Spring 1997 Gems & Gemology - GIA · Geophysics in Gem Exploration GEMS & GEMOLOGY Spring 1997 5 Gemology and geophysics are similar in many ways. The nondestructive techniques that

T A B L E O F C O N T E N T S

ABOUT THE COVER: Geophysics has the potential to play an increasingly importantrole in gemstone exploration. Advances in geophysical technology and computer imag-ing now permit mapping of geologic features at various depths beneath the Earth's sur-face. Seismic-reflection profiling and georadar are particularly promising for high-reso-lution mapping of near-surface pegmatites, similar to the one in the Mesa Grande dis-trict of San Diego County—the Queen mine—that yielded these beautiful tourmalinecrystals. Extremely rare, these "blue tops" measure about 12.5 cm (5 inches) across;the tallest crystal is about 17.5 cm (7 inches) high. Courtesy of Pala International,Fallbrook, California.

Photo © Harold & 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.© 1997 Gemological Institute of America All rights reserved. ISSN 0016-626X

11 EEDDIITTOORRIIAALL

TThhee DDrr.. EEddwwaarrdd JJ.. GGüübbeelliinn MMoosstt VVaalluuaabbllee AArrttiiccllee AAwwaarrddAlice S. Keller

FFEEAATTUURREE AARRTTIICCLLEESS

44 AApppplliiccaattiioonnss ooff GGeeoopphhyyssiiccss iinn GGeemmssttoonnee EExxpplloorraattiioonn

Frederick A. Cook

2244 RRuubbiieess aanndd FFaannccyy--CCoolloorr SSaapppphhiirreess ffrroomm NNeeppaallChristopher P. Smith, Edward J. Gübelin, Allen M. Bassett, and Mache N. Manandhar

4422 GGeemmoollooggiiccaall PPrrooppeerrttiieess ooff NNeeaarr--CCoolloorrlleessss SSyynntthheettiicc DDiiaammoonnddss

James E. Shigley, Thomas M. Moses, Ilene Reinitz, Shane Elen, Shane F. McClure, and Emmanuel Fritsch

RREEGGUULLAARR FFEEAATTUURREESS

5544 GGeemm TTrraaddee LLaabb NNootteess6600 GGeemm NNeewwss7711 GGeemmss && GGeemmoollooggyy CChhaalllleennggee7733 GGeemmoollooggiiccaall AAbbssttrraaccttss

pg. 5

pg. 33

pg. 43

pg. 64

It is with great pleasure that we announce that Dr. Edward J. Gübelin has agreed to have his name associated withGems & Gemology's most valuable article competition. One of the industry's most respected scientists, Dr.Gübelin's name has been synonymous with gemology for over 60 years. He has written numerous books and arti-cles (including the paper on Nepal rubies in the present issue), published a comprehensive map on world gem local-ities, and even produced films on the historic Burmese ruby and jadeite deposits. Photoatlas of Inclusions inGemstones, co-authored with John I. Koivula, is perhaps his best-known contribution; the more than 1,400 pho-tomicrographs are unparalleled in both their quality and the information they convey. In Dr. Gübelin's honor, theaward has been renamed the Dr. Edward J. Gubelin Most Valuable Article Award.

Beginning this year, as a result of Dr. Gübelin's generosity, the first-place winner of the Dr. Edward J. GübelinMost Valuable Article Award will receive a plaque, and winners of the second- and third-place awards will be givencertificates. Awards of $1,000, $500, and $300, respectively, will also be given to the authors of the three articlesthat our readers selected as the most valuable papers published in 1996.

Receiving the Gübelin plaque for the 1996 first prize is the article "De Beers Natural versus Synthetic DiamondVerification Instruments," by De Beers researchers Christopher M. Welbourn, Martin Cooper, and Paul M. Spear.This Fall issue article describes the research behind, and the operation of, two ground-breaking new instruments toseparate natural from synthetic diamonds. Another article about diamonds, although from a completely differentperspective, won second place: A. J. A. "Bram" Janse's "A History of Diamond Sources in Africa: Part II." (Part I ofMr. Janse's article won third place in last year's contest.) John I. Koivula, Robert C. Kammerling, Dino DeGhionno,Ilene Reinitz, Emmanuel Fritsch, and Mary L. Johnson co-authored this year's third-place winner, "GemologicalInvestigation of a New Type of Russian Hydrothermal Synthetic Emerald." Both the second- and third-place arti-cles appeared in the Spring 1996 issue.

Photographs and brief biographies of the winning authors appear below. Congratulations also to MargaretAlexander of Mimbres, New Mexico, whose ballot was randomly chosen from all submitted to win the five-yearsubscription to Gems & Gemology.

F I R S T P L A C ECCHHRRIISSTTOOPPHHEERR MM..WWEELLBBOOUURRNN MMAARRTTIINN CCOOOOPPEERR PPAAUULL MM.. SSPPEEAARR

CChhrriissttoopphheerr MM.. WWeellbboouurrnn isprincipal scientist in the PhysicsDepartment at De Beers DTCResearch Centre, Maidenhead,United Kingdom. Dr. Welbourn,who joined the De BeersResearch Centre in 1978, has aPh.D. in solid state physics fromthe University of Reading. He has published several papers on the use of optical spectroscopy and X-ray andcathodoluminescence topography to study diamonds. Research director at De Beers DTC Research Centre, MMaarrttiinnCCooooppeerr has a B.Sc. in physics from the University of London and a M.Sc. in materials science from BristolUniversity. PPaauull MM.. SSppeeaarr is a research scientist in the Physics Department at De Beers DTC Research Centre.With De Beers since 1986, he has a Ph.D. from King's College, University of London.

THE DR. EDWARD J. GÜBELIN MOST VALUABLE ARTICLE AWARD

ALICE S. KELLER, EDITOR

Christopher M. Martin Cooper Paul M. SpearWelbourn

Gubelin Award GEMS & GEMOLOGY Spring 1997 1

S E C O N D P L A C E

AA.. JJ.. AA.. ""BBRRAAMM"" JJAANNSSEE

AA.. JJ.. AA.. ""BBrraamm"" JJaannssee is manager of his own geological consulting company—ArchonExploration Pty Ltd, in Perth, Western Australia—and a director of KWG Resources,Montreal, Canada. During the 38 years he has been involved with diamond explo-ration, he has worked on projects in Australia, Brazil, Canada, India, and SouthAfrica. He has a B.Sc. in geology and a M.Sc. in petrology and mineralogy from theUniversity of Leiden in the Netherlands, and a Ph.D. in petrology from theUniversity of Leeds in England. He is currently developing an extensive database ondiamond and kimberlite occurrences.

T H I R D P L A C E

JJOOHHNN II.. KKOOIIVVUULLAA RROOBBEERRTT CC.. KKAAMMMMEERRLLIINNGG DDIINNOO DDEEGGHHIIOONNNNOO IILLEENNEE RREEIINNIITTZZ EEMMMMAANNUUEELL FFRRIITTSSCCHH MMAARRYY LL.. JJOOHHNNSSOONN

JJoohhnn II.. KKooiivvuullaa,, chief researchgemologist at the GIA Gem TradeLaboratory (GIA GTL), is co-editor ofthe Gem News section of Gems &Gemology. World renowned for hisknowledge of inclusions, Mr. Koivulaholds bachelor's degrees in chemistryand mineralogy from EasternWashington State University. The lateRRoobbeerrtt CC.. KKaammmmeerrlliinngg was vice-presi-

dent of research and development at GIA GTL, then in Santa Monica, California. Aprolific author, he was also an associate editor of Gems & Gemology and co-editor ofthe Gem Trade Lab Notes and Gem News sections. DDiinnoo DDeeGGhhiioonnnnoo is senior staffgemologist at GIA GTL, Carlsbad, California, and a contributing editor to the GemNews section of Gems & Gemology. Before he joined GTL, Mr. De Ghionno workedfor 14 years as a teacher and later manager of GIA's resident student program in colored stone identification.IIlleennee RReeiinniittzz is a research scientist at GIA GTL in New York. A regular contributor to the Gem Trade LabNotes section and a co-author of several Gems & Gemology articles, she has a Ph.D. from Yale University.EEmmmmaannuueell FFrriittsscchh is a professor of physics at the Gemology Laboratory, Physics Department, University ofNantes, France. He has a Ph.D. from the Sorbonne in Paris. His research specialties are spectroscopy in gemolo-gy, the origin of color in gem materials, and treated andsynthetic gems. MMaarryy LL.. JJoohhnnssoonn,, also a co-editor ofGems & Gemology's Gem News section, is a researchscientist at GIA GTL. Dr. Johnson, has co-authored sev-eral articles in Gems & Gemology, is a Gem Trade LabNotes contributing editor, and is a frequent contributorto the journal's abstract section. She holds a Ph.D. inmineralogy and crystallography from HarvardUniversity.

From left, John I. Koivula, Dino DeGhionno, and Mary L. Johnson

A. J. A. (Bram) Janse

Ilene Reinitz

Robert C. Kammerling Emmanuel Fritsch

2 Gubelin Award GEMS & GEMOLOGY Spring 1997

uch of the exploration for gemstones (even dia-monds, in some regions) is conducted with rela-tively simple, often primitive, techniques. In

many areas of known gem occurrences, development and min-ing involve little more than shovels and screens for alluvialdeposits, and explosives plus mechanical tools (such as pneu-matic drills) for hard-rock deposits (e.g., pegmatites; figures 1and 2). In areas where gems are not known to occur, most dis-coveries are made by chance. This contrasts with the explo-ration procedures routinely used for many other Earthresources, such as oil and gas or metallic minerals, in whichyears of regional geologic and geochemical mapping are usuallyfollowed by intensive periods of acquiring and analyzing geo-physical data. Once target areas are located, more detailed—and usually more expensive—exploration methods (such asdrilling or excavation) can be undertaken. However, geophysi-cal techniques have seldom been used to delineate gemstone,particularly colored stone, deposits.

This article describes the principles behind geophysicalexploration, together with some applications in diamond explo-ration and some potential applications in colored stone explo-ration. Different contemporary geophysical exploration tech-niques are highlighted, and suggestions are made with regard tohow these techniques could be applied to certain types of gemoccurrences. As consumer demand for gems continues to rise,and older deposits are depleted, geophysical prospecting hasconsiderable potential for identifying new resources or newareas of gem mineralization at existing occurrences.

GEOPHYSICS AND GEMOLOGYGeophysics is the application of physical principles (e.g., usingmagnetism, gravity, or electrical properties) to study the inter-nal structure and geologic evolution of the Earth and otherplanets. In particular, physical principles are used to study thecomposition and geologic structure of rocks below the surface.Thus, geophysics includes a number of techniques that canhelp identify promising localities for mineral exploration.

4 Geophysics in Gem Exploration GEMS & GEMOLOGY Spring 1997

APPLICATIONS OF GEOPHYSICS INGEMSTONE EXPLORATION

By Frederick A. Cook

ABOUT THE AUTHOR

Dr. Cook is professor in the Department ofGeology and Geophysics, The University ofCalgary, Alberta T2N 1N4, Canada.

Acknowledgments: This work was supported bythe Natural Sciences and Engineering ResearchCouncil of Canada. The author thanks Dr. A. A.Levinson of the University of Calgary for manyhelpful discussions, constructive criticisms, andeditorial suggestions. Dr. A. G. Green of the SwissFederal Technical Institute in Zurich kindly providedsupport for recording the georadar data in theGotthard area of the Swiss Alps. The georadardata used to produce figure 9 were processed atthe University of Calgary, and the assistance of A.Van der Velden and Dr. K. Vasudevan in process-ing those data is greatly appreciated.

Gems & Gemology, Vol. 33, No. 1, pp. 4–23.

© 1997 Gemological Institute of America

Advances in geophysical technology andcomputer imaging now permit mapping ofgeologic features at various depths beneaththe Earth’s surface in more detail than everbefore. Geophysical methods, which for someyears have been used to explore for diamonds(and many other minerals), also hold promiseas important tools in both exploration fornew colored stone deposits and assessment ofknown deposits. Particularly exciting are theuse of seismic-reflection and radar-imagingtechniques for high-resolution mapping ofnear-surface pegmatites, veins, metamorphiclayering in crystalline rocks, and alluvialdeposits in sedimentary rocks. These tech-niques can produce images of regions only afew centimeters to a few meters in size downto several meters to tens of meters below thesurface. Thus, it is now possible to locatedirectly some buried structures containinggem deposits.

M

Geophysics in Gem Exploration GEMS & GEMOLOGY Spring 1997 5

Gemology and geophysics are similar in manyways. The nondestructive techniques that a gemolo-gist uses to determine the identity of a gemstonetypically include determination of refractive index(speed of light ratios), specific gravity (density), ther-mal conductivity, and magnetism (occasionally), aswell as describing or mapping internal characteris-tics (e.g., inclusions, internal growth structure, etc.).A geophysicist also uses generally nondestructive(sometimes called remote sensing) methods. Someof the most common techniques (table 1) includethose that measure variations in the speed of vibra-tional waves (seismic refraction), in rock density(gravity), in the thermal state of the Earth (heat flow),and in rock magnetism (magnetics); as well as thosethat use imaging to map subsurface structures (seis-mic-reflection profiling, ground-penetrating radar).

There are essentially two branches of geo-physics that apply to studies of rocks: (1) whole-Earth geophysics, and (2) exploration geophysics. In

whole-Earth geophysics, geophysical methods areused to map deep and large-scale variations in theEarth’s properties, such as the depth and configura-tion of the Earth’s core, the internal characteristicsof the Earth’s deep mantle (the region of the earthbetween the crust and the core), the depth andstructure of the lithosphere (the outer shell of theEarth, approximately 100 km deep on average), andthe thickness and properties of the crust (the outer-most layer of the lithosphere, about 10–40 kmthick; see figure 3). The lithosphere is generally rigidand consists of a series of “plates” that shift relativeto one another (i.e., plate tectonics), thus producingearthquakes, volcanoes, and uplift of mountains.The lithosphere usually reaches its greatest thick-ness (as much as 300–400 km) beneath ancient con-tinents.

In exploration geophysics, on the other hand,geophysical techniques are used to search for hydro-carbon (oil and gas), mineral (gold, silver, lead, zinc,

Figure 1. The discovery ofpegmatites, and of gem-

bearing pockets in a peg-matite dike, is one of the

greatest challenges in gemexploration. These bicol-

ored tourmalines from theHimalaya mine, San

Diego County, California,represent some of the fine

gems waiting to be discov-ered in pegmatites world-

wide. Courtesy of PalaInternational, Fallbrook,

California; photo ©Harold & Erica Van Pelt.

TABLE 1. Common geophysical exploration methods and potential applications to gemstone explorationa.


etc.), and other economically significant deposits.Hydrocarbon and mineral deposits are usually locat-ed in geologic structures that are large enough(sometimes several square kilometers) to be easily

mapped by geophysical imaging. Although geophys-ical results rarely can identify whether economical-ly significant hydrocarbons or minerals are present,they often can pinpoint geologic features that may

Figure 2. This pegmatite pocket fromCalifornia’s Himalaya mine yieldedthousands of carats of fine tourmaline.Unfortunately, the discovery of suchpockets typically requires hundreds ofmeters of tunneling and a great deal ofluck. Rapid improvements in geophysi-cal prospecting techniques offer thehope that this situation will change inthe near future. Photo © Harold & EricaVan Pelt.

Effective resolution

100 m to 10s of km(depends on spacing ofmeasurements)

A few m to 10s of km(depends on spacing ofmeasurements)

10 m to 100s of km(depends on spacing ofmeasurements)

A few m to 10s of km(depends on knowledge ofproperties at depth)

A few cm to 100s of m (dependson spacing of measurements)

10s of km(depends on frequency of signaland wave velocity)

10 m to 100s of m(depends on length of profiles;energy source)

A few m to 100s of m(depends on frequency of signaland wave velocity)

About 10 cm to 1 m(depends on frequency of signaland wave velocity)

Effective depth

Meters to 100s of km(depends on spacing of mea-surements)

Meters to 50 km(depends on spacing of mea-surements)

Meters to 100s of km(depends on measurementtime)

1 km to 100s of km(depends on knowledge ofproperties at depth)

A few cm

To center of earth

10 m to 100s of km(depends on length of profiles;energy source)

100 m to 100s of km(depends on length of profiles,energy source)

Meters to 100 m

Usual geologic application

Mapping subsurface structurebased on variations in density

Mapping subsurface structurebased on variations in magneticproperties

Mapping fluids (e.g., groundwater) and metal deposits

Mapping variations in the outflowof the Earth’s heat

Mapping concentrations ofradioactive minerals

Mapping large-scale (lithosphere)and deep-Earth structure

Mapping crustal thickness andregional structure; depth tobasement

Mapping detailed geologicstructure and stratigraphy

Mapping detailed geologicstructure and stratigraphy.

