the
Chelyabinsk region. A production company was
formed, and some small-scale organized mining
began in 1994; however, no commercial production
has yet been reported. Reserves at the KorlzodinChrisolitlza deposit are estimated at 5,000 kg.
The Kamenuschka deposit, about 5 km north
of Korkodin, is unexploited, unlicensed, and
remains available for development. Trenches cut
by "bandit diggers" expose the bedrock and reveal
demantoid in drusy cavities, which are valued both
for jewelry and as mineral specimens. No commercial production has been officially reported, and the
Figure 10. These five stones (0.24-0.35 ct) were
selected from a collection of about 50 small cut
stones to represent the range of demantoid colors.
On the basis of qualitative EDXRF analysis, the
bright green stone on the far left contains much
more chromium than any of the others, and the
near-colorless stone on the far right contains much
less. The three center stones all have about the
same chromium content, between that of the previous two samples. By comparison, a medium-green
tsavonte garnet analyzed revealed more chromium
than all but the brightest green demantoid here,
but i t also had 10 to 100 times more vanadium
than chromium. Photo by David W. Hawkinson,
BYU Museum of Art.

Figure 9. In the Urals, most of the demantoids are

recovered b y local villagers who illegally search
the river gravels. Here, a Russian woman examines gravels from the Bobrovka River.

reserves at the Kamenuschlza deposit are estimated

at roughly 2,000-3,000 kg.
Unofficial estimates place total demantoid rough
production for 1995 at about 8 kg (Elizavetinslzoye5 kg, Poldnevaya-3 kg). About 40% of this total,
some 16,000 carats, is recovered as cut stones. Less
than 10% of the latter group, about 1,600 carats, is
represented by stones of one carat or more.
Little more can be said at this time about the
present status or future development of demantoid
garnet in the Ural Mountains of Russia, except that
interest is growing, foreign investors are welcome,
and proposals to develop the deposits are under consideration. However, all of the areas containing
demantoid, except Korkodin, supply drinking water
to local cities, so the serious development of any one
of them could become an environmental concern.
MATERIALS AND METHODS
The senior author selected from his collection of
about 50 Russian demantoids five small cut stones
(0.24 to 0.35 ct.) that ranged from the deepest green

106

Russian Demantoid Garnets

GEMS & GEMOLOGY

Summer 1996

to almost colorless (figure 101. In addition, for comparative spectral analysis he included one yellowish brown andradite from Coyote Front Range,
Inyo, California; one medium green tsavorite from
east Africa; and one colorless grossular from
Wakefield, Canada.
Qualitative energy-dispersive X-ray fluorescence (EDXRF)analysis was performed on these
samples at Brigham Young University, solely to
determine the presence or absence of chromium in
the five samples. The UV-visible spectra were
obtained by the senior author with a HewlettPackard HP8452A diode array spectrophotometer,
also at BYU, on one medium-green demantoid as
well as on each of the tsavorite, andradite, and
grossular samples described above. Details of analyses are available on request from the senior author.
DESCRIPTION OF THE DEMANTOIDS
Demantoid garnet is gem-quality green andradite
(Ca3Fe2,[SiO4I3),
usually very near the ideal andradite composition, 97.02 wt.% to 99.67 wt. % andradite (Stockton and Manson, 19851, with minor
chromium contributing the valued green color, and
traces of aluminum, titanium, vanadium, and
sometimes manganese. Demantoid ranges from
yellowish or brownish green to "golden" green (figure 1l), and-the rarest-"emerald" green (again,
see figure 1).The gemological properties are consistent with those for other garnets (see table A-1 in
Box A), with the exception of the unusually high
R.I. (1.89)and dispersion (0.57).As noted earlier, for
the most part demantoids are small, less than 1 ct.
Although the authors have heard of at least one
faceted stone over 21 ct (S. Fesenlzo, pers. comm.,
1996),this is extraordinarilyrare.

Figure 11. One of the most distinctive demantoid

colors is this bright yellowish or "golden" green
color. These two Russian den~antoids,1.36 ct
(round brilliant) and 1.08 ct, are courtesy of Mayer
a ) Watt, Maysville, Kentucky. Photo 0 G1A and
Tino Hammid.

Cause of Color. Cr3+ substitution for Fe^ in octahedral Y sites (again, see Box A) is responsible for
the rich "grass" green of demantoid, which is
superimposed over the yellow overtone contributed
by the intrinsic ferric iron of andradite (figure 12).
Cr-bearing demantoid shows red through the

Russian Demantoid Garnets

Figure 12. These W-Visiblespectra for a

medium green tsavorite garnet from East
Africa and a medium yellowish green
demantoid garnet from the Central Urals
clearly separate the two green garnets and
also illustrate the greater influence of
chromium in the demantoid. The spectra
shown for samples of andradite and
grossulai that lack significant chromi~un
and vanadium, illustrate the effects of
these chromogens on these garnet types.
The andradite is a brownish yellow sample from Coyote Front Range, Inyo,
California, and the colorless grossular is
from Wakefield, Canada.

GEMS & GEMOLOGY

Summer 1996

107

BOX A: Demantoid's Place in the Garnet Group

The general formula for the garrtet g m p is
X3Y:;[$iO4]^ where X = divalent ions with eightfold
coordination, primarily Ca2+, Mg2+, I'e2-1-,and Mn%

arid Y trivalent ions with sixfold coordination, primarily A^+,FGt,

V3*, and Oft*. Nature combines X
and Y elernenta to form red gamete depending on the
avallabtlity of elementsand the pressure and temperature of formation; it seldom, if fever, fomaa pure, endmember garnet {k,
one in which only one ion m u pies th6 X fite and one the Y elte].'within the garnet
group, thesixmost Important end membera, grouped
by series, aye; [I]uvaxCmte (C.iiCr2[Si04~l,
grossular
ICaaAh.[Siwland
, andradite [C&FwSiO.&] 1the
calcium series);and {1\ pyrope [MggAlilSQIaJ, aiman.
dine [PC^~A~+S~O&I,
and spessartlne [MnaAIA
ISiQil.il Recently, a seventh garnet species in the .Cia
series, goldmaniti (CagVj[Si04b],has gai&.d significance for gemologists, as it forms solid solution with
grossulat to produce teawrite.
Demantotd is anandradite garnet. Complex solid
solution is p$ibIe between grossular and ahdraditc,
and individual members of this series (notably, dernantold) may con& minor CriOi in partial solid solution with ~varavite,theif chrottiiuni analog.
The grccn gem varieties tsavorite anddemantoid
'

both He near the idci~lend-member coinp6sltlons,

respectivelyt of fitosealar arid andfadtte [Stocktop and
Manson, 192351. However, tritemediate (yellow, not
Cr-containiflg)gcossulai-andradltegems recently have
beep noted (Hurwit et al., 1994; Johns<met al., 1995).

Chelsea filter in proportion to the amount of

chromium present. It typically shows no color zoning; however, the color is patchy (i.e., not homogeneous), and the senior author has observed intense
green surrounding tiny chromite grains that appear
to act as nucleation points for the deinantoid crystals (figure 13).This is also graphic evidence that
dissolution of chromite frees Cr to contribute to
the green color of the demantoid.
EDXRF analysis of the stones shown in figure
10 confirmed that the intense color was an indication of high chromium content, although the
stones of the medium color range showed less variability in chromium than in apparent color. None
of these five stones showed measurable vanadium.
The single East African tsavorite revealed the presence of both chromium and vanadium, but 10 to
100 times more of the latter than the former.
Both the demantoid and the tsavorite showed
two major absorption pealzs (again, see figure 12)-

108

Russian Demantoid Garnets

Selected optical and physical properties of gamete in

theCa series are presented in table A.1. Tlie individual
specles, and their relationship to demaniold, are discussed below.
Andradite f p m a complete solid solution with ~VXWIlar, and intermediate compositions are common
(Stockton and Manson, 1983; Griffen, 1992)^however,
gem-quality specimens of both andrddite and grossular
oapri in nature, usually with compositions near their
ideal end-members!Andradite is intrinsically colored as
a consequence of ferric iron [@+I in sixfold cooidination IY ~ i t e )which
,
produces yellow (Loefflerand
Burns, 19761, Yellow or yeUow'peen andradite gems
(once refen-ed to "topazolite," a term now in disfa~ rare.l attractive but rare (Gill, 1978; Webstet, 1983).
Demantoid results when Cr3+from solid solution with
uvarovite $ubstitutes for F& in the Y sites and superimposes a green color over the intrinsic yellow,
Schorlomlte [ C a m , F e ) s O u \ is a black yinet colored by Ti4+and Mn2+ in combination with
intrinsic ferric iron ( ~ h i l l l and
~ a Griffen, 1981) that
occurs in.solid solution with andradite to yield

lrmelanUe,"a Ti-rich black andradite. These black

garnets have little gem application except in mourning jewelry, which was popular with the late
Victorians, as an alternative to jet or black onyx.
Uwovite ia a beautiful green gamet In small sizes;
with larger specimens so dark as to appear almost

one in the blue and violet wavelengths, with lesser

absorption in the orange and yellow range-and
transmit green and red wavelengths. Also evident
in figure 12, the tsavorite showed strong absorption
pealzs at 426 and 608 nm, which are attributed
largely to V3+ modified by lesser Cr3+. Unfortunately, the two vanadium absorption peaks
almost coincide with the chromium absorption
doublet (Loefflerand Bums, 1976);the slight shoulder on the right side of the 426 nm peak and the
asymmetry of the 608 nm peak may be due to
chromium. The transmission valley is centered on
green. Demantoid typically shows very sharp,
strong absorption at 438 nrn (blue-violet)and weak
absorption at about 594 nm (orange);these represent the characteristic absorption of C9+, which
normally consists of two strong absorption pealzs
(Loeffler and Burns, 1976).The ferric iron (Fe3+)
inherent to andradite amplifies the 438 nm chromium peak in demantoid and, when Cr^ is minor,

GEMS & GEMOLOGY

Summer 1996

Table A-1. Selected properties of garnets in the ugrandite (uvarovite-grossular-andradite)

series.
Garnet species/

R.I.

S.G.

Hardness

Dispersion

1.880-1.886
1.880-1.889
1,798-1.868
1,731-1.760
1.739-1.744
1.821
1.752-1.782

3.77-3.88
3.80-3.88
3.71-3.01
3.40-3.78
3.57-3.65
3.75
3.63-3.70

6<A
6%
7%
7
7

0.057
0.057

gem variety
Andradite a h c
Demantoid a
Uvaroviteaob
GrossulaP*
Tsavoritea
Goldmanited
Intermediate (Mali)e

'From Rouse (7986).

^Calculated value from McConnell(1964).
"From Stockion and Manson (1985).

black. It occurs mostly as drusy layers of tiny dodecahedral crystals in fractures in chromite. It is sometimes used for jewelry, showing tiny, bright green
scintillation points. Uvarovite is typically associated
directly with chromite spinel (FeCr204).

Grossular is colorless when pure; the addition of transition-element impurities IF$+, Mn3+, C$+,V^) produces a wide variety of colors. For example, the hessonite variety (colored by F$+) ranges from brownish
yellow to orange-red, and manganese varieties are
pink.
Tsavorite (also lltsavolite")is gem-quality transparent green grossular in partial solid solution with goldmanite, which contributes vanadium as the primary
source of color. It is a recent addition to the garnet family, having been discovered in the 1960s in Kenya and

demantoid transmits essentially all colors except

violet. This yields strong yellow (red and green add
to form yellow) with lesser blue, so the stone is
green with a strong yellow overtone. As the Cr3+
content increases, the absorption peak at 594 nm
becomes both higher and broader, absorbing orange
and most of the yellow and red wavelengths. A
similarly enlarged peak at 438 nm may absorb
more short wavelengths, so that only green wavelengths are transmitted. Thus, Cr-rich demantoid
may be "emerald" green without any modifying
hue. A third high-energy Cr^ absorption in the
ultraviolet may be responsible for the red fluorescence of most Cr-colored gems.
Internal Features. The most distinctive internal
feature of demantoid is its characteristic "horsetail" inclusion (again, see figure 13)which was present, whole or in part, in most of the Russian
demantoids examined by the authors and which is

Russian Demantoid Garnets

0.028
0.028

"From Moench and MeyrwiIz (1964).

"From Johnson et a/. (7995).

introduced to the gem community in 1974 (Bridges,

1974). As much as 3 wt.% V203 has been recorded in
some stones (Muijeet al., 1971).Some specimens are colored solely by vanadium, but others contain a subordinate contribution from chromium and may even appear
red through the Chelsea filter. Tsavorite ranges from
intense "emerald" green (themost valuable) through yellow-green to almost colorless. Good-colorgem tsavorites
over 3 ct are very rare (Rouse, 1986, p. 100).

Goldmanite is a dark green to brownish green garnet

that is much rarer (and much less attractive) than any
of the other Ca garnets. It was first described in 1964
(Moench and Meyrowitz) from specimens found in
Laguna, New Mexico, where it derives from the contact metamorphism of a uranium-vanadium ore
deposit in a sandstone-limestone host rock.

found in no other green gem. In the gemological literature, the "horsetail" generally is referred to as
hair-like byssolite (an obsolete name for asbestiform amphibole, usually of actinolite-tremolite
composition) fibers that diverge from a focal point,
usually a tiny opaque crystal of a spinel-group mineral, probably chromite or magnesiochromite.
Recently, however, the horsetail fibers emanating
from the chromite grains have been identified as
serpentine (chrysotile) by Dr. A Peretti (pers.
comm., 1996).Sometimes the fibers form a dense,
eye-visible bundle (figure 14)or cone stained orange
or brown by ferric oxide. Each demantoid crystal
appears to nucleate on the tiny chromite crystal
that also serves as a nucleation point for the serpentine "horsetail," so the demantoid and serpentine fibers must grow simultaneously. As the crystals are small, seldom large enough to cut more
than one gem, each cut gem is likely to contain one
horsetail, or the part thereof not removed by cut-

GEMS & GEMOLOGY

Summer 1996

109

Figure 13. The most distinctive internal feature of

demantoids is the "horsetail" of fibrous serpentine (chrysotile) that is present, i n whole or i n
part, in most pebbles or cut gems. Note the tiny
crystal (probably chromite) from which all the
fibers radiate. Photomicrograph by John I.
Koivula; magnified 20x.

ting. In the authors' experience, cut specimens

lacking all traces of a horsetail are exceedingly
scarce. Rarely, the fibers are uniform and parallel,
and the gem may be cut to show chatoyancy
("Cat's-eye demantoid," 1960).Some of the demantoids that have emerged most recently on the international gem market contain randomly oriented
acicular inclusions that do not appear to be related
to a "horsetail" (figure 15).
MARKETING AND DISTRIBUTION
Because of its extreme rarity, the small size of the
crystals, and its relatively low hardness, demantoid
has not enjoyed widespread recognition outside
Russia. It was popular in the Western world for a
brief period during the Edwardian age at the turn of
the century, but it slipped back into relative obscurity after World War I, when little emerged from
the young Soviet Union. Many of the demantoids
seen in the market today are set in estate jewelry
from the Edwardian period. Typically, demantoids
appeared in Edwardian jewelry as a field of green
melee or as an encircling band of melee or calibrated stones highlighting a much larger central stone.
They also appeared in whimsical jewelry of the era,
as bright green frogs, lizards, snakes, dragonflies
(again, see figure 5), and the like (Misiorowski and
Hays, 1993, p. 164).
Much of the modern production represents the
secret caches of villagers and the work of "bandit
diggers." Each team of "thieves" has its own lapidaries, who cut all the rough in some corner of a
small apartment in Elzaterinburg. Many stones are

Russian Demantoid Garnets

Figure 14. In some demantoids, like this recently

mined stone, the "horsetail" takes the form of a
cone-like bundle of fibers. Photomicrograph by
Shane F. McClure; magnified 20x.

poorly cut, although some fine ones are produced.

Unfortunately, many are slightly scratched or
chipped in handling. Essentially no rough demantoid appeared in the Russian "gem fairs" in 1995,
and very little rough reaches the outside market.
Although many cut stones (about 15%-20% ] are
sold in Russia, where there is historic appreciation,
the bulk of present production goes "unofficially"
abroad, most of it to Germany with lesser amounts
finding their way to Israel, the United States, the
United Arab Emirates, Spain, Italy, and Canada.
The authors' experience is that German dealers,
like most Russians, tend to prefer the pale "golden
green" stones that best display demantoid's inherFigure 15. In some of the newer demantoids to
emerge from Russia (see, for example, figure
I), large quantities of randomly oriented acicular inclusions are found throughout the cut
stones, with no evidence of a nucleating crystal. Photomicrograph by Shane F. McClure;
magnified 20x.

GEMS & GEMOLOGY

Summer 1996

ent brilliance and fire. American dealers usually

seek out the much rarer "emerald1'-green stones, in
which the darkness of the color largely masks the
fire (again, see figure 1).The best prices are found in
Elzaterinburg at the Urals Exhibition of Colored
Stones, held each year in the fall.
CONCLUSION
Demantoid is a beautiful green to yellow-green
gem variety of andradite garnet, colored by minor
chromium and intrinsic iron. It has great brilliance
and "fire" (dispersion), but it is slightly lower in
hardness than tsavorite.
Historically, demantoid has been found almost
exclusively in Russia's Central Urals. Discovered
in the mid-19th century, this bright green gem
soon became popular among the nobility of Czarist
Russia and even enjoyed brief exposure in Western
Europe at the turn of the century. With the onset of
the Bolshevik Revolution in 1917, however, exportation of this rare gem came to a halt, and little
new material emerged until the economic restructuring (pezestroilw)of the late 1980s.
Two important demantoid districts are known
in the Central Urals, one about 115 km northnorthwest and one about 75 lzm south-southwest
of Elzaterinburg; each of these districts has two or
more alluvial or in situ deposits. All of the locali-

ties appear to have significant reserves, as to date

there has been no systematic or sustained mining of
this gem material. Although two deposits are
presently leased t o Russian companies by the
Russian government, neither provides a significant,
continuous, or dependable supply of demantoid rough.
Neither company has the finances to develop the properties they hold, and thus they are eagerly seelung foreign investment. The I<amenuschlza deposit is of special interest, as it yields stones of the most desirable
color and presently is not licensed. Unauthorized
recovery from river gravels by illegal miners provides
most of the present production; unfortunately, there
are no reliable statistics for demantoid productionhistoric or recent-from this area.
In the next few years, the Russian government
is likely to assume control of the production, cutting,
and distribution of demantoid, as it has recently done
with emerald and alexandrite. Demantoid may also
be declared a gem of the "first group," which would
put it in the company of diamond, emerald, ruby,
sapphire, and alexandrite and thus result in more
stringent purchase and export regulations.
Political and economic stability in Russia may
bring about the development of these world-class
deposits. Until then, demantoid will remain a rare
and costly gemstone, sought primarily by wealthy
collectors and those with a taste for the exotic.

REFERENCES
Cat's-eye demantoid (19601.Gems a)Gemology, Vol. 10, No. 3, p. 92.
Bridges C.R. (1974)Green grossular garnets ("tsavorites") in East
Africa.Gems a)Gemology, Vol. 14, pp. 290-295.
Church A.H. (1879) On so-called green garnets, from the Urals.
Mineralogical Magazine, Vol. 2, pp. 191-193.
Clark A.M. (1993)Hey's Mineral Index, 3rd ed. Chapman & Hall,
London.
Dana E.S. (1958)A Textbook of Mineralogy, 4th ed. Rev. by W. E.
Ford, John Wiey &Sons, New York.
Eichmam E. (1870)Nils von Nordenskiold. Transactions of the St.
Petersburg Emperor's Mineralogical Society, Series 2, bd. Vol. 5,
pp. 189-192 (inGerman].
Eremeyer P.V. (1871) Conference Proceedings of the Russian
MineralogicalSociety, March 17, 1870, pp. 391-392.
Fritsch E., Rossman G.R. (1987)An update on color in gems, Part 1:
Introduction and colors caused by dispersed metals. Gems a )
Gemology, Vol. 23, No. 3, pp. 126-139.
Gill J.O. (1978)Demaiitoid-the complete story. Lapidary [ournal,
,
Vol. 32, No. 7, pp. 1542-1545.
Gnffen D.T. (1992)Silicate Crystal Chemistry. Oxford University
press, New York.
Hurwit K., Johnson M.L., Fritsch E. (1994)Gem trade lab notes:
Grossular-andradite garnet from Mali, West Africa. Gems ss)
Gemology, Vol. 30, No. 4, pp. 265-266.
Johnson M.L., Boehm E., Krupp H., Zang J.W., Kamrnerling R.C.
(1995)Gem-qualitygmssular-andradite:A new gamet from Mali.
Gems &> Geniology, Vol. 31, No. 3, pp. 152-166.

Russian Demantoid Garnets

Loeffler B.M., Bums R.G. (1976)Shedding light on the color of gems

and minerals. American Scientist, Vol. 64, No. 6, pp. 636-647.
McConnell D. (1964) Refringence of garnets and hydrogarnets.
Canadian Mineralogist, Vol. 8, Part 1, pp. 11-22.
Misiorowski E., Hays N. (1993)Jewels of the Edwardians. Gems ss)
Gemology, Vol. 29, No. 3, pp. 152-171.
Moench R.H., Meyrowitz R. (1964)Goldmanite, a vanadium gamet
from Lagiina, New Mexico. American Mineralogist, Vol. 49,
NOS.516, pp. 644-655,
Miuje P., Muije C.S., Miuje L.E. (1979)Colorless and green grossular
from Tanzania. Gems e>)Gemology, Vol. 16, No. 6, pp. 162-173.
Payne T. (1981)The andradites of San Benito County, California.
Gems ed Gemology, Vol. 17, No. 3, pp. 157-160.
Phillips W.R., Griffen D.T. (1981) Optical Mineralogy, The Nonopaque Minerals. W. H. Freeman and Co., San Francisco.
Rouse J.D.(1986)Garnet. Butterworths, London.
Sarnsonov J.P., Turingue A.P. (1985)Gems of [he U.S.S.R, Nedra,
Moscow.
Stockton C.M., Manson D.V. (1983)Gem andradite garnets. Gems
&>Gemology,
Vol. 19, No. 4, pp. 202-208.
Stockton C.M., Manson D.V. (1985)A proposed new classification
for gem-quality garnets. Gems ed Gernology, Vol. 21, No. 4, pp.
205-218.
Webster R. (1983)Gems, their Sources, Descriptions and identification, 4th ed. Rev. by B.W. Anderson, Butterworths, London.
Wilson W.E. (1985)What's new in minerals? Tucson show 1985.
MineralogicalRecord, Vol. 16, No. 4, pp. 303404.