Principle involved

Responds to variations in rockdensity

Responds to variations in rockmagnetism (usually ironcontent)

Responds to variations in elec-trical conductivity (metals orfluids)

Responds to variations in tem-perature at depth; thermal con-ductivity

Responds to variations inradioactivity of rocks

Responds to variations in seis-mic-wave velocity (varies withrock density)

Responds to variations inacoustic impedance (density ×velocity)

Responds to variations inacoustic impedance (density ×velocity)

Responds to contrasts in elec-trical conductivity

Method

Gravity

Magnetics

Electrical

Heat flow

Radioactivity

Earthquake seismology

Seismic refraction

Seismic-reflectionprofiling

Ground-penetratingradar (Georadar)

aNot all of these methods are described in detail in the text as their application to gemstone exploration may be limited (e.g., electrical, heat flow).Nevertheless, they may be useful for reconnaissance work under favorable conditions. For additional information on the applications of these techniques,see Telford et al. (1976).


be promising targets. Unlike mineral and hydrocar-bon deposits, however, many primary gemdeposits—such as those in veins or pegmatite cavi-ties (figure 2)—are no larger than a few meters indiameter. Because of their small size, these depositsare easily missed with conventional mining meth-ods (e.g., trenching or tunneling), and most geophys-ical techniques cannot produce the resolution(detail) necessary to detect them.

Today, though, new methods of data acquisitionand analysis (including sophisticated computer pro-cessing) enable some geophysical techniques toresolve geologic features that are as small as a frac-tion of a meter to a few meters in longest dimensionand very shallow, within a few meters of the surface.In addition, improvements in our knowledge of deepgeologic structures (tens to perhaps hundreds of kilo-

meters below the surface) now make it feasible touse large-scale reconnaissance geophysical tech-niques to identify regions of the Earth within whichcertain kinds of gem deposits might be found. Thus,the science of geophysics has the potential to play anincreasingly important role in the exploration anddevelopment of some gemstone deposits. It is evenpossible that specific exploration strategies could bedeveloped using geophysical techniques to optimize:(1) opportunities of for success in exploring for gems,and (2) knowledge of existing deposits (e.g., theirextent and thus, potentially, their value) .

THE “RESOLUTION PROBLEM”The objective of all geophysical methods is to pro-duce an image of the Earth’s interior, usually bymaking measurements at the surface. The type of

Potential applicationin gem exploration

Reconnaissance; thickness of crust andlithosphere

Delineating kimberlites; mappingregional geologic structures

Reconnaissance; thickness of litho-sphere; mapping kimberlites

Reconnaissance; thickness of the crustand lithosphere

Might help to indicate presence ofaccessory radioactive minerals

Reconnaissance; thickness of the litho-sphere in diamond exploration

Reconnaissance; thickness of the crust;depth to basement beneath alluvialdeposits

Reconnaissance; thickness of the crust;detailed structure; mapping of alluvialdeposits

Alluvial deposits; delineating pegmatitecavities

Comments

Used for reconnaissance work and as ameasurement for density estimate

Used to map kimberlite pipes; may alsoassist in mapping regional geologicstructures in basement

Used to explore for metal deposits; mayassist mapping lithosphere thicknessand basement structures

Used to calculate temperature at depth,geothermal energy

Used to explore for radioactive mineraldeposits

Used for mapping variations of seismicvelocity in the mantle

Used for mapping variations in depth tomantle and depth to basement

Most widely used geophysical methodin hydrocarbon exploration due to goodimage quality

Used in mapping near-surface featuressuch as archeological sites, pipelines,tunnels, etc.

Limitations

Inherent ambiguity between rockgeometry and density limits usefulnessin exploration

Inherent ambiguity between rockgeometry and magnetism limitsusefulness in exploration

Inherent ambiguity between rockgeometry and conductivity limitsusefulness in exploration

Requires drillholes, thus limitingnumber of measurements; inherentambiguities in calculation

Limited penetration; not useful forregional studies

Low resolution; not useful for detailedstructure

Low resolution; not useful for detailedstructure

Relatively expensive; not easy to imageshallower than 50 m

Signal can only penetrate up to 50 or100 m; quality may be severely affectedby surface material

Relative cost

Low ($100s per day);cost increases for greater detail



Low, but usually requires costly drill-holes at $1000s per drillhole


Low to moderate;cost increases with numberof receiving stations

Moderate ($100s–1000s per day);cost per kilometer increases withnumber of receiving stations

High ($1000s per day);cost increases for greater detail

Low ($100s per day)


image depends on what parameter(s) are measuredby the particular method being used (again, see table1). The value of a technique depends on its resolu-tion, that is, its ability to distinguish responses fromdifferent types and sizes of geologic features.

In all cases, the measured response includes thecombined effects of rocks between the target zone(area of interest) and the surface; in some cases, italso includes features below the zone of interest.Consider the following analogy. In measuring thedensity (specific gravity) of a gem, a gemologistweighs a sample first in air and then in water, andcalculates the specific gravity according to howmuch water is displaced. In fact, however, the spe-cific gravity of the sample is affected by the specificgravity of the gem material plus the specific gravi-ties of any inclusions, some of which may belighter (e.g., air or water) or heavier (e.g., heavy

minerals) than the host gem. For example, an emer-ald that has many pyrite (S.G. ~ 5.00) inclusionswill have a specific gravity that is greater than forberyl alone (2.70). The inclusions thus add noise(i.e., an unwanted signal) to the result, which caus-es the measured specific gravity to deviate from thetheoretical value of the pure gem material or evenfrom the value typical for relatively inclusion-freematerial.

The geophysical method analogous to specificgravity is the measurement and analysis of theEarth’s gravity field (Box A; table 1). After account-ing for large-scale variations caused by the Earth asa whole (such as its shape and size), the remainingsignal (e.g., the pull of gravity) is a function of varia-tions in the densities of rocks at depth. When a geo-physicist measures the Earth’s gravity field at alocation on the surface, the reading includes the

Basin

Faults

Descending Plate Magma

�� Archon

200 km0

Lith

osph

ere

Asthenosphere

Subcrustal lithosphere

Crust

��

7

8

4

Ocean CratonOrogen

Ocean basins Continents

Lith

osph

ere

Asthenosphere

Subcrustal lithosphere

Lower crust

Orogen Basin

Crus

t

Figure 3. In this schematic diagram, normal lithosphere is segmented into the tectonic plates of the Earth’s surface,with unusually thick lithosphere (combined crust and subcrustal) under archons (portions of cratons with an age offormation exceeding 2,500 million years [Janse, 1994]), where diamonds may form. Cratons are areas of the continentthat have been stable for long periods of geologic time, and orogens are regions that have been subjected to foldingand faulting. Geophysical techniques may be applied in different areas and at different scales to delineate gemdeposits. For example, some methods are useful on a reconnaissance scale to delineate thick lithosphere, where mostprimary diamond deposits are found, as well as to identify regions of thin lithosphere, where economic primary dia-mond deposits are very unlikely to occur. Other geophysical methods are more useful for intermediate-scale explo-ration. Magnetic surveys, for example, respond to differences in properties between kimberlites and surrounding rocksand assist in locating specific areas for exploration (figure 4), whereas seismic profiles assist in delineating the deepgeometry of mountains (figure 7) and basins (figure 8). Approximate positions of data in figures 4, 7, and 8 are shownas circled numbers.


combined effects of all rocks below the area beingmeasured, from great depths to the surface. Theanalysis must account for as many effects as possi-ble, so that resultant anomalies (in this context,deviations from uniformity or normal values in anexploration survey) can be interpreted. However, itis often difficult to separate the effects of differentmasses of rock, because their shapes and densitiesare not usually known. Because such detailedknowledge of subsurface rock types and geometriesis not yet available, geophysicists cannot resolve

variations in rock density to a few meters in size, aswould be necessary for many gem exploration appli-cations. Similar difficulties exist in measuring theEarth’s magnetic field (table 1), although high-preci-sion instruments and detailed measurements nowpermit greater resolution than ever before.

In fact, interpretations of virtually all geophysi-cal measurements require careful understanding ofthe possible causes of the background noise thataffects resolution. Usually more than one method ofexploration must be applied before an economic

Subtle variations in the Earth’s gravitational pull arebrought about by differences in the density of largemasses of rock. These variations can be identified bythe gravity method, which is analogous to measuringspecific gravity. In it, a gravity meter, or gravimeter—asensitive mass-spring system that weighs about 5–7 kg(10–15 pounds)—is carried (usually on the ground,although sometimes in an airplane) from location tolocation, measuring the pull of the earth’s gravity field(figure A-1). When a gravity measurement is taken overan object with a high density, the object pulls on themass and is recorded as an increase in gravity (figure A-1); decreases in gravity occur over low-density objects.

Once appropriate corrections for elevation and lat-itude are made to the recorded gravity data, the result-ing values (anomalies) represent variations in mass(density) at depth. A problem inherent in interpretingthese kinds of data is that the actual measured value isthe sum of responses of all mass variations at depth(e.g., A and B in figure A-1). This makes it difficult todetermine the subsurface structure with much preci-sion from this method alone, so gravimeter data areusually used in conjunction with magnetic and otherdata.

The magnetic field at any location is a combina-tion of the field produced by the Earth’s core pluseffects from nearby magnetic objects, such as iron-bearing rocks. Magnetic data are acquired in much thesame way as gravity data, except that a magnetometeris used to measure the magnetic field on or above thesurface. Once the Earth’s field is removed from themeasured signal, the resulting anomalies can be inter-preted in terms of geologic features. Normally, rocksthat contain large amounts of iron (e.g., kimberlites)produce large magnetic anomalies (see text figure 4).One advantage that the magnetic method has overmost other methods is that a magnetometer can be car-ried behind an airplane, thus allowing the acquisitionof large amounts of data in a short time. Interpretationof magnetic anomalies is similar to interpretation of

gravity anomalies. Aeromagnetic surveys have beenbroadly used in the exploration for diamond depositsin Australia, Africa, and, most recently, Canada.

Gravity Measurements

A

Gravity effect of A

Total measured gravity

Gra

vity

Distance along surface

MassMass

Mass

Spring

Distance

Gravimeter

Depth

Deflection due to pull from A and B

B

Gravity effect of B

Figure A-1. In this schematic illustration of thegravity method, a gravity meter (gravimeter),which is essentially a mass-spring system, is placedat a series of locations along the surface. The dis-placement of the mass is a function of the strengthof the gravity at each location. Following correc-tions for elevation and latitude, the remaining sig-nal consists of a series of anomalies that representthe combined gravitational attraction of local geo-logic features (A and B in this figure).

BOX A: GRAVITYAND MAGNETIC METHODS


deposit can be clearly identified. The problem ofbackground noise is even greater for conditions ofanalysis that might provide the detailed resolutionnecessary to delineate most types of gem deposits.

Nevertheless, when much larger-scale varia-tions are considered (such as changes in the litho-sphere or locating some large features that may con-tain gems), geophysics can and does play an impor-tant role. Such is the case in diamond exploration.

CURRENT USE OF GEOPHYSICSIN DIAMOND EXPLORATIONPrimary Deposits. Geophysical techniques are com-monly used to locate promising regions for primarydiamond deposits (Atkinson, 1989; Smith et al.,1996), which are found in two types of rare igneousrocks, kimberlites and lamproites. The magmasfrom which these igneous rocks crystallized origi-nated in the Earth’s mantle at depths of more than150–200 km. As they ascended to the surface, thesemagmas traveled through the lower part of the con-tinental lithosphere, where the diamonds probablyformed (again, see figure 3). Sometimes, these mag-mas carried some of the diamonds along with them.We know, then, that primary diamond depositsform in regions where the continental lithosphere isthick (150–200 km or more); to be economic, thekimberlites or lamproites must be present at or nearthe surface (Levinson et al., 1992).

The presence of kimberlites or lamproites canbe detected by geophysical methods, particularly bymagnetic and electrical techniques. Kimberlites andlamproites typically have high iron concentrations,so they often are more magnetic and more electri-cally conductive than the rocks they have intruded(Atkinson, 1989; Hoover and Campbell, 1994;Smith et al., 1996). Furthermore, they usually formcircular or irregularly shaped structures (in mapview), so that they can be pinpointed by magnetic

(figure 4) and electrical measurements (Smith et al.,1996), even if they are covered by a thin blanket ofsedimentary rocks.

However, only a small fraction of kimberlitescontain diamonds in economic amounts, and thesetend to be in areas with thick lithosphere. Thicklithosphere is more likely to have economic kim-berlite or lamproite deposits because diamondsform from carbon (or carbon-containing com-pounds such as methane and carbon dioxide) atpressures corresponding to 150–200 km depth. Thelithosphere is usually about 100 km thick, so a150–200 km thick lithosphere is uncommon andwould appear as a “keel” or “root,” much like thelower part of an iceberg. Thick lithosphere tends tooccur beneath old continental cratons (those partsof the earth’s crust that have attained stability, andhave been little deformed for prolonged periods),because that is where there has been the greatestamount of time for the lithosphere to cool andthicken (billions of years, as compared to a fewhundred million years for oceanic regions). Thus,methods that measure the thickness of the litho-sphere can be valuable in assessing whether aregion might contain diamondiferous kimberlites.Figure 5, for example, illustrates the various thick-nesses of the lithosphere under North America,with specific reference to the recently discovereddiamond deposits in Canada. Lithospheric roots arehigher in density than the surrounding mantlematerial, so geophysical techniques that respond todensity are useful (e.g., gravity and earthquake seis-mology; again, see table 1).

Measurements of seismic-wave velocity providesome estimation of variations in rock density as wellas of the rigidity (strength) of the rocks. Becauserocks in the mantle part of the lithosphere usuallyare denser and more rigid, they have higher seismic-wave velocities than adjacent rocks of the partly

300 m

Jwaneng Area

Kimberlite pipe

Kolo Area

50 m0 0

Kimberlite pipe

Figure 4. Magnetic anomalies from kimber-lites tend to be roughly circular in mapview (modified from Atkinson, 1989). Onthese maps of two kimberlites in Africa,Kolo and Jwaneng, the red lines representmagnetic highs (regions of high iron con-centrations), the blue contours representmagnetic lows, and the gray stippled areasare outlines of the kimberlite pipes. Mapinformation taken from Burley andGreenwood (1972, for the Kolo area) andLock (1985, for Jwaneng).


a. 0–140 km Depth Range

c. Cross-section A-A'

b. 235–320 km Depth Range

0∞ W

30∞ N

30∞W60∞W90∞ W

0∞ N

A

A'

60∞N

0∞ W

30∞ N

30∞W60∞W90∞ W

0∞ N

A

A'

60∞N

400

A A'Surface

Dept

h(k

m)

Canadian ShieldGreenland

Iceland

Map b

Map a

800

1200

1600

0.0 Velocity Perturbation

-3.0% (Slow)

+3.0% (Fast)

Figure 5. The thickness of the lithosphere beneath continents may be estimated by measuring the velocities of seismicwaves. Two maps of North America (a and b) show estimates of seismic-wave velocity for different depth ranges. Areasof increasingly high velocity (and probably increasingly rigid material) are represented by light to dark blue, whereasareas of lower velocities (and probably less rigid, partly molten material) appear pink to red. (a) For depths from the sur-face down to 140 km, this map shows high velocities under the center of the continent, where the lithosphere is thick.The velocities tend to decrease slightly near the edges of the continent, where some partly melted rock is closer to thesurface. (b) For the depth range 235–320 km, only the central part of the continent (Canadian Shield) has high velocitiesand, thus, more rigid rock; this is the area where major primary diamond deposits have recently been found (diamondsymbol). In both maps, dotted lines represent the boundaries between tectonic plates, and the scale at the bottom indi-cates perturbation of seismic-wave velocity from expected values. A-A’ is the location of the cross-section of (c).Diagonal-lined areas have no data. (c) Cross-section A-A’ through the North American continent illustrates seismic-wave velocity as a function of depth. Again, the dark blue (high velocity) region extends to 300–600 km depth beneaththe central part of the continent (Canadian Shield). The depth ranges of maps a and b are shown on the right of the dia-gram. Figures are modified from Grand (1987).


molten asthenosphere (the part of the Earth’s mantlethat is located below the lithosphere and that actssomewhat like a fluid; figure 3). The top of theasthenosphere is found typically between about 100and 200 km, but it is sometimes much deeperbeneath ancient cratons (again, see figure 3). Thus,maps of wave velocities (measured from earthquakewaves) at different depths have anomalously highvalues (hence, high densities and greater strength)deep below the thick cratons (Grand, 1987). Thedark blue areas of figure 5 represent regions withhigh density and are interpreted as the lithosphericroots. Once the lithospheric roots have been located,higher-resolution magnetic and electrical data mayallow identification of anomalies associated withindividual kimberlite or lamproite structures. Theseare the kimberlite structures that should receive themost intensive study (Atkinson, 1989; Smith et al.,1996). Of course, identification of kimberlite fieldsin areas of thick lithosphere does not guarantee thateconomic diamond deposits will be found. Withoutthese conditions (kimberlites/lamproites and thicklithosphere), though, the possibility of economic pri-mary deposits is severely diminished.

Atkinson (1989) states that geophysical tech-niques, specifically magnetics, were used as early as1932 to locate the boundaries of kimberlite pipes.Gravimetric and resistivity surveys were also usedfor similar purposes. All these early attempts wereconducted on the ground. Following World War II,however, aeromagnetic surveys were flown as partof exploration for kimberlite, with the Russiansbeing the first to use this method extensively inYakutia. Since the early 1970s, aeromagnetics havebeen used in many places, such as Australia andBotswana, resulting in the discovery in the late1970s of the Ellendale lamproites in WesternAustralia. Today, geophysical techniques, particu-larly aeromagnetics, are used throughout the worldin exploration for kimberlites, such as in theNorthwest Territories, Canada (Smith et al., 1996).

Secondary (Including Alluvial) Deposits. Explorationfor secondary diamond deposits (alluvial, beach,marine deposits etc.) involves identification of thesedimentary layers within which diamonds occur.Diamonds may be transported long distances (hun-dreds of kilometers) from their primary source loca-tions, collecting in deposits that are no longer situ-ated on thick lithosphere. Consequently, geophysi-cal exploration for secondary diamond depositsrequires techniques that allow mapping of detailed

features at shallow depths. For example, in SouthAfrica and Namibia, diamond deposits in river grav-els are found at the edge of, or even offshore from,the African continental craton (Gurney et al., 1991),where the lithosphere is probably thin. In suchcases, it is necessary to explore with methods thatallow mapping of the shallow subsurface geometryof the diamondiferous sediment layers. Because thisapproach is essentially the same for all alluvialgems, colored stones as well as diamonds, it is dis-cussed in greater detail below.

USES OF GEOPHYSICS INCOLORED STONE EXPLORATIONColored stones are found in a much greater varietyof geologic environments than are diamonds, andmany of the most valuable gem deposits occurwhere the stones are concentrated into secondary(alluvial) environments after redistribution fromtheir primary source locations. Also unlike dia-mond deposits, primary colored stone deposits—such as those found in pegmatites, in igneous intru-sions in metamorphic or sedimentary rocks, in vugsin volcanic rocks, and the like—tend to be associat-ed with geologic features on the order of a fewmeters or less. As is the case with diamonds, how-ever, for a colored stone deposit to be economic itmust be located within a few meters or tens ofmeters from the surface. Hence, whether the col-ored stone deposits being investigated are primaryor secondary, the methods chosen must be able todelineate the near-surface geology in some detail.

In the past, geophysical techniques generallyhave not been as useful for mapping the extremelyshallow part of the Earth (upper few meters to tensof meters) as they have been for greater depths; themore established techniques—such as gravity andmagnetic measurements or electrical and seismicmethods—simply have not had the resolution nec-essary to map the near-surface in detail. This is whywe rarely hear of geophysics being used in coloredstone exploration. However, this situation haschanged in the past five to 10 years.

Some geophysical techniques are now beingadapted for application to near-surface geologicinterpretation, primarily to address geologic factorsin environmental problems such as waste disposal(Beres et al., 1995; Lanz et al., 1994). Some of thesemethods are also being used in a limited way toexplore for gemstones (e.g., Patterson, 1996; WilliamRohtert, Kennecott, pers. comm., 1997). Becausemost geophysical exploration techniques (such as


gravity and magnetics; see Box A) can be acquiredand processed with high resolution, they are appro-priate for certain specific, near-surface, explorationtargets. For example, magnetics may be useful inmapping locations of some gem-bearing igneousdikes that have intruded into sedimentary strata(because igneous rocks usually contain largeamounts of iron compared to most sedimentaryrocks), or in mapping iron-rich sedimentary rocks. Infact, magnetics have been used to map concentra-tions of ironstone deposits that contain preciousopal in Queensland, Australia (Senior et al., 1977).Magnetic anomalies coupled with anomalies inradioactivity have also been helpful in outliningpossible areas for exploration of red beryl in Utah(William Rohtert, pers. comm., 1997; figure 6).

Nevertheless, it is rare for these (gravity, mag-netics) and most other geophysical methods to beused for mapping alluvial deposits or other geologicfeatures (e.g., veins, cavities, pegmatites, metamor-phic layering) that contain gemstones. The reasonfor this is exactly the same as that for the difficul-ties that arise in interpreting specific gravity read-ings: The ambiguities inherent in the analyses pre-clude obtaining the necessary subsurface geometryand resolution. Because these methods have some-what limited use in colored stone exploration, theywill not be discussed further.

There are, however, two methods that haveproved valuable for mapping subsurface geology—seismic-reflection profiling (Box B) and ground-pene-trating radar (georadar; Box C)—that may also havebroad applicability to gemstone exploration.Although these respond to different physical proper-ties, the images of subsurface geometry produced byboth of these geophysical exploration techniquesare consistently superior to those of the other meth-ods for mapping geologic layering. Illustrations ofthese methods clarify their potential value.

Seismic-Reflection Profiling. The Method. Seismic-reflection profiling was first used in the 1920s tomap subsurface geologic structures in the search foroil and gas. Since then, it has become the mostimportant tool in exploration geophysics; morethan 90% of geophysical exploration for hydrocar-bons is accomplished with seismic profiling. This isbecause the technique produces results that aresimilar to geologic cross-sections (figure 7).Although interpreting these data often requires spe-cific skills, particularly knowledge of how seismicwaves propagate through rocks, with these skills

the geometry of the rock layers beneath the surfacecan often be determined with precision (again, seeBox B).

The layers on the image in figure 7 representreflections from boundaries between different rocksor, more specifically, between rocks with differentproperties (seismic velocity and density). By deter-mining how fast the waves travel in the rock layersbelow the surface, we can convert the travel timesto the reflecting surfaces to the depths of those sur-faces (figure 7b). Thus, if a wave travels at about 5km (3 miles) per second, a two-way (round-trip)time of 5 seconds represents 12.5 km depth (5km/second × 5 seconds ÷ 2). With the data displayedthis way, we are essentially looking at a cross-sec-tion of this portion of the Earth (about 50 km long ×20 km deep in the example in figure 7), in a mannersimilar to a CAT scan or an X-ray image of a por-tion of the body.

The profile in figure 7, which was taken in theRocky Mountains of southwestern Canada, revealsa cross-section of some deformed (faulted and folded)

Figure 6. Kennecott geologists used a combination ofmagnetic anomalies and radioactivity anomalies to out-line potential deposits of red beryl in Utah’s Wah WahMountains. The 10 × 9 × 8 mm crystal and 0.76 ct cutstone shown here represent some of the fine red berylsfound at this locality. Courtesy of Michael M. Scott;photo © Harold & Erica Van Pelt.

In seismic-reflection profiling, vibrational waves aregenerated on the Earth’s surface and travel into the sub-surface; there, they reflect off rock layers and return tothe surface, where they are received by a row of sensors.It is the most common geophysical technique in hydro-carbon exploration and has been developed to a verysophisticated level. As for all of these techniques, thebasic components for seismic-reflection profiling arefield acquisition, data processing, and interpretation(figure B-1). In data acquisition, a source of elastic(vibrational) energy such as a small explosion or a largevibrator truck (figure B-2) produces the signals that pen-etrate into the Earth and are reflected back. A series ofsensors (geophones; figure B-3) are positioned along thesurface and are connected to a recording truck by cableor radio communication. These sensors measure varia-tions in arrival times and amplitudes of waves, whichrelates to the densities and velocities of the differentrock types. At the recording truck, the received signalsare recorded through a computer onto magnetic tape.

The geometry of the field recording allows signalsfrom a single point on a boundary (P in figure B-1) to berecorded at different geophone positions as the processis repeated at subsequent locations. This means thatthe separate reflections from P can be added together indata processing to enhance the signal from the layerboundary and thus identify the rock formation (or typeof formation) that boundary represents.

Data processing is usually time consuming, andrequires that a highly trained individual make judg-ments about parameters as different computer pro-grams are applied (figure B-1). The basic sequencerequires inputting the data from a field tape to the com-puter facility, displaying the field data (step 1 in figureB-1); editing bad traces (step 2; a trace is the series ofsignals recorded at a specific geophone for a singlesource vibration); collecting the signals of vibrationsthat were reflected from a single point from differentgeophone locations (step 3); removing noise from the

waves that travel along the ground surface, rather thanfrom those that reflected from depth (step 4); lining upthese signals (step 5) and summing them into a singlecomposite trace (step 6); applying different filters toenhance the pulse (step 7); and displaying the final sec-tion for each reflection point (step 8). Data presented inthis article were processed through these steps. Thefinal procedure is to interpret the data (step 9 in figureB-1), which requires that the analysts use as much geo-logic information as possible, as well as any other rele-vant geophysical information, to optimize the result.Because seismic-reflection profiling has potential foridentifying relatively small geologic features (see dis-cussion in the text), it has promise for gem exploration.

BOX B: SEISMIC-REFLECTION PROFILING

Figure B-1. This schematic drawing shows the manysteps (described in the box) that are required to col-lect and analyze data in the seismic-reflection pro-filing technique (modified from Cook et al., 1980).

Figure B-2. Vibrator trucks—typically three or morein tandem—are commonly used as a source of ener-gy for seismic-reflection work. Each of these trucksvibrates a signal of known frequency and energy forseveral seconds. These signals travel into the Earthuntil they are reflected back to the surface by a rocklayer. Photo courtesy of K. W. Hall.

Figure B-3. Geophone sensors, each about 4 cm acrossthe top, are placed on the ground surface to record thevibrations returned from beneath the surface. Photocourtesy of K. W. Hall.


sedimentary rocks. A drill hole located 10 km to theleft of the section intersected the prominent layersobserved; thus, these layers could be interpretedwith some certainty. Furthermore, the drill holealso intersected a zone of metallic (lead, zinc, silver)mineralization that has not been exploited. Thismineralization zone is associated with a stratumthat can be followed as a recognizable layer, some-times even across faults, for tens of kilometers onthese and related seismic-reflection data. Thus,while the method does not actually produce imagesof the lead, zinc, or silver minerals, it does allow usto map the geometry and extent of the layer thatmay contain the minerals.

The large-scale profile (about 100 km long) infigure 8 illustrates an application of the method in areconnaissance survey. This seismic-reflectioncross-section of an area in western Canada requiredabout two to three weeks of field work and anotherfour to six weeks of computerized data analysis(again, see Box B). Layers are visible from near the

surface to the base of the section (about 30–40 kmdepth). In this instance, the method provided animage of a previously unknown, ancient (Precambrian,or older than 570 million years) basin that lies abovewestward-thinning crust and below young, flat-lying sedimentary rocks (figure 8b; Cook and Vander Velden, 1993).

While details of the interpretation are not ofmajor importance here, it is easy to see how seis-mic-reflection profiling can be used to map the deepgeologic boundaries, and thus how it could be avaluable reconnaissance tool. But is this techniqueapplicable to the very shallow part of the Earth,where gemstones might be recoverable? To a largeextent, the answer to this question depends notonly on the depths in question, but also on howsmall a feature the technique can detect; that is, onthe resolution of the signal.

Resolution of the Data. The resolution of seismic-reflection data depends on the wavelength (L) of the

0

5

Trav

eltim

eof

refle

cted

wave

s (s

econ

ds)

Ground surface

Zone of metallic mineralization

Seismic Reflection Profile

0 0

5

5

20

10

15

Appr

oxim

ate

dept

h(k

m)

Trav

eltim

eof

refle

cted

wave

s (s

econ

ds)

Ground surface

10 km0 Basement (not deformed by faults)

Undeformed sedimentary rocks

Deformed sedimentary rocks

Faults (arrows indicate movement direction)

Layer boundaries

b

aFigure 7. (a) This seismic-reflectionprofile, taken in the RockyMountains of southwestern Canada,is plotted with the horizontal axis asdistance along the ground surfaceand the vertical axis as time (in sec-onds) for waves to travel from thesurface (0) to a reflecting boundary(linear features on the data) and backto the surface. (b) Geologic interpre-tation of the data in (a) reveals differ-ent rock types, as shown by the col-ored areas, and faults. In this inter-pretation, sedimentary rocks weredeformed by folding and faultingabove undeformed basement rocks.A drill hole (not shown) about 10 kmto the left of this profile penetrated azone that was found to have largeconcentrations of metallic minerals(lead, zinc, silver). See figure 3 for theposition of this profile relative toother geologic features. Adapted fromCook (1995).


signal, which is a function of the frequencies (f) ofthe waves that are used, as well as on their velocities(V) through the rocks. The following equation sum-marizes the relationship between these parameters:

L = V / f

In general, the smaller the wavelength of the signal,the smaller the feature that can be detected and thehigher the resolution of the measurement. In prac-tice, it is possible to obtain seismic frequencies upto about 100 cycles per second in the shallow sub-surface (first 20–50 m), where the seismic-wavevelocity varies from about 6.0 km/sec in many crys-talline (e.g., granite) rocks to about 2.0 km/sec inloose sedimentary rocks (e.g., gravels). Using f = 100in the equation above, we find that L = 60 m (0.06km) for granites and L = 20 m (0.02 km) for loosesedimentary rocks. The resolution of the method isdetermined by the lower limit of size of the featurethat can be imaged. In seismic profiling, geophysi-cists have found that this limit is about one-fourthtimes the wavelength of the signal, or about 15 m

for granitic rocks and about 5 m for loose sedimen-tary rocks. This means that we can, in principal,map features on the order of 5 m thick in someareas (e.g., alluvial deposits), which approaches theresolution necessary for effective gem exploration.For example, it might be possible to map the loca-tions of large dikes (e.g., pegmatites) or the geome-try of important geologic layers, such as a gem-bear-ing gravel. However, it would still be difficult toidentify pockets or cavities smaller than 15 m indiameter.

We cannot change the speeds of the seismicwaves in the rocks (which largely depend on thekinds of minerals the rocks contain), but we canvary the frequency of the source of seismic waves toimprove the resolution. However, the Earth doesnot always cooperate, because it absorbs high-fre-quency signals very easily. (This is why the boom-ing low frequencies of a stereo sound system can beheard for a long distance, while the higher- frequen-cy signals are readily absorbed.) Additional techni-cal problems make it difficult to use the seismic-

Layered reflections

Seismic Reflection Profile

50 km0

b

a

0

5

10

15

Trav

eltim

eof

refle

cted

wave

s (s

econ

ds)

Ground surface Southeast

0

5

10

15

Appr

oxim

ate

dept

h(k

m)

Trav

eltim

eof

refle

cted

wave

s (s

econ

ds)

Ground surface Southeast0

15

30

45

Figure 8. (a) This seismic-reflection pro-file taken in northeastern BritishColumbia shows a decrease in thick-ness of the crust (depth to the top of themantle, or Moho). (b) Interpretation ofthe data reveals that the flat layersbetween 0 and about 2 seconds are rel-atively young sedimentary rocks,which rest on Precambrian crystalline(granite and metamorphic) rocks(orange—right) and Precambrian sedi-mentary basin rocks (yellow—left). TheMoho (Mohorovicic discontinuity), thetransition from the crust to the mantle,lies at about 45 km on the southeastand about 35 km on the northwest.Inasmuch as some gem deposits maydepend on variations in crustal thick-ness (e.g., some gem corundumdeposits appear to be located in areaswhere the crust is 40 or more km thick;Levinson and Cook, 1994), mappingthe thickness of the crust could provevaluable for reconnaissance. (For thisfigure, the approximate depth valueshave been calculated assuming anaverage seismic-wave velocity of 6km/sec, rather than 5 km/sec as in fig-ure 7, because of different characteris-tics of the rocks.) Figure is modifiedfrom Cook and Van der Velden (1993).


reflection technique for depths less than about 20m. Nevertheless, as technical advances improveseismic-reflection profiling for applications to geo-logic problems in the shallow subsurface, it maybecome an important tool in gem exploration.

Cost. Another important consideration is that seis-mic-reflection profiling can be relatively expensive;exact costs depend on the parameters (e.g., the spac-ing of the receiver points) that are used in the field.Although acquiring high-resolution shallow data isusually less expensive than acquiring deep reflec-tion data (because the area surveyed is smaller), sur-vey costs of several thousand dollars per day are notunusual, and a proper survey usually takes one ormore weeks. Thus, even if the technique is refinedto the point where meter-scale resolution is possiblein the uppermost few meters, the costs may be pro-hibitive for many gem-exploration applications.Fortunately, there is a less-costly alternative forsome situations—georadar.

Georadar. The Method. Also known as ground-pen-etrating radar, or GPR, georadar is finding a numberof geologic applications in the shallow subsurface(Box C). Radar has been used for about 15 years tomap thicknesses of ice sheets and glaciers. Morerecently, it has been applied in mapping under-ground pipes and building foundations, shallowarcheological sites (Imai et al., 1987), the geometryof landfill sites (Lanz et al., 1994), the geometry ofsediment deposits (Beres and Haeni, 1991; Smithand Jol, 1992; Beres et al., 1995), and fracture sys-tems in some crystalline (metamorphic andigneous) rocks (Piccolo, 1992; Grasmueck, 1996).Based in part on work by Grasmueck (1995, 1996),and in part on results from the author’s own experi-ments (figure 9), it is proposed here that, if carefullyperformed, radar may be applicable to gemstoneexploration in many instances.

When most of us think of radar, we think of airtraffic controllers or military personnel monitoringthe positions of airplanes or missiles. However, if aradar (radio wave) signal is directed into the ground,it can penetrate some distance and then reflect offrock layers beneath the surface (again, see Box C).The distance that a radar signal penetrates dependson the absorptive properties of the rocks throughwhich it is traveling. The property that most affectspenetration is probably electrical conductivity, ameasure of how easily an electrical signal travelsthrough a material. The more electrically conduc-

tive a material is, the more it absorbs a radar signal.Metals, as well as clay-rich and saltwater-rich mate-rials, are very electrically conductive; consequently,they prevent radar signals from being transmittedvery far because they absorb much of the energy.Like ice, fresh water, and air, granite is largely trans-parent to radar signals; thus, the signals travelthrough it readily. This characteristic is key to thesuccess (or failure) of the method. When a radar sig-nal impinges on a boundary between nonconduc-tive and conductive materials, some of the energyreflects back to the surface, where a receiver canmeasure it. Where there are large contrasts in elec-trical conductivity, large amounts of signal energymay be reflected.