GEMS & GEMOLOGY

Summer 1996

111

NOTESAND NEWTECHNIQUES

OPALFROM SHEWA
PROVINCE,
ETHIOPIA
By Mary L.Johnson, Robert C. Kammerling, Dino G. DeGhionno, and John I. Koivula

Opal occurs as nodules in volcanic rocks at a new gem opal locality in the
Menz Gishe district of S h m Province, Ethiopia. The opalfield,still in the
early stages of exploration, extends over several square kilometers. Opals
from Shewa. Province have near-colorless to white, yellow, orange, ray, or
brown body colors; some show face-up play-of-color, and many have contra
luz play-of-color. The ~emologicalproperties are consistent with those of
other natural opals, and small particles are common inclusions.
Preliminary stability tests indicate that much of the material absorbs
waterund some crazes when exposed to light and heat,

Gem-quality opals, for the most part, come from

two types of deposits: volcanic and sedimentary
(Frondel, 1962). Although the most significant
deposits of gem opal-those in Australia-are sedimentary in nature (see, for instance, Wise, 1993))
other important opal deposits are related to siliceous
volcanic rocks. These volcanic occurrences include,
among others, Querktaro, Mexico (see, e.g., Koivula
et al., 1983; Gubelin, 1986; and Spencer et al., 19921,
and Opal Butte, Oregon (Smith, 1988),with the latABOUT THE AUTHORS
Dr. Johnson is a research scientist, the late Mr. Kammerling was
vice president for research and development, Mr. DeGhionno is
senior staff gemologist, and Mr. Koivula is chief research gemologist at the GIA Gem Trade Laboratory, Santa Monica, California.
Acknowledgments: The authors wish to thank Telahun Z.
Yohannes, of the Ethics-AmerCan Resource Development
Corporation, Wheat w e , Colorado, for providing photographs and mining information, Jbaning fashioned stones
and rough, and allowing some destructive testing of the rough
material. Dr. Paul Downing, of Majestic Opal, Estes Park,
Colorado, and Klaus G. Klement, of Rough Times, San Diego,
California, provided additional stones and information. Dr. N. R.
Barot, Nairobi, Kenya, sent some early information about opals
from Ethiopia. Dijon Doupliner, of the GIA Gem Trade
Laboratory, performed the FTIR and EDXRF measurements.
Gems & Gemology, Vol. 32, No. 2, pp. 112-1 20
0 1996 GemologicalInstitute of America

Notes & New Techniques

ter's related deposits in Idaho (Broughton, 1972) and

British Columbia (Downing, 1993).
Recently, opals were discovered at Yita Ridge
in the Menz Gishe district of Shewa Province,
Ethiopia. The opal-bearing rock is a nodular rhyolite (similar to that at Opal Butte). Material examined from the Ethiopian deposit included stones
resembling "contra luz" (that is, play-of-color only
visible with transmitted light) and "crystal" opal,
as well as a fire agate-like opal that shows play-ofcolor on a dark brown body color (figure 1).As the
deposit is still in the preliminary evaluation stage
(T. Yohannes, pers. comm., 1996), its full production potential is unknown. However, early sampling results indicate that the opal-bearing rocks
extend over several square kilometers.

BACKGROUND
To the best of our knowledge, the first report on
gem opals from Ethiopia appeared in the February
1994 ICA Gazette (Barot, 1994). According to that
report, Ethiopian opals were first seen in the
Nairobi gem market in mid-1993. Some of these
opals (obtained in Nairobi as being of Ethiopian origin, but with the precise locality unconfirmed)
were subsequently examined and reported in the
Spring 1994 Gem News section ( K o i v ~ ~et
l a al.,

GEMS & GEMOLOGY

Summer 1996

1994a).The Summer 1995 Gem News section contained a short entry specifically describing opals
from the Yita Ridge area of Shewa Province (Kammerling et al., 1995).
LOCATION AND ACCESS
Yita Ridge, in the Menz Gishe District of Shewa
Province, is about 240 km (150 miles) northeast of
Ethiopia's capital, Addis Ababa (Kammerling et al.,
1995; figure 2). The opal field lies approximately
15-20 lzm north of Mezezo, about 8 lzm off the road
between Mezezo and Hirute, which branches off
the main road between Addis Ababa and Dese. The
road between Yita Ridge and the Mezezo-Hirute
road has been washed out for about 10 years, however, so the opal area is currently accessible only by
mule trail or helicopter (T.Yohannes, pers. comm.,
December 1995).
The opal-bearing roclzs outcrop along the north
flank of Yita Ridge, at about 2,450 m elevation.
The surrounding area is predominantly agricultural
(figure3).
GEOLOGY AND OCCURRENCE
The opal-bearing area lies in a large volcanic field
just west of the northern Great Rift Valley of
Ethiopia (the valley of the Awash River, in this
area), according to Mr. Telahun Yohannes, of the
Ethio-American Resource Development Corporation,
a joint Ethiopia-U.S. firm that is mining the area.
The opal nodules (figure 4) occur in a continuous
layer of welded tuff (approaching obsidian in character), about 3 m thiclz, that lies between more
weathered (decomposing)rhyolite layers. The beds
have been uplifted and tilted slightly since deposition, but they are still more-or-less horizontal. The
entire sequence of volcanic rocks, about 300400 in
thiclz, is probably part of the Amba Alaji rhyolites,
which are Miocene in age (8 to 27 million years
old; Merla et al., 1979). The opal nodules average
about 10 cm (4 inches) in diameter.
PROSPECTING AND MINING
Only a small amount of material (about 200 1%)has
been produced so far, all from surface and near-surface occurrences. Although the opal-bearing layer is
primarily horizontal, the steep, gulch-like nature of
the topography means that this layer outcrops in
many places. The decision to begin commercial
production is pending completion of surveys of the
area to determine the economic feasibility of the
deposit. It is expected that these surveys-tracking

Notes & New Techniques

Figure 1. Gem-quality opal was recently discovered

in Shewn Province, Ethiopia. These four opals
demonstrate the range in body color and play-ofcolor in stones from that locality. Clockwise
from top, 7.52 ct cabochon, 2.15 ct round brilliant, 4.04 ct cabochon, and 2.31 ct triangls modified brilliant. Photo 0 GIA and Tino Hammid.

the opal-bearing outcrops, and assessing the quality

and quantity of opal present-will be completed
before the end of 1996. The few (less than 20) ininers now working in the area use hand tools only, as
the opal is too fragile for blasting (figure 5). In the
future, however, road-grading equipment may be
used to remove the decomposed rhyolite above the
opal-bearing layer.
As of December 1995, on the basis of these outcrops of opal-containing rock, the gem field had been
estimated visually to extend over an area of at least 7
x 7 l a . It has also been estimated, from opal nodules
randomly gathered from 12 area sites, that about
15% of the opal recovered is gem quality. About 1%
of this gem opal shows distinct play-of-color.

GEMS & GEMOLOGY

Summer 1996

Figure 2,Opals are found along Yita Ridge in the Menz Gishe
district of Shewa Province, Ethiopia. Location map on the
lower left modified from Ethiopian Tourist Commission map
(1992);opal locality map modified from one provided by
Telahun Yohannes.

MATERIALS AND METHODS

We examined 19 fashioned stones-six faceted
(0.55480 ct) and 13 cabochons (0.83-19.71 ct)and over 250 grams of rough, including sections of
several different nodules. We viewed the face up
colors using both fluorescent and incandescent illumination. We observed play-of-color with the stone
against a dark background and spot illumination
placed above and then perpendicular to the viewing
direction (the latter to observe any contra luz
effect). Refractive indices were measured with a
Duplex II refractometer and a near-monochromatic, Na-equivalent light source. Specific gravity was
determined by hydrostatic weighing; in those several cases where the stone absorbed water, we
reported only the first result, which should be
regarded as a minimum S.G. for t h a t stone.
Ultraviolet fluorescence was observed in a darlzened room using a controlled viewing environment
and a short-wavellong-waveW lamp. Polarization
behavior was noted using a GIA GEM Illuminator
polariscope, and absorption spectra (for the body

114

Notes & New Techniques

color of the stones) were observed using a Beck

prism-type spectroscope. The Chelsea color filter
reaction was determined with illumination from a
spectroscope base. Internal features were observed
with a standard gemological microscope and brightfield, darlzfield, and oblique fiber-optic illumination, as well as polarizing filters.
Trace-element chemistry was determined qualitatively by energy-dispersive X-ray fluorescence
[EDXRF) spectrometry, using a Tracer X-ray
Spectrace 5000 unit with a rhodium-target X-ray
tube. Mid-infrared FTIR absorption spectra were
taken with a Nicolet Magna-IR Model 550 spectrometer, with data collected in the range between
6000 and 4000 cm-1. We employed X-ray powder
diffraction analysis to identify some included materials, using a Debye-Scherrer camera mounted on a
Siemens Kristalloflex X-ray generator.
Eight partially polished pieces were tested for
durability. We exposed four samples to light and
heat by placing them on a black surface, 8 cm from
a 100-watt, high-intensity incandescent light, for

GEMS & GEMOLOGY

Summer 1996

Figure 3. The area in

which the opal nodules are found is
very rural, consisting primarily of
numerous small
farms and villages.
Photo courtesy of
Telahun Yohannes.

24-hour periods up to a total of 192 hours (the temperature of the blaclz surface reached 155~-about
6 8 ~to) simulate shop-window conditions. Another
four pieces were immersed in water for a few hours
and dried in air to a constant weight (about 24
hours), for two cycles, to test the effects of drastic
humidity changes.
APPEARANCE AND
GEMOLOGICAL PROPERTIES
Color. The fashioned stones were near-colorless,
milky white, and yellow through orange to brown
(again, see figure 1); one cabochon was pinkish
brown. T h e two darkest stones had brown and
blaclz body colors; however, the darkest stone was a
doublet that had been backed with blaclz obsidian,
which affected the apparent body color. Most of the
material was suitable for jewelry (figure 6).

minated from the side or rear; figure 7.) The blaclz

opal doublet had good play-of-color face up.
Diaphaneity. The fashioned samples were transparent to opaque. The darker they were, the less transparent they became.
Refractive Index. Refractive indices ranged between
1.40 and 1.45. In six cases, a second R.I. could be
seen, usually a t 1.45; this effect resembled the
anomalous R.I. seen in some tourmaline (illustrated
in Koivula et al., 1994b).
Figure 4. Opal nodules (the split one on the dark rock
is about 10 c m in diameter) are abundant in the rhyolitic welded tuff, a rock that approaches obsidian i n
character. Photo courtesy of Telahun Yohannes.

Play-of-Color. Two dark brown opals showed good

play-of-color face-up, and the lighter stones had
good (one example), moderate (four examples),
weak (one example), or n o play-of color when
viewed face-up. Nine stones showed contra luz
play-of-color. (The play-of-color is designated "contra l ~ i z "if i t is seen with transmitted light-as
described by Koivula and Kammerling [I9881 and
Smith [I9881 for the Opal Butte material. For our
nine contra luz samples, we saw play-of-color when
we viewed the stones face up while they were illu-

Notes & New Techniques

GEMS & GEMOLOGY

Summer 1996

even yellow-green (with no phosphorescence);

slightly chalky faint (even] to weak (uneven) yellow-green; slightly chalky faint-to-weak blue; inert;
faint even blue-white; faint uneven blue and yellow; and faint even orange. The doublet was inert
to long-wave W radiation, but it fluoresced a weak
even yellowish green to short-wave W.
Specific Gravity. The hydrostatic measurements
ranged between 1.35 and 2.03; samples that did not
soak up water were in the 1.87-2.03 range. Most of
the near-colorless opals had specific gravities of 2.0
or higher. Seven stones changed their weights
noticeably by soaking up water during the S.G.
measurements: two near-colorless opals, two milky
white opals, and one each orangy brown, light
pinkish brown, and mottled yellow.
Microscopy/Inclusions. Solid inclusions were relatively common in the fashioned Ethiopian opals.

Figure 5. Workers remove opal nodules from a sample pit in this outcrop of the opal-bearing welded tuff
zone. Photo courtesy of Telahun Yohannes.

Figure 6. As these three pendants indicate, the

Ethiopian opals (2.31-7.52 ct) make attractive
jewelry stones. Jewelry courtesy of (he Gold Rush,
Northridge, California;photo by Shane F. McClure.

Polariscope Reaction. Twelve pale-colored stones

showed weak-to-moderate anomalous double refraction (ADR)when viewed between crossed polarizing
filters. Two darker stones showed no ADR.
Optical Absorption Spectrum. No spectrum was
seen in the 14 lightest-color stones with the hand
spectroscope. Three stones showed lower cutoff
edges at 510 (yellow stone), 530 (yellowish orange),
and 600 nm (brown).One mottled brown stone
showed a band between 550 and 590 nm, a lower
cutoff at 520 nm, and an upper cutoff at 670 nm.
Color filter. Two brown stones appeared red when
viewed through the Chelsea color filter.
Fluorescence to UV Radiation. We observed the following reactions to long-wave UV radiation: faintto-weak, even yellow-green (with weak phosphorescence); slightly chalky faint blue (moderate
whitish blue); inert; faint uneven blue and yellow;
and faint even orange. When the stones were
exposed to short-wave UV,we saw: faint-to-strong,

116

Notes & New Techniques

GEMS & GEMOLOGY

Summer 1996

Figure 7. This piece of

rough Ethiopian opal
illustrates the contra luz
effect:With darkfield illumination (left)no play-ofcolor is seen; with fiberoptic illumination from
the side (right),strong
play-of-coloris visible.
Photomicrograph by John
I. Koivula; magnified 4x.

Most of the transparent-to-translucent gems contained tiny crystals of unknown composition.

Perhaps the most distinct inclusions (because of
their color) were randomly scattered red-brown
grains (figure8).
Intermixed with the red-brown grains were
numerous small, opaque, blaclz grains (figure 9).
Some of these appeared a dark, brassy yellow when
they were examined from above with a strong pinpoint fiber-optic light. Although most of the mineral grains were extremely small, a few of the largest
were surrounded by small tension cracks (figure
101, which were probably caused by a difference in
volumetric expansion between the inclusions and
the enclosing opal.
Many of the red-brown and blaclz inclusions
appeared to have a square cross section or outline,
indicating that they might be isometric. Even the
largest of these inclusions was too small for X-ray
diffraction analysis, however, so their identity is
still unlznown. Although there is no direct proof,
we speculate that the blaclz grains might be pyrite
Figure 8. Tiny red-browngrains were visible with
magnification in most of the Ethiopian opals. In
some stones, small near-colorless crystals were
also seen. Photomicrograph by John I. Koivula;
magnified 5 0 ~ .

Notes & New Techniques

(on the basis of their brassy metallic luster and

square outline). Similarly, the red-brown grains
might be pyrite altered to hematite.
Also present, although much less common,
were tiny white to near-colorless grains that
revealed no distinct form (visiblein figures 8, 9, and
10; see also figure 11).Like the black and red-brown
inclusions, these particles were too small to be
identified by the methods available to us.
The most distinctive inclusion type-and the
rarest-was observed in only one stone. These
inclusions had the appearance of elongated, roughsided "voids" or "tube systems" (figure 11). All
reached the surface of their host and were partially
packed with what apparently was a light brown-towhite, semi-translucent epigenetic matter. At the
interface with the surrounding opal, the surfaces of
these "tubes" were randomly spotted with tiny redbrown and blaclz grains, similar to those previously
described. In comparable inclusions seen in opals
from Mexico (Koivula et al., 1983), the voids once
contained hornblende crystals that had dissolved
away and been replaced by, or partially filled with,

Figme 9. These small opaque grains (in a 4.77 ct

Ethiopian opal) may be pyrite crystals.
Photomicrograph by John I. Koivula; magnified 20~.

GEMS & GEMOLOGY

Summer 1996

Figure 10. Some of the largest black mineral grains

were surrounded by tension cracks. Photomicrograph by John I. Koivula; magnified 4 0 ~ .

limonite and a white kaolinitic clay. Because we

could not perform destructive tests on the sample
loaned to us for this study, however, we could not
determine the identity of the filling material in this
Ethiopian opal.
When opal nodules are freed from the rhyolite
matrix, a black crust is sometimes visible on both
the surfaces of the rough opals and on the rhyolite
itself (figure 12); it may also be present in the
cracks in some gem opals. An X-ray powder diffraction pattern, obtained from a scraping taken from
one of these blaclz crusts, matched that of the black
manganese oxide ramsdellite.
Only the two opaque dark brown oval cabochons showed any obvious flow structure. This
resulted from varying degrees of iron pigmentation,
and it was manifested in the form of light brown
veins and fingers extending into and through a
much darker brown opal groundmass (figure 13).In
one of these stones, a cell-like structure had formed
where the light brown areas surrounded the darker
areas. This was particularly obvious in oblique
reflected light, in which the dark brown "cells"
showed strong play-of-color(figure 14).

Figure 11. Rough-sided epigenetically filled

"tubes," similar LO those seen in some opals from
Querbtaro, Mexico, were observed in this 3.57 ct
Ethiopian opal cabochon. Photomicrograph by
JohnI. Koivula; magnified 1 5 ~ .

rubidium (17 stones), or niobium (15 stones); and

some stones also contained trace amounts of copper (10 stones), yttrium (8 stones), zinc (7 stones),
lead (6 stones), titanium (5 stones), manganese (3
stones), chromium (2 stones), barium (2 stones), or
gallium (1stone).
STABILITY AND DURABILITY TESTING
Very little has been published about testing opals
for durability. Pearson (1985)tested opals for crazing by refluxing them in flammable organic liquids;
however, we did not try to duplicate this test
because the materials involved are hazardous. A
Figure 12. Black flakes of ramsdellite were sometimes visible at the interface between the opal and
its matrix and may occur in the cracks in some
rough and fashioned Ethiopian opal.
Photomicrograph by JohnI. Koivula; magnified 5x.

SPECTRAL AND CHEMICAL ANALYSES

Infrared Spectrum. All the stones showed a typical
opal spectrum in the range between 6000 and 4000
cm-1 (seeFritsch and Stockton, 1987, for comparison).

Chemistry. Qualitative EDXRF analyses were performed on all 19 fashioned stones and on two
pieces of matrix-free rough. For each stone, the
only major element seen was silicon. Every stone
also contained detectable trace amounts of calcium, iron, strontium, and zirconium; most stones
contained trace amounts of potassium (20 stones),

118

Notes & New Techniques

GEMS & GEMOLOGY

Summer 1996

Figure 13. Flow structure is clearly visible in this

2.36 ct Ethiopian opal. Photomicrograph by John
I. IZoivnia; magnified 25x.

practical "field test" for opal stability consists of

exposing the material in a sunlit location for a few
months, then recovering the undamaged remnants,
according to Smith (1988),who also recommended
a slow-drying technique to stabilize Oregon opal.
He estimated that about 20% of the "fire opal"
from Opal Butte, Oregon, was stable.
The lack of consensus in the literature led us to
develop a few simple tests based on the assumption
that gradual or sudden dehydration/rehydration was
the process most likely to affect opals. Although we
could not subject the fashioned stones we had borrowed to destructive testing, Mr. Yohannes kindly
allowed us to do durability testing on some of the
rough material. We had a flat surface polished on
each of eight pieces of rough-two each of seinitransparent yellow contra luz, semi-transparent gray,
semi-translucent yellow, and semi-translucent gray
opal. We then divided the stones into two groups for
testing, one for gradual dehydration and the other for
sudden changes in water-vapor pressure. Because of
the limited amount of material tested, we regard our
results as informative but not definitive.
Gradual dehydration was simulated under "shop
window" conditions: a bright light source and a
black background (see "Materials and Methods").

Figure 14. Play-of-colorin this 4.04 ct dark brown

Ethiopian opal is confined,to distinct "cells," separated by lighter-colored brown opal, as seen here
in oblique reflected light. Photomicrograph by
John 1. Koivula; magnified 1 5 ~ .

Under these conditions, the contra luz and the sernitranslucent gray opals did not change. The two other
stones had developed craclzs when first examined
after 24 hours (figure 15); these craclzs increased in
size with subsequent exposure, but no new ones
formed. It is interesting to note that during the first
24-hour test period, the semi-transparent gray stone
developed a transparent outer layer (figure 16) that
did not extend further into the stone on subsequent
exposure, although craclzs formed during the same
period did grow. No additional changes were noted
in any stone after 96 hours' exposure, which suggests that an exposure test such as this one may be
sufficient to separate durable from easily crazed
Ethiopian opals.
Sudden changes in water-vapor pressure were
accomplished by repeated hydrostatic specific-gravity determinations. Between each test, we allowed
the stones to dry completely (until the weight did
not change). The contra luz and semi-transparent
gray opals showed weight fluctuations of about
0.5-0.6 wt.% between wet and dry conditions, with
significant cracking after the second humiditychange test. The two semi-translucent opals (yellow

Figure 15. This 7.74 ct

piece of rough yellow
Ethiopian opal (left)
cracked during 24 hours'
exposure to light and
heat (right).Photos by
John I. Koivula.

Notes & New Techniques

GEMS & GEMOLOGY

Summer 1996

119

Figure 16. A 9.59 C Lpiece of semi-transparent gray Ethiopian opal (left, before testing) showed "clarification"of
the less transparent areas and a fewincipient cracks after 24 hours' exposure to "shop window" conditions
(center);by 48 hours (right),the clacks had extended through the stone. Photos by John I. Koivula.

and gray) developed less-pronounced cracking after

two cycles, despite far greater weight fluctuations of
about 11 and 18 wt.%, respectively. The weight
gains (with soaking) and losses (as the stones dried)
did not vary much with repeated soaking and drying, indicating no permanent gain or loss of water.
Note, however, that we consider this test inherently destructive and do not recommend any "real
world" equivalent-such as wearing an opal ring
while washing dishes-for any opal.
In addition, the fashioned stones were exposed
to variable temperature and humidity in open plastic bags for six months in our laboratory. During
this time, external (relative)humidity fluctuated
between about 10% and 70%. One stone-the
black opal doublet-crazed, but none of the other
stones was affected.

REFERENCES
Barot N.R. (19941New precious opal deposit found in Ethiopia,
ICA Gazette, February 1994, p. 2.
Broughton P.L. (19721 Precious opal mining in the Snake River
Plain rhyolites, Idaho, / o ~ ~ r nof
a lGemmology, Vol. 13, No. 3,
pp. 100-104.
Downing P.B. (19931Okanagan
Lapidary
- opal.
. .Journal, Vol. 46,
No. 11, pp. 63-66.
Ethiooian Tourist Commission (19921
,
, Ethiopia. Scde 1:2 000
000. Addis Ababa, Ethiopia, 1 p.
Fritsch E., Stockton C.M. (1987)Infrared spectroscopy in gem
identification. Gems a1 Gemology, Vol. 23, No. 1, pp. 18-26.
Frondel C. (1962) The System of Mineralogy of fames Dwight
Danu a n d Edward Salisbury Dana, Yale University
1837-1892, 7th ed., Volume ffl: Silica Minerals. John Wiley
& Sons, New York, 334 pp.
GObelin E. (1986)Opal from Mexico. Australian Gemmologist,
Vol. 16, No. 2, 1986, pp. 45-51.
Kammerling R.C., Koivula J.I., Fritsch E. (19951 Gem news:
Update on opal from Ethiopia. Gems a) Gemology, Vol. 31,
No. 2, p. 132.
Koivula J.I., Fryer C., Keller P.C. (1983) Opal from Querktaro,
Mexico: Occurrence and inclusions. Gems a) Gemology,

120

Notes & New Techniques

CONCLUSION
A new source for precious opal is being evaluated at
Yita Ridge, in the Menz Gishe district of Shewa
Province, Ethiopia. The nodules occur in a broad
range of body colors and play-of-color, including
excellent contra luz material. Tiny red and black
particles are pervasive throughout even the most
transparent opals; among the other inclusions noted
were hollow tubes and black platy manganese
oxides. As with opals from other regions (including
the similar volcanic environments of Querktaro,
Mexico, and Opal Butte, Oregon), durability is a
potential cause for concern. Only time will tell if
full-scale mining at Yita Ridge will prove to be economically feasible. However, current indications are
that some material will continue to reach the world
market even if only small-scale mining continues.

Vol. 19, No. 2, pp. 87-96.

Koivula J.I., Kammerling R.C. (1988) Gem news: Opal from
Oregon. Gems o)Gemology, Vol. 24, No. 3, pp. 179-180.
Koivula J.I., Kammerling R.C., Fritsch E. (1994a) Gem news:
Opal from Ethiopia. Gems a! Gemology, Vol. 30, No. 1, pp.
52-53.
Koivula J.I., Kammerling R.C., Fritsch E. (1994b) Gem news:
Tourmaline with atypical R.I. readings. Gems a! Gemology,
Vol. 30, No. 3, p. 198.
Merla G., Abbate E., Azzaroli A., Bruni P., Canuti P., Fazzuoli
M., Sagri M., and Tacconi P. (19791 A Geological Map of
Ethiopia and Somalia (1973) 1:2.000.000 and Comment
with a Map of Major Landforms. Consiglio Nazionale delle
Richerche, Florence, Italy. 95 pp. plus maps.
Pearson G . (19851 T h e role of water i n cracking of opal.
Australian Gemmologist, Vol. 15, No. 12, pp. 435-445.
Smith ILL. (1988).Opals from Opal Butte, Oregon. Gems it)
Gemology, Vol. 24, No. 4, pp. 229-236.
Spencer R.J., Levinson A.A., Koivula J.I. (1992) Opal from
Querktaro, Mexico: Fluid inclusion study. Gems a)
Gemology, Vol. 28, No. 1, pp. 2834.
Wise R.W. (19931Queensland boulder opal. Gems it) Gemology,
Vol. 29, No. 1, pp. 4-15.

GEMS & GEMOLOGY

Summer 1996

BACK

Spring 1987

ISSUES

O F

Summer 1992

Modern" Jewelry: Retro to Abstract

Infrared Spectroscopy in Gem Identification
A Study ol the General Electric Synthetic Jadeite
Iridescent Orthoamphibole from Greenland

Gem Wealth 01 Tanzania

Gamma-Ray Spectroscopy and Radioactivity
Dyed Natural Corundum as a Ruby Imitation
An Update on Sumitomo Synthetic Diamonds

Summer 1987

Fall 1992

Gemstone Durability: Design to Display

Wessels Mine Sugilite
Three Notable Fancy-Color Diamonds
The Separation of Natural from Synthetic
Emeralds by Infrared Spectroscopy
The Rutilated Topaz Misnomer

Ruby and Sapphire Mining in Mogok

Bleached and Polymer-Impregnated Jadeite
Radiation-Induced Yellow-Green in Garnet

Winter 1992

Lirfiited quantities of these issues are still available

Fall 1987
An Update on Color in Gems. Part I
The Lennix Synthetic Emerald
Kyocera Corp. Products that Show Play-of-Color
Man-Made Jewelry Malachite
Inamori Synthetic Cat's-Eye Alexandrite

Sprlng 1993
Queensland Boulder Opal
Update on Dillusion-Treated Corundum:
Red and Other Colors
A New Gem Beryl Locality: Luumaki, Finland
De Beers Near Colorless-to-Blue Experimental
Gem-Quality Synthetic Diamonds

Winter 1987
The De Beers Gem-Quality Synthetic Diamonds
Queen Conch 'Pearls"
The Seven Types of Yellow Sapphire

Summer 1993
Flux-Grown Synthetic Red and Blue Spinels
from Russia
Emeralds and Green Beryls of Upper Egypt
Reactor-Irradiated Green Topaz

Summer 1988
The Diamond Deposits of Kalimantan, Borneo
An Update on Color in Gems: Part 3
Pastel Pyropes
Three-Phase Inclusions in Sapphires from Sri Lanka

Fall 1993
Jewels of the Edwardians
A Guide Map to the Gem Deposits of Sri Lanka
Two Treated-Color Synthetic Red Diamonds
Two Near-Colorless General Electric Type lla
Synthetic Diamond Crystals

Fall 1988
An Economic Review of Diamonds
The Sapphires of Penglai, Hainan Island, China
Iridescent Orthoamphibole from Wyoming
Detection of Treatment in Two Green Diamonds

Winter 1993

Spring 1989

Russian Gem-Quality Synthetic Yellow Diamonds

Heat Treating Rock Creek (Montana) Sapphires
Garnets from Allay, China

The Sinkankas Library

The Gujar Killi Emerald Deposit
Beryl Gem Nodules from the Bananal Mine
Opalite:" Plastic Imitation Opal

Spring 1994
The Anahf Ametrine Mine, Bolivia
Indaia Sapphire Deposits ol Minas Gerais, Brazil
Flux-Induced Fingerprints in Synthetic Ruby

Summer 1989
Filled Diamonds
Synthetic Diamond Thin Films
Grading the Hope Diamond
Diamonds with Color-Zoned Pavilions

Summer 1994
Synthetic Rubies by Douros
Emeralds from the Mananjary Region,
Madagascar: Internal Features
Synthetic Forsterite and Synthetic Peridot
Update on Mining Rubies and Fancy Sapphires in
Northern Vietnam

Fall 1989
Polynesian Black Pearls
The Capoeirana Emerald Deposit
Brazil-Twinned Synthetic Quartz
Thermal Alteration ol Inclusions in Rutilated Topaz
Chicken-Blood Stone from China

Fall 1994
Filled Diamonds: Identification and Durability
Inclusions of Native Copper and Tenorite in
Cuprian-Elbaite Tourmaline, Paralba, Brazil

Spring 1990
Gem Localities of the 1980s
Gemstone Enhancement and Its Detection
Synthetic Gem Materials in the 1980s
Nevi Technologies of the 1980s
Jewelry of the 1980s

Winter 1994
Color Grading of Colored Diamonds in the GIA
Gem Trade Laboratory
Ruby and Sapphire from the Ural Mountains, Russia
Gem Corundum in Alkali Basalt

Winter 1990

%ql&

A Suite o
Emeraldoli

Spring 1995

Sapphires
IIUIIU

Rubies from Mong Hsu

The Yogo Sapphire Deposit
Meerschaum from Eskisehir Province,Turkey

d Jewelry

Spring 1991

Summer 1995

Age, Origin, and Emplacement of Diamonds

Emeralds of Panjshir Valley, Afghanistan

Summer 1991
Fracture Filling of Emeralds: Opticon and "Oils"
Emeralds from the Ural Mountains, USSR
Treated Andamooka Matrix Opal

Fall 1991
Rubies and Fancy Sapphires from Vietnam
New Rubies from M O ~ O Q OTanzania
~O,
Bohemian Garnet-Today

Winter 1991
Marine Mining of Diamonds off Southern Africa
Sunstone Labradorite from the Ponderosa Mine
Nontraditional Gemstone Cutting
Nontransparent "CZ" from Russia

Spring 1992
Gem-Quality Green Zoisite
Kilbourne Hole Peridot
Fluid Inclusion Study of Queretaro Opal
Natural-Color Nonconductive Gray-to-Blue Diamonds
Peridot as an Interplanetary Gemstone

Determining the Gold Content of Jewelry Metals

Diamond Sources and Production
Sapphires from Changle, China

Complete your back issues of

Gems & Gemology NOW!
Single Issues"

$ 9.95 ea. U.S.

$ 14.00 ea. elsewhere

$ 36.00 ea. vol. U.S.

$ 45.00 ea. vol. elsewhere

$ 99.00 U.S.

$125.00 elsewhere
$200.00 elsewhere

Complete Volumes:'

1987, 1991,1992,
1993,1994,1995
Three-year set
Five-year sel
'10% discount

$160.00 U.S.

IM GIA Annual Fund donorsillShe Boosleis Circle leveland above.

TO ORDER: Call toll free (800) 421-7250, ext. 202 or

(310) 82S-2991, ext. 202
FAX (310) 453-4478 OR WRITE: G&G Subscriptions GIA
1660 Stewart Street, Santa Monica, CA 90404 USA

History of Pearling in La Paz Bay

An Update on the Ural Emerald Mines
AVisual Guide to the Identification
of Filled Diamonds

Fall 1995
Gem-Quality Grossular-Andradile: A New Garnet
from Mali
Sapphires from Southern Vietnam
Ti-Sapphire": Czochralski-Pulled Synthetic Pink
Sapphire from Union Carbide

Winter 1995
A History ol Diamond Sources in Africa: Part 1
A Chart for the Separation of Natural and
Synthetic Diamonds
Some issues from the 1984-1986
volume years are also available. Please contact the
Subscriptions Office lor details.

ORDER

NOW!

Editor
C. W, Fryer, GM Gem Trade Laboratory
Contributing Editors
GIA Gem Trade Laboratoy, East Coast
G.Robert Crowningshield

LAB NQTES
DIAMOND
Some Unusual Cuts

Most modern diamond cuts are symmetrical i n shape and in facet

arrangement, with proportions chosen to take maximum advantage of
diamond's optical properties. By contrast, diamonds cut before the turn of
the century were more likely to be
characterized by "lumpy" shapes and
an irregular placement of facets.
Although modern-cut diamonds are
generally more pleasing to the eye
than the older styles, these earlier
styles-such as the Mogul, table,
rose, old mine, and old European
cuts-often have considerable histone and aesthetic appeal.
A few months ago, t h e East
Coast lab had the opportunity to
examine some of these older cuts.
One client submitted two Mogul-cut
stones (figure 1). The Mogul cut is
characterized by a broad, often asymmetrical base; a crown with either a
table or four shallow facets in place

Figure 1. These Mogul-cut diumonds (9.27 ct, left, and 9.54 ct)
mu y be from India's Golconda
region, possibly fashioned several
centimes ago.

122

Gem Trade Lab Notes

of a table; and two or more rows of

facets between the table and the girdie. (The 115.60 ct Taj-i-Mah [Crown
of the Moon] diamond in the Iranian
crown jewels is a famous example of
a Mogul cut.) The client mentioned
t h a t these Mogul-cut diamonds
might be several centuries old and
may have come from the Golconda
region of India, which gained broad
fame as a source of spectacular diamonds after French gem merchant
Jean-Baptiste Tavernier chronicled
his 17th-century travels there.
A third example was a drilled
diamond (figure2) that our client said
was old and came from India. It was
cut and polished to follow and retain
the external crystal shape, in contrast
to modem cutting techniques which
typically leave no evidence of the
original external morphology. The
hole, which ran along the longest
dimension, had been drilled at an
angle from each end to meet in the
middle of one side of the stone. In the
early days of diamond cutting,
drilling holes in diamond was very
time consuming. The worker could
either use a diamond "sharpu-a
small, sharp-edged crystal fragrnentor continuallv load diamond dust
onto the steel point of a bow drill.
We suspect that the latter method
was used for this stone because the
hole was so deep. In outline, this
stone resembles t h e Shah Jahan
Table Cut diamond, which also was
pierced. However, those holes were
much shallower, and they probably
were drilled with a "sharp," perpendicular to the longest dimension [see,

Karin Hurwil
Thomas Moses
Ilene Reinitz
GM Gem Trade Laboratory, West Coast
Mary L. Johnson
Shane F. McClure
Cheryl Y. Wentzell

Figure 2. This diamond, which

measures 11.65 x 9.75 x 5.60
mm and weighs about 6.5 ct,
was cut along the original crystal shape and drilled at an angle
from each end for mounting.
e.g., E. A. Jobbins et al., "A Brief Description of a Spectacular 56.71 carat tabular diamond,^' Journal of Gemmology,
Vol. 19, No. 1,1984, pp. 1-7).
Figure 3 shows a 12-sided tablet
cut, a variety of the table cut. This
stone (16.24x 15.73x 3.63 mm thick)
was probably cut recently, as indicated by its symmetrical facet arrangement and the lack of abrasions. The
tablet cut resembles a rondelle (a
thin, flat round bead) that has not
been drilled. Historically, such tablet
cuts, if used to cover watch faces or
miniature portraits set in rings, were
called portrait diamonds.

Editor's note: The initials at the end of each item identity

the contributing editor(s) who provided that item.
Gems & Gemology, Vol. 32, No. 2, pp. 122-127
@ 1996 Gemological Institute of America

GEMS & GEMOLOGY

Summer 1996

Figwe 3. Despite the antique cutting style of this 10.74 ct tabletcut diamond, the facet symmetry and good polish indicate that
it was probably fashioned within
the last several years.

The last example was a thin,

irregular but roughly oval-shaped
tablet-cut diamond on which an
inscription had been engraved (figure
4). The stone was probably cut from a
cleavage piece. Our client informed
us that this was an old cut, and we
determined that the inscription was
in Arabic. Because of diamond's
extreme hardness, pre-modem exainpies of inscriptions are relatively rare.
Perhaps the largest known inscribed
stone is the Darya-i-Nur, also in the
Iranian crown jewels, which has been
variously reported to weigh between
175 and 195 ct.
Nicholas DelRe
Figure 4. Historic engraved diamonds, such as this 2.65 ct
example, me relatively rare
because of diamond's hardness.
The inscription, in Arabic, reads
"Yo Allah "and is an invocation
of God.

Gem Trade Lab Notes

Imitation Jade
A stone in a closed-back ring (figure
6), seen in the East Coast lab, reinforced the fact that bleached, polymer-impregnated jadeite (so-called
"B" jade)-although an ever-growing
concern worldwide-is not the only
rn
B challenge in the identification of
translucent green gems. We must
Figure 5. These natural-color
remain aware of the earlier attempts
jadeite cabochons appeared red
to imitate fine jadeite and be on the
when examined with a Chelsea
lookout for these as well.
color filter, which could lead a
Normally, a spectroscope will
gemologist to falsely suspect that
quickly prove die identity of naturalthey were dyed.
color jadeite. However, because of
the closed-back mounting, we could
JADEITEJADE
not use transmitted light. Instead, we
Natural Color
reflected light off and into the surface
Recently seen in the East Coast lab of the stone. Since we saw no jadeite
were two slightly dark, but otherwise
spectrum, this test was inconclusive
very attractive, jadeite cabochons (fig- (sometimes reflected light will not
ure 5). Routine gemological testing reveal a spectrum that might have
established that the stones were
been seen easily with transmitted
jadeite; and chrome lines in the red
light).
end of the spectrum proved that the
The 1.63 soot refractive index
color was natural.
was too low for jadeite, which is usuHowever, when we examined
ally about 1.66. Close examination
the stones with a Chelsea color filter, with magnification quickly revealed
they appeared red. Usually, a red
the piece's true identity-an iinitacolor under the Chelsea filter indi- tion. A fern-like pattern (figure 7) discates that the stone has been dyed. tinctive of partially devitrified glass
was present. This type of glass is
However, we have seen that jadeite
that is very rich in chromium (espe- known in the trade as "meta-jade" or
GRC and TM
"Iimori stone."
cially so-called "Yunnan jade"), may
appear red under this filter, as well as
show strong absorption in the red end
of the spectrum. This should serve to A PEARL Mystery
remind readers that under some cirWhat started off as a routine investicumstances, the Chelsea color filter
gation into whether a pearl was natutest is not reliable, and other means
ral or cultured turned into a challengshould be used to determine whether
a stone has been dyed. GRC and TM
Figure 6. The green stone in this
closed-back zing proved to be a
jadeite imitation.

Figure 7. Microscopic examination of the jade imitation in figwe 6 revealed the fern-like pattern that results from the devitrification (partial crystallization)
of glass. Magnified 20x.

GEMS & GEMOLOGY

Summer 1996

123

In the past, pearls with similar

X-radiographs have turned out to be
hollow (see, e.g., Spring 1984 and
Spring 1994 Lab Notes, pp. 48-49 and
45, respectively). However, most of
those pearls were irregular in shape.
Since our findings were still
inconclusive, the client removed the
pearl from the mounting so we could
examine inside the drill hole.
Although the hole was narrow (1.1
mm), which made examination difficult, a combination of strong overhead and oblique illumination
revealed that the pearl was filled with
a substance (figure 10). A minute
amount of this material, which
appeared to be conchiolin, was
removed and tested with a thermal
reaction tester [hot point). It gave off
an odor of burnt hair, proving that it
was organic and the pearl was natural.
KH

Figure 8. This 14-mm-longdrop-shaped pearl, here set as the pendant on

a Retro-style necklace, was proved to be natural only after it was
removed from its mounting.
ing task for our East Coast lab. The
very light gray, fairly symmetrical
drop shape measured about 10.5 mm
in diameter by 14 mm long (figure 8).
The yellow-metal Retro-style neclzlace from which it was suspended
had a scroll motif that was popular in
the 1940s and '50s. The necklace was
also embellished with diamonds,
blue sapphires, and smaller pearls.
The drop-shaped pearl did not

Figure 9. The X-radiograph of the

drop-shaped pearl in figure 8
shows that ihe center is either h01low or contains a substance that
is transparent to X-radiation

fluoresce to X-rays, which indicated a

saltwater origin but not whether it
was natural or cultured. Usually an
X-radiograph readily reveals a pearl's
mode of growth and, hence, its oriein. However, this one was peculiar:
Except for the thin outside nacreous
surface layer, the interior appeared
con~pletelyand uniformly black (figure 9). This indicates that either
there was no material in this area or
the substance present was transparent to X-radiation.

Figure 10, When we examined the

pearl in figure 8 inside the drill
hole, we found that it contained a
natural organic substance.

Late last year, a 3.14 ct oval mixed

cut arrived in the East Coast lab for
identification. The stone's overall

Figure 11. This 3.14 ct pale

brownish pink garnet was identified as pyrope.

d
Gem Trade Lab Notes

pYROPEGARNET

color was a pale brownish pink (figure 11). This 9.53 x 7.29 x 5.79 mm
stone (reportedlyfrom one of the new
deposits in Tunduru, Tanzania) had
the following gemological properties:
diaphaneity-transparent; R.1.1.735; optic character-singly refractive; S.G.-3.69; and fluorescencefaint red to long-wave ultraviolet
radiation (inert to short-wave UV). A
faint Fe2+ spectrum, with features

GEMS & GEMOLOGY

Summer 1996

similar to those seen ( i n a m u c h

more pronounced fashion) in rhodolite and almandine, was visible in the
handheld spectroscope: absorption
below 430 nm, and lines at 505, 527,
and about 576 nm. With magnification, w e s a w s o m e i n t e r s e c t i n g
coarse needles and interrupted needles. From these gemological properties, we readily identified the stone as
a garnet. But which species? T h e
spectrum was not consistent with
the R.I. and S.G.
T h e m o s t c o m m o n pale pink
garnets are grossular a n d hydrogrossular. However, pale p i n k
pyropes are also known (see C. M.
Stockton's
P ~ r o ~ e s t "
Gems
GemologY~ S u m m e r l9881 PP.
104-106). T o obtain conclusive evidence, the stone was sent to the West
Coast laboratory for X-ray diffraction
analysis. The resulting pattern confinned that the material had a garnet
structure, with a unit-cell spacing of
less than 11.49 A. (Unit-cell spacing,
t h e distance across one structural
unit of a mineral, is measured by Xray diffraction.) End-member pyrope
has a unit-cell spacing of 11.459 A,
while grossular and hy~rogrossular
have quite different unit-cell spacings
of 11 .85 1 A
11, 8 5 1 2 . 1 6 A,
respectively (W. A. Deer et al., An
Introduction to the Rock Forming
Minerals, 1974, Longman Group,
~ o n d ~ pp.
n , 21-31). To further confirm that this was pyrope, we also
performed qualitative energy-dispersive X-ray fluorescence (EDXRF)
analysis on the stone, which revealed
s i (consistent with
major M ~AJ,
,
pyrope garnet); minor Mn, Fe, Ca,
and Zn; and trace amounts of Cr, Ti,
V, K, Ga, Ge, Y, and Zr. (The chromiurn was probably responsible for the
faint red long-wave UV fluorescence,
rarely seen in garnets, and the iron
was probably responsible for t h e
spectral features). Note that although
R.1 a n d S. G . d o n o t d i s t i n g u i s h
grossular or hydrogrossular from pale
pyropes, Stoclcton~s(1988)criterionthe spectrum seen with the handheld
spectroscope-still c a n s e p a r a t e
them.
M I , and TM
@
)

Gem Trade Lab Notes

Figure 12. Note the uniformity of color in these cabochons (1-4 ct), all of
which proved to be quench-crackled synthetic rubies.