As with a seismic section, the travel time toand from each reflector is measured and, if thewave velocity is known, the depth can be deter-mined. Reflected signals received by instrumentson the surface are recorded in the field and storedfor later computer enhancement. Many of the com-puter-enhancement techniques that have beendeveloped for seismic-reflection images are alsoapplicable to radar images (Fisher et al., 1992;Grasmueck, 1996). The result, when recording asingle profile or line, is a cross-section—much likea seismic-reflection section—that has a series ofcoherent signals, each of which corresponds to areflection from a rock interface (or an abruptchange in rock properties) at depth.

There are, however, two major differencesbetween seismic-reflection data and georadar data.First, the frequencies and velocities of radar signalsare much higher than those of seismic data: Radarsignals travel at the speed of light in air (300,000 kmper second) and about one-third of that speed ingranite (Davis and Annan, 1989). Second, radar sig-nals respond to the electrical properties of the mate-rial, whereas seismic signals respond to the elasticproperties of the material. (Elastic properties aremeasures of how easily a material deforms, such aswhen it vibrates.) Both of these characteristics ofradar (higher frequencies and electromagneticwaves), prevent radar signals from penetrating verydeeply into the rocks; in figure 9, for example, theradar section corresponds to only 60 m of rock, ascompared to the 45 km depth for the seismic sec-tion of figure 8b. Even more importantly, radar sig-nals in the shallow subsurface often have wave-lengths of about one meter or less, providing a reso-lution to 0.25 m (25 cm), which is potentially usefulfor gem exploration.


North

Cracks

Cavities

Quarry floor

Quarry wall

10 m

2D Profile (figures a, b)

Photograph (figure d)

0

0.0

30

60

Appr

oxim

ate

dept

h(m

eter

sbe

low

datu

m)

Georadar Profile Northeast

Ground surface

Cavities?

Cracks

(datum)

Ground surface

Cracks?

10 m0

X

Y

0.0

0.5

1.0Tr

avel

time

ofre

flect

edwa

ves

(103

nano

seco

nds)

Northeast

(datum)

Georadar Profile a

c

d

b

Ground surface

Cavity

Cavities

Cracks

Cracks

Figure 9. This georadar profile (a) was recorded by the author in a gneiss in southern Switzerland. Here, reflections arecaused by differences in electrical properties, and the depth of penetration is about 50–60 m (compare with figures 7 and8). Thus, the resolution is significantly more detailed than that for seismic profiling. Note that the datum (0.0 time line)is above the northeast-sloping ground surface. Interpretation of the data (b) reveals that prominent reflections at a depth(below the ground surface) of about 3 m (reflection X) and 6 m (reflection Y) on the right side of the section are probablycracks that are only a few centimeters thick (as interpreted on another dataset in the area by Grasmueck, 1995; see also[d] below). Red dots represent discontinuities, such as cavities or small faults, that give rise to the arcuate features (bluelines). The schematic diagram in (c) illustrates the relationship between the location of the georadar profile of (a) and (b)and the photograph of (d), an outcrop where a zone of mineralization (including cavities and small faults) is exposed.The outcrop is in a quarry wall immediately beneath the ground surface on the right side of the profile in (b).


Georadar in Three Dimensions. Most georadarrecordings, as with most seismic-reflection record-ings, are made along lines that result in a two-dimensional profile (horizontal distance in onedirection and vertical depth, or distance through afeature, in the other; see Box C and figure 9).However, pockets or cavities that are small (e.g.,one meter or less in diameter), but nevertheless pro-ductive, would be easy to miss unless profiles wererecorded very close together. In such a situation, itmight be desirable to acquire data in three dimen-sions (two horizontal and one vertical; figure 10), asdescribed, for example, by Beres et al. (1995).Although three-dimensional data require somewhatmore effort in both recording and data processing,the resulting information provides images of theinternal structure of a volume of rock, rather than asingle cross-sectional profile; features such as cavi-ties or mineralized zones may be much more appar-ent and less likely to be missed. Preliminary resultsfrom an experiment conducted by the author and acolleague (J. Patterson) within a pegmatite haveshown that it is possible to outline the three-dimen-sional geometry and size of a pocket.

Georadar is not very expensive relative to seis-mic profiling. A three-dimensional survey, withaccompanying profile lines like the one shown infigure 10, requires about two to four person days offield time and perhaps two to four weeks of com-puter work. A similar three-dimensional seismic-reflection survey would have required 10 times asmuch effort.

Possible Uses of Georadar Profiling in GemstoneExploration. Three characteristics of georadar makeit a potentially useful method for gemstone explo-ration:

1. Radar provides very high resolution (to as smallas 25 cm) images of the near-surface environ-ment (in some cases, in the first few meters totens of meters beneath the surface).

2. Georadar data are comparatively inexpensive toacquire and process.

3. Crystalline (granitic and metamorphic) rocks, inwhich many gem deposits typically occur, arerelatively transparent to radar signals. However,the cavities or pockets that contain the gems(e.g., fluid-filled fractures, hydrothermal veins,clay-filled pockets in pegmatites, etc.) are likelyto have electrical properties very different fromthe surrounding rocks. They could therefore pro-

duce very prominent reflected signals (figures 9aand b).

As noted earlier, most economic gemstonedeposits presently are found at the surface or withina few meters to tens of meters of the surface.Examples include various alluvial deposits (wherediamonds, rubies, sapphires, spinels, and manyother gems are found), silica-rich sedimentary orvolcanic rocks (such as some opal deposits inAustralia and Mexico, respectively), pegmatitedikes (where quartz, beryl [figure 11], tourmaline,and a variety of rare gems are found), and lampro-phyre (ultramafic) dikes (such as Yogo Gulch,Montana, where sapphires are found). Each of thesedeposit types may, if conditions are appropriate, beamenable to exploration using georadar or, in somecases, other geophysical tools.

For example, exploration for gemstones in sedi-mentary rocks, such as alluvial and some opaldeposits, requires mapping of the subsurface geome-try of the sedimentary layers. Georadar, and some-times high-resolution seismic-reflection data, can bevery effective tools for constructing such maps. Thedata illustrated in figure 10 were recorded to map thethree-dimensional geometry of sedimentary layersnear the surface (Beres et al., 1995). Detail is provid-ed to 15 m depth. If one of the layers in such a sur-vey (e.g. that labeled R2) were known to have highconcentrations of gems (diamonds, corundum,spinels, etc.), it would be easy to follow that layer inthe subsurface to predict its extent, its continuity,and thus the potential value of the deposit.

Exploration for primary deposits in granitic andmetamorphic rocks is equally promising. Radar isalready being used to locate high-quality (unfrac-tured) ornamental rocks in quarries (Piccolo, 1992);thus, there may be applications for locating zones ofhigh-quality jade or other massive gem materials.Highly fractured, lower-quality material would havemeasurably different properties from solid material.

Radar has been tested in some pegmatitedeposits in California (Patterson, 1996), and at theAlma, Colorado, rhodochrosite deposits (Brian Lees,pers. comm., 1996) with limited success. In bothareas, georadar has been useful for mapping subsur-face structures. At the Colorado rhodochrositedeposits, for example, the technique has been suc-cessfully used to map faults associated with thedeposit. However, it has been less successful in pro-ducing images of crystal-bearing pockets, probablybecause the pockets are very small (commonly lessthan 25 cm; Brian Lees, pers. comm., 1996).


Radio waves reflecting from contrasts in electrical con-ductivity can be detected by sensitive antennas. Thegeoradar method is similar to seismic-reflection profil-ing in that a source of energy (radio waves in this case)is sent through a transmitting antenna (antenna T infigure C-1 and C-2) into the ground, and the returnedsignal is detected and collected at a receiving antenna(R in figure C-1 and C-2). The received signal is storedon a computer for later display and data processing.The simplicity and portability of georadar systems (fig-

ure C-3) allow the instruments to be used horizontally,such as into or through walls (e.g., figure C-2).Anomalous objects such as cavities (e.g., in a peg-matite) would appear as traces with unusual shapes orarrival times (figure C-2).

Figure C-1. This schematic diagram illustrates thegeoradar method for profiling along the Earth’s sur-face. The transmitting antenna (T) sends a pulse intothe ground, where it is reflected back to a receivingantenna (R). The signal is sent through cable to acomputer, where it is stored on disk for later displayand processing.

Figure C-2. This schematic diagram illustrates thegeoradar method for recording waves that are trans-mitted through an object (a wall) to detect anomalouszones such as cavities. Here the transmitting andreceiving antennas are placed on opposite sides of thewall, and the anomalous zone appears in the data asan unusual waveform or arrival time.

BOX C: GEORADAR

Figure C-3. The equipment required for geo-radar is very portable, and the procedure isnot labor intensive. In this photo of a geo-radar field crew working in western Canada,the person in the foreground is setting theantennas, one for transmitting the radarpulse and the other for receiving the returnedsignal. The person in the background is car-rying the instruments that control the sig-nals and record them onto magnetic tape ordisk. The wire along the ground connects theantennas to the recording instruments.Photo courtesy of D. G. Smith.


Limitations of Georadar in Gem Exploration.Georadar is not a panacea for gemstone explorationat shallow depths. Some gem deposits in sedimen-tary rocks, such as some alluvials and some opaldeposits, may have large quantities of clay near thesurface. In such cases, the radar signal might notpenetrate more than a few centimeters (Davis andAnnan, 1989), and thus georadar would not beappropriate.

Pegmatite deposits in which the granitic rockshave weathered to produce a layer of surface clayare also problematic for radar, again because theclay absorbs much of the signal. This is commonlythe case in southern California (Jeffrey Patterson,pers. comm., 1996). However, exposed pegmatitesand deposits in relatively dry, sandy regions or inglacially scoured areas could be excellent environ-ments for radar detection of underground cavitiesand fractures (figure 9). Hence, georadar may be anappropriate tool for the first 20–40 m depth if thereis not a layer of clay or saline fluid near the surface.If the desired target is at a greater depth, or if thesurface is not conducive to radar-signal penetration,other methods such as reflection seismology ormagnetics (see table 1) may be appropriate.Therefore, surface and near-surface conditionsmust be investigated, usually with field observa-tions or other tests, before deciding which tech-nique(s) to use.

STRATEGIC APPLICATION OF GEOPHYSICSIN GEMSTONE EXPLORATIONApplication of geophysical methods to gemstoneexploration begins with establishing the nature ofthe target. In some cases, geophysical programs thataddress continental-scale measurements may beuseful to identify regions appropriate for moreintensive exploration. This approach is commonlyused in exploring for diamonds, but it may also beapplicable in exploring for gem corundum. Somegem corundum is found in alkali basalts that origi-nated as magma below 50 km depth and picked upcorundum from the lower crust (30–50 km) beneathcontinents as the magma traveled to the surface(Levinson and Cook, 1994). Thus, methods thatallow the thickness of the crust to be mapped fromplace to place could be valuable. The seismic sec-tion of figure 8 is an example of a change in thethickness of the crust from about 40 km on the eastto about 30 km on the west. Alkali basalts foundeast of this change would be more likely to containcorundum than those found west of it, if the theory

Figure 10. (top) These georadar data were acquiredin northern Switzerland to map sedimentary (allu-vial) structures in the near-surface (modified fromBeres et al., 1995). R1, R2, and R3 are positions ofreflected signals. Horizontal “slices” at 2.8 and 2.0m are shown above the cube to illustrate howreflections R1 and R2 can be followed in threedimensions. If R2 were a gem-bearing gravel, itcould be mapped easily over the region to deter-mine both its lateral and vertical extent. R3 isanother boundary at about 10 m depth.Knowledge of gem concentrations within thisstructure could thus be used to predict the totalvalue of the deposit. The colors represent reflectedwaves from different layer boundaries. (Bottom) Atwo-person radar team works on sedimentary stra-ta exposed in cross-section. The right-dipping lay-ers are about the same size as layers R2 above,thus illustrating the kind of detail that may bepossible with this georadar application. Photocourtesy of D. G. Smith.


proposed by Levinson and Cook (1994) is correct;that is, that alkali basalt–hosted gem corundumoriginates in thick crust.

For other types of colored stone deposits, geo-logic mapping and serendipity will necessarily con-tinue to play major roles. However, geophysics maybe useful in delineating the extent of a newly dis-covered deposit, as well as in identifying specifictargets (e.g., cavities) to investigate. In these cases, itis likely that high-resolution seismic reflection and,especially, georadar will soon become importantexploration tools.

Following is a strategy to evaluate the potentialusefulness of geophysical methods in a particularexploration venture:

1. Define the target (potential gem deposit, orregional structure) in terms of its:a. Size: If the target is the entire thickness and

lateral extent of the lithosphere, and the desireis to establish likely regions for detailed work(e.g., diamond exploration), then only low-res-olution techniques may be required.

b. Depth: The greater the depth of a target is, thelower the resolution that can be provided bygeophysical techniques.

c. Geologic environment: Alluvial deposits maybe imaged by techniques that provide strati-graphic information; primary deposits mayrequire more direct images, such as of cavities.

d. Physical properties: Knowledge of physicalproperties allows predictive models to be con-structed. Such properties include seismic-wavevelocity, density, magnetic characteristics, andelectrical conductivity.

2. Acquire existing “test” data. “Regional” data sets(e.g., gravity, magnetics, occasionally seismic)often are available from government agencies.Such data can provide a considerable amount ofbackground information, such as the thickness ofthe crust, the lithosphere, and so on.

3. Evaluate the costs versus the potential return.

Initial field testing is advised, particularly withhigh-resolution techniques such as radar. For exam-ple, it may be worthwhile to record a single radarprofile to determine if near-surface clay layers are aproblem. Such a test could also indicate whether alarger-scale survey, or even a three-dimensional sur-vey, might be appropriate.

SUMMARYToday, geophysical methods are sometimes used indiamond exploration, but they are not commonlyapplied for either exploration or exploitation of col-ored gemstone deposits. Part of the reason for this ishistorical: People searching for gems have not previ-ously used these techniques and often are not famil-iar with them. Much of the reason, however, istechnical: Geophysical methods simply have nothad the resolution to be effective in either finding ordelineating gem deposits, which are typically small.In addition, many geophysical techniques areexpensive and require sophisticated equipment andtechnical expertise.

As technology progresses, however, the resolu-tion of these techniques is being enhanced. Seismic-reflection data, for example, may now resolve fea-tures as small as a few meters; 10 years ago, it wasdifficult to resolve features of even 20 m. Georadar,

Figure 11. Georadar has great potential in pegmatiteexploration, with the potential to locate gem pocketssuch as the one in which this superb aquamarinecrystal (7.5 cm long) was found. Courtesy of MichaelM. Scott; photo © Harold & Erica Van Pelt.


too, was in the early stages of development as anapplication to subsurface studies a decade ago. Sincethen, the methods for acquiring and processing suchdata have improved measurably. It is now possibleto analyze georadar data in the field with smallrecording systems and computers; more than fiveyears ago, the equipment was much more cumber-some, if it existed at all.

In the future, as resolution continues toimprove, and as data collection and analysis becomeeven easier, it is likely that these and other geophys-ical techniques will become valuable tools for gem-stone exploration. Even today, georadar holds poten-tial for locating gem-bearing structures in peg-matites, alluvial beds, and alkali basalts, amongother types of occurrences.

REFERENCESAtkinson W.J. (1989) Diamond exploration philosophy, practice,

and promises: A review. In J. Ross, Ed., Kimberlite andRelated Rocks, Vol. 2, Proceedings of the FourthInternational Kimberlite Conference, Geologic Society ofAustralia Special Publication 14, Blackwell ScientificPublications, Oxford, England, pp. 1075–1107.

Beres M., Haeni F. (1991) Application of ground-penetrating radarmethods in hydrogeologic studies. Ground Water Research,Vol. 29, pp. 375–386.

Beres M., Green A., Huggenberger P., Horstmeyer H. (1995)Mapping the architecture of glaciofluvial sediments withthree-dimensional georadar. Geology, Vol. 23, No. 12, pp.1087–1090.

Burley A.J., Greenwood P.G. (1972) Geophysical surveys overkimberlite pipes in Lesotho. Institute of Geologic Sciences ofLondon, Geophysical Division, Reference 540 1069/72 (citedin Atkinson, 1989).

Cook F.A. (1995) Lithospheric processes and products in thesouthern Canadian Cordillera: A Lithoprobe perspective.Canadian Journal of Earth Sciences, Vol. 32, No. 10, pp.1803–1824.

Cook F.A., Brown L.D., Oliver J.E. (1980) The southern Appal-achians and the growth of continents. Scientific American, Vol.234, No. 4, pp. 126–139.

Cook F.A., Van der Velden A.J. (1993) Proterozoic crustal transi-tion beneath the Western Canada sedimentary basin.Geology, Vol. 21, No. 9, pp. 785–788.

Davis J.L., Annan A.P. (1989) Ground-penetrating radar for high-resolution mapping of soil and rock stratigraphy. GeophysicalProspecting, Vol. 37, No. 5, pp. 531–551.

Fisher E., McMechan G.A., Annan A.P. (1992) Acquisition andprocessing of wide-aperture ground-penetrating radar data.Geophysics, Vol. 57, No. 3, pp. 495–504.

Grand S.P. (1987) Tomographic inversion for shear velocitybeneath the North American plate. Journal of GeophysicalResearch, Vol. 92, No. B13, pp. 14,065–14,090.