RmY,
Q^ch

Crackled

Occasionally over the years, we have

tested synthetic stones that have
been quench crackled to produce natural-appearing fractures. In s o m e
i n s t a n c e s , t h e s t o n e s had been
quenched in dye or (we have been
told) in tincture of iodine. The coloring agent penetrated the induced fractures and crystallized to form very
realistic "fingerprints."
We have seen numerous other
stones that have been quench crackled in various dyes. Quartz that has
green
been quench crackled and
to h i t a t e enlerald is the nlost cornmoll. W e also h a v e e n c o u n t e r e d
q u a r t z dyed purple t o i m i t a t e
amethyst and quartz dyed red to imitate r ~ l b ~ - Ye;lrs
we were
a
quench-crac1c1ed
red stones that had been purchased,
as "Swiss Jade"
the East
identify a group
lab was
red cabochons that the
were
in
Ifipre
12)gemological tests
were ruby.
proved that t h e
they
as cabochon rubies
these
seemed
uniform. In fact, on the surfaces of
s o m e of t h e cabs, t h e fractures

I
--,

..-

Fislre 13, A distinctive lloneycomb-like pattern was evident

the fractures reached the
surface of the synthetic ruby
cabochons shown in figure 12.
Magnified 1 3 ~ .
f o r m e d a distinctive h o n e y c o m b
appearance (figure 13). Some of the
fractures resembled fingerprint inclusions (figure 14), and some seemed to
contain a substance which gave them
a more believable appearance than a
f r a c t u r e a l o n e w o u l d have had.
Fortunately, subtle curved striaecrossing the fractures-were visible
when the cabochons were examined
in the microscope with darkfield illumination. T h i s proved t h a t these
rubies were actually synthetic.
GRCand TM

GEMS & GEMOLOGY

Summer 1996

125

Because of the nature of sapphire

crystals, we more commonly see
color-zoned stones in which the color
is confined to the pavilion (see, e.g.,
figure 17). Although such stones
appear evenly colored face up, they
are much less likely to be mistaken
for an assembled stone than this
stone, in which the apparent color
zone separation was straight and lay
in the girdle plane.
GRC and TM
Figure 14. Some of the fractures
induced in the synthetic ruby
cabochons shown in figure 12 had
the appearance of natural fmgerprint inclusions. Magnified 40x.

SAPPHIRE
A Natural Stone Mistaken
for a Doublet
Shortly after issuing a laboratory
report on a ring-mounted natural sapphire [figure 15))staff members at the
East Coast lab received a call from
the client, who insisted that an error
had been made on the report. In fact,
he was so certain of our error that he
had already sent the ring back to us
for reexamination. Eventually, we
were able to convince him of the
accuracy of our conclusion.
is confusion was due to a distinct separation of colors in the girdle
plane that made the sapphire appear
as if it had been assembled, with a
blue crown and a colorless pavilion.
We explained to him that this separation was caused bv well-defined blue
and colorless zoning, not by t h e
boundary between two pieces of an
assemblage. Our original observations revealed fluid-filled "fingerprints" that were perpendicular to
the girdle plane and extended across
it into both the crown and the pavilion. Furthermore, when the ring was
immersed in methylene iodide, a routine procedure for all corundum that
we examine, the side view [figure 16)
showed both color zoning and the
lack of a cement plane, the latter an
unavoidable feature of any sapphiresapphire doublet.

126

Gem Trade Lab Notes

With an Unusual Star

As heat-treated rubies and sapphires
becoming increasingly common, it is
a pleasure to see inclusions that
prove that a natural sapphire has not
been treated. The gemologist gets an
added bonus when the arrangement
of these diagnostic inclusions also
creates a beautiful internal scene.
Such was the case when a 2.22 ct

Figure 16. A side view of the

stone in figure 15 shows that the
color is confined to the crown,
but there is no cement plane,
As can be seen with darkfield
illumination (figure 18)) the rutile
formed a six-spoked stellate pattern
of thin, white-appearing needles that
are oriented in a planar arrangement
perpendicular to the optic axis. There
is little or no rutile between the
spokes. The six wedge- or V-shaped
areas that are devoid of rutile needles
extend outward from the center of the
Figure 17. Unlike the stone in figures 15 and 16, color-zoned sapphires more typically reveal color
confined to the pavilion, toward
the culet, as illustrated by the
stone shown here (immersed in
methylene iodide). Such stones
usually appear evenly colored
when viewed face up.

Figure 15. Our client mistakenly

thought that this 8.2 x 6.6 mm natural sapphire was an assemblage.

cushion-shaped, modified brilliantcut sapphire was submitted to the

West Coast Gem Trade Laboratory for
identification.
The transparent blue gem contained both fluid inclusions, which
proved that the stone had not been
heated to the temperature needed to
alter its color, and a bold display of
rutile needles that painted an obvious
picture. The specific arrangement of
the rutile needles and their orientation to t h e table facet made this
stone unique.

GEMS & GEMOLOGY

Summer 1996

Figure 18. The spoked arrangement of the mtile n e e d l d i g u r e 19. The star pattern becomes iridescent when
i n this natural sapphire creates a pleasing star pattern,
illuminated by a strong fiber-optic light directed
but not true phenomenal asterism. Magnified l o x .
through the table facet from above. Magnified W x .
"star" formation, which results in the
isolated rutile star seen in this stone.
Another interesting feature of
this stone was visible because the
plane of the table facet was cut just
off-parallel to the plane of the rutile
star. This orientation results i n a
beautiful iridescent display by the

rutile needles when they are illuminated through

- the table from above by
intense fiber-optic lighting (figure 19).
iridescent needles have
been observed many times in the
past, the combination in this stone of
the vibrant iridescence and the
unusual spoked star pattern make

this a most pleasing inclusion display.

fohn I. Koivula
PHOrOmBUTS

Nicholas DelRe supplied the pictures used in figures

,_,o and 12-16 ~i~~~~11 was taken by ~~h~
DeMaggio. Figure 17is courtesy om GIAslide library,
photographer unknown. The photomicrographs in figures18andl9weretakenbyJohnl.Koiwla.

I N R E S P O N S E T O P O P U L A R D E M A N D ... 1

Learn-At-Home Course #52

GIA is proud to bring you one of
the most eagerly awaited LearnAt-Home courses in recent
memory. Now you can get the
training you need to properly
prepare insurance replacement
appraisals,
h

For more information or tai enroll

Call (800) 421-7250 6xt. 292
Outside the U.S.
(310) 829-2991 ext. 292
Fax (3 10) 453-7674

Gem Trade Lab Notes

Videotape included!

GEMS & GEMOLOGY

ij

Summer 1996

GEM
I

Editors a Mary L. Johnson and John I. Koivula

Contributing Editors
Dino DeGhionno, Robert C. Kammerling, Shane F, McClure,
GIA GTL, Santa Monica, California
Henry A. Hanni, SSEF, Basel, Switzerland
Karl Schmetzer, Petershausen, Germany

DIAMONDS
Diamond with a fluorescent phantom crystal. Bob Lynn,
of Lynn's Jewelry, Ventura, California, found an unusual
faceted diamond in his stock and loaned it to the editors
for examination. The 1.00 ct near-colorless round brilliant
looked normal under standard fluorescent and incandescent illumination; however, when the stone was illuminated by (long- or short-wave)ultraviolet radiation (figure
I), a fluorescent internal octahedral crystal was easily
seen. Higher magnification revealed that the fluorescence
in the phantom crystal was zoned [figure 1, right).
In other respects, the diamond was not notable.
With the standard fluorescent lighting used for grading
purposes, the included crystal was visible only as transparent graining and did not set the clarity grade of the
stone. In polarized light, the inclusion appeared as a
region of concentrated strain lines; it did not form an
obvious phantom (unlike the strain phantoms illustrated
in, for example, the Fall 1993 [pp. 199-2001 and Summer
1995 [pp. 120-1211 Lab Notes sections].
Tairus synthetic diamonds in jewelry. Among the many
synthetic materials produced by Tairus (a joint venture
between the Siberian Branch of the Russian Academy of
Sciences and Pinky Trading Co., Bangkok, Thailand) are
synthetic diamonds grown in Novosibirsk. Tairus yellow

synthetic diamonds are now being marketed in jewelry

(figure 2) by Superings, Los Angeles, California. To the
best of our knowledge, this is the first example of synthetic diamonds being marketed as such in jewelry.
Although die jewelry is not yet available in mass-market
quantities, advertising brochures that were distributed by
Superings this past spring generated 2,000 calls for further information on these goods in one week, according
to Walter Barshai, president of Pinky Trading. Individual
samples (mounted in rings) seen at the JC-K Show in Las
Vegas this past June ranged between 0.15 and 0.25 ct.
Mr. Barshai also provided details about Tairus's production of synthetic diamonds. Near-colorless, yellow,
and blue synthetic diamonds are being produced at their
factory in Novosibirsk using the BARS method; also, the
yellow synthetics can be treated [by irradiation followed
by heat treatment) to produce pink and red colors (see,
e.g., Gem Trade Lab Notes, Spring 1995, pp. 53-54). As of
early June 1996, Tairus was producing 10-20 colored
synthetic diamond crystals per month, but the company
hopes to increase production to 100-300 crystals per
month by winter, and possibly to 1,000 crystals per
month by the summer of 1997. The largest crystal grown
by Tairus thus far weighs about 2.5 ct, with most crystals
weighing between 1 and 2 ct; by late 1997, synthetic diamond crystals up to 5 ct may be produced. Growth times

Gem News

Figure 1. This 2.00 ct diamond

(about 6.5 mm in diameter) contained an octahedral phantom
crystal that was visible only
when the stone was exposed to
UVradiation. With higher magnification (right, 7x), the inclzzcled crystal (also seen reflected in
the table facet) showed a zoned
fluorescence pattern. Photomicrographs, taken with shortwave UV illuminaiion, by
Shane Elen, GIA Research.

GEMS & GEMOLOGY

Summer 1996

are surprisingly fast: in theory, a 5 ct crystal could be

grown in three days. Most synthetic diamonds as grown
are light yellow-brown to saturated yellow, but blue crystals can be grown if boron is added. Near-colorless stones
require additives such as zirconium and special growing
conditions; the largest near-colorless crystals grown by
Tairus so far weigh only about 0.10 ct.
Approximately 100 carats of Tairus synthetic diainonds-of any color-have been placed in the market to
date. Mr. Barshai believes that the market will be much
more tolerant of colored than near-colorless synthetic
diamonds, so Tairus is concentrating on the production
of colored synthetic diamonds.

COLORED STONES AND

ORNAMENTAL MATERIALS
Anorthosite rock with uvarovite garnet "speckles." In
September 1994, contributing editor Henry Hanni
received some pieces of rough from Tay Thye Sun, a
Singapore gemologist. One of Mr. Sun's clients acquired
the material in the Philippines because it looked like
jade. On the weathered skin of the predominantly white
samples, deep green idiomorphic crystals were visible.
These 0.2-2.0 m m crystals showed excellent rhombic
dodecahedron shapes (figure 3, left). Microscopic examination of a polished surface in reflected light (figure 3,
right) revealed a granular groundmass that was occasionally interspersed with idiomorphic crystals (about 5% of
the volume of the rock). The green crystals had a much
higher luster than the matrix.
Raman spectroscopy quickly revealeqt-hat the
groundmass was plagioclase and the crystals w 're garnets. Examination with an electron microprobe (
lyst J. Partzsch) determined the exact chemical composition. In fact, the groundmass was two different plagioclases: The main mass consisted of almost pure anorthite
( A b m h 8 :2 mol.% albite [Ab] and 98 mol.% anorthite
[An]),but oligoclase (AbT7Anz3)was present between
these grains. The garnet crystals were strongly zoned;
they were composed primarily of uvarovite with varying
amounts of andradite in solution.
These pieces of rough may be the same material as
the snuff bottle that was described and illustrated in the

I.aria-

Figure 2. These three platinum rings contain fashioned yellow synthetic diamonds (0.30-0.40 ct).
Courtesy of Superings, Los Angeles; photo by
Shane F. McClure.
Spring 1994 Lab Notes section (pp. 42-43). The specific
gravity of the rough material was 2.80, compared to the
2.76 reported for the snuff bottle.
Update on porous chrysocolla-colored chalcedony. In the
Spring 1992 Gem News column, we reported on chalcedony colored by chrysocolla that changed appearance
when soaked in water for several hours (iiChrysocollacolored Chalcedony from Mexico," pp. 59-60). In that
material, the soaking caused the blue color to intensify,
the material to become more transparent (less opaque),
and the piece to gain weight.
Late last year, Chris Boyd of CB Gems &. Minerals,
Scottsdale, Arizona, brought to our attention chalcedony
colored by chrysocolla from Arizona, which showed a
similar change in color and transparency that can also be
produced by humidity alone. According to Mr. Boyd, this
transition happens at about 55% humidity. Stones are "a
nice semi-transparent blue" when exposed to humidity

Figure 3. The green hexagons in

this polished section of rock from
the Philippines (left)are uvarovite
garnets; the largest garnet is 2 mm
across. In reflected light (right),
the high-relief garnets (brightest
reflectance) can be distinguished
easily from the groundmass of
anorthite feldspar (medium gray)
and oligoclase feldspar (dark
gray). Photomicrographs courtesy
of the Swiss Gemmological
Institute (SSEF), Basel.

Gem News

GEMS & GEMOLOGY

Summer 1996

129

(Mg,Fe+2]5Al(Si3Al)Oio(OHi8,
a member of the chlorite
mineral group. The source of this clinochlore is the
Korshunovslzoye deposit, on the east side of the Bratsk
Reservoir, east of Lake Baikal, in Siberia, according to an
article by A. A. Evseev in the Russian mineralogical magazine World of Stones (1994, No. 4, p. 44). Standard mineralogical references give the hardness of clinochlore as
2-2.5. One sample had been wax treated.

Figure 4.This cabochon and the tablet are made

of clinochlore, a chlorite mineral. Photo by
Maha DeMaggio.
of 55% or greater, or if soaked in water; when these same
stones are exposed to lower humidity, or left under a
heat lamp, they lose color intensity and transparency.
This process is repeatable.
Mr. Boyd also said that, in his experience, almost
90% of Arizona chrysocolla-colored chalcedony will dry
out if exposed to less than 55% humidity.
Chatoyant clinochlore from Russia. One of the interesting "phenomenal" materials seen at the 1995 and 1996
Tucson shows was a chatoyant chlorite-group mineral
from Russia. Two sample7.65 ct triangular tablet and
a 6.06 ct cabochon (figure4)wer obtained from Heaven
and Earth, Marshfield, Vermont, which was marketing
this material as "Seraphinite." Gemological properties
were as follows: color-green; diaphaneity-semi-translucent; color distribution-uneven (mottled, fibrous);
pleochroism (visiblein some more transparent areas) yellow-greenbluish green; optic characterdoubl refractive
with an aggregate structure. Individual fibrous crystals in
the aggregates appeared chatoyant, with a pearly luster,
One stone had poorly discernible R.1.k of 1.580 and 1.585,
and the other gave a spot reading of 1.58. Specific gravities
were 2.62 and 2.66. Uneven luminescence was seen to
both wavelengths of UV radiation: weak yellowish white
to whitish yellow for long-wave UV,and very weak white
to short-wave UV.In the spectroscope, one stone showed
a cutoff edge at about 460 nm and a weak band centered
at 500 nmjthe other showed these bands as well as a faint
band at 470 and 545 nin, and a cutoff edge at about 690
nm. With magnification, both stones showed a radial
fibrous structure. Qualitative energy-dispersive X-ray fluorescence (EDXRF)spectroscopy revealed numerous elements in both: Mg, Al, Si, K, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu
i n one sample),Zn, and Ga.
X-ray powder diffraction patterns for both samples
were consistent with that of the mineral clinochlore,

130

Gem News

Cat's-eye diopside from Southern India. Many materials

that contain parallel acicular inclusions can be fashioned
into cabochons to show a cat's-eye effect. Among the
green cat's-eye stones are apatite, chrysoberyl, demantoid
garnet, emerald, opal, tourmaline, zircon, and zoisite. We
were delighted to see recently an unusually bright green
example of another material-cat's-eye diopside (figure
5)-from a relatively new locality.
The 4.59 ct oval cabochon (12.70 x 5.70 x 6.57 mm),
which was loaned to the editors by Maxam Magnata of
Fairfield, California, was a translucent dark green, with
weak pleochroism of green and slightly brownish green. Its
optic character was aggregateshowing the presence of
many individual grains. The refractive index measured
1.67, and a "blink" was seen at 1.70. The stone was inert
to both long- and short-wave UV.With the spectroscope,
we saw weak general absorption between 400 and almost
500 nm, with two lines at 495 (weaker)and 505 (stronger)
run. Three "chrome lines" were also seen at 640,660, and
680 nm; despite this evidence of chromium, there was no
reaction to the Chelsea color filter. These properties are
consistent with those previously reported for chrome diopside. Magnification revealed a fibrous texture, with some
coarse channels (or,possibly, needles)present.
The material comes from a relatively new mine near
the small town of Iddipadi, in southern India, close to the
Figure 5. This 4.59ct cat's-eye diopside is from a
relatively new find in southern India, near the
town of Iddipadi. Stone courtesy of Maxam
Magnata; photo by Maha DeMaggio.

GEMS & GEMOLOGY

Summer 1996

Figure 6. Mozambique is the country of origin for

these two garnets (1.26 and 1.38 ct), identified as
pyrope-almtindine. Photo by Mciha DeMaggio.

Figure 7. Acicular inclusions were the only internal feaHire seen in the Mozambique garnets shown in figure 6.
Photomicrograph by John I. Koivula; magnified 30x.

cutting center of Karur Kangayam, according to Mr.

Magnata. He believes the find was first made about six
years ago, with the bulk of the cutting material sent to
Hong Kong until recently. The mine is reportedly within
40 m of star diopside and cat's-eye enstatite deposits.

Russia. Mr. Shade acquired the material through an

(undisclosed)American company that imports stones for
the building industry; it was originally imported for use
as kitchen and bathroom tiles.
The gemological properties of our test sample were: clarity-semi-translucent to opaque; aggregate structure; R.1.about 1.66 (some areas, 1.655 to 1.670; others, 1.48-1 ,501;
S.G.4.29; luminescence-moderate mottled blue-white to
long-wave UV radiation, and very weak to weak mottled
blue-white to short-wave UV; typical green jadeite spectrum
when viewed with a handheld spectroscope.
Further tests helped explain some of these properties. X-ray powder diffraction analysis of scrapings from
two areas gave patterns indicating jadeite (predominantly) and minor natrolite. The natrolite, which accounted
for the 1.48-1.50 R.I. readings and the slightly low S.G.,
occupied about 5 % of t h e sample by volume. T h e
infrared absorption spectrum of a thin slice that had been
cut and polished from this material showed no evidence
of polymer impregnation. However, two small "wax
peaks" were found at 2926 and 2855 cm-1. These probably resulted from the oil used in cutting the thin slice.

Garnets from Mozambique. African garnets were very

much in evidence at the 1996 Tucson shows, including
brownish yellow to bright green grossular-andradites
from Mali (see, e.g., "Gem-Quality Grossular-Andradite:
A New Garnet from Mali," Gems e) Gemology, Fall
1995, pp. 152-167), orange spessartines from Namibia
(Gem News, Spring 1996, pp. 56-57), and various colorchange garnets from Tanzania (Gem News, Spring 1996,
p. 53). A number of dealers also offered dark orangy red to
slightly brownish red stones from still another locality,
the southern African country of Mozambique.
Although this locality is not new, garnets from
Mozambique have not been characterized in the gemological literature. T o do just that, we purchased two
brownish red modified triangular brilliant cuts (1.26 and
1.38 ct; figure 6) on which we determined the following
properties (where there are two sets of values, the value
for the 1.26 ct stone is first): R.1.-1.770, 1.776; S.G.
(determined hydrostatically)-3.96, 3.89; singly refractive
with strong anomalous birefringence noted between
crossed polarizers; inert to both long- and short-wave UV
radiation; and spectra (determined with a desk-model
prism spectroscope) exhibiting strong bands at about 504,
520, and 573 nm. Magnification revealed a few acicular
rutile crystals in one stone (figure 7). On the basis of
these properties, and using criteria established by C. M.
S t o c k t o n a n d D. V. M a n s o n ("A Proposed N e w
Classification for Gem-quality Garnets," G e m s ed
Gemology, Winter 1985, pp. 205-218)) we identified the
garnets as pyrope-almandine.
Green-and-white jadeite from Russia. Ross Shade of the
Addexton Company, Auburn, California, provided us
with a 113.90 c t sample of green-and-white mottled
jadeite (figure 8 ) that reportedly c a m e from central

Gem News

Visit to Mvanma Gems Museum. The October 1993

Emporium was the first such event held in the new
Myanma Gems Enterprise (MGE) exhibition hall in
Yangon (Rangoon), which was built solely for these biannual sales. In the summer of 1995, the Myanma Gems
Museum opened in a building adjacent to the exhibition
hall. One of the Gem News editors (RCK) visited this
museum in November 1995 and provided the following
report on this new facility.
Each of the first three floors consists of a large open
showroom containing retail sales outlets for firms licensed
by the Myanmar government to sell gems and jewelry (all
sales require MGE receipts like those used at the emporia).
Ornamental objects fashioned from silver were offered, in
addition to loose gems and gem-set jewelry.
The top floor of the building houses the actual inuseuin, which contains many interesting displays. Jadeite is
very well represented. Along one wall are sawn jadeite

GEMS & GEMOLOGY

Summer 1996

I
Figme 8. This 113.90 ct block of jadeite (here cut
into three pieces) reportedly came from central
Russia. Photo by Muha DeMaggio.
boulders, each identified by its locality. A display of cabochons illustrates the three categories of jadeite-"imperiall" "commercial," and "utilityn-used in
markets. An exceptional 4.7 kg boulder of imperial jadeite
has its own display case. There are also attractive displays
of fashioned jadeite, including a plate with chopsticks
(total weight 2,390 g) and the remaining 106 kg "water
jade" boulder from which they were cut.
Not surprisingly, there are a number of interesting
ruby displays, including material from Mong Hsu, from
the famous marble quarry at Sagyin in Mandalay
Division, and from Pyin Lon in northern Shan State.
Among the noteworthy gem-set items is a peacock fashioned from 61.86 g of gold and 256.84 ct of sapphires atop
a jadeite base. Also on display is a large gold pendant set
with a 329 ct oblong peridot cabochon (labeled as the
world's largest cut peridot) and many smaller peridots.
Other cases highlight t h e range of gems found i n
Myanmar, including zircon, iolite, danburite, diopside,
tourmaline, garnet, lapis lazuli, moonstone, enstatite,
albite, and fluorite. Cultured pearls from the gold-lipped
oyster are also well represented.
A recent report provided by U T i n Hlaing of
Taunggyi University, Myanmar (pers. coinm., May
1996), lists the following large gems as being among
those currently on display at the museum: a 21,450 ct
(17.5 cm long x 11 cm wide x 10 cm high) rough ruby
from Block No. 14 of the Mogolz Stone Tract, which
Myanmar officials believe to be the world's largest; a
63,000 ct sapphire from Mogolz; and a 23,500 ct piece of
jade from the Hlzamti area.

anma mar gem

Variegated translucent jasper from Mexico. One of the

more notable trends at the Tucson gem shows this year
was the rising interest in ornamental materials. In addition to the exotic materials seen-such as drusy vanadinite and clinochlore-varieties of quartz and chalcedony

132

Gem News

were quite prominent, especially those with aesthetic

inclusions or patterns. One interesting example was socalled Imperial jasper (figure 9), reportedly mined in
Mexico. Cathy Cwynar, of Minneapolis, Minnesota,
loaned us three fashioned pieces for characterization.
The gemological properties of the three free-form
cabochons (15.43, 16.28, and 17.86 ct) were as follows:
color and clarity-variegated between translucent-toopaque brownish (olive) green to green in some regions
and to opaque greenish white in others; optic characteraggregate reaction; R.I. -1.54 in the green area, 1.53 to
1.55 in the white areas; S.G.-2.43 to 2.51 (the white
regions were porous, and the more white material there
was in the piece, the lower the S.G. was). For the most
part, this material was inert to both long- and short-wave
UV radiation, but some whitish areas showed a weak
blue fluorescence to long-wave UV. Only the darkest
(olive) green regions showed a distinct absorption spectruin in the handheld spectroscope, with a lower-wavelength cutoff at about 460 nm, and faint bands at 500 and
520 nm. Other than the typical chalcedony structure, no
inclusions were evident with magnification,
Of particular note was the material's unusual behavior in transmitted light: The dark green areas were
translucent, with yellow "rims" between the green and
opaque areas (figure 9, right)-an effect not evident in
reflected light. Qualitative EDXRF spectroscopy revealed
no significant differences in chemistry between the
white and green regions of one sample examined; both
areas showed major silicon, minor potassium, calcium,
iron, and titanium, and trace amounts of manganese,
zinc, rubidium, and strontium.
An unusual morganite. In fashioned gemstones, fractures
are generally considered very undesirable. One exception,
however, is found in the variety of rock-crystal quartz
known as "iris quartz" (not to be confused with "iris
agate"), in which numerous air-filled fractures behave
like thin films, imparting a rainbow-like iridescence
when the gem is illuminated from above.
Recently, Edward Swoboda, a gem and mineral
miner and dealer from Beverly Hills, California, showed
the editors another form of "iris" gem. The light pink
stone, a 157.19 ct oval step-cut morganite beryl, measured 42.69 x 32.34 x 19.28 mm.
During cutting, instead of removing the fractures
and making a number of smaller stones from the rough,
the lapidary went against convention and fashioned a
single large stone, positioning the fractures so that their
plane was just off parallel to the table facet. In this way,
light reflected from the fractures was not simultaneously
reflected from the table facet, which would otherwise
mask or diminish the "iris" effect. As figure 10 shows,
this cutting technique produced a morganite with a table
that is vividly iridescent when the stone is viewed from
above in virtually any direction. As the stone and/or light
source is moved, both the pattern and position of the various colors change dramatically.
GEMS & GEMOLOGY

Summer 1996

Figure 9. Among the m a n y

ornamental gem materials
seen at the 1996 Tucson
gem shows was this 17.86
ct cabochon of variegated
jasper, fashioned b y Cathy
Cwynar, shown on the left
i n reflected light. In transmitted light (right),the
dark green regions are
translucent, revealing yell o w rims that are not visible in reflected light.
Photos b y Maha
DeMaggio.
Scottish pearls. One of the most interesting items seen at
the February Tucson shows was a double strand of graduated pearls that ranged from 4 to 9 mm on one strand, and
4 to 10 m m on the other (figure 11). The pearls were
accompanied by a report from t h e Gemmological
Association and Gem Testing Laboratory of Great Britain
in which they were identified as natural. Of particular
interest, however, was their provenance: All were collected from freshwater mollusk; in the rivers of Scotland, by
veteran pearl fisher Bill Abernethy. These pearls, reportedly the result of 40 years of pearl fishing, are particularly
well-matched rounds and show the attractive luster that is
characteristic of Scottish pearls (see, e.g., E. A. Jobbins and
I<. Scarratt, "Some Aspects of Pearl Production . . . ,I'
fournal of Gemmology, Spring 1990, pp. 5-7). The most
important Scottish pearl recovered to date, called the
Abemethy pearl, is 11.5 mm in diameter.
Rubies from Madagascar. Blue sapphires from
Madagascar have received attention in the trade press
since their 1992 arrival on the gemstone market (see the
article by D. Schwarz et al. in this issue). However,
southern Madagascar is also producing gem-quality
rubies, according to Christopher P. Smith of the Gubelin
Gemmological Laboratory, Lucerne, Switzerland.
Although most fashioned stones weigh 1 ct or less, some
as large as 20 ct have been encountered. Recently, the
Gubelin laboratory had the opportunity to examine more
than 40 faceted Madagascar rubies. These rubies had a
richly saturated, even color, most resembled Thai or
Cambodian rubies in this respect, although some lightertoned stones looked like rubies from Burma or Vietnam.
Mr. Smith noted a distinctive combination of internal features in these rubies. Most had dense concentrations of fine needles and platelets that, for the most part,
were dull in luster and slightly blue or white in color;
they lacked the highly iridescent appearance of rutile
needles and platelets observed in rubies from Burma, Sri
Lanka, or Vietnam. (It is, therefore, more likely that they
are ilmenite or hematite.) Doubly truncated hexagonal
dipyramidal negative crystals, surrounded by equatorial

Gem News

thin films or fractures, were another common feature

(figure 12). These inclusions are reminiscent of those
seen in Thai and Cambodian rubies. In addition, the
Madagascar rubies revealed very prominent internal
growth structures that gave a roiled appearance with l o x
magnification. Many of the rubies also had a distinctive
series of linear growth boundaries that created the
appearance of a "cellular structure" (figure 13).
Semi-quantitative chemical analysis performed on
30 samples revealed the following contents: 97.3-99.1
wt.% A1203; 0.27-1.68 wt.% Cr203; 0.45-0.74 wt.%
Fe203; 0.0-0.07 wt.% Ti02; 0.0-0.02 wt.% V20r,; and
0.0-0.01 wt.% Ga203.
From his extensive 1922-23 studies of the gem
deposits of Madagascar, Lacroix described corundum
deposits (ruby and sapphire) in the northern and central
parts of t h e island [Mineralogie d e Madagascar,
Challamel Ed., Paris, 3 Vols.). Later, additional deposits
were noted by Besairie in 1956 (Carte g o l o g i q u e au
1/200000 de Sakoa-Benenitra, 2nd ed.) and Noizet in

Figure 10. The lapidary fashioned this 157.19 ct morganite to display a pleasing iridescent, or "in's,"
effect. Photo b y Maha DeMaggio.

GEMS & GEMOLOGY

Summer 1996

Figure 11. The 154 pearls (4-10 mm) in this double strand were reportedly fished from freshwater mussels in Scottish waters by Bill Abernethy,
Courtesy of the Susan Clarli Gallery of Gem Art,
Vancouver, BC, Canada.

1958 (Carte gdologique azz 1/200000 de AmpanihyBekily, 2nd ed.], in the southern portion of the island. In
a 1992 geologic report, Dr. Alain Mercier placed the ruby
deposits east-southeast of the town of Tulear, in the
Fotadrevo-Vohibory region of southern Madagascar, concentrated along northeast-trending faults. Some of the
principal deposits include Anavoha, Vohitany, Volubory,
Marolinta, and Ianapcra.
Sapphire mining in Kanchanaburi. The Winter 1990
Gem News section contained an entry on S.A.P. Mining
Co., a large, mechanized sapphire-mining operation in
Thailand's Kanchanaburi Province, about 100 lzm (60
miles) west of Bangkok. Mining takes place near Bo
Phloi, a town about 40 lzm (25 miles] north of Kanchanaburi City. Last fall, one of the editors (RCK)revisited S.A.P. Mining and obtained updated information
about that operation.
S.A.P. Mining has determined that blaclz spinel is the
best indicator mineral for finding sapphires, according to
exploration and mining manager Rouay Limsuvan. In general, the larger the spinels found in an area are, the larger
the sapphires recovered will be. Dark brown (to blaclz)
pyroxene also has proved to be a good indicator mineral.
All of the sapphires recovered to date have been from secondary (alluvial)deposits. The actual gem-bearing gravel
Gem News

layers are typically about 1 m thick. They occur as shallow as 5-7 in, or as deep as 13-15 m, below the surface.
Once a prospected area has been selected for mining,
modem excavators strip the overburden to reach the gembearing gravels, which are then loaded on large dump
trucks and talzen to the firm's modern processing plant. In
a single day, some 250 trucldoads of gravel-1,000 in3 of
material-are transported to the processing plant. The
gravels are first dumped into a sheet metal-walled pen.
High-pressure water cannons help wash material out of
the trucks, remove the clay and sand from the gravels,
and force the gravels via chute into a large rotating trommel, which removes the bigger rocks. The gravels then
pass through one of two rotating tromn~elswith a sinaller mesh size (figure 14), after which they are processed in
one of three vibrating jig systems [each of which consists
of six jigs). The resulting heavy concentrate is then collected and taken to an on-site facility for hand sorting. The
amount of rough sapphire recovered daily varies widely,
from 200 grams to 2 kg. While the goal is to recover corundum, also found are sanidine, blaclz spinel, garnet, magnetite, pyroxene, and olivine.
Although S.A.P. Mining is still very active, with a
claim area covering 250-300 km2, the overall level of
mining activity in the area has decreased in recent years.
In the past decade, no fewer than 47 companies have
been involved in mining for sapphires in the area.
Currently, however, only four firms work the Bo Phloi
area; a year earlier, there were seven active operations,
according to Mr. Limsuvan. As S.A.P. Mining plans for
the eventual end of mining operations on its holdings, it
is carrying out an extensive reclamation program. It has

Figure 12. This type of negative crystal-a doubly

terminated hexagonal &pyramid with a surrounding equatorial thin film/fracture planewas a common feature in the Madagascar rubies
examined. Similar inclusions are found in rubies
)?om Thailand and Cambodia. Photomicrograph
by Christopher P. Smith; magnified 20x.

GEMS & GEMOLOGY

Summer 1996

already developed mined-out portions of the lease area

into a golf course.
Update on Tanzanite mining. Gem dealer Michael
Nemeth, of San Diego, California, has provided an
update on tanzanite mining at Merelani, Tanzania, following a visit there in late 1995. At that time, the D
bloclz contained about 200 mines and over 2,000 miners
(see Gem News, Winter 1991, p. 262, for a description of
the four main "blocks" for which mining rights have
been granted at Merelani). "Vietnam Camp," a mine in
the D bloclz, was being worked by 28 miners and extended to 3,000 in in length. Mr. Nemeth learned that overall
production had been good, and a "7 kg crystal" with
excellent color and clarity had been found recently.
However, local prices seemed high, probably inflated by
the presence of foreign buyers.
The Vietnam Camp mine was one of the few mines
with a wooden ladder instead of a rope for access (figure
15).In general, though, mining conditions were primitive
(the typical miner's headgear was a flashlight strapped on
with a thick band of rubber). Throughout the visit, Mr.
Nemeth heard and felt numerous dynamite blasts, after
which glittering, graphite-rich dust filled the air of the
mine shaft. He was told that "a few miners" are killed
weekly by poor ventilation or by the blasts themselves.
Some miners stay underground for two to three days at a
time, eating only salt and zigali (a cornmeal-like porridge
compressed into cakes in one's hand).
Figure 16 was talzen at a depth about 250 m, at which
point the mine shaft became less vertical. Further descent
was accomplished by crawling head first down the steeply
sloping adit. Over the next 50 m, the shaft narrowed from
1 m to about 60 cm (three to two feet),and turning around

Figure 13. A distinctive cellular structure is apparent in this ruby from Madagascar. The linear
growth features separate zones filled with fine
needles and platelets (probably ilmenite or
hematite). Photomicrograph b y Christopher P.
Smith; magnified 25x.

Gem News

Figure 14. A t S.A.P. Mining's modem washing plant near

Bo Phloi in Kanchanabwi Province, Thailand, a huge
rotaiing trommel (foreground,right) removes the larger
rocks. The gravels then pass through one of two other
trommels with smaller mesh sizes, one of which is a1
left. Photo by Robert C. Kammerling.

was impossible. Mr. Nemeth and his party stopped at this

point, since the miners above ground had talzen a lunch
break and stopped pumping air below.
Mr. Nemeth was told that the miners were paid
only with stones found in the mine, not currency. The
age-old practice of high-grading by miners to supplement
their meager incomes was tolerated as long as amounts
were kept to a "single handful," as in general only the
mine manager, not the owner, was present.
"Golden" tourmaline from Kenya. Tourmaline from a
new find in Kenya is being described as "savannah" or
"golden" tourmaline. Gemologist Yianni Melas recently
sent us two samples of rough material for examination
(figure 17). The larger (4.50 ct) piece was transparent
brownish orangy yellow, with some green areas. The
smaller (3.06 ct) piece was a crystal section with a brownish orangy yellow core and a green rim. External morphology of the (broken)crystal section consisted of prism faces
only, which were not striated parallel to the c-axis.
We recorded the following gemological properties on
the two samples: pleochroism-brownish orange (0) to
yellow (e)in yellow regions, and dark green (0) to light
yellowish green (E) in green regions; optic character-uniaxial; color-filter reaction-none (yellow regions), red
(green regions); R.1.-1.630 to 1.650, and birefringence0.020 (measured on green rim only); S.G.4.05; fluorescence-inert to long-wave UV radiation (both colors),
moderately chalky, moderately strong yellow (to shortGEMS & GEMOLOGY

Summer 1996

revealed a small anhedral crystal and a needle (or tube) in

the larger piece; the smaller piece contained two-phase
(liquidlgas)inclusions in healed fractures ("fingerprints")
and hollow tubes.
Qualitative EDXRF spectroscopy revealed Mg, All
Si, K, Ca, Ti, Fe, Zn, Ga, and Sr in both samples, with a
small M n peak in one spectrum. Cr was also detected in
the green crust of the smaller piece, and the color-filter
reaction and absorption spectrum were consistent with
the presence of Cr, which causes the green color.
Three possible sources have been cited in the trade
press for this or similar material: (11the Kasigau Mine, 450
km from Nairobi in Kenya's Tsavo West National Park
["What's in a NarneSale for Brown Tourmaline" I.C.A.
Gazette, April 1995, p. 71, (2)the Mangari area of Tsavo
West National Park, and (3)M'gama Ridge, south of the
Taita Hills in south eastern Kenya (the latter two described
in "New Find: East African Golden Tourmaline," Jewelers'
Circular-Keystone,February 1995, p. 66).

F w e 15. Gem dealer Michael Nemeth cautiously

descends the upper shaft of the Vietnam Camp tanzanite mine, in block D at Merelani. Miners use the
piece of wood in the foreground to rest during their
long descents and ascents. Photo courtesy of
Michael Nemeth.
wave W),with stronger fluorescence in the green rim of
one sample. Using a handheld spectroscope, we saw a
general absorption in both stones to about 480 nm; the
green rim also showed an absorption band between 590
and 630 nm, and lines at 650 and 680 nm. Magnification
Figure 16. The shaft of the Vietnam Camp mine continues to narrow as i t approaches 300 m depth.
Photo courtesy of Michael Nemeth.

TREATMENTS
Novel opal doublets from Brazil. In the editors' experience, two basic types of doublets using natural opal are
encountered in the trade. The first type uses a blaclz
material-such as dyed chalcedony ("blaclz onyx") or
glass-as the backing material. (When the opal is fairly
thin and translucent, the backing provides a dark baclzground that accentuates the opal's play-of-color.) The
second type of opal doublet uses as the backing material
sandstone or ironstone matrix, typical of that seen at
opal deposits, to reinforce the opal. If examined carelessly (or if mounted to obscure the demarlzation between
the two components), this second type can be mistaken
for natural boulder opal. The opal in most doublets of
both types that we have seen is usually Australian.
While examining gems at the Tucson booth of a
Brazilian firm, one editor came across a type of opal doublet that he had not seen previously. These stones had
been assembled from two Brazilian gem materials: milky
white opal from Piaui State that was backed with
sodalite from Bahia State (figure 181. The dark violetish
blue of the sodalite provided a good contrast for the
opal's weak play-of-color. It also produced a background
color similar to that seen in many opal triplets.
As at many other gem shows, the editors also
encountered assemblages made with synthetic opal. For
example, Manning International, of New York, was offering both doublets and triplets constructed from recentproduction Gilson synthetic opal. The backing material
on the doublets was ironstone matrix from Australia.
Sapphires with "induced" stars. At the February 1996
Tucson show, one of the editors (DDG)saw several star
sapphires. Accompanying literature claimed that the
asterism was "indu~ed,'~
but without "additional chemicals." Since no "chemicals" were involved, the literature
maintained, this process was different from diffusion
treatment. Curious, the editor obtained some of these

136

Gem News

GEMS & GEMOLOGY

Summer 1996

sapphires for examination (see, eg., figure 19).One end of

a stone was polished flat so that the properties of the
interior could be compared to those of the "skin" region.
Gemological testing confirmed that the material
was sapphire. With n~agnification,"angulargrowth layers
and two-phase and "fingerprint" inclusions were visible.
When the stone was examined with fiber-optic illumination, silk in the outer layer looked similar to the silk
seen in synthetic star sapphires. When the stone was
immersed in a heavy liquid, a shallow brownish surface
layer was evident.
To determine whether the treatment that induced
the asterism had changed the stone's chemistry, we
examined both the surface and the polished, exposed
interior of the stone with quantitative EDXRF spectroscopy. (The penetration depth for X-rays is about 0.1
mm under our standard operating conditions.)We found
five times more titanium on the surface of the stone than
in its interior (0.19 wt.% Ti as Ti02 on the surface; 0.04
wt.% TiOl in the interior.)
Because of the significant difference in titanium content between the two regions, we concluded that the
TiOi "silk" must have been produced by diffusing titanium into the surface (despite the seller's claim to the contrary). However, diffusion treatment may not have been
the intended result. Dr. John L. Emmett (co-author of
"Heating the Sapphires of Rock Creek, Montana," Gems
&1 Gemology, Winter 1993, pp. 250-272) remarked
recently that he knew of a treater who packed some sapphires in a white titanium-dioxide-bearing powder for
heat treatment, hoping that the powder would slow cooling and prevent breakage. When he examined the finished stones, however, the unnamed treater was surprised to discover that his blue sapphires were now both
heat treated and diffusion treated.

SYNTHETICS AND SIMULANTS

Manufactured composite material imitating jadeite. A

great many materials are used to imitate jadeite. These
include such natural, untreated gem materials as massive

Figure 17. These two pieces of tourmaline rough (4.50

and 3.06 ct [the latter 12.60 x 6.67 x 5.63 mm])are
from a new find in Kenya. Photo by Maha DeMaggio.

Gem News

Figure 18, Opal from Piaui has been backed with

sodalite from Bahia to produce these Brazilian
opal doublets. Photo by Maha DeMaggio.

grossular garnet and serpentine; such wholly manufactured materials as glass and plastic; and such treated, natural gems as dyed massive quartz and-more relevant to the
following discussion-n~assive calcite that has been selectively dyed to resemble mottled green-and-white jade.
Last fall, one of the editors (RCK)acquired, from a
street vendor in a Chinese neighborhood of Los Angeles,
a 239.37 ct statuette (figure 20) that appeared quite siinilar to some of the mottled-colored, dyed calcite mentioned above. Gemological testing revealed the following
properties: diaphaneity-semi-translucent; polariscope
reaction-aggregate (appears light when rotated between
crossed polarizers); Chelsea color-filter reaction-none
(appears same color as the filter); specific gravity (determined by hydrostatic weighing)-1.98. Refractometer
testing produced a birefringence blink. With a handheld
spectroscope, green areas of the statuette showed a strong
band centered at 650 nm, a weak band at 600 nm, and a
faint absorption at 500 nm. The statuette luminesced a
chalky, moderate-to-strong blue-white to long-wave UV
radiation, and a chalky weak blue to short-wave UV. (The
green areas had a weaker fluorescence to both wavelengths than the white areas had.)
With a microscope, we saw that the material consisted of white grains in a groundmass of transparent,
colorless (or green) material with a resinous luster. X-ray
powder diffraction analysis identified the white grains as
calcite. The transparent groundmass burned readily
when touched with a thermal reaction tester, producing
an acrid odor and some liquid. Fourier-transform infrared
(FTIR) spectroscopy, performed on a pellet of material
scraped from the statuette, gave a spectrum that was
very similar to that of an alkyd resin polymer, a synthetic resin used, for example, in house paint.
On the basis of this information, we identified the
statuette as a calcite/plastic composite that had been
selectively dyed to resemble jadeite. It may have been
GEMS & GEMOLOGY

Summer 1996

137

produced in a mold that was filled with the pre-polymerized resin and finely ground calcite. Although the color
and color distribution made it a fairly convincing imitation of some mottled jadeite, both its resinous luster and
"heft" (1.98 S.G.) would raise the suspicions of anyone
familiar with jade.
"Drusy" silicon, a computer-industry by-product. The
Summer 1992 Gem News section (p. 133) discussed and
illustrated meteoritic iron used as a gem material. A manufactured material with a similar appearance is elemental
silicon (figure 21); at the Tucson shows this year, we saw
both fashioned and rough examples of this material.
Bill Heher, of Rare Earth Gallery, Trumbull,
Connecticut, loaned us a 31.5 x 40.3 x 5.54 mm elemental silicon doublet for examination. Only a few gem properties from this assembled triangular tablet could be
determined: color and luster-metallic gray (backing
material-yellow and brown); color distribution-even;
diaphaneity-opaque; fluorescence-inert to both longand short-wave UV radiation. The R.I. was over the limit
of our refractometer (greater than 1.81), and no absorption spectrum was detected using a desk-model spectroscope. We also purchased three chunks of elemental silicon from Gems Galore, Mountain View, California. One
had a hardness of about 6V2 and a specific gravity of 2.34
(hydrostatic method).
Some elemental silicon used in jewelry is different
from that used to make computer chips; for instance, one
distinguishing feature is the druse-like surface, consisting of many nearly flat-lying crystal faces, on some
pieces. (This is the feature that resembles the Widinanstatten pattern observed on some etched iron meteorite
surfaces, as in the Summer 1992 illustration.) To the best
of our knowledge, the "drusy" elemental silicon is
Figure 19. The star in this 3.73 ct (9.18 x 6.82 x
5.39 mm) sapphire was produced by a surface diffusion treatment. Photo by Maha DeMaggio.