Grasmueck M. (1995). Development of a georadar system forthree-dimensional imaging of the subsurface and its applica-tion to studies of crystalline rock bodies. Ph.D. dissertation,Swiss Federal Institute of Technology, Zurich, 104 pp.

Grasmueck M. (1996). 3D ground-penetrating radar applied tofracture imaging in gneiss. Geophysics, Vol. 61, pp.1050–1064.

Gurney J.J., Levinson A.A., Smith H.S. (1991) Marine mining ofdiamonds off the west coast of southern Africa. Gems &Gemology, Vol. 27, No. 4, pp. 206–219.

Hoover D.B., Campbell D.L. (1994) Geophysical model of dia-mond pipes. In W. D. Heran, Ed., Codicil to the GeophysicalExpression of Selected Mineral Deposit Models. UnitedStates Geologic Survey Open File Report 94–174, pp. 32–36.

Imai T., Sakayama T., Kanemori T. (1987) Use of ground-probing

radar and resistivity surveys for archaeological investigations.Geophysics, Vol. 52, No. 1, pp. 137–150.

Janse A.J.A. (1994) Is Clifford’s rule still valid? Affirmative exam-ples from around the world. In H. O. A. Meyer and O. H.Leonardos, Eds., Proceedings of the Fifth InternationalKimberlite Conference, Vol. 2, Diamonds: Characterization,Genesis and Exploration, Companhia de Pesquisa deRecursos Minerais (CPRM), Rio de Janeiro, Brazil, pp.215–235.

Lanz E., Jemmi L., Mueller R., Green A., Pugin A. (1994)Integrated studies of Swiss waste disposal sites from georadarand other geophysical surveys. In J. D. Redman, A. P. Annan,J. P. Greenhouse, T. Klym, and J. R. Rossiter, Eds.,Proceedings of the Fifth International Conference on GroundPenetrating Radar, Waterloo Center for GroundwaterResearch, Kichener, Canada, pp. 1135–1150.

Levinson A.A., Gurney J.J., Kirkley M.B. (1992) Diamond sourcesand production: Past, present, and future. Gems &Gemology, Vol. 28, No. 4, pp. 234–254.

Levinson A.A., Cook F.A. (1994) Gem corundum in alkali basalt:Origin and occurrence. Gems & Gemology, Vol. 30, No. 4,pp. 253–262.

Lock N.P. (1985) Kimberlite exploration in the Kalahari region ofsouthern Botswana with emphasis on the Jwaneng kimberliteprovince. In Prospecting in Areas of Desert Terrain, Instituteof Mining and Metallurgy, London, pp. 183–190 (cited inAtkinson, 1989).

Patterson J.E. (1996) Modeling miarolitic cavities in a pegmatitedike using subsurface interface radar, Little Three mine,Ramona, California. In The Spirit of Inquiry, undergraduateresearch sponsored by the Honors Center, University ofArizona, p. 17.

Piccolo M. (1992) GPR applications for the definition of uncon-formities in a Carrara marble quarry (Massa Carrara—Italy).In P. Hanninen and S. Autio, Eds., 4th InternationalConference on Ground Penetrating Radar, Geologic Surveyof Finland, Special Paper 6, pp. 223–228.

Senior B.R., McColl D.H., Long B.E., Whitely R.J. (1977) Thegeology and magnetic characteristics of precious opaldeposits, southwest Queensland. BMR Journal of AustralianGeology and Geophysics, Vol. 2, pp. 241–251.

Smith D.G., Jol H.M. (1992) Ground-penetrating radar investiga-tion of a Lake Bonneville delta, Provo level, Brigham City,Utah. Geology, Vol. 20, No. 12, pp. 1083–1086.

Smith R.S., Annan A.P., Lemieux J., Pederson R.N. (1996)Application of modified GEOTEM system to reconnaissanceexploration for kimberlites in the Point Lake area, NWT,Canada. Geophysics, Vol. 61, No. 1, pp. 82–92.

Telford W.M., Geldart L.P., Sheriff R.E., Keys D.A. (1976)Applied Geophysics. Cambridge University Press,Cambridge, 860 pp.

24 Nepal Rubies GEMS & GEMOLOGY Spring 1997

RUBIES AND FANCY-COLORSAPPHIRES FROM NEPAL

By Christopher P. Smith, Edward J. Gublein, Allen M. Bassett,and Mache N. Manandhar

GGem-quality rubies and fancy-color sap-phires have been recovered from dolomitemarble lenses located high in theHimalayan mountains of east-central Nepal(Ganesh Himal). First discovered in theearly 1980s, these deposits have had onlylimited and sporadic production because oftheir isolated locations, high altitudes, andharsh seasonal weather conditions. The pre-cise locations and geology of the two best-known deposits are described. In addition,27 polished samples and 19 rough crystalsand mineral specimens were thoroughlyinvestigated to document their gemologicalcharacteristics and crystal morphology. Themost distinctive internal features are theircolor-zoning characteristics, various cloud-like and other patterns, as well as a widevariety of mineral inclusions, some ofwhich have not been seen in rubies fromother sources.

ABOUT THE AUTHOR

Mr. Smith is a research gemologist residing inLucerne, Switzerland. Dr. Gübelin is a gemologistand co-author of the Photoatlas of Inclusions inGemstones and various other books and publica-tions. Dr. Bassett is retired professor of geology,California State University, San Diego, and technicaldirector of Himalayan Gems Nepal; he resides muchof the year in Katmandu, Nepal. Mr. Manandhar isgeneral director of Himalayan Gems Nepal,Katmandu, Nepal.Gems & Gemology, Vol. 33, No. 1, pp. 24–41© 1997 Gemological Institute of America

N epal is a small mountainous country situatedalong the Himalayan mountain chain in southernAsia.Landlocked by Tibet (China) to the north

and India to the east, south, and west, Nepal's surfacearea covers only 140,000 km2; it has approximately 20million inhabitants. Southern Nepal is the birthplace,more than 2,500 years ago, of Siddartha Gautama, thefounder of Buddhism. For many centuries, Nepal wasclosed off from the rest of the world. Only since itemerged from a feudal state in 1950, have the mineralsand gemstones of Nepal begun to enter the world econo-my. Although Nepal has a reigning monarch, a multi-party democracy was established in 1990.

Most widely recognized for having the highest eleva-tion of any country in the world, Nepal is the home ofMount Everest (8,848 m/29,028 feet above sea level). Inall, seven of the eight highest peaks in the world can befound in this country's rugged terrain.

Although Nepal has limited industrial developmentand relatively meager mineral wealth, it is blessed witha broad range of gem minerals that includes tourmaline,beryl, garnet, quartz, spinel, danburite, hambergite,kyanite, apatite, sodalite, zircon, sphalerite, epidote,diopside, iolite, feldspar, pyrite, hematite, andalusite,lepidolite, and corundum (Bassett, 1979; 1984; 1985aand b; 1987). Gems have been recovered in Nepal since1934, when tourmaline and aquamarine were first dis-covered there (Bassett, 1979). However, not until theearly 1980s did rubies and sapphires begin to appear (figure 1).

As the story was told to authors AMB and MNM,goat-herders of the Tamang ethnic group first spottedred crystals in 1981, high in the Ganesh Himal massif ofthe Dhading District in east-central Nepal. These stoneswere brought to the capital city (Katmandu) to sell forshop displays, where they were identified as rubies. Thefirst official report of corundum in Nepal was made byBaba (1982), who erroneously described

Nepal Rubies GEMS & GEMOLOGY Spring 1997 25

Figure 1. Since the early1980s, rubies and fancy-

color sapphires have beenrecovered in the

Himalaya Mountains ofeast-central Nepal. TheNepalese rubies shownhere range from 0.50 to

2.50 ct. Jewelry courtesyof the Gold Rush,

Northridge, California;photo by Shane F.

McClure.

ruby/corundum coming from the TaplejungDistrict, in eastern Nepal. Soon after, however, itwas determined that the source was actually in theGanesh Himal (Dhading District), and many stakedclaims there, hoping to make their fortunes. For themost part, however, these earlier miners lacked anyunderstanding of gemstone recovery methods or ofthe geology and challenging topography of themountainous region, so these early efforts met withlimited success. It was not until 1984 that one ofthe authors (AMB) made the first geologic study ofthe corundum-producing region at Ganesh Himal(Bassett, 1984). After concluding this investigation,AMB joined with MNM to found Himalayan GemsNepal, which acquired the leases to the two mostcommercially viable claims, known as Chumar andRuyil. Himalayan Gems Nepal began officially rec-ognized mining activities during 1985. They alsoset up a cutting factory to bring these rubies andfancy-color sapphires to the gem and jewelry mar-kets. Several other parties also mined unofficiallyin the surrounding region.

None of these mining ventures lasted long, how-ever, because of the isolated locations, harsh weatherconditions, and difficulties encountered in efforts todevelop larger-scale mining. All of the claims wereeventually abandoned by their owners. Today, localTamang farmers continue rudimentary mining; theybring their production to Katmandu for sale and dis-tribution. As a result, throughout the mid- to late1980s and early 1990s, the production of gem-qualityrubies and fancy-color sapphires was sporadic.Although Himalayan Gems Nepal also ceased activemining after about two years, they continue to buygems from the local farmers and cut them in theirfactory.

Since mid-1996, two other deposits—Shelgharand Shongla—have produced corundum that is com-parable in quantity and quality to Chumar andRuyil. During studies in 1985, AMB determined thatthe geology of these two deposits is identical to thatof Chumar. Besides these four producing deposits,two others (Pola and Sublay) are known to havepotentially significant amounts of corundum,


Figure 2. The ruby-bear-ing occurrences may be

found in the northernDhading District of cen-tral Nepal. The Chumar

and Ruyil deposits arethe two largest deposits

and the ends of a geolog-ic chain that includes

more than a dozensmaller deposits.

but they have not yet been mined. However,Chumar and Ruyil remain the most recognizeddeposits of commercial-grade and finer qualityrubies and pink sapphires in Nepal.One of the authors (AMB) was told that blue sap-phires are currently being recovered from theTaplejung District of eastern Nepal. The authorshave examined a few non-gem-quality specimens,but there is no additional information on theamount or quality of the material being unearthed.The claim holder reports, however, that the sapphirecrystals occur in outcrops on the slopes of the moun-tain known as Sapphire Hill, approximately 1.5 kmnorthwest of the small village called Khupatal, northof the Tamur River in the Taplejung District.Contrary to Baba's 1982 report, our investigationsyielded no evidence that rubies have ever been foundin this district.

The first in-depth study of ruby from Nepal wasprovided by Harding and Scarratt (1986). Soon after,researchers in Germany also described fancy-color(pink and violet) sapphires from unspecified depositsin Nepal (Kiefert and Schmetzer, 1986, 1987). Adetailed investigation of one very fine gem-qualityruby from Nepal was reported in 1988 (Bank et al.,1988). Robinson et al. (1992) were the first to men-

tion a specific deposit, named Chumar, for thecorundum occurrences. Overviews on gems fromNepal also described corundum deposits (e.g.,Niedermayr et al., 1993). The most recent observa-tions on the internal features of this material wereprovided by Henn and Milisenda (1994). The presentstudy supports the findings of these previousresearchers, and adds both new characteristics andspecific source information to these earlier reports.

LLOOCCAATTIIOONN AANNDD AACCCCEESSSSThe corundum deposits occur on the southwestflank of the Ganesh Himal (a high mountain massifconsisting of several major peaks clustered together),68 km north-northwest of the capital city ofKatmandu, and 40 km west of Trisuli Bazar. Anoverview of the region reveals rice paddies in thelowlands, with primarily corn and barley crops inthe hills. Above approximately 1,830 m (6,000 feet),one finds forests of rhododendron trees, then grass,and, eventually, bare rock. The landscape at thedeposits is that of extremely steep, cliff-like domesof dolomitic marble about 150 m (500 feet) highlying in lenticular bodies that are surrounded bysteep slopes of schist. There are deeply entrenched,rocky gorges, choked with dense vegetation that


reaches almost to the corundum deposits, whichcrop out in the dangerously craggy cliffs.

Travel to the deposits from Katmandu usuallytakes about six days. The trek begins with a cir-cuitous bus ride from Katmandu to Trisuli Bazar,and then five days of rather strenuous hiking up theAnkhu River Valley and its tributaries. Travel to thedeposits is hampered not only by the high eleva-tions, remote locations, and rugged terrain, but alsoby the severe seasonal weather conditions. Wintersnow usually covers the higher trails from mid-November to the end of April, and monsoons buffetthe countryside from mid-June through September.Therefore, the best times to reach the deposits areusually from early May to mid-June and from earlyOctober to mid-November. A knowledgeable guideis essential.

Both the Chumar and Ruyil deposits lie withinwhat was formerly known as the Laba Panchayat ofthe northern Dhading District, Bagmati Zone, incentral Nepal (figure 2). The Chumar deposit lies at28° 13' 20" N, 84° 58' 52" E, at an altitude of 3,800 m(12,500 feet); it is approximately 1.2 km south of theMandra Danda mountain peak (elevation 4,358m/14,300 feet), near the Tamang village of Burang.The Ruyil deposit lies about 6 km northeast ofChumar, at 28° 15' 05" N, 85° 02' 07" E, and an ele-vation of about 4,200 m (13,800 feet).

GGEEOOLLOOGGYY AANNDD OOCCCCUURRRREENNCCEEThe geology of Nepal is dominated by theHimalayan Mountains, which formed as a result ofthe collision between the northward-drifting Indian

plate and the Asian plate starting about 30 to 50 mil-lion years ago. The initial line of collision (called theIndus Suture Zone) is now in Tibet. The Indian platecontinued in a northward drift, pushing beneath theAsian plate, whereupon the Indian plate broke intonorthern and southern segments, as the southernsegment was forced beneath the northern segment.This major structural breakage is known as the MainCentral Thrust (MCT) and extends east-west, thefull length of Nepal (figure 3). At least twice subse-quently, the continuing northward drift of the Indianplate caused a major breakage along progressivelyyounger and more southerly thrusts, known as theMahabharat Thrust and the Main Boundary Thrust(MBT). Each continental slice was pushed beneaththe higher and more northerly slices of the Indianplate to form the imbricate (shingle-like) stack thatis responsible for the elevation of the HimalayanMountains. Each of these various continental sliceshas a different lithology and grade of metamorphism;yet they are all roughly the same geologic age(Precambrian and lower Paleozoic), with the highestgrade above the MCT and lower grades below theMCT (there is no metamorphism below the MBT).This sequence is inverted from what is normallyencountered, because of the subduction thrusts.

The Ganesh Himal corundum occurs near the topof the lower-grade metasediments of the NawakotSeries (again, see figure 3), below the MCT. Thegems are found in what was a limestone formation(the Malekhu Limestone, lower Paleozoic) that has

Figure 3. This simplifiedgeotectonic map ofNepal indicates themain geologic structures(tectonic zones) and themain rock-forming seriesand groups in Nepal. Thestudy area is near thetop of the Nawakotseries, below the MCT.Source: MineralResources Map of Nepal,United Nations publica-tion, 1993.

been converted to a dolomite marble because of itsproximity to the tectonic forces of the MCT. Thisonce-continuous dolomite marble layer has beentorn into a dozen isolated bodies, about 1 km fromone another. These dolomite bodies occur asthickened beds and isolated pods 60–150 m thickand up to a kilometer long. The bodies are separat-ed by black schist with quartzite interlayers.

Within these isolated dolomite pods are miner-al seams that were originally aluminous clayinterlayers in the limestone,- these have been con-verted, again due to the intense shearing pressures

Figure 4. The remote Chumar ruby deposits arevisible along the steep slopes of the Himalayas,where the white marble outcrops overlook thevalley below. Photo by Ted Daeschler.

of the MCT, into a suite of metamorphic mineralsthat includes corundum. Each seam has a slightlydifferent suite of minerals, depending on variationsin the original composition of the clay interlayers.In one notable variation (the Ruyil deposit), abun-dant graphite has formed, presumably due to a localabundance of marine organic material in the origi-nal interlayer. The dolomite marble host rock isconsistent throughout the dozen pods.

At the western end of these dozen dolomitebodies, which extend for 15 km, they are more dis-tinctively pod-like lenses,- in the central section,the dolomite bodies are more like a thickenedstratigraphic bed with parallel walls less than 100m thick. The continuation eastward has not beenexplored in the very deep, densely forested chasms,but corundum is known to occur at the Sublayclaim at the far eastern end. Of the dozen dolomitebodies, the westernmost four—Pola, Chumar,Shongla, and Shelghar—have produced corundum,only Ruyil in the central part, and only Sublay atthe eastern end, are corundum bearing. Of these sixcorundum-bearing bodies, Chumar and Ruyil havebeen the most productive, with Shongla andShelghar only recently developed and Pola andSublay not yet consistently worked. This studyfocuses primarily on the Chumar and Ruyildeposits.

CChhuummaarr DDeeppoossiitt.. First studied in 1984 by AMB, theChumar deposit extends east-west for about 550 m,with a central thickness of approximately 150 m(figure 4). At the west end, the dolomitic body isbounded by a sharp fault trending N60°W, dippingsteeply to the northeast. The east end of the bodytapers down to a tail, with no fault line at theboundary. Distinct bedding in the dolomite body,which now consists of large eroded domes, strikeseast-west throughout all but the extremities, whereit is bent to the northwest, creating a sigmoidshape. The seams, which are about 8 m apart anddip 32°, vary widely in thickness but average about20 cm; they are sometimes folded on a small scale.