Figure 20. This 239.37 ct imitation jadeite statuette (47.35 x 32.90 x 32.75 mm) is composed of
finely ground calcite in a gronndmass of allzyd
resin polymer (plastic). The green areas were
probably produced by selective dyeing, Photo by
Maha DeMaggio.

formed from t h e same starting material as t h e

Czochralslii-pulled elemental silicon boules that are the
source material for computer chips; the drusy crystals
form in the impurity-rich material left remaining in the
"pot" after the purified boule is extracted. The trace-element content of the clrusy material is thus higher than
that of the boules; the silicon surface of the doublet we
examined contained trace amounts of arsenic, as determined by EDXRF. Our one boule contained only silicon.
The fashioned tablet had been backed with another
material because elemental silicon is brittle. The material is popular in jewelry for people who work in computer
fields, according to a Gems Galore representative.
A convincing tanzanite substitute. Bill Vance, a gemologist from Denver, Colorado, recently loaned us one of the
most superficially convincing imitations of gem-quality
tanzanite that we have ever seen. This new tanzanite
substitute is now being sold at gem and mineral shows
across the United States. Mr. Vance purchased it at just
such a show in Denver in September 1995.

138

Gem News

GEMS & GEMOLOGY

Summer 1996

Figure 21. This tablet is assembled from drusy elemental silicon and an unidentified backing material. Elemental silicon (shown in its "raw" form on
the right) is a byproduct of computer-chip manufacturing. Tablet courtesy of Bill Heher, Rare Earth
Gallery;photo by Maha DeMaggio.

Natural tanzanite crystallizes in the orthorhombic

crystal system; it is biaxial with strong pleochroisin. In
fact, strong pleochroism is one of tanzanitefs most distinctive characteristics. It would seem unlikely, then,
that an isotropic material with no pleochroism could convincingly masquerade as tanzanite. Nevertheless, as figure 22 shows, this new material is a most convincing violet-to-blue.
EDXRF analysis revealed that this manufactured
material contains yttrium, aluminum, and europium.
The sample we examined was a 7.07 ct transparent oval
mixed cut (12.89 x 8.82 x 6.27 mm). The only internal
characteristics that we could resolve microscopically
were very weak curved striae visible at a slight angle
when we looked down the long direction. Through a
Chelsea filter, the sample appeared orangy red. Its R.I. was
over the limits of a standard refractometer, and its S.G.
was 4.62. With a handheld spectroscope, we saw five distinct absorption lines: the strongest at 530 nm, another
strong one at 468 nm, two lines of moderate strength at
589 and 472 nm, and a weak one at 480 nm.The sample
fluoresced a moderately chalky, strong reddish orange to
short-wave UV radiation, with a similar but weaker reaction to long-wave UV. In fact, while all the above-mentioned properties will help separate this substitute from
natural tanzanite, the fluorescence to short-wave UV radiation is the easiest test to distinguish the two materials
"Ta~alite,~'
cubic zirconia colored by an optical coating.
Deposition Sciences Inc., of Santa Rosa, California, is
marketing unusually colored cubic zirconia under the
trademarked name "Tavalite." Recently, we examined
several examples in the six available colors (figure 23).
Color descriptions for this material are complicated, as
the color is due to a thin optical coating that produces a
different appearance in reflected and transmitted light.
Gem News

Six round-brilliant-cut samples-one of each colorwere studied: one 6.5-mm-diameter sample weighing
1.79 ct (not shown in figure 23)) and five 4-mm-diameter
samples weighing between 0.45 and 0.47 ct. The gemological properties were consistent with cubic zirconia.
The color distribution was 011 the surface only, both
crown and pavilion. With magnification, the surface
coatings on two samples appeared "dusty," and a third
showed a build-up of the coating material at the facet
junctions. EDXRF spectroscopy of all six samples
revealed yttrium, zirconium, and hafnium; these three
elements are consistent with a base material of yttriumstabilized cubic zirconia. Also discovered were tantalum
and silicon, two elements that were probably components of the optical coating. In two examples, titanium
was also detected, again probably as part of the optical
coating. These coatings may also contain oxygen or
nitrogen, but neither can be detected by our EDXRF
instrumentation.

Figure 22. This 7.07 ct tanzanite imitation-which

contains yttrium, aluminum, and europium-has
a very convincing color. Courtesy of Bill Vance;
photo by Maha DeMaggio.

As for durability, the surface coating was responsible

for all of the color in these samples. The distributor cautioned that strong soaps might affect the surface of
"Tavalitel'; abrasives and polishing compounds should
not be used. However, nonaqueous cleaners (such as rubbing alcohol and "gentle" cleaning with a mild soapeven in an ultrasonic cleaning unit] are acceptable,
according to product literature supplied with the samples.
ANNOUNCEMENTS
Sixth annual Cutting Edge Awards. Judges chose 18 winners and 11 honorable mentions from 160 entries in this
year's sixth annual Cutting Edge competition, sponsored
by the American Gem Trade Association (AGTA) in
Dallas, Texas, on April 27 and 28.
GEMS & GEMOLOGY

Summer 1996

Mark Gronlund, of the Custom Jewelry Shop in

Deltona, Florida, took three honors, including first place
in the Faceting division with his 23.92 ct trillion brilliant-cut citrine. The citrine, when surrounded by specially cut amethysts, also won second place in the Pairs
and Suites category (figure 24). First place in Carving
went to William Cox, of William H. Cox Enterprises,
Prove, Utah, for a 28.67 ct teardrop-shaped sunstone. A
79.10 ct ametrine-fashioned by Steve Walters, of
Ramona, California-was awarded first place in the
Combination category. Justina De Vries, of BettenmannDe Vries, Medford, Oregon-who also took second place
in the Classic Cuts category-won first place in Pairs and
Suites with 7.76 ct and 8.07 ct tourmalines. Objects of
Art first place (for the second year in a row) went to Gil
Roberts, of North American Gem Carvers, Pilot,
Virginia, for a carved chartreuse quartz perfume bottle;
Stephen H. Newberg, of CCCC Company, Houston,
Texas, was awarded first in Classic Cuts for his 2.52 ct
pink spinel.
The competition was open to all colored gemstones of
natural origin that were fashioned in North America by a
professional lapidary artist. Entries were evaluated on originality of design, quality of work, technique, quality and
rarity of the gem material, and overall beauty. The winning gemstones were displayed at the June JC-K Show in
Las Vegas, Nevada; winners will be honored at a January
29, 1997, reception at the AGTA Gem Fair in Tucson.
Gem-related exhibits at American Museums. Several
natural history museums currently have exhibits with

Figure 23. These six 4-mm-diameter samples

(about 0.46 ct each) of a cubic zirconia marketed
as "Tovalite" owe their colors to thin optical coatings. Photo by Maha DeMaggio.

Figure 24. This trillion-cut citrine took first place in

the Faceting division of the 1996 AGTA Cutting
Edge competition. The ciirine with the amethysts
tooli second place in Pairs and Suites. Stones
faceted by Mark Gronlmd; photo 0Tino Hammid.

gem and jewelry themes. In New York City, t h e

American Museum of Natural History has "Amber:
Window into the Past," which runs through September
2, 1996. It includes major sections on the natural history
of amber (including a simulation of an ancient amberproducing forest environment), and on the use of amber
in jewelry and decorative objects in many ancient and
modern cultures; a catalog is available. In Cambridge,
Massachusetts, the Museum of Cultural and Natural
History at Harvard University is presenting an exhibit on
birthstones that is scheduled to remain on display for the
next five years. Finally, a traveling exhibit of more than
400 items from the studio of Peter Carl Faberg6, including several Fabergk eggs, is visiting the Metropolitan
Museum of Art in New York City; the M. H. de Young
Museum in San Francisco; the Virginia Museum of Fine
Arts in Richmond; the New Orleans (Louisiana)Museum
of Art; and the Cleveland (Ohio)Museum of Art.
Special synthetics exhibit at the Sorbonne. The Mineral
Collection of the Universite Pierre et Marie Curie, 34 me
Jussieu, Paris, features a special exhibit of historical and
contemporary synthetic gem materials. (The first synthetic crystals were grown in France in the late 19th century.) This special exhibit, set among the fabulous specimens of this world-class collection, will be open through
October 31, 1996.

140

Gem News

GEMS & GEMOLOGY

Summer 1996

THANK You, DONORS

The Treasured Gifts Council, chaired by Jeanne Larson, and co-chaired internationally by Martin
Harman, has been established to encourage individual and corporate gifts-in-kind of stones, library
materials, or other non-cash assets which can be used directly in GIA's educational and research
activities. Gifts-in-kind help GIA serve the gem and jewelry industry worldwide while offering
donors significant philanthropic and tax benefits. We extend a most sincere thank you t o those
who contributed t o t h e Treasured Gifts Council i n 1995.
1995 Treasured Gifts
Award Recipients*
Byard F. Brogan, Inc.
Chatham Created Gems
Eric Engel
A. F. Greenwood Co., Inc.
Kimberley-Created Emeralds, Inc.
Kurt Nassau
OroAinerica
Hyman Savinar
John Sinlzanlzas
Tiffany & Company
Zale Corporation
1995 Treasured Gifts Donors
A. Levian & Company
David M. Baker
Claude Barbera
Gerhard & Andreas Beclzer
K. C. Bell
Gordon Bleclz
William E. Boyajian
George Brooks
Cabochon Gems & Designs
Donald Clary
Jo Ellen Cole
Colonial Jewelry Co.
Richard W. Corirossi
Crystural Corporation
Martin B. De Silva
Omer S. Dean
Patrick J. Delaney
Robert Dunnigan
Ayman Dweilz
L. W. Ford
Emmanuel Fritsch
Barry Fruchtinan
Wang Fuquan
Robert I. Gait

Gem Bank Group

Gem Reflections of California
Gem Stones
Gemstones International
Giraux Fine Jewelry Ltd.
Aura M. Godoy
Golay Buchel USA Ltd.
Diane C. Goodrich
Laurence Graff
Patricia Gray
Wang Guo-Fang
Liu Guobin
Manoj Gupta
Larry Hall
Martin I. Harman
Ingeborgs Stenar AB
Isotope Products Laboratories
J. Clayton Spears
Zeng Jiliang
John Anthony Jewelers
Mary L. Johnson
Josephs Jewelers
Robert C. Kammerling
Larry P. Kelley
Linda Kha
King Stone Gems
Alexander J. Kissin
John I. Koivula
Guy Langolier
Lelia L. Larson
Gail B. Levine
Holz-Shing Liu
P. C. Lunia
M. Fabrilzant & Sons
Marlz Maim
Manning International
D. Vincent Manson
Wolfgang Mayer
Talzahilzo Mizuno
Moses Jewelers

Thomas Moses
Kusum S. Naotunne
Alvin M. Natlzin
Saw Naung U
Gaylon OIDell
Michael OIDonoghue
Keith R. Olivas
Parkers' Karat Patch
Franz Pulver
Maurice D. Quam
Heather S. Quayle
Wanda Radziwill
Reiner's Inc.
River Gems & Findings
Ramiro M. River0
Miguel Romero
Hiroshi Saito
Thomas M. Schneider
Henri-Jean Schubnel
Ann Seregi
Menahem Sevdermish
Robert J. Shagy
Vladimir Shatslzy
James E. Shigley
Shinju Pearl Ltd.
Marlz H. Smith
Iouliia P. Solodova
Amirali Soofi
Margherita Superchi
Supergems Trading Corp.
D. Swarovski & Company
Tech Gem Corp.
Thomas J. Terpilalz
U.S. Jem, Inc.
Vardi Stone House, Inc.
Joseph C. Wildhagen
Yu Xueyuan
Ou C. Yang
Mollie Yasuda-Abramitis
Joaquim W. Zang

'Gifts valued at $10,000 or more.

In its efforts to serve the gem and jewelry industry, GIA can use a wide variety ofsifts. These include natural and synthetic stones, media resources for the Richard T. Liddicoat Library and Information Center, and equipment and instruments for ongoing research support. If you are interested in making a donation, and r e c e i w tax deduction information,
please call A m Johnston, Manager, Treasured Gifts Council, at ext. 274 at (800)421-7250 01 from outside the U.S. (310)
829-2991, fax at (310) 829-2269, or e-mail [email protected]. Every effort has been made to avoid errors in tlus listing. If
we have accidentally omitted or misprinted your name, please notify us at one of the above numbers.

141

Thank You, Donors

GEMS & GEMOLOGY

Summer 1996

THE ART OF JEWELRY

DESIGN, Volume I

COLLECTIBLE BEADS:
A UNIVERSAL AESTHETIC

By Maurice P. Galli in collaboration

with Nina GiambeUi, 224 pp., illus.
by Mamice P. Galli, Dominique
Rivihre, and Fanfan Li, publ. by
SchifferPublishing, Atglen, PA,

By Robert K. Liu, 256 pp., illus.,

publ. by Ornament, Inc., Vista, CA,
1995. US $49.95"

1994.*

Promoted in the foreword by the

Fashion Institute of Technology's
Samuel Beizer as the first comprehensive reference for the jewelry design
process, The Art of Jewelry Design
accomplishes this objective visually,
with minimal text. The approach is
not only refreshing, it is also appropriate, given that the essence of a subject should be presented in its most
effective and understandable format.
This colorful volume is filled with
illustrations of rendering techniques
as well as fanciful paintings of jewelry designs that would be of interest to
students, designers, artisans, and anyone who appreciates fine jewelry.
There is a fascination with iewelry design that is unique and Personal. It is exciting to envision the creation of a work for human adornment, exquisite in detail and movement, from materials that are themselves beautiful, valuable, and rare.
However, acquiring the ability to
communicate this vision of a jewel
through a striking illustration is' even
more compelling. This is the primary
reason for the book's appeal. Much of
the mystery behind the illustrative
process is revealed, as the observer is
invited to look over the shoulder of a
professional jewelry designer and
watch as the works unfold on paper.
The collaborating illustrators
share their expertise and experience,
which was honed in some of the trade's
most prestigious jewelry design firms.
Mr. Galli, who manages Tiffany's jewelry design department and teaches
jewelry design at FIT in New York, has
designed for Harry Winston, David
Webb, and Van Cleef & Arpels. Ms. Li,
an award-winning designer, currently
creates for Van Cleef & Arpels in New
York. Mr. Rivikre, who has been associated with the top jewelry salons in the
world, now designs for GemVeto of
New York.
In Part I of this volume, the
designers introduce the basics for

142

Book Reviews

experimenting with a classic rendering style. The sequential presentations of the general painting methods
will be clearly understood by the
novice but also offer alternative styles
of rendering for the experienced illustrator. Several illustrations compare
and contrast the application of various design principles.
Part I1 supports the principles
touched on in Part I with a series of
specific methods to paint gemstones
and metal. Of importance to student
designers are the helpful bits of information, such as the shapes and cuts
of those gemstones that are most frequently used, the standard setting
styles, guidelines for stone layout,
and variations on basic shapes. The
practical information on working
with costing parameters, terminology, and history is also useful.
Unfortunately, I found the quality of the book itself-the printing,
resolution of images, binding, and
particularly the graphic design-to
be incongruent with the authors' fine
illustrations. Although each plate is
thoughtfully arranged, with numerous and diverse examples, throughout the book there are inconsistencies in the graphics (c.g., some details
are numbered while others are not)
and the text layout, which often
makes for difficult reading.
While the work of Galli, Riviere,
and Li is worthy of a more professional presentation, I am reminded that
the book was "conceived as a handbook and reference"; as such, it retains
the overall impact of its content. The
A n of Jewelry Design would be a
beautiful and pertinent addition to
any jewelry library.
LAINIE MANN
Czuriczilz~mDesign Specialist
Course Development Department
CIA Santa Monica

Here is a book that every contemporary

bead enthusiast should own. Robert
Liu takes the reader through the actual
process of collecting these portable, perforated artifacts-from those that date
as far back as 40,000 years to those
made by present-day artists.
Throughout history, beads have
been an integral part of human existence. They reflect the development
of societies and of the skills, art, and
economies of various civilizations.
They are also useful in tracing the
interactions between civilizations. To
illustrate these themes, Liu takes the
reader on a country-by-country tour of
major cultural and geographic bead
regions: Africa, China and Taiwan,
Japan, Korea, the Himalayan countries, Indonesia, Thailand and the
Philippines, the Middle East, the
Americas, pre-Columbian America,
and Europe. It is more of a travelogue
of beads than a strict chronology.
In his narrative, Liu covers
mainly those beads that have gained
appreciation as collectibles during the
past 20 years of what he considers to
be the rise of the bead movement.
The actual process and strategies of
collecting are discussed in depth in
this beautifully illustrated reflection
of the author's enthusiasm for the
subject. The book also includes chapters o n necklace design, various
stringing techniques, and clasp design;
a glossary of bead terminology, as well
as a listing of bead societies, research
organizations, and international bead
publications; and many well-referenced quotations about beads that
have appeared in the literature. This
publication is definitely a prized addition to any "beadophile's"library.
ELAINE FERRARI
Instructor
GIA Santa Monica

"This book is available for purchase through

the GIA Bookstore, 1660 Stewart Street,
Santa Monica, CA 90404, Telephone: (800)
421-7250, ext. 282; outside the U.S. (310)
829-2991, ext. 282. Fax: (310) 449- 1161.

GEMS & GEMOLOGY

Summer 1996

GEMOLOGICAL
C. W. FRYER,EDITOR

REVIEW BOARD
Charles E. Ashbaugh Ill
Isotope Products Laboratories
Burbank, California
Andrew Christie
GIA, Santa Monica
Jo Ellen Cole
GIA, Santa Monica
Maha DeMaggio
GIA Gem Trade Lab, Santa Monica

Emmanuel Fritsch
University of Nantes, France

Mary L. Johnson
G I A G Trade
~ ~ Lab, Santa Monica

Himiko Naka
Pacific Palisades, California

Michael Gray
Missoula, Montana

A. A. Levinson
University of Calgary
Calgary, Alberta, Canada

Gary A. Roskin
European Gemological Laboratory
Los Angeles, California

Patricia A. S. Gray
Missoula, Montana

Loretta B, Loeb
Visalia, California

James E, Shigley
GIA, Santa Monica

Professor R. A. Howie
Royal Holloway
University of London
United Kingdom

COLORED STONES AND

ORGANIC MATERIALS
Captured in amber. D. A. Grimaldi, Scientific American,
Vol. 274, No. 4, April 1996, pp. 70-77.
This short article about Dominican and New Jersey
amber does not really furnish much essential news for
gemologists, but it is impressive because of its many fascinating color photos of different insects in amber frozen
forever in different acts (even sex). Also pictured are a
gecko, frogs, and a feather (the latter the oldest terrestrial
record of a bird in North America). David Grin~aldi
(amber expert and professor of entomology at the
American Museum of Natural History in New York)
describes how the excellent preservation of plant and animal tissues in amber allows examination of minute
details by electron microscopy (illustrated here by a series
of photomicrographs). Creating dinosaurs from DNA preserved in amber, as was done in the novel/movie Jurassic
Park, may still be the stuff of fiction. However, it has
been possible to analyze the DNA preserved in some cells
and thus establish the evolutionary relationship between
termites, cocluoacl~es,and praying mantises.
RT
Corundum from basaltic terrains: A mineral inclusion
approach to the enigma. J. Guo, S. Y. O'Reilly, and
W. L. Griffin, Contributions to Mineralogy and.
Petrology, Vol. 122, No. 4, 1996, pp. 368-386.
From over 1,000 corundum (predominantly sapphire)
crystals and fragments associated with basaltic rocks,
Gemological Abstracts

Elise B. Misiorowski
GIA, Santa Monica
Jana E. Miyahira
GIA, Santa Monica

Carol M. Stockton
Alexandria, Virginia
Rolf Tatje
Duisburg University
Duisburg, Germany

mostly from eastern Australia and China, 82 primary

mineral inclusions were identified. Most abundant (65%
of the total) were niobium (Nb) and tantalum (Ta) oxide
minerals, of which columbite, ilrnenorutile, and pyrochlore
are the most important. Silicates (30%)were predominantly zircon and feldspar, and rare sulfide (pyrrhotite)
and phosphate (brockite) minerals made up most of the
remainder.
On the basis of known geochemical characteristics
and geologic associations of these mineral inclusions, the
authors postulate that at least two magmas are involved
in the formation of the corundum crystals. For example,
feldspar, zircon, and ilmenorutile are associated with
alkaline felsic rocks, whereas the types of columbite and
pyrochlore that occur as inclusions in the corundums are
associated with carbonatitic rocks. From this they develThis section is designed to provide as complete a record as practical of the recent literature on gems and gemology. Articles are
selected for abstracting solely at the discretion of the section editor and his reviewers, and space limitations may require that we
include only those articles that we feel will be of greatest interest
to our readership.
Inquiries for reprints of articles abstracted must be addressed to
the author or publisher of the original material.
The reviewer of each article is identified by his or her initials at the
end of each abstract. Guest reviewers are identified by their full
names. Opinions expressed in an abstract belong to the abstracterand in no way reflect the position of Gems & Gemology or GIA.
0 7996 Gemological Institute of America

GEMS &. GEMOLOGY

Summer 1996

143

op a "mixinghybridisation" process to explain the origin

of corundun~at these localities.
The proposed process involves the interaction
between a high-silica magma and a carbonatitic (or some
other silica-poor) magma. From the resulting hybrid
magma, corundum precipitates and gives rise to locally
distributed lenses of corundum-bearing roclz. Subsequent
volcanic eruptions of basaltic magma, from mid-crustal
levels within the Earth, bring the corundum as
xenocrysts to the Earth's surface (the corundum-bearing
rocks having been disintegrated by the basaltic magma
and the sapphires released). Regrettably, the authors have
not considered other recently proposed mechanisms for
the origin of corundums from the same localities. AAL
Notes from the Gem and Pearl Testing Laboratory,
Bahrain-5. A. B~ibshaitand N. Sturman, Journal
of Gemmology, Vol. 25, No. 1, 1996, pp. 20-23.
This issue's Notes section focuses on amber and emeralds. An amber box, reportedly of Russian origin, was
constructed primarily of natural amber, but darker-colored inlays in the lid proved to be pressed amber. The latter was identified on the basis of interference colors and
typical inclusions-that is, minute brown to black impurities forming swirled, smoke-like patterns. The identification was made more challenging by the presence of a
reddish brown dye in the adhesive. A string of prayer
beads that were represented as amber proved to consist of
amber particles embedded in molded plastic. The nature
of the beads became evident with magnification, which
revealed typical seam lines running around the midsection of each bead.
Recently, the lab has encountered a number of resinfilled emeralds, primarily of Colombian origin. One
notable cxample exhibited a pronounced flash effect from
numerous filled fractures, as well as moderate reactions
to both long-wave and short-wave UV radiation-both
fluorescence and (to short-wave) phosphorescence. Staff
members at the Bahrain lab had previously not seen such
luminescence in resin-filled emeralds. This entry also
addresses the ongoing debate on disclosure, and closes by
stating the lab's policy on such identifications: The report
simply states that a filler is present (because of the difficulties in determining the exact identity of these fillers).
When lab staff agree on a more exact identification for a
filler, they report it verbally to their clients.
CMS

DIAMONDS
1995 a record year for diamond imports, exports. M. I<.
Golay, National Jeweler, April 1, 1996, p. 76.
In 1995, American merchants imported 11.9 million
carats of cut, unset diamonds valued at more than $5.3
billion, compared to 10.6 million carats valued at more
than $4.9 billion in 1994, according to U.S. government
information interpreted by the American Diamond
Industry Association, New York. Both figures established
year-to-year records-an
11.7% increase in carats and

144

Gemological Abstracts

7.6% in value. More than 39% of the diamonds by value

($2.1 billion) came from Israel. India ($1.4 billion) and
Belgium ($1.3billion) followed. Together, the three countries supplied more than 95% of the total weight in diamonds imported into the U.S. The average price of
imported loose, polished diamonds dropped 3.7% in
1995, from $463 to $446 per carat. Of the top eight
sources by value, the average price of diamonds imported
from Switzerland was highest, at $6,978 per carat, whereas stones from India averaged $183 per carat.
Exports of cut, unset diamonds from the United
States in 1995 set a record. Total shipments of 2.1 million
carats ($2 billion) were all-time highs for both weight and
value. Exports were up 62.2% by weight and 15.6% by
value over 1994. Israel, Belgium, and Hong Kong were the
United States' top export trading partners by value, at
$478 million, $463 million, and $433 million, rcspectively. Shipments increased in value to each of these countries over 1994 by 22.9%, 15%) and 8.5%, respectively.
Percentage increases were also seen in the value of diamonds exported to Singapore (58.4%),Japan (18.2%),and
Canada (12.6%).U.S. diamond exports averaged $932 per
carat in 1995, compared to $1,317 per carat in 1994. This
29% drop is a reflection of larger shipments of lowerpriced goods to each of the top eight trading partners,
MD
except the United Kingdom and Canada.
Canadian diamond booni. S. Voyniclz, Rock &> Gem, Vol.
26, No. 6, June 1996, pp. 21, 22, 24, 26, 68.
It is now about six years since the Lac de Gras kimberlite
field in the Northwest Territories of Canada was discovered. It is anticipated that, in a few years, this locality will
be producing about 3,000,000 carats of diamonds annually. This timely, well-written article repeats the saga of
how Charles Fiplze followed the trail of diamond-indicator minerals (pyrope garnet, ilmenite, chrome diopside)
dispersed from lzimberlites by glaciers, persevering over
great adversities until he found the first of the many pipes
in this inhospitable region. Mr. Fipke is now estimated to
have a personal wealth of more than $300 million.
This article, however, differs significantly from most
others written on this topic in that it begins with a summary of the basic facts of diamond formation and occurrence that were undoubtedly used by Mr. Fipke in his
exploration concepts and decisions. These include the
depth (at least 120 lzm) within the Earth at which diamonds form (diamond stability zones); the rapid rate at
which diamonds must be carried to the surface by lumberlite magma (or else they will not survive the trip,
because they will dissolve in the lumberlite]; and the critical recognition that lumberlite is emplaced in specific
geologic regions [archons).
The article also discusses the importance of geophysics (e.g. airborne magnetics) in exploration, how diamonds are recovered from lzimberlite roclz (e-g., densemedium separation, X-ray sorters), and other topics of a
AAL
general nature.

GEMS & GEMOLOGY

Summer 1996

Discard small diamonds-they're not worth the effort.

Diamond Industry Week, Vol. 2, No. 39, October 9,
1995, p. 4,
The per-carat value of large diamonds compared to small
diamonds is much greater than can be accounted for by the
difference in carat weight alone. In recognition of this fact,
Western Australia's Argyle diamond mine will no longer
process ore to extract diamonds smaller than 1.5 mm.
Although revenue is expected to increase as much as 10%
because of this decision, the projected open-pit life of the
Argyle mine will decrease by 10 months (ending in 2004).
Argyle's decision is likely to result in a slight increase
in the price of small diamonds, which may hurt the &amend industry in India, where many such stones are cut.
However, ignoring smaller dian~ondsis an action consistent with a "true market economy for diamonds." ML]
Inclusions in precious stones. I. Solid inclusions in diamonds (Polishwith English abstract). M.Plaszynslca,
Mineralogia Polonica, Vol. 26, No. 1, 1995, pp.
79-86.
Mineral inclusions identified in diamonds arc listed and
linked to their probable paragenesis. The author concludes (1)that the cl~en~ical
nature of most of these inchsions does not depend 011 the mode of occurrence of the
diamonds, and (2) that two distinct suites of primary
inclusions in diamonds can be distinguished (on the basis
RAH
of X-ray and microprobe analyses),
New places to look for diamonds. B. Cordua, Rocks
Digest, Vol. 8, No. 1, 1995, pp. 5, 8.
Very small diamonds (about 0.01 mm) have been found as
perfect cubes and octahedra in garnet and zircon crystals
recovered from metamorphic gneisses in the Itukchetin
(or Kokchetev) block in Kazakhstan, about 800 kin west
of Novosibirsk. These rocks were first laid down as sediments about 2 billion years ago, and they were metamorphosed about 530 million years ago. In 1990, Sobolev and
Shatsky concluded that the diamonds grew during the
metamorphism of the rock, which occurred at teinperatures about 900-10000 and pressures above 40 lcbars
(depths greater than 100 km).The diamonds survived the
uplift interval, which should have converted them to
graphite, because they were encased in the garnet and zircon; a short time-scale for uplift is also implied.
ML]
South Africa diamond plant success. Mining Journal,
London, September 22, 1995, p. 216.
Dowding Reynard and Associates, in Johannesburg,
South Africa, design mineral processing plants. In 1995,
they reported sharply increased interest in modular diamond-recovery plants for delivery to Africa and elsewhere. A new product is a full-scale dense media separation plant that is exported in modular form. Traditional
modular diamond plants can process up to 25 tons per
hour of diamondiferous material; however, a 150-ton-perhour plant has been ordered for De Beers in Namibia

Gemological Abstracts

(Namdeb), a 120-ton-per-hour plant for De Beers in

Tanzania, and two 100-ton-per-hourplants for Auridiain's
River Ranch project in Zimbabwe.
Smaller plants require fewer resources. A jig plant for
exploration in Mozambique, which processes 8 tons per
hour, has an onboard diesel generator for stand-alone use;
and a miniature DMS plant has been designed for field
sampling and on-the-spot appraisal of resources in northML/
ern Canada.
Southern Era reportssignificant values at SUF-1.
Diamond Industry Week, Vol. 2, No. 40, October
16, 1995, p. 3.
Southern Era Resources announced the most recent valuations of diamonds recovered from its kimberlite dike
project in South Africa: The two lowest averaged
US$85.29 per carat. As of mid-October 1995, 263.39
carats had been recovered from the main pit (average
grade was 5.49 carats per metric ton); an additional 22.05
carats were recovered from 17.66 metric tons of fault
breccia. The largest diamond weighed 10.51 carats, and
about 94% of the diamonds were "cuttable gems."
Reconnaissance sampling indicates that the main fissure zone contains kimberlite and/or diamonds along its
length for at least 30 km. Diamonds have also been recovered as far as 1150 m from the main fissure
ML1
Thinking the unthinkable: Argyle ponders a break from
De Beers. R. Shor, Jewelers' Circular Keystone, Vol.
167, No. 5, May 1996, pp. 98-105,
Australia's Argyle diamond mine is weighing the pros and
cons of going independent, as De Beers's Central Selling
Organisation successfully concludes negotiations to continue distributing and marketing Russia's diamond production. An independent Argyle could fundamentally change
the way rough diamonds are sold in the international inarketplace. Foremost, De Beers's price-control policies would
face the reality of actual supply and demand. An independent Argyle could help restructure control of the diamond
trade, in essence allowing several large corporate mining
companies to control the core trade in the future. Possible
future "diamond czars" (alongwith De Beers]include RTZCRA, a British conglomerate and now the world's largest
mining company; BHP, a large Australian nunmgroup that
controls the major Canadian diamond discoveries; and
Ashton, a smaller Australian mining company with a 40%
share of the Argyle Mine.
Argyle executives realize the importance of their
decision and have been planning a possible break with De
Beers for a long time. To this end, they have worked (successfully) to enlarge their customer base and improve the
efficiency of their current mining operations. Even with a
recent revamping of surface mining at Argyle, however,
the future economic success of the mine will depend on
whether the company can afford to develop a11 underground shaft, a decision that may have to be made as soon
as next year.
TEC

GEMS & GEMOLOGY

Summer 1996

145

A year of overfeed. L. Rombouts, Diamond International,

No. 39, January-February1996, pp. 61,62,64-66,68.
In 1995, more rough diamonds-130 million caratsentered the market than in any other year in history (108
million carats in 1994).Production statistics for 1995 are
presented for 21 countries (which accounted for 111.5
million carats). The remaining 18.5 million carats are
believed to have originated from Russia (as diamonds
brought to the market via non-CSO channels, either from
Russia's stockpile or through polishing joint ventures)
and from artisanal mining in Africa and South America.
Notwithstanding the large increase in "outside" rough,
the diamond market absorbed the "overfeed" without
serious problems. Statistics are also presented for the
average value (US$per carat) of each country's production
and for production from the world's 18 major mines
(mostly pipes, but also some secondary deposits, such as
in Namibia).In 1995, CSO sales amounted to 63% of the
world's total rough diamond supply.
The status of diamond exploration in numerous
countries is reviewed with the purpose of predicting
future diamond reserves. For the next 20 years, according
to Dr. Rombouts, 98% of the world's diamond reserves by
weight will be found in eight countries: Botswana (32%))
Russia (21%), Australia (11%), Angola (lo%), South
Africa (9%),Zaire (8%),Canada (5%),and Namibia (2%).
However, by value, the sequence is (for 97% of the total):
Botswana (33%)) Russia (26%), Angola (11%), South
Africa (lo%),Namibia (7%),Canada (5%),Zaire (3%],and
Australia (2%).Relative to 1995 production statistics, the
above 20-year reserve and value estimates show major
reductions in the role of Australia and Zaire and, for the
first time, a significant role for Canada (where production
is expected to start in 1998).
AA L

GEM LOCALITIES
Atop the Andes-mining Chile's mountain-high lapis. F.
Ward, Lapidary Journal, Vol. 50, No. 3, June 1996,
pp. 36-40.
I had always considered Chilean lapis lazuli to be somewhat "second-rate" until I read this enlightening article
by a popular and well-known gemological author. Eight
beautiful color photos illustrate quality lapis from this
locality, now available in sculpture, accent tile, and a
variety of other decorative objects. Covered are the geologic forces that form the lapis, the history of the Las
Flores de 10s Andes lapis mine and methods of mining
there, nomenclature, and differences between the classic
Afghanistan deposit and its Chilean counterpart,
JEC
Chad: Discovering new mineral wealth Mining Journal,
London [advertisement supplen~ent],September
22, 1995, 12 pp.
The geology and mineral potential of this land-locked
African nation have not been investigated to any great
extent; however, alluvial diamonds have been found in
many regions. The main areas of alluvial diamond pro-

146

Gemological Abstracts

duction are Abeche, Biltine, Am Zoer, and Adre in

Ouadday, and Melfi-Bitlune in Guera. Some alluvial
stones also have been found along the Lim River in
Baibokoum (on the Central African border), but the lack
of diamond-indicator minerals implies that these diamonds may have come from sources in neighboring
countries. The Oadday and Guera stones are far from
other drainages, however, and the primary sources,
although not yet found, are believed to be in Chad. MLJ
The gems of Mont Saint-Hilaire, Quebec, Canada. W.
Wight, Journal of Gemmology, Vol. 25, No. 1, 1996,
pp. 24-44.
The common gravel quarry at Mont Saint-Hilaire is one of
the world's richest sources for rare minerals, many of
which have been faceted or polished as collectors' gems.
Within this small (about 0.5 lu-1-12)
area, more than 300
species of minerals have been found, many for the first
time and others in colors or qualities found nowhere else.
Remarkably, this rich source has attracted the attention of
mineral collectors only since the 1960s and lapidaries only
since the 1970s. Among the notable (and facetable) minerals found at Mont Saint-Hilaire are serandite, willemite,
siderite, rhodochrosite, carletonite, hackinanite, and shortite. These and others are illustrated. In addition, two tables
list notable minerals recovered from this deposit, as well as
identification properties for a considerable number of
them. The author cautions that many of these minerals
CMS
and gems can be very difficult to identify.
A guide to new fossicking legislation. B. Neville,
Queensland Government Mining Journal, Vol. 96,
No. 1123, June 1995, pp. 5-9
'Rocldiounding" in the United States is called "fossicliing" in Australia. Many Australian states have laws to
regulate this activity. In 1994, Queensland revised its fossicking law. Although fossicking is defined as purely an
amateur activity, a renewable license is now required.
Certain public lands are designated as Fossicking Lands
and Fossicking Areas, collecting is permitted on private
lands (with the owner's permission), and some areasincluding national parks and lands under native title-are
closed to collecting. There are limitations on the types of
tools allowed, and fossiclzers must pay royalties to
Queensland for quantities removed above exempted
amounts. Among the places set aside especially for fossiclzers are some sapphire and opal fields. Vertebrate fossils and meteorite-related materials are not covered by
fossicking regulations; however, this article does not say
whether such materials may be collected legally. MLf
India: A rich mining heritage. Mining Journal, London
[advertisement supplement], September 15, 1995,
12 PP.
This review article contains some information (from
India's Ministry of Mines via Mining Journal resource services) about diamonds and other gem materials found in
GEMS & GEMOLOGY

Summer 1996

India. Primary diamond deposits (lzimberlites and/or lamproites) are found in the states of Madhya Pradesh (the
Panna district) and Andhra Pradesh (the Anantpur,
Krishna, and Karnool districts); reserves are estimated at
one million carats. One mine-the.Majhgawan operation
in Madhya Pradesh-produces 19,000 carats annually.
Diamonds are also found in the "extensive" conglomerates i n die Kurnool and Vindhyan basins, and i n gravels
along the Krishna and Mahanadi Rivers. India has the
"world's largest diamond cutting and polishing industry,"
employing 600,000 people and producing 70% of the
world's finished diamonds. Colored stones (such as emeralds, rubies, garnets, and cat's-eye stones) are mined in
the Kalahandi and Bolangir districts of Orissa.
ML!
River pearls from Bavaria and Bohemia. H. Hahn, Journal
of Gemmology, Vol. 25, No. 1, 1996, pp. 45-50.
The freshwater mussel Margaritifera margorjtifera has
long lived in the streams and rivers of northern Europe.
During the Middle Ages, these mussels were exploited
heavily for their pearls, and exquisite examples of their
use-in both religious and secular pieces-can be seen i n
the museums of Germany and Austria in particular.
Bohemia developed both a thriving pearl trade and an
extensive jewelry manufacturing center in Prague. By the
19th century, M. margaiitifera had become virtually
extinct, largely as a result of industrial pollution of the
rivers. In fact, recent interest in propagating M.murgunlifer0 is due primarily to the fact that these mussels are
so sensitive to water pollution that they are one of the
best indicators available to ecologists concerned with
river and stream conditions. Various government and
educational groups throughout northern Europe are cooperating to expand the distribution of M margaiitifera.
The author also provides some fascinating information on these mussels, including their dependence on a
particular species of trout for propagation. As this once
nearly extinct animal again begins to thrive in European
waters, gemologists can look forward to the bonus of seeing more of their exceptional by-product-pearls.
CMS
The role of fluorine in the formation of colour zoning in
rubies from Mong Hsu, Myanmar (Burma). A.
Peretti, J. Mullis, and F. Mouawad, Journal of
Gemmology, Vol. 25, No. 1, 1996, pp. 3-19.
The growth patterns and color zoning typical of Mong
Hsu rubies suggest a complex growth history with compositional fluctuations in the formation environment.
When both Ti and Cr are abundant, violet-to-black sapphire forms, when Cr is moderateto high and Ti virtually absent, ruby forms; and when Cr is moderate and Ti
low, pinli-to-violet sapphire forms. The great number of
mineral inclusions found in these rubies confirm the
complexity of their growth environment and provide
information about that environment. Specifically, the
presence of fluorite inclusions is evidence that fluorine (F)
was an element of the growth solution. Likewise, fluid
inclusions indicate the presence of water. O n the basis of
Gemological Abstracts

a preliminary model, the authors conclude that F concentrations played a major role in the development of the
color zoning in Mong Hsu rubies.
CMS
Shenzi: The politics of mining tanzanite. E. Blauer,
Lapidary Journal, Vol. 50, No. 3, June 1996, pp.
42-45.

This beautifully illustrated, concise article chronicles die

fascinating history of tanzanite mining in the country
after which this distinctive blue gem was named.
According to the author, shenzi is a Swahili word that
translates roughly to "sleazy." After reading this article,
it is not difficult to sec why the word might apply to the
subject at hand.
Discovered in 1967, tanzanite (a gem variety of the
mineral zoisite) has suffered through periods of erratic
supply because of government intervention. Since 1971,
the Tanzanian government has tried to control and profit
from mining and sales of tanzanite. The net result of
these efforts has been rampant illegal mining and smuggling that, at times, only served to drive prices down, as
large quantities of the gem entered the market with no
regard to the sizes and qualities that were really needed.
Although goods are still plentiful, especially the big
stones that have made tanzanite so popular with designers, there is concern that some areas at the Merelani mining district are worked out. Smaller stones appear to be in
short supply, especially in light of the demand created by
the home shopping networks. Reported new discoveries
in Kenya and elsewhere in Tanzania, however, may help
to keep the supply steady for now.
The final page of the article describes how Narottam
Pattni discovered how to heat treat tanzanite, turning a
near-worthless off-color crystal into the sought-after deepblue treasure that it is today.
T-v'
Treasures of Glen limes. M. Kellv.
,, Australian Gold Gem
el Treasure, Vol. 11, No. 2, February 1996, pp.
32-37.
The New England region of New South Wales, Australia,
has many different gem materials for the amateur collector. Buried alluvial beds at Wellingrove contain sapphire,
topaz, and zircon; beryl is found in Torrington, and nearly
all the streams in the Glen Innes area contain sapphires.
Persistent searchers continue to turn up stones missed by
commercial miners. Unlike Queensland, however, there
are no large zones of free access for gem mining, so collectors must ask local landowners for permission to dig.
Although "cornflower blue" sapphires are easily recognized in washed gravels, other colors (and shades of
blue) can be overlooked. One technique to spot these sapphires is the "old mirror trick": Take a large spoonful of
gravel concentrate-washed "as clean as the water available to you will allowH-and spread i t across the surface
of a mirror in the sunlight. The sapphires become translucent from the light reflecting back through them.
The town of Glen Lines boasts "the best country
museum in Australia," the Land of the Beardies Museum.
GEMS & GEMOLOGY

Summer 1996

It contains displays of minerals, gems, and old mining

equipment.
MLJ
All unusual ruby-sapphire-sappliirinc-spinel assemblage
from the Tertiary Barrington volcanic province,
New South Wales, Australia. F. L. Sutherland and
R. R. Coenraads, Mineralogical Magazine, Vol. 60,
No. 4, 1996, pp. 623-638.
Sn~all,corroded, crystalline aggregates of ruby-sappl~csapphirine-spinel occur in corundum-bearing alluvial
deposits derived from the Barrington shield volcano. The
sapphirine has a near 7:9:3 (MgO:Al2O3:SiOi)composition
and, together with the corundiim, shows reaction rims of
pleonaste. Spinel in these aggregates has a composition of
Sp60.73H~7.29Cin0.3.
Potential origins for these aggregates
include metamorphic recrystalhzation of aluminous inaterial (<1460C
or high-temperature, high-pressure crystallization reactions related to lamprophyric or basaltic
magmas ( ~ 1 3 0 0 ' Cand 20 kbar). Sapphirine-spinel thermometry suggests final crystallization for the aggregates
at - 780"-94O0C and reaction with the host magmas at
>100O0C.The Barrington gemficld includes two distinct
corundum sites. One, typical in eastern Australia, is donunated by blue-green, well-crystallized, growth-zoned sapphire, commonly containing nitile silk and Fe-rich spinel
inclusions. The other, an unusual site, is dominated by
ruby and pastel-colored sapphires; these have little crystal
shape or growth zonation, and they have Limited mineral
inclusions of clxomian pleonaste and pleonaste. Chemical
analyses are given for sapphirine and a range of spinels. It is
tentatively suggested that sapplkine may be of use as an
alluvial indicator mineral for ruby.
RA H

INSTRUMENTS AND TECHNIQUES

Hanneman-Hodglzinson synthetic emerald filter. T.
Linton and A. Shields, Australian Gemmologist,
Vol. 19, NO. 2, 1995, pp. 65-68.
The authors tested the Hanneinan-Hodgkinson Synthetic
Emerald Filter and found that it performed as stated in the
instruction booklet. This new emerald filter is designed to
be used with the well-established Chelsea filter, not as a
substitute for it. While jewelers are always hoping for the
so-called black box that will easily
natural from
, separate
*
synthetic emeralds, the Hanneman-Hodgldnson Synthetic
Emerald Filter is not the all-inclusive answer. It is a gemological tool that works when used as specified. However, it
does have certain limitations.
For instance, table 1 shows that six observers differed
about what color they saw when examining the same test
stone(s1. They variously described the filter reaction for
an emerald from Colombia as colorless, green, blue, and
pink. The filter colors reported for some of the synthetic
emeralds tested also varied somewhat. While all
observers agreed that the Regency hydrothermal synthetic emerald appeared red through the filter, the visual color
opinions on the Chatham and Inamori stones were far
less consistent.