RRuuyyiill DDeeppoossiitt.. The first studies of the Ruyildeposit were conducted in 1985 by AMB. Thisdeposit extends east-west for approximately 128 m,with a general thickness of 60 m. The western por-tion of the dolomite body appears to end abruptlyin a curved outline of shattered dolomite withoutdistinct bedding; yet there is no discernible faultline. No corundum has been found in the continu-ous outcrops of bedded dolomite on the easternside of this body.


Figure 5. The Nepalese corundum deposits have aninteresting assemblage of mineral occurrences,including tabular stacks of green fuchsite, orange-brown phlogopite, and colorless to white margaritemicas, as well as blades of blue kyanite and dipyra-midal forms of corundum in shades of red or pink(some containing large blue color zones). The hostrock is white calcite and dolomite. Photo by Dr. H.A. Hänni.

AASSSSOOCCIIAATTEEDD MMIINNEERRAALLSS OOFF TTHHEE CCOORRUUNNDDUUMMDDEEPPOOSSIITTSS

Six specimens of corundum in host rock from theChumar mine (ranging from 2.3 × 1.3 × 1.8 cm to 7.3 ×3.5 × 4.6 cm) were examined. We identified the seammaterial as primarily calcite and dolomite. Associatedminerals include blue tabular blades of kyanite, smallbrownish orange euhedral crystals of rutile, differentmicas (stacked sheets of transparent green fuchsite,orange-brown phlogopite, and whitish margarite), andcolorless apatite and scapolite, in addition to the vio-letish blue-to-red crystals of corundum (figure5). One of the authors also identified brucite, tremolite,talc, pyrite, and sphalerite during his geologic studies atthe Chumar mine (Bassett, 1985a). This list expands onthe ones published previously; other researchers haveadditionally identified zoisite, epidote, muscovite, andanor-thite in mineral samples containing corundumfrom Nepal (Harding and Scarratt, 1986). Mineral speci-mens from this locality are often very colorful, withgreen fuchsite, orange phlogopite, blue kyanite, and redcorundum, all occurring together in a white calcitematrix.

Rubies from the Ruyil deposit, unlike those fromChumar and the other deposits in Ganesh Himal, areembedded in abundant graphite with a less completesuite of associated minerals. In all other respects,though, they appear to be similar to the rubies fromelsewhere in the district. Wel-formed gem-qualitycorundum crystals are more common from Ruyil thanfrom Chumar.

MMIINNIINNGG MMEETTHHOODDSSThe actual mining for gemstones is usually performedby local Tamang villagers using crowbars, picks, andshovels (figure 6), with occasional blasting. Blasting fol-lowed by hammering is used more commonly at theChumar deposit than at Ruyil, and incline tunnels havebeen started at the latter deposit (Chakrabarti, 1994). Allof the ruby and fancy-color sapphire deposits are prima-ry (i.e., the stones are found in the host rock); no sec-ondary deposits have been located to date. The minersusually excavate where the marble outcrops can be seenat the surface; the corundum itself is the prime indica-tor mineral for new deposits.

PPRROODDUUCCTTIIOONN,, QQUUAALLIITTYY,, AANNDD SSIIZZEESSIt is very difficult to estimate how many tens of thou-sands of carats of corundum have been produced fromthese deposits over the past 15 years or so. Except for

Figure 6. At the Chumar deposit, local farmers usebasic hand tools to remove the corundum crystalsfrom the host rock. Photo by Allen M. Bassett.

the brief period that Himalayan Gems Nepal offi-cially controlled the mining efforts, most mininghas been unofficial and uncontrolled, so no

Nepal Rubies


of testing 27 polished samples (0.21 to 9.47 ct), noneof which had been heat treated. Seven of these werefrom the Chumar deposit and 11 were from Ruyil;the exact deposits for the remaining nine are notknown. Also included in this study were 13 crystalsand crystal fragments, ranging from 0.35 to 106.42ct, and the six Chumar mineral specimensdescribed earlier. This collection of gemstones andcrystals represents the full range of colors—toneand saturation, as well as hue—observed inNepalese corundums to date.

We used standard gemological instrumentationto record the refractive indices, birefringence, opticcharacter, pleochroism, optical absorption spectra(desk-model spectroscope), and reaction to ultravio-let radiation (365 nm long-wave, 254 nm shortwave)on all fashioned samples. Specific gravity was deter-mined by hydrostatic weighing with an electronicbalance equipped with the appropriate attachments.A binocular microscope, incorporating fiber-opticand other lighting techniques, was used to docu-ment the internal features of all fashioned samplesand single crystals. The results are given in table 1.We used a Perkin Elmer Lambda 9 spectrometer,with a beam condenser and polarizing filters, forpolarized spectroscopy in the UV-visible

Figure 7. Many well-formed crystals have beenrecovered from the corundum deposits in east-cen-tral Nepal. While most crystals are small or medi-um (usually 5 ct or less) in size, some-such as this106.42 ct crystal-exceed 100 ct. Photo by Shane F.McClure.

complete records are available. On the basis of ourexperience in Nepal and the material we have seen inKatmandu, we estimate that current yearly produc-tion is less than 1,000 kg. Most of the corundumcrystals are small to medium in size (typically 5 ct orless). However, crystals over 100 ct are occasionallyrecovered (figure 7). The vast majority of the produc-tion is in the low-to-medium "commercial quality"range, which is best suited for polishing en cabochon.Only a small portion of the material is of fine to veryfine "gem quality," which is appropriate for faceting(figure 8).

MMAATTEERRIIAALLSS AANNDD MMEETTHHOODDSSAll of the samples in this study were collected bytwo of the authors (MNM and AMB) in Nepal, fromindependent miners and the mines managed directlyby Himalayan Gems Nepal. We originally selected atotal of 29 polished gemstones (22faceted and sevencut en cabochon) that were fashioned from rough atthe cutting factory of Himalayan Gems Nepal. Twofaceted samples were subsequently removed from thestudy. We identified one as a flame-fusion syntheticruby, and the second yielded properties and charac-teristics that were not totally consistent with theother samples. Thus, this article presents the results

Figure 8. The Nepal deposits have produced somevery high quality rubies, such as the fine 1.40 ctfaceted ruby and 3.08 ct ruby crystal shown here.The crystal has a habit of c, r, n and z crystalfaces. Also note the narrow, blade-like formwhich is characteristic of much of the rough fromthese unique deposits. Photo by Shane F.McClure.


through near-infrared region (between 280 and 880nm) on 17 fashioned samples. For the region between400 and 6000 wavenumbers (cm-1), we used a Pye-Unicam Fourier-transform infrared (FTIR) 9624 spec-trometer with a diffuse reflectance unit for samplemeasurement. We performed a total of 49 analyseson different regions, color zones, and orientations of27 samples. Energy-dispersive X-ray fluorescence(EDXRF) chemical analyses were performed on 25samples (and on different color zones on some sam-ples, for a total of 34 analyses) on a SpectraceTN5000 system, using a program specially developedby Prof. W. B. Stern for the semi-quantitative analy-sis of corundum. Prof. Stern's program uses chemi-cally pure element standards and three spectra focus-ing on the light, medium, and heavy elements, sothat the results can be interpreted to three decimalplaces (i.e., 0.001). He also used a beam condenser tomeasure small areas or zones of an individual stone.To analyze the internal growth structures of the pol-

ished stones, we used a horizontal microscope, a spe-cially designed stone holder, and a mini-goniometerattached to one of the oculars on the microscope,employing the methods described by Schmetzer(1986a and b), Kiefert and Schmetzer (1991), andSmith (1996). We analyzed more than 45 inclusionsusing a scanning electron microscope with an ener-gy-dispersive X-ray spectrometer (SEM-EDS), X-raydiffraction analysis, and a Raman micro-spectrome-ter. Some of the inclusions were identified by a sin-gle method and others by a combination of tech-niques (see table 1).

GGEEMMOOLLOOGGIICCAALL CCHHAARRAACCTTEERRIISSTTIICCSS OOFF TTHHEERRUUBBIIEESS AANNDD FFAANNCCYY--CCOOLLOORR SSAAPPPPHHIIRREESS FFRROOMMNNEEPPAALLThe data revealed no significant differences from onesample to the next-including those samples identi-fied as originating from Chumar and Ruyil—

TABLE 1. Gemological characteristics of the rubies and fancy-color sapphires from Nepal.

Propererties No. samples Observations

Color 27 polished Ranging from red to pink to purplish pink, often with visible violetish blue color banding

Clarity 27 polished Very clean to heavily included, most in the range slightly to heavily included13 crystals

Refractive 20 faceted ne = 1.760-1.762

Index no = 1.769-1.770

7 cabochon n = 1.76-1.77 (spot method)

Birefringence 20 0.008-0.009

Optic 20 uniaxial negative

character

Specific gravity 27 polished 3.94-4.01

13 crystals

Pleochroism 27 Moderate to strong dichroism

Red zones: yellowish orange to orangy red parallel to the c-axis, and reddish purple to purple-red, or purplish pink to purple-pink, perpendicular to the c-axis Violetish blue zones: greenish blue to blue parallel to the c-axis, and violetish blue to violet-blue perpendicular to the c-axis.

UV 27 Red to pink zones. Long-wave (365 nm): Moderate to very strong red Short-wave (254 nm):Iuminescence Faint to moderate red

Violetish blue zones: Generally inert

Optical 27 General absorption up to approximately 450 nm, 468 (sharp, narrow), 475 nm (sharp, absorption weak to moderate), 476 nm (sharp, narrow), 525-585 nm (broad band), 659 nm (faint, narspectrum (nm) row), 668 nm (faint, narrow), 675 nm (very faint, narrow; when present), 692 nm (sharp,

narrow), 694 nm (sharp, narrow)

Internal features 27 Short rutile needles, various cloud patterns, stringer formations, strong color zoning, weak to prominent growth structures, laminate twinning and parting, "fingerprints," negative crystals and crystalline inclusions of (with elements identified and mode of identification in parentheses):

margarite mica (Al, Si, Ca; SEM-EDS and XRD)a apatite (Ca, P; SEM-EDS, XRD and Raman)a rutile (Ti; SEM-EDS) diaspore (XRD, Raman, and FTIR)a dolomite (Raman)

phlogopite mica (K, Mg, Al, Si, Fe; SEM-EDS and XRD)a calcite (Ca, C; SEM-EDS and FTIR) uvite tourmaline (XRD) anorthite feldspar (Raman) boehmite (FTIR)

aOften containing black inclusions that presumably are graphite.


VViissuuaall AAppppeeaarraannccee.. Face up, most of the polishedsamples ranged from a "pure" red or pink to a pur-plish pink. Both the rough and polished samples com-monly had strong, eye-visible color zoning (seeGrowth Characteristics below). A few could even bebetter described as bicolored (red and very dark vio-letish blue) corundum; some had additional colorlessor pink zones (figure 10). The diaphaneity of the crys-tals and polished samples ranged from transparent totranslucent, depending on the nature and number ofinclusions present, as well as on the color saturationand tone of the violetish blue zones.

RReeffrraaccttiivvee IInnddiicceess,, BBiirreeffrriinnggeennccee,, OOppttiicc CChhaarraacctteerr,,aanndd SSppeecciiffiicc GGrraavviittyy.. These standard gemologicalproperties were found to be consistent with corun-dum in general (see, e.g., Webster, 1983; Liddicoat,1989; Hughes, 1990; Hurlbut and Kammerling, 1991)and, more specifically, with the rubies and fancy-color sapphires from Nepal described by pastresearchers (Harding and Scarratt, 1986; Kiefert andSchmetzer, 1986, 1987; Henn and Milisenda, 1994;see table 1).

RReeaaccttiioonn ttoo UUllttrraavviioolleett RRaaddiiaattiioonn.. The various colorzones of the sample corundums had different reac-tions to UV radiation. The red to pink zones fluo-resced red to both long-wave (moderate to very strongintensity) and short-wave (faint to moderate intensi-ty) UV. The dark violetish blue zones were generallyinert to both long- and short-wave UV radiation.

PPlleeoocchhrrooiissmm.. All samples exhibited moderate tostrong dichroism when viewed perpendicular to the c-axis with a dichroscope. Within the red-to-pink colorzones, we observed yellowish orange to orangy red orpink parallel to the c-axis and reddish purple to pur-ple-red or purplish pink to purple-pink (i.e., for rubiesor pink sapphires, respectively) perpendicular to the c-axis. In the dark violetish blue color zones, we notedgreenish blue to blue parallel to the c-axis and vio-letish blue to violet-blue perpendicular to the c-axis.

GGrroowwtthh CChhaarraacctteerriissttiiccss.. Internal Growth Structures.Weak-to-prominent growth structures were seen inessentially all the polished gemstones examined.Most common were straight and angular sequences ofthe dipyramid z planes (figure 11). Less common weresequences of the dipyramids n or ??, as well as thebasal pinacoid c and the positive rhombohedron r.

in terms of standard gemological properties, crys-tal morphology, inclusion patterns, UV-Vis-NIRspectra, or chemical composition. Therefore, thecomplete data collected on all of the samples—those with specific deposit designations, as well asthose without—will be presented as a singlegroup.CCrryyssttaall MMoorrpphhoollooggyy.. The rough crystals were pre-dominantly euhedral, with little or no evidence ofchemical dissolution on their surfaces. Two maincrystal forms dominate the morphology of thecorundum found in Nepal (figure 9). The first aredipyramidal crystal habits composed of larger,dominant hexagonal dipyramid z (2241) faces andsmaller, subordinate basal pinacoid c (0001) andpositive rhombohedron r (1011) faces. The secondis a modification of this basic habit, where there isan addition of subordinate hexagonal dipyramid n(2243) faces. Rarely, crystal forms consisting ofdominant hexagonal dipyramid ?? (14 14 28 3)faces, with subordinate c (0001), r (1011), and occa-sionally n (2243) faces, were also encountered.

Many of the crystals had a blade-like appear-ance when viewed parallel to the c-axis (again, seefigure 8). Groups of intergrown crystals were alsofrequently encountered.

Figure 9. Two crystal habits were most typical inthe rabies and fancy-color sapphires from Nepal.(A) The primary crystal form was dominated bydipyramid z (2241) crystal faces, with more sub-ordinate basal pinacoid c (0001) and positiverhombohedron r (1011) crystal faces also present.(B) This primary crystal form was frequentlymodified by subordinate to intermediate dipyra-mid n (2243) crystal faces.


Figure 10. A wide range of gem-quality rubiesand fancy-color sapphires have been recoveredfrom Nepal. Distinctly bicolored—red or pinkand violetish blue—stones illustrate some of theunusual zoning in this material. The non-heat-treated Nepalese rubies and sapphires shownhere range from 0.87 to 3.86 ct. Photo by ShaneF. McClure.

Figure 11. Weak to prominent internal growthstructures were frequently observed in the sam-ple rubies and sapphires. This 4.81 ct rubyreveals moderate zonal structures parallel totwo series of dipyramid z (2241) planes, alongwith distinct blue color banding. Immersion,magnified 8 ×.

Observation of the gem perpendicular to the c-axisoften enabled us to trace the progression of crystal-habit formation (figure 12).

Color Zoning. Chemical fluctuations in the growthenvironment produced obvious color zoning in mostof the samples examined. Oscillations between con-secutive periods of crystal growth, as well as a pref-erential crystallographic orientation of the color-causing mechanisms Cr3+ (ruby) or Fe2+ ? Ti4+ (bluesapphire), are responsible for the inhomo-geneity ofcolor observed in the samples (also refer to UV-Vis-NIR Spectroscopy). When the color was not homo-geneous, red-to-pink and near-colorless zones weretypically noted parallel to the dipyramid z planes,whereas the dark violetish blue zones were concen-trated along the dipyramid z, n, or ?? planes, in addi-tion to the positive rhombohedron r (again, see fig-ure 12).

The most distinctive color-zoning characteristicswere related to the dark violetish blue zones. Thedark appearance of these zones is attributed to thepresence of red (i.e., chromium) and blue (i.e., ironand titanium) chromophores in the same growthphases. These color zones could be very narrow orvery thick; the latter were observed only parallel tothe dipyramid z planes (see figures 11 and 13). Theyalso appeared as distinct "wedges" cutting into thegemstone or even dominating it to the point of abicolor (again, see figure 13). In these zones, we alsoobserved a texture that might best be described aswispy or smoke-like. In some samples, an irregularcolor concentration had an undulating nature thatresulted from near-colorless areas, or halos, sur-rounding very small, black-appearing mineral grainsof presumably rutile (figure 14). In the case of rutile(TiO2) inclusions, Ti would be absorbed from thehost corundum, thereby depleting an essential com-ponent of the Fe2+ ?Ti4+ charge-transfer necessaryfor the blue coloration..

Twinning. We saw twinning parallel to the positiverhombohedron r (1011) in several of the stones.Typically, we noted only one direction of laminatedtwinning, parallel to a single series of r (1011)planes; occasionally, however, there were as manyas three twinning systems, parallel to additionalpositive rhombohedral planes. Parting parallel to r(1011) was also prominent in a few samples (figure15).

IInncclluussiioonnss.. A rich diversity of inclusions were notedin the sample rubies and fancy-color sapphire.


Figure 12. When viewed perpendicular to the c-axis, several samples revealed a great deal aboutthe sequence of crystallographic growth. ThisNepalese ruby shows a crystal habit composed ofthe basal pinacoid c (the faint horizontal growthplanes), the dipyramid n (the angled growthplanes), and the dipyramid z (the vertical growthplanes). Immersion, magnified 12×.

Most commonly, clouds of very fine, short rutileneedles, present throughout, gave some of thegemstones a slightly "hazy" appearance to theunaided eye (figure 16). Bright orange-to-blackcrystals of rutile were observed singly or in small

Figure 14. Some Nepalese rubies revealed irregularblue color zones that had paler patches or halos. At66× magnification, it can be seen that these halosactually surround small mineral grains of what arepresumed to be rutile. Darkfield illumination; pho-tomicrograph by Edward f. Gübelin.

irregular clusters (figure 17); these displayed a metal-lic luster when they were polished at the surface.Other stringer-type inclusion patterns consisted ofnearly parallel, slightly diverging "sprays," extendingessentially perpendicular to growth planes or in anantenna-like pattern (figure 18).

Apatite took on a variety of forms, includingeuhedral hexagonal columns (figure 19, left) andslightly curved rods (figure 19, right). Although notcommonly noted in apatite, basal cleavage was

Figure 13. Distinct color zoningwas present in most of thesample rubies and sapphiresfrom Nepal. Colorless,red/pink, and dark violetishblue zones filled large sectionsof the stones, traversed them inbands, or appeared as largewedge-like forms. Such color-zoning characteristics gavemany of the gemstones a bicol-ored appearance.


Figure 15. Parting planes parallel to one, two, orthree series of positive rhombohedron r (1011)crystal faces, were seen in a few samples. In thisstone, the parting planes-which form a checker-board pattern-were lined with the aluminumoxy-hydroxide boehmite, AIO(OH). Fiber-opticillumination, magnified 20×.

unusual was the "halo" of minute rutile needles sur-rounding an unidentified small black mineral grain(figure 24).

We observed a wide range of fingerprint-likeinclusions, all involved in various stages of the "heal-ing" process. Two-phase (liquid and gas) inclusionswere common. "Intersection tubules" at the junctionof two or three twin planes were frequently penetrat-ed by alteration products such as boehmite.Boehmite was also identified lining the partingplanes. Irregular "veins" of AlO(OH)—mostlyboehmite, but also diaspore—were also notedtraversing several of the polished gemstones. Inreflected light, the reduced luster of the AlO(OH)"vein," as compared to the higher luster of the hostcorundum, could be mistaken for the glass-like

Figure 17. Clusters of small rutile crystals werealso common in the rubies from Nepal. Typically,they were bright orange to black and displayed ametallic luster where polished at the surface.Fiber-optic illumination, magnified 50×.

Figure 18. Antenna-hke stringer formations were fre-quently seen in the Nepalese samples. Althoughsuch inclusions have been noted in rubies from LucYen (Vietnam) and Mong Hsu (Myanmar), theytended to be more densely concentrated in theNepalese samples. Fiber-optic illumination, magni-fied 32×.

Figure 16. Clouds of very fine, short rutile needleswere present in nearly all of the rubies and fancy-color sapphires from Nepal. However, none of thesestones revealed the nest-like concentrations that aretypical of rubies from Mogok. Fiber-optic illumina-tion, magnified 22×.

observed in several of the apatites we identified,especially in the rod-shaped forms. The calcium-rich mica margarite was often present in irregularmasses within which additional inclusions werenoted, such as individual crystals of apatite orrutile, or masses of graphite, sometimes formingone complex mineral assemblage (figure 20). Lightbrown crystals of phlogopite mica occurred insmall, mostly irregular, rounded forms (figure 21).Transparent colorless crystals of calcite anddolomite were seen infrequently (most of thetransparent colorless crystals in these stonesproved to be apatite or margarite). Although rare,black uvite tourmaline (figure 22) and transparentcolorless anorthite feldspar (figure 23) were iden-tified. Also


fillings observed in some heat-treated rubies, butcareful examination will establish that the gem hasnot been heated.

AAbbssoorrppttiioonn SSppeeccttrraa.. All spectra were dominated byCr3+ absorption features, with the bands being weak-er in the lighter red (i.e., pinkish) stones and moreintense in the deeper red stones. Occasionally a sec-ondary absorption influence was seen as a result ofthe Fe2+ ? Ti4+ intervalence charge transfer responsi-ble for the blue color component in the dark vio-letish blue zones. These results are consistent withthose reported earlier by Kiefert and Schmetzer(1986,1987).

Desk-Model Spectroscope. In the visible range, a gen-eral absorption to approximately 450 nm was appar-ent, along with weak to distinct lines at 468

Figure 20. Complex inclusion assemblages presentedsome intriguing identification challenges. In thisNepalese ruby, a large mass of the calcium-rich micamargarite (M) played host to other mineral inclu-sions of its own. The larger, granular-appearing massproved to be a graphite and margarite combination(G/M), while several of the brighter spots were iden-tified as apatite (A) and rutile (R). SEM photo cour-tesy of SUVA, Lucerne, Switzerland.

nm and at 475 and 476 nm (a doublet). The width of amoderate-to-distinct absorption band from approxi-mately 525 to 585 nm was related to the chromiumcontent of the gemstone. We also noted faint lines at659 and 668 nm, plus two strong lines at 692 and 694nm, which appear as a bright emission line at 693 nm.

UV-Vis-NIR Spectroscopy. The general shape of thespectral curve also varied considerably depending onthe chromium content of the zones measured. Thetwo broad bands at about 405 and 550 nm, as well asthe weak to distinct sharp peaks recorded at 468, 475,476, 659, 668, 692 and 694, are all ascribed to Cr3+. Afaint absorption peak observed at 675 nm in somestones has also been recorded in rubies from othersources (e.g., Mong Hsu, Myanmar, and certaindeposits in east Africa), but the cause is still unclear(Peretti et al., 1995).

Infrared Spectroscopy. In addition to the dominantabsorption characteristics of corundum, between 300and 1000 cm-1 (peak positions at about 760, 642, 602,and 450 cm-1; Wefers and Bell, 1972), the rubies

Figure 19. Common in theNepalese samples, apatitetook on several forms, includ-ing euhedral, hexagonalcolumns (left, magnified 50 ×)and slightly curved "rods"(right, 32×). In most cases,these inclusions displayedbasal cleavage, a propertyinfrequently seen in apatite.Fiber-optic illumination.

Figure 21. Predominantly small, rounded lightbrown crystals of phlogopite mica also were seenfrequently in the Nepalese rubies. Fiber-optic illu-mination, magnified 45×.


Figure 23. Raman micro-spectroscopy identified thismineral aggregate in a Nepalese ruby as anorthitefeldspar. This is the first report of this mineral inruby from any source. Photomicrograph by EdwardJ. Gübelin; darkfield illumination, magnified 66×.

the next most significant trace elements recorded,followed by measurable amounts of vanadium (V)and gallium (Ga), as shown in table 2.

One interesting observation related to the trace-element concentrations of areas that were "pure" redas opposed to ones that were "pure" dark violetishblue. In every instance, the concentrations in abso-lute and relative values showed no consistent varia-tion between Cr, Fe, or Ti within the two colorzones.

DDIISSCCUUSSSSIIOONNCorundum from Nepal has received sporadic men-tion in the gemological literature over the pastdecade or so (see, e.g., Harding and Scarratt, 1986;Kiefert and Schmetzer, 1986 and 1987; Bank et al.,1988; Niedermayr et al., 1993). The results of thiscurrent investigation support and expand on thefindings of such earlier researchers and, for the firsttime, provide detailed locality information. We iden-tified three inclusions in the rubies from Nepal thathad not been described before—uvite tourmaline,anorthite feldspar, and diaspore—as well as severaldistinctive inclusion and color-zoning patterns.However, we did not observe the three-phase inclu-sions described by Kiefert and Schmetzer (1986,1987) in any of our samples. Also, we did notencounter the identification difficulties experiencedby Bank et al. (1988); that is, all of our samples wereeasily identified as natural corundum.

In general, the rubies and fancy-color sapphiresfrom Nepal are similar to corundum from other mar-ble-type sources found around the world, includingMyanmar (Burma), Vietnam, Afghanistan, Pakistan,and Tanzania. The most distinctive features of theseNepalese corundums are their inclusions.

Figure 22. A series of unusual granular masseswere identified as uvite tourmaline, which has notbeen documented as an inclusion in rubies fromany other source. It thus provides a useful indica-tion of Nepalese origin. Fiber-optic illumination,magnified 35×.

and sapphires in this study revealed two dominantbands at about 3320 cm-1 and 3085 cnr-1, with anadditional pair of weaker bands at 2100 and 1980cm"1 (figure 25). These absorption bands are relat-ed to OH stretching frequencies and identify thepresence of boehmite (Farmer, 1974; Wefers andMisra, 1987). Although different color zones didnot reveal any statistical differences in the pres-ence or absence of boehmite, different areas andorientations of the same stone did show variationsin the absolute and relative intensities of theabsorption bands, as well as a slight shift in theposition of the absorption maximum. To a muchlesser degree, diaspore was also indicated in theinfrared spectra of some samples, with bands atapproximately 1990, 2040, 2885, and 3025 cm-1

(Farmer, 1974; Wefers and Misra, 1987; Smith,1995).

The presence of AlO(OH)—boehmite and dias-pore—was generally traced to locations along part-ing planes or irregular seams. Not all samplesshowed AlO(OH)-related absorption bands, whileseveral displayed such strong AlO(OH) absorptionfeatures that a distinction between boehmite anddiaspore was not possible. Absorption bands associ-ated with mica and calcite were also occasionallyrecorded.

CChheemmiiccaall AAnnaallyyssiiss.. The most significant varia-tions were recorded in Cr concentration, whichagain correlated to the depth of red-to-pink color inthe area measured. Titanium (Ti) and iron (Fe) were


The dense concentrations of very fine, short rutileneedles are unlike the long, highly iridescent rutileneedles observed in most rubies from marble-typedeposits. Dense "nests" of short rutile needles, sotypical of rubies from the historic Mogok stonetract in upper Myanmar, were not seen in ourNepalese study samples. The various stringer pat-terns we saw in these Nepalese stones have alsobeen seen in rubies from other deposits, includingMong Hsu in Myanmar (Smith and Surdez, 1994;Peretti et al., 1995) and Vietnam (Kane et al., 1991;Smith, 1996). However, the Nepalese stonesappeared to have much denser concentrations ofsuch patterns.

Apatite crystals of various forms may be seen inrubies from many sources, including Mogok (e.g.,Gübelin and Koivula, 1986), Vietnam (Kane et al.,1991), and various deposits in East Africa.However, the rod-shaped apatites observed in thesamples from Nepal have not been described inrubies or fancy-color sapphires from any othersource. Kane et al. (1991) described similar transpar-ent colorless rod-shaped minerals in rubies fromVietnam, but these were identified as calcite. Acommon mineral inclusion in rubies from marble-type deposits, calcite was encountered only rarelyin the rubies from Nepal.

Mica is common in corundum from otherdeposits (e.g., Myanmar, Sri Lanka, and Tanzania),but the sheer number of inclusions of the varietymargarite in our test samples is unlike anything we

have seen in rubies from other sources. Anorthitefeldspar and uvite tourmaline, which were observedin a few of our sample rubies and fancy-color sap-phires, have not been identified in rubies from anyother locality. Therefore, they may also be instru-mental in establishing Nepalese origin. Nor havethe authors seen inclusion patterns such as thesmall black mineral surrounded by a spherical"halo" of minute rutile needles in rubies from othersources. On the other hand, inclusions such aspyrrhotite, zircon, pyrite, and spinel have not beenidentified to date in rubies from Nepal; these are fre-quently seen in rubies from East Africa, Vietnam,and Myanmar, as well as other sources.

In their natural (not heat-treated) state, theserubies, pink sapphires, and bicolored sapphiresshould not be difficult to distinguish from rubies,including those with blue color zones, from othersources. No heat-treatment studies were conductedon these samples. However, when such stones areheat treated, the inclusion features and color zoningwill be greatly altered, making it more difficult toidentify corundum from Nepal (see, e.g., Peretti etal., 1995). Although it is usually difficult to detect agemstone's source by the inclusion features, obser-vation of any of the above-described mineral inclu-sions will clearly separate a natural ruby or fancy-color sapphire from a synthetic counterpart pro-duced by any of the known manufacturingprocesses.

The linear and angular sequences of non-color-zoned growth structures in our samples—followingthe crystal planes z, n, ?, r, and c—did not reveal anyunique or diagnostic features or patterns. Similarsequences of growth structures have been observedby one of the authors (CPS) in natural rubies fromother sources (e.g., Vietnam or East Africa).However, they can be useful in separating

TABLE 2. Semi-quantitative EDXRF chemical analyses of major-to-trace elements in the rubies and fancy-color sapphires fromNepal.

Oxide Wt.%

Al203 98.9-99.8Cr2O3 0.013-0.383TiO2 0.016-0.224Fe2O3 0.004-0.069V2O5 0.004-0.034G a2O3 0.010-0.023

Figure 24. Still another feature that has not beennoted in rubies from other sources was present inmany of the rubies and fancy-color sapphires fromNepal: small black mineral grains surrounded byhalos of very fine, minute rutile needles. Fiber-optic illumination, magnified 50×.


Figure 25. The non-heat-treated rubies and fancy-color sapphires from Nepal revealed additionalabsorption features in the infrared region of thespectrum. Distinct absorption bands at 3320 and3085 cm-1, and weaker peaks at 2100 and 1980 cm-1,indicated the presence of boehmite, which was seenconcentrated along veins or lining parting planes.Such absorption characteristics are helpful not onlyin identifying foreign mineral phases that may bepresent, but also for indicating that the gem has notbeen heat treated.

natural from synthetic rubies (Schmetzer, 1986a andb; Kiefert and Schmetzer, 1991; Smith, 1996). Kiefertand Schmetzer (1986, 1987) also identified growthplanes consisting of the second-order hexagonalprism a (1120) and hexagonal dipyramid v (4481).Bank et al. (1988) described an unusual sequencewith these two growth planes that also was notencountered during our study. This unusualsequence appears to relate directly to the blade-likeforms of some rough crystals. Our samples did notreveal any "swirled" growth structures, such as thoseseen in rubies from Mogok (e.g., Gübelin andKoivula, 1986) or Vietnam (Kane et al., 1991). Nordid they have centralized "cores," such as those pre-sent in rubies from Mong Hsu (Smith and Surdez,1994; Peretti et al., 1995).

Blue color zones and color banding have beennoted in rubies from several sources, includingAfghanistan (Bowersox, 1985; Bowersox andChamberlin, 1995), Myanmar (Mong Hsu: Smith andSurdez, 1994; Kammerling et al., 1994; Smith, 1995;Peretti et al., 1995), Vietnam (Kane et al., 1991), aswell as Sri Lanka and Tanzania (Tunduru), by one ofthe authors (CPS). However, the blue color zones and

banding in Nepal rubies have some distinctivecharacteristics. These include the successions ofstraight and angular, thin-to-thick bands parallelto the dipyramid z (2241) planes; the large red anddark violetish blue zones (i.e., bicolored stones);the "wedge-shaped" color zones,- and the wispy tosmoke-like textures observed in the color bandsand "halos" of nonblue zoning surrounding miner-al inclusions.

Chemically, these rubies and fancy-color sap-phires are also similar to their counterparts fromother marble-type sources (Tang et al., 1988;Hughes, 1990). Nevertheless, their chemistry willprovide a ready means of separating them fromhigh-iron natural rubies from sources such asThailand (Tang et al., 1988), Cambodia (Jobbinsand Berrange, 1981), Madagascar (Smith, 1996),and certain deposits in East Africa (see, e.g., Hänniand Schmetzer, 1991). In addition, the collectionof chemical data is also valuable in separating nat-ural from synthetic corundum (e.g., Stern andHänni, 1982; Muhlmeister and Devouard, 1991).

Infrared spectroscopy may provide additionalproof that the gemstone was not heat treated,when AlO(OH) is present, as well as a very goodindication of whether it is natural or synthetic(Volynets et al., 1972; Beran, 1991; Smith, 1995,pp. 326–328). One of the authors (CPS) hasobserved that rubies from other natural sources,such as certain deposits in Vietnam or East Africa,sometimes also reveal dominant absorption fea-tures relating to boehmite, as well as other miner-als.

No consistent variations were noted in thesamples from Chumar as compared to those fromRuyil, with respect to their standard gemologicalproperties, UV-Vis-NIR and FTIR spectral charac-teristics, crystal morphology, chemical make-up,or the inclusions identified. One inclusion feature,however, that may indicate that a ruby came fromthe Ruyil mine is the presence of graphite (in asso-ciation with certain mineral inclusions), whichappears to be much more prevalent in the corun-dum from this deposit.

It is interesting to note that "trapiche" corun-dum, with six "arms" extending from a centralcore, has been found at the Ruyil deposit (figure26). Trapiche-like rubies and pink sapphires havebeen reported from Vietnam and Mong Hsu(Müllenmeister, 1995; Schmetzer et al., 1996).However, the arms in the Nepalese samples arevery different in composition—predominantly phl-ogopite, apatite, calcite, and graphite—from thatidentified in samples from other sources.