148

Gemological Abstracts

Another limitation, as stated by the filter's inventors,

is that it does not work on Biron or Russian hydrothermal
synthetic emeralds. These appear to have the same color
through the filter (green) that most natural emeralds
have.
With the Hanneman-Hodgkinson Synthetic Emerald
Filter, if one observes a red color, then the stone is not a
natural emerald. If a pink color is noted, then the stone is
probably synthetic, but caution is in order. If the stone
appears green, then other avenues of gemological testing
are needed.
As with all gemological testing techniques that
appear outwardly simple, the Hanneman-Hodglunson filter actually requires a great degree of interpretive expertise and experience. The usefulness of this filter would be
greatly enhanced if a set of reference emeralds of known
natural and synthetic origin were available for direct
comparison.
John I. Koivula
New methods of photography through the microscope:
Application to gem materials. I<. A. Kinnunen,
Geological Survey of Finland, Special Paper 20
(Current Research 1993-19941, 1995, pp. 185-187.
New techniques are described for the photon~icrography
of macroscopic specimens, including crystal faces and
gold nuggets, in their natural state. An acetate peel is
obtained from the cleaned surface of die specimen, and
the peel is mounted and studied in transmitted light and
photographed. Color-filter shadowing is used to enhance
the specimen's three-dimensional aspects. Examples
given include photographing the rhombohedra1 faces of
amethyst crystals. The method has been used to identify
faked specimens and in criminological studies.
RAH

PRECIOUS METALS
U.S. mines more gold. M. I<. Golay, National Jeweler,
April 1, 1996, p. 24.
The United States could top South Africa as the world's
largest gold producer by the year 2000, if the current trend
continues, according to the Gold Institute. It is projected
that by 1998, U.S. gold production will supply 16% of the
world's demand, as compared to 2% for 1979. However,
South Africa's contribution has dropped from 53% to
21%. Royal Oak Mines, Kirkland, Washington, expects
its 1996 gold production to reach approximately 425,000
ounces, about 15% more than its 1995 production of
371,151 ounces. Australia's gold production is also rising,
from 1% of the world total in 1979 to an estimated 13%
in 1998. Canadian output remained roughly the same, up
MD
just slightly from 4% to 6%.
World silver use. M. K. Golay, National Jeweler, April 1,
1996, p. 24.
Jewelry and giftware manufacturers in the United States
used approximately 10.8% more silver in 1995 (a total of
4.1 million ounces) than in 1994, according to the Silver
Suivey 1996, by CPM Group, a precious metals research
GEMS & GEMOLOGY

Summer 1996

firm. The increase was due to greater consumer interest

in silver jewelry, reports CPM. Worldwide silver use in
jewelry and decorative objects was relatively stable last
year, rising only 0.4% from 222 million ounces in 1994 to
222.9 million ounces in 1995. The major reason for the
sluggish market was a 15% decline in silver use in India
because of severe supply constraints there. U.K. jewelers
and silversmiths continued to expand, consuming 3 million ounces of silver in 1995, an l l % increase. Italian jewelers increased their silver use by an estimated 14.5% to
46.6 million ounces last year, CPM said. Silver prices
year-to-year ended up fairly even, 1995 was the sixth
straight year in which silver fabrication demands exceeded total new silver supplies.
MD

SYNTHETICS AND SIMULANTS

[New type of] synthetic alexandrite we encounter in gem
ID lab lately (in Japanese).National Gemmological
Association Technical Laboratory, Gemmology,
Vol. 26, No. 308, May 1995, pp. 4-5.
This article describes a new type of synthetic alexandrite,
which contains needle-like inclusions that are very similar to those seen in natural alexandrite. This material first
caught the authors' attention in November 1994. Since
then, they have tested several mixed-cut loose synthetic
alexandrites, from 1 to 3 carats. All of the samples examined were of high clarity and showed a strong change-ofcolor from bluish green to purplish red, similar to that
seen in natural alexandrite from Brazil.
Their refractive index was 1.738-1.746 0.001,
which were low values for their color (birefringence,
0.008). All of the samples fluoresced strong red to longwave ultraviolet radiation; some showed strong red, and
others showed moderately strong chalky yellow, to shortwave UV. With magnification, the authors observed randomly placed, needle-like metallic inclusions that were
sometimes accompanied by gas bubbles. Immersed in
methylene iodide, the stones showed curved color bands.
The spectroscope revealed absorption at 680, 665, 655,
645, and 465 n1111, but it did not reveal the 365 and 375
m m features almost always shown by natural alexandrites. In addition, the absorption cut-off in the violet was
shifted toward the lower wavelenghts. From these testing
results, the authors concluded that the samples were
manufactured by the (Czochralski)pulled method. HN

TREATMENTS
Heat treating sapphires from the Anakie District,
Australia. T. Themelis, Australian Gemmologisi,
Vol. 19, No. 2, 1995, pp. 55-60.
This article vaguely discusses the colors that resulted
when 20,000 carats of rough sapphires were heat treated
at various temperatures and atmospheres. The article
looks very scientific and informative at first glance, but
few practical details axe given. For example, the author
says that "careful controlled rates of heating and cooling"
were used, but he does not say what the rates were. Nor
Gemological Abstracts

does he elaborate on the "gas mixtures used [that] were

poisonous." However, I did glean a few facts: A 2O0C-perminute rate of heating and cooling prevented recrystallization of iutile; strong reducing atmospheres during
heating produced muddy green or black colors; blue-togreen sapphires turned a better color than green-to-yellow
ones; and final color was established one to two days after
treatment.
CEA

MISCELLANEOUS
1995: The geosciences in review. [Various Authors],
Geotimes, Vol. 41, No. 2, February 1996, pp. 17-54.
This compendium reviews trends and new discoveries in
the geologic sciences during 1995; some of these insights
may be relevant for gemologists. Perhaps the most significant change in recent years has been the drastic decrease
in government funding for the geologic sciences, as evidenced by the recent closure of the U.S. Bureau of Mines
Also included in the many topics covered:
Information-science trends (reviewed by B. E.
Haner) reflect the increasing importance of communication through computer networks, especially-but not
limited to-the rising prominence of the World Wide
Web. The invaluable GeoRef CD-ROMs (which this
abstracter consults frequently) continue to be updated,
and high-resolution scanning is being developed, leading
to the preservation as digitized information of large-format illustrations, such as fragile geologic maps.
Although planetary geology (reviewed by J. R.
Zimbelman) leans toward the esoteric, one result this last
year is that there are a calculated 72,000 tons of (admittedly tiny) 0.2 mm diamonds in the suevite rocks within
the Ries impact crater, Germany. Similarly, meteoritics
(reviewed by H. Y. McSween Jr.) reveals still smaller diamond, silicon carbide, and corundum grains (older than
our solar system] found in chondritic meteorites.
An exciting trend in exploration geophysics
(reviewed by W. H. Dragoset) is the development of realtime, three-dimensional seismic characterization of subsurface rocks in the field.
Exploration geochemistry (reviewed by J. E. Gray)
continues to be invaluable in the search for diamonds i n
Australia, Brazil, China, Guyana, India, Indonesia,
Namibia, South Africa, Tanzania, Venezuela, Zimbabwe,
and several provinces of Canada.
Satellite remote sensing data (reviewed by E.D.
Paylor I
l and M. Baltuck) have been used for observation
of paleodrainages (former river channels] along the Nile
River.
In the field of mineral chemistry (reviewed by C.
Shearer), a research effort approaching fruition is the
understanding of the rules governing trace-element distributions in minerals, especially for the rock-forming
clinopyroxenes (such as diopside and jadeite).
The growing development of microbeain analytical techniques (such as ion microprobes and laser ablation/mass spectrometry) has enabled the study of light
GEMS & GEMOLOGY

Summer 1996

149

elements (H, Li, Be, B) in tourmahne, vesuvianite, and

other minerals. The ion microprobe is also being applied to
the study of changes in roclzs and minerals induced by
metamorphism (reviewed by G. E. Bebout). Computed Xray tomography, another new technique, is currently being
used to study the porosity of carbonate roclzs (reviewed by
C. Kerans, R. K. Goldhammer, and J. L. Banner).
In clastic sedimentology (reviewed by M. H.
Gardner), fluid dynamics models are being applied to
stream flow, in the hopes of determining original depositional environments (e.g., valleys versus channel-fill
deposits).
MLJ
GPS: Useful, with care. L,R. Ream, Mineral News, Vol.
11, No. 4, April 1995, p. 10.
Global Positioning Satellite (GPS) receivers, which can
pinpoint a position on the basis of data received from several orbiting satellites, are a wonderful way to find out
where you are on the Earth's surface. Now they are coinmercially available at modest ($300) to immodest
($15,000) prices. However, some factors limit their usefulness. These include interference from dense foliage,
narrow views of the sky, and signals bounced off mountainsides and canyon walls (not to mention the fact that
the U.S. Department of Defense purposely degrades the
accuracy of publicly accessible satellite signals). The
more expensive units are generally more accurate.
The author recounts several anecdotes regarding
false GPS readings. He notes that a GPS unit is generally
more accurate than precise (i.e., although the readings
may not be the same each time, they will be "in the
vicinity" of correct) and is subject to long-term drift in its
information. He suggests that, if you buy one of the
lower-end GPS units, you first take several readings and
average them, to learn the limitations of the device and
how to work with them.
ML1

Minerals in rock mass hold clues to 400-kilometer

ascent. R. A. Kerr, Science, Vol. 271, March 1996,
p. 1811.
Mineralogists are agog over the discovery of a large chunk
of garnet peridotite in the mountains of southern
Switzerland, 400 km above the strata in the Earth's mantie where all mineralogical knowledge says it should be.
The Alpe Arami massif, which measures 800 x 500
m, has traveled to the surface from the mantle transition
zone, 400 to 670 l a down. Researchers determined its
original depth of formation from the presence of ilmenite
inclusions in olivine-some rod-shaped and some of a

150

Gemological Abstracts

structure previously unknown-all indicating formation

below 300 lzm.
Researchers can only tentatively explain this phenomenon as "deep subduction"; that is, the peridot massif was surrounded by much lighter sedimentary crustal
roclz and buoyed to the surface. To the question of why
this lightweight sedimentary roclz from the Earth's crust
was in the deep mantle zone in the first place, some mineralogists have proposed "the Ivory soap principle."
According to this theory, light crustal roclz was driven
downward bv the collision of continents, then warmed
up and bobbed back to the surface (like Ivory soap), carrying chunks of heavier mantle with it.
Dian~onds, which occur in the only other rocks
known to have pulled off the trick, arrived by a different
mechanism (having been blasted up from more than 100
lzm depth through narrow volcanic conduits, or pipes).
There are, however, similarities between the odd mixture
of minerals in these rocks and that in the one from the
Alpe Arami massif. Although "the Ivory soap principle"
to explain the massif roclz is controversial, the discovery
has major implications for scientists' understanding of
the subd~ictionprocess, and of the mineralogy and chemistry of the Earth's mantle.
AC
Twisted and contorted plants. E. B. Heyln~un,
International California Mining Journal, Vol. 65,
No. 1, September 1995, pp. 30-31.
This short summary of the 1983 textbook Biological
Methods of Prospecting for Minerals, by R. R. Brooks,
describes plants that are found preferentially in regions
where the soil contains economically valuable elements.
California poppies indicate copper, alyssum indicates silver-lead-zinc or niclzel (and niclzel is often associated
with platinum), pennycress can indicate silver-lead-zinc,
wild buckwheat is associated with silver, and wild rye
grass occurs with sulfur. Equisetum (horsetails), wild
onions, and "miner's moss" arc claimed by some to be
associated with gold. In general, any poisonous, hallucinogenic, or foul-smelling plant (except poison ivy and
poison oak) may indicate that the soil is rich in unusual
elements.
Another indicator of mineralization is unusually
"stressed" plants-especially
con~posites (daisies)showing features such as stunted growth, twisted or contorted growth, yellow foliage, or unusually colored flowers; or plants that are unusually early or late blooming.
Boron, radioactivity, and bitumenous hydrocarbons may
cause gigantism in plants.
MLI

GEMS &. GEMOLOGY

Summer 1996

GUIDELINES FOR AUTHORS

Authors must use the following guidelines when submitting a manuscript to
Gems o)Gemology, or the paper may be
returned unreviewed. Please contact the
Editor's office if you have any questions
about these guidelines or any other
aspects of your topic.

Gems eJ Gemology publishes original

articles concerning the study of gemstones and research in gemology and
related fields. Appropriate topics include
(but are not limited to) colored stones,
diamonds, gemological instruments and
identification techniques, gem localities,
gem enhancements, gem substitutes (simulants and synthetics), gemstones for the
collector, jewelry manufacturing arts,
jewelry history, and contemporary trends
in the trade. Remember that these are
general guidelines only. Your best source
of appropriate topics is Gems a)
Gemology itself. Read the journal to
acquaint yourself with the breadth of
material covered.
Manuscripts may be subnlitted as:
Feature Articles-full-length articles
describing previously unpublished studies
and laboratory or field research. Such articles should be no longer than 6,000 words
(24 double-spaced, typed pages) plus
tables, illustrations, and references.
Review Articles~comprehensivereviews
of topics in the field. Length of text
should not exceed 8,000 words (32 double-spaced, typed pages).
Notes & New Techniques-brief preliminary communications of recent discoveries or developments in gemology and
related fields (e.g., new instruments or
identification techniques, gem minerals
for the collector, and lapidary techniques).
Articles for this section should be approximately 1,000-3,000 words (4-12 doublespaced pages).
Gem News-very brief (100-500 words,
one-half page to two double-spaced pages)
items on current events in the field or
unusual gem materials.
To be considered for publication, all wntributions to Gems o) Gemology must be
original and not previously published in

Guidelines for Authors

English. We will consider articles already

published in languages other than English
only on a case-by-case basis and only if
the author(s)inform us at the time of submittal when and where the article was
first published. It is our policy not to publish feature articles, review articles, or
Notes on single stones or research based
on a single sample (unless of extraordinary historical importance, such as the
Hope cliamoncl).Authors may not publish
the same material elsewhere for at least
three months after the mail date of the
issue in question, without express written
permission from the Editor's office.
Gems a) Gemology also includes these
regular sections: Gem Trade Lab Notesreports of interesting or unusual gemstones, inclusions, or jewelry encountered
in the GIA Gem Trade Laboratory; Book
Reviews-as
solicited by the Book
Reviews editor (publishers should send
one copy of each book that they wish to
have reviewed to the Editor's office);
Gemological Abstracts-summaries of
important articles recently published in
the gemology literature.

All material (including tables, figure legends, and references) MUST be typed,
double spaced (no exceptions), one side
only, on 8.5 by 11 inch (21 by 28 cm)
white paper (no colors) with 1.5-inch (3.8
cm] margins. Each page should be consecutively numbered-including the first
and those containing figure legends, references, etc. Use a fresh typewriter or printer ribbon; one that prints black, not gray.
Please avoid sending draft-quality dotmatrix printouts. It is strongly recommended that the article also be submitted
on a floppy disk in: Microsoft Word 4
through 6 (DOS),Wordstar 3 through 5.5,
Wordperfect 5+, Rich Text Format (RTF),
Microsoft Word for Windows (2 or 61,
Windows Write, Microsoft Word 4+ for
Macintosh, or ASCII (DOS and Macintosh). Please indicate which format is
used.
Identify the authors on the title page
only, not in the body of the manuscript or
figures, so that author anonymity can be
maintained with reviewers (the title page
is removed before the manuscript is sent

for review). The various components of

the manuscript should be prepared and
arranged as follows:
Title Page. Page 1 should include: (a)the
article title; (b) the full name of each
author; (c)each author's affiliation (the
institution, city, and state or country
where he/she was working when the article was prepared); and Id) aclznowledgmerits of persons who helped perform the
research, prepare the report, or do the
photography, etc.
Abstract. Page 2 should repeat the title of
the article, followed by an abstract. The
abstract [no more than 150 words for a
feature or review article, 75 words for a
Note) should state the purpose of the
article, what was done, and the main
conclusions.
Text. Papers should follow a clear outline
with appropriate headings. For example,
for a research paper, the headings should
be: Introduction, Previous Studies,
Materials and Methods, Results,
Discussion, and Conclusion. Use other
heads and subheads as the subject matter
warrants. Also, when writing your article,
please avoid jargon, spell out the first
mention of all abbreviations, and present
your material as clearly and concisely as
possible. For general style (grammar, etc.),
see The Chicago Manual of Style
(University of Chicago Press, Chicago).
Papers that describe original research
must include a Materials and Methods
section that contains, at a minimum, the
numbers and descriptions of all samples
examined, and the techniques and instiumentation used to obtain the data.
References. References should be used for
any information that is taken directly
from another publication, to document
ideas and facts attributed to another writer, and to refer the reader to other sources
for additional information on a particular
subject. References must be cited in the
body of the text (in parentheses), with the
last name of the author(s) and the year of
publication; add the appropriate page
number when citing a direct quote or a
specific illustration or set of numbers or
data. An example would be: (Kammerling
et al., 1990, p. 33). The list of references
at the end of the paper should be typed

GEMS & GEMOLOGY

Summer 1996

151

double spaced in alphabetical order by

the last name of the senior author; if
there is more than one paper by that
author or same group of authors, list
those papers in chronological order, starting with the oldest paper first. List only
those references actually cited in the text
(or in the tables or figures). A paper that
lacks a reference section or is missing
information in the reference sectionpublishers, dates, authors, etc.-may be
returned unreviewed.
Include the following information,
in the order given here, for each reference:
(a) all author names (surnames followed
by initials); (b) the year of publication, in
parentheses; (c)for a journal, the full title
of the article or, for a book, the full title of
die book cited; and (d) for a journal, the
full title of the journal plus volume number, issue number, and inclusive page
numbers of the article cited or, for a book,
the publisher of the book and the city of
publication. Sample references are:
Kammerling R.C., Koivula J.I., Kane R.E.
(1990) Gemstone enhancement and
its detection in the 1980s. Gems ed
Gemology, Vol. 26, No. 1, pp. 32-49.
Armstrong J.T. (1988) Accurate quantitative analysis of oxygen and nitrogen
with a Si/W multilayer crystal. In
D.E. Newbury, Ed., Microbeam
Analysis-1988, San Francisco Press,
San Francisco, CAI pp. 301-304.
Liddicoat R.T. (1989) Handbook of Gem
Identification, 12th ed., 2nd rev.
printing. Gemological Institute of
America, Santa Monica, CA.
Personal communications (for information obtained from someone witli particular expertise) should be cited in the
body of the text only, as follows-(G.
Rossman, pers. comm., 1994). Permission
must be obtained from the people cited to
use their names for this purpose. Such
resources should also be listed, with their
affiliations, in the Aclcnowledgments section.
Tables. Tables can present a large amount
of detail in a relatively small space. Consider using one whenever the bulk of
information in a section threatens to overwhelm the text.
Type each table double spaced on a
separate sheet. If tlie table must exceed
one typed page, please duplicate all headings on the second sheet. Number tables
in the order in which they are cited in the
text. Every table should have a title; every
column (including die left-hand column)
should have a heading. Please make sure

152

Guidelines for Authors

terms and figures used in the table are

consistent with those used in the text.
Line Illustrations. We prefer that all line
art (graphs, charts, etc.) be sent to us on
computer disk, accompanied by hard
copy. (Please contact us about acceptable
formats.) If this is not possible, line art
should be professionally drawn and the
original sent to us.
S u b m i t black-and-white photographs and photomicrographs in the
final desired size if possible. Where appropriate, please use a bar or other scale
marker on the photo, not outside it.
O n the back of each illustration,
affix a label witli the appropriate figure
number and the article title. Do not trim,
mount, clip, or staple illustrations.
Color Photographs. For originals, 35-mm
slides or 4- by 5-inch transparencies are
ideal. For review purposes (see below) we
need an additional three complete sets of
color prints, slide dupes, or high-quality
color Xeroxes submitted with the copies
of the paper. We reserve the right to reject
photographs that are not in keeping with
the production standards of Gems eJ
Gemology. Please contact the Editor's
office if you need help finding appropriate
photographers or speciniens.
"Call Outs." Figures and tables must be
called out at the appropriate place in the
text. Figures and tables must be numbered consecutively, starting with the
first mention in the text.
Figure Captions. Type figure captions
double spaced on a separate page(s).Each
caption should clearly explain, in complete sentences, the significance of the
figure and any symbols, arrows, numbers,
or abbreviations used therein; it should
be consistent with the text. Where a
magnification is appropriate, please
include it in the caption. [For the purposes of this journal, the magnification for a
photomicrograph should refer to the
magnification at which the image was
photographed.)

Because your manuscript will be reviewed

by at least three separate reviewers, we
must have three complete copies of the
paper and three complete, properly labeled
sets of all illustrations. Color Xeroxes of
the color illustrations are acceptable as
long as they are of reasonable quality. We
cannot consider any submission that does
not include these three copies of the
manuscript and figures.

Send all submissions to:

Alice S. Keller, Editor
Gems 01 Gemolosy
1660 Stewart Street
Santa Monica, CA 90404
Copyright. In view of U.S. copyright law,
we must ask that submitted manuscripts
be accompanied by the following statement, signed by all of the authors:
'Upon publication of (title of article)
in Gems &> Gemology, I (we) transfer to
the Gemological Institute of America all
electronic-use, serialization, and ancillary
rights, titles, and interest to the work,
including copyright, together witli full
right and authority to claim worldwide
copyright for tlie work as published in
this journal. As author(s),I [we)retain tlie
right to excerpt (up to 250 words) and
reprint the material o n request to the
Gemological Institute of America, to
make copies of the work for use in classroom teaching or for intenial distribution
within my (our) place of employment, to
use-after publication-all or part of this
material in a book I (we) have authored,
t o present this material orally at any
function, and to veto or approve permission granted by the Gemological Institute
of America to a third party to republish
all or a substantial part of the article. I
( w e ) also retain all proprietary rights
other than copyright (such a s patent
rights]. I (we) agree that all copies of the
article made within these terms will
include notice of the copyright of the
Gemological Institute of America. This
transfer of rights is made in view of the
Gemological I n s t i t u t e of America's
efforts in reviewing, editing, and publishing this material.
As autlior(s], I (we)also warrant that
this article is my (our)original work. This
article has been submitted in English to
this journal only and has not been published elsewhere."
No payment is made for articles published in Gems i') Gemology. However,
for each article die author(s) will receive
50 free copies of that issue.

REVIEW PROCESS
Manuscripts are examined by the editor,
editor-in-chief, technical editor, and at
least three reviewers. Authors will remain
anonymous to the reviewers. Decisions of
the editor are final. All material accepted
for publication is subject to copy editing;
authors will receive galley proofs for
review and are held fully responsible for
the content of their articles.

GEMS &. GEMOLOGY

Summer 1996