CCOONNCCLLUUSSIIOONNIn recent years, a number of ruby and sapphiredeposits have been discovered in a wide variety ofsources around the world. While Nepal is neitherthe newest nor most significant corundum sourcein the trade today, it offers an interesting array ofruby and fancy-color sapphires for jewelers, gemolo-gists, and consumers alike. These gems from theGanesh Himal appear to be concentrated along asingle geologic "belt" within the northern DhadingDistrict of east-central Nepal, and in two mines inparticular, Chumar and Ruyil. Since they were firstdiscovered in the early 1980s, these gemstones havebeen entering the world markets. The isolated loca-tions, high altitudes, harsh seasonal weather condi-tions, and other difficulties have contributed to thesporadic mining activities and the relatively smallamounts of gem material produced to date.However, research by two of the authors (MNMand AMB) indicates that larger reserves of theserubies and fancy-color sapphires are yet to be dis-covered. Modernized mining equipment and meth-ods (including tunneling or "benching" techniques)will be necessary for these deposits to reach theirfull potential.

Acknowledgments: The authors are grateful toDr. L. Kiefert and Dr. H. A. Hänni of the SwissGemmological Institute (SSEF), Basel,Switzerland, for performing Raman spectral anal-yses; to Dr. H.-D. Von Schulz of SUVA, Lucerne,Switzerland, for performing scanning electronmicroscopic analyses; to Professor S. Graeser ofthe Institute of Mineralogy and Petrography,University of Basel, for X-ray diffraction analy-ses; and to Prof. W. B. Stern of the Institute ofMineralogy and Petrography, University of Basel,Switzerland and Ms. N. Surdez of the GübelinGemmological Laboratory, Lucerne, for providingthe semi-quantitative chemical analyses. Unlessotherwise noted, photomicrographs are byChristopher P. Smith.

Figure 26. Reminiscent of trapiche emeralds, avery small number of rubies from the Ruyildeposit display a series of rays that intersect at thecenter of the crystal, along the three-fold axis andparallel to the prism crystal faces. Each ray con-sists of a linear concentration primarily of phlogo-pite, apatite, calcite, and graphite. Such unusualforms of "trapiche" corundum have also beenrecovered from Vietnam and Myanmar. Photo byEva Strauss-Paillard.

Although similar in general to rubies and fancy-color sapphires from other marble-type deposits,the Nepal corundums may be distinguished on thebasis of the entire collection of the gem's individu-al properties and characteristics. Certain featuresthat may prove helpful in the identificationinclude dense concentrations of very fine, shortrutile needles throughout the stone; rod-shapedforms of apatite; masses of transparent colorlessmargarite,-opaque black masses of uvite tourma-line; transparent colorless anorthite feldspar; andsmall black mineral grains surrounded by "halos"of minute rutile needles. Nepal may also be indi-cated for a particular ruby if any of the transparentmineral inclusions are associated with black mass-es of graphite. Nepal rubies and fancy-color sap-phires may also have distinctive color-zoning char-acteristics, such as large red and dark violetish blueportions in a single stone (i.e., bicolor), or dark vio-letish blue zones in thick bands or wedge shapes.Especially distinctive is the presence of a wispy orsmoke-like texture or near-colorless "halos" sur-rounding a mineral inclusion.

Such an ensemble of internal features also willhelp separate a Nepalese corundum from a synthet-ic corundum of any of the various production tech-niques. As a reminder, even in the remotestregions of the world, one should never take forgranted that a gemstone is natural, as the syntheticruby crystal fragment purchased in Nepal by one ofthe authors (MNM) illustrates.

Nepal's mineral wealth—consisting of tourma-line, beryl, garnet, quartz, spinel, danburite, kyan-ite, apatite, sodalite, zircon, sphalerite, epidote,diopside, iolite, and andalusite, among others—isalready recognized widely in the trade. Now rubies,pink sapphires, and bicolored corundum can alsobe added to the list.


RREEFFEERREENNCCEESSBaba T. (1982) A gemstone trip to Nepal. Gemmological

Review, Vol. 4, No. 12, pp. 2–5. Bank H.; Gübelin E., Harding H.H., Henn U., Scarratt K.,

Schmctzer K. (1988) An unusual ruby from Nepal. Journal ofGemmology, Vol. 21, No. 4, pp. 222–226.

Bassett A.M. (1979) Hunting for gemstones in the Himalayas of Nepal. Lapidary Journal, Vol. 33, No. 7, pp.1492–1520.

Bassett A.M. (1984) Rubies in the Himalayas of Nepal. Reportsubmitted to the Nepal Department of Mines and Geology,Katmandu, 19 pp.

Bassett A.M. (1985a) Application for mining license.Submitted to the Nepal Department of Mines and Geology,Katmandu, 14 pp. Bassett A.M. (1985b) The tourmalines ofNepal. Mineralogical Record, Vol. 16, No. 5, pp. 413–418.

Bassett A.M. (1987) Nepal gem tourmalines. Journal of NepalGeological Society, Vol. 4, No. 1–2, pp. 31–41.Beran A. (1991) Trace hydrogen in Verneuil-grown corun-dum and its colour varieties-an IR spectroscopic study.European Journal of Mineralogy, Vol. 3, pp. 971–975.

Bowersox G.W. (1985) A status report on gemstones fromAfghanistan. Gems & Gemology, Vol. 21, No. 4, pp.192–204.

Bowersox G.W., Chamberlin B.E. (1995) Gemstones ofAfghanistan. Geoscience Press, Tucson, AZ.

Chakrabarti C.K. (1994) A preliminary report on the GaneshHimal ruby occurrences. Submitted to the NepalDepartment of Mines and Geology, Katmandu, 8 pp.

Farmer V.C. (1974) The infrared spectra of minerals.Mineralogical Society Monograph 4, Mineralogical Society,London.

Gübelin E.J., Koivula J.I. (1986) Photoatlas of Inclusions inGemstones. ABC Edition, Zurich, Switzerland.

Hänni H.A., Schmetzer K. (1991) New rubies from theMorogoro area, Tanzania. Gems & Gemology, Vol. 27, No.3, pp. 156–167.

Harding R.R., Scarratt K. (1986) A description of ruby fromNepal. Journal of Gemmology, Vol. 20, No. 1, pp. 3–10.

Henn U., Milisenda C.C. (1994) A microscopical study ofNepalese ruby and sapphire. Journal of the GemmologicalAssociation of Hong Kong, Vol. 17, pp. 82–84.

Hughes R.W. (1990) Corundum, 1st ed. Butterworths GemBooks, Butterworth-Heinemann, London.

Hurlbut C.S., Kammerling R.C. (1991) Gemology, 2nd ed.Wiley Interscience, New York.

Jobbins E.A., Berrangé J.P. (1981) The Pailin ruby and sapphiregemfield, Cambodia. Journal of Gemmology, Vol. 17, No. 8,pp. 555–567.

Kammerling R.C, Scarratt K., Bosshart G., Jobbins E.A., KaneR.E., Gübelin E.J., Levinson A.A. (1994) Myanmar and itsgems-An update. Journal of Gemmology, Vol. 24, No. 1, pp.

3–40.Kane R.E., McClure S.F., Kammerling R.C, Khoa N.D., Mora

C, Repetto S., Khai N.D., Koivula J.I. (1991) Rubies andfancy sapphires from Vietnam. Gems & Gemology, Vol. 27,No. 3, pp. 136-155.

Kiefert L., Schmetzer K. (1986) Rosafarbene und violetteSapphire aus Nepal. Zeitschrift der DeutschenGemmologischen Gesellschaft, Vol. 35, pp. 113–125.

Kiefert L., Schmetzer K. (1987) Pink and violet sapphires fromNepal. Australian Gemmologist, Vol. 16, No. 6, pp. 225–230.

Kiefert L., Schmetzer K. (1991) The microscopic determina-tion of structural properties for the characterization of opti-cal uniaxial natural and synthetic gemstones, part 1:General considerations and description of the methods.Journal of Gemmology, Vol. 22, No. 6, pp. 344–354.

Liddicoat R.T. (1989) Handbook of Gem Identification, 12ththed. Gemological Institute of America, Santa Monica, CA.

Muhlmeister S., Dcvouard B. (1991) Determining the naturalor synthetic origin of rubies using energy-dispersive X-rayfluorescence (EDXRF). In A.S. Keller, Ed., Proceedings of theInternational Gemological Symposium, GemologicalInstitute of America, Santa Monica, CA, pp. 139–140.

Müllenmeister H.J. (1995) Ein Trapiche-Rubin aus Myanmar(Burma). Lapis, Vol. 12, No. 95, pp. 50.

Niedermayr G., Brandstätter P., Hammer V.M.F. (1993)Edelund Schmucksteinvorkommen in Nepal. Zeitschrift derDeutschen Gemmologischen Gesellschaft, Vol. 42, No. 2–3,pp. 69–89.

Peretti A., Schmetzer K., Bernhardt H.J., Mouawad F. (1995)Rubies from Mong Hsu. Gems & Gemology, Vol. 31, No. 1,pp. 2–26.

Robinson G.W., King V.T., Asselbom E., Cureton F.,Tschernich R., Sielecki R. (1992) What's new in minerals?Mineralogical Record, Vol. 23, No. 5, pp. 423–437.

Schmetzer K. (1986a) An improved sample holder and its usein the distinction of natural and synthetic ruby as well asnatural and synthetic amethyst. Journal of Gemmology,Vol. 20, No. 1, pp. 20–33.

Schmetzer K. (1986b) Natürliche und synthetische Rubine–Eigenschaften und Bestimmung. Schweizerbart, Stuttgart,Germany.

Schmetzer K., Hänni H.A., Bernhardt H-J., Schwartz D. (1996)Trapiche rubies. Gems & Gemology, Vol. 32, No. 4, pp.242–250.

Smith C.P., Surdez N. (1994) The Mong Hsu ruby: A new typeof Burmese ruby. JewelSiam, Vol. 4, No. 6, pp. 82–98.

Smith C.P. (1995) Contribution to the nature of the infraredspectrum for Mong Hsu rubies. Journal of Gemmology, Vol.24, No. 5, pp. 321–335.

Smith C.P. (1996) Introduction to analyzing internal growthstructures: Identification of the negative d plane in naturalruby. Gems & Gemology, Vol. 32, No. 3, pp. 170–184.

Stern W.B., Hänni H.A. (1982) Energy dispersive X-ray spec-trometry: a non-destructive tool in gemmology. Journal ofGemmology, Vol. 18, No. 4, pp. 285–296.

Tang S.M., Tang S.H., Tay H.S., Retty A.T. (1988) Analysis ofBurmese and Thai rubies by PIXE. Applied Spectroscopy,Vol. 42, No. 1, pp. 44–18.

Volynets F.K., Sidrova E.A., Stsepuro N.A. (1972) OH-groupsin corundum crystals which were grown with the Verneuiltechnique. Journal of Applied Spectroscopy, Vol. 17, pp.1088–1091.

Webster R. (1983) Gems: Their Sources, Descriptions andIdentification, 4th ed. Butterworths, London.

Wefers K., Bell G.M. (1972) Oxides and Hydroxides ofAlumina, Alcoa Research Laboratories Technical Paper No.19. Alcoa Research Laboratories, St. Louis, MO.

Wefers K., Misra C. (1987) Oxides and Hydroxides ofAlumina, Alcoa Research Laboratories Technical Paper No.19, Revised. Alcoa Research Laboratories, St. Louis, MO.

42 Near-Colorless Synthetic Diamonds GEMS & GEMOLOGY Spring 1997

EExamination of 51 colorless to near-col-orless synthetic diamonds from all knownsources of production confirms that theycan be distinguished from similar-appear-ing natural diamonds on the basis of theirgemological properties. Although some maycontain opaque mtallic inclusions, themost distinctive feature of near-colorlesssynthetic diamonds is their luminescenceto ultraviolet radiation and to an electronbeam (cathodoluminescence). In particular,almost all fluoresce yellow or yellow-greento short-wave UV and, when the ultravioletlamp is turned off, they continue to phos-phorese for 60 seconds or more. These dis-tinctive reactions to ultraviolet radiationare very useful in identification, becausemany diamonds can be checked at onetime.

ABOUT THE AUTHOR

Dr. Shigley is director of GIA Research, Carlsbad,California; Mr. Moses is vice president ofIdentification Services, and Dr. Reinitz is a researchscientist, at the GIA Gem Trade Laboratory, NewYork; Mr. Elen is a research associate in GIAResearch, Carlsbad; and Mr. McClure is supervisorof Identification Services in the GIA Gem TradeLaboratory, at the Gemological Laboratory, PhysicsDepartment, University of Nantes, France.Gems & Gemology, Vol. 33 No. 1, pp. 42–53© 1997 Gemological Institute of America

T he past few years have witnessed the entry intothe trade of synthetic diamonds (mainly yellow)

from various research institutes in Russia and elsewhere(Shigley et al., 1993a; Scarratt et al., 1994). In 1993, how-ever, Thomas Chatham, of Chatham CreatedGemstones, made the first of several public announce-ments of his intention to market Russian-grown syn-thetic diamonds, including "colorless" ones, for jewelrypurposes ("Chatham to sell 'created' diamonds," 1993;Nassau, 1993). So far, however, there are only a few doc-umented cases of faceted "colorless" synthetic diamondsin the jewelry trade. In 1996, a near-colorless 0.16 ctround-brilliant-cut synthetic diamond was submitted tothe GIA Gem Trade Laboratory in New York by a localdiamond dealer, and was quickly identified by laborato-ry staff. Nevertheless, gem-testing laboratories, individ-ual diamond dealers, and jewelers need to prepare forthe appearance of small amounts of this material in themarket, possibly represented as natural stones.A brief description of the distinctive features of color-less to near-colorless synthetic diamonds was providedin a chart and article by Shigley et al. (1995). However,the information in that chart was based on the examina-tion of only 22 near-colorless synthetic diamonds, thetotal data base at the time the chart was prepared. Thepresent article expands on that description by presentingdetailed information on these 22 synthetic diamondsplus 29 examined since then (see, e.g., figure 1), for atotal of 51 colorless to near-colorless synthetic dia-monds tested by GIA researchers from 1984 through1996 (see the list in table 1). All known manufacturersare represented.

BBAACCKKGGRROOUUNNDDIn 1971, Robert Crowningshield published the firstgemological description of the faceting-quality synthetic

GEMOLOGICAL PROPERTIES OFNEAR-COLORLESS SYNTHETIC

DIAMONDSBy James E. Shigley, Thomas M. Moses, Ilene Reinitz, Shane Elen,

Shane F. McClure, and Emmanuel Fritsch

Near-Colorless Synthetic Diamonds GEMS & GEMOLOGY Spring 1997 43

Figure 1. These near-col-orless synthetic diamonds

(0.41 to 0.91 ct) werefashioned from crystalsgrown for experimental

purposes at the De BeersDiamond Research

Laboratory inJohannesburg, South

Africa. Examination ofthese and many other

near-colorless syntheticdiamonds, representing

all known methods ofproduction, has revealedseveral distinctive gemo-

logical properties thatallow then identificationby standard gem-testing

methods.

diamonds that had been grown by General Electric(G.E.) scientists. Among the samples he examinedwere two faceted synthetic diamonds that werenearly colorless (0.305 ct and 0.260 ct; "J" and "G"color grades, respectively).

Although a number of diamond simulants(such as strontium titanate, yttrium aluminumgarnet [YAG], and synthetic spinel) were availablein the jewelry trade at that time (Nassau, 1980;Hobbs, 1981), these could be easily and readily dis-tinguished from diamond on the basis of theirthermal conductivity and other gemological fea-tures. In contrast, the possibility of faceted "color-less" synthetic diamonds entering the gem marketcaused great concern among diamond dealers atboth wholesale and retail levels. If these synthet-ics could not be identified readily and practicallywith simple gem-testing equipment (since thermalconductivity meters would read the same for bothnatural and synthetic diamond), they could under-mine consumer confidence in natural gem dia-monds.

Following the article by Crowningshield (1971)and an update on the G.E. synthetics by Koivulaand Fryer (1984), we know of only a few othergemological articles that mention "colorless" syn-thetic diamonds. In the early 1990s, we reportedon two experimental, isotopically pure carbon-12,diamond crystals (1.04 and 0.91 ct) grown atGeneral Electric Superabrasives (Anthony et al.,1990; Shigley et al., 1993b). Shigley et al. (1992)described a 5.09 ct synthetic diamond crystal pro-duced by Sumitomo researchers that was yellowand blue at the outer portions, and colorless in thecenter. Rooney et al. (1993) described a small,

experimental, boron-doped, near-colorless (slightlygray) De Beers synthetic diamond that weighed0.049 ct. In their Fall 1996 Gems & Gemology arti-cle, Welbourn et al. describe two diamond-verifica-tion instruments developed by De Beers scientists,the DiamondSure™ and the DiamondView™. TheDiamondSure, which is based largely on a spectralfeature inherent to most near-colorless type la nat-ural diamonds, but not their colorless to near-col-orless synthetic counterparts, will refer near-color-less synthetic diamonds for further tests. However,the Diamond View uses the pattern of ultravioletfluorescence, which is very different for natural ascompared to synthetic diamonds, as the basis forseparating both near-colorless and colored dia-monds.

MMAATTEERRIIAALLSS AANNDD MMEETTHHOODDSSSSyynntthheettiicc DDiiaammoonnddss EExxaammiinneedd.. Table 1 lists the 51near-colorless synthetic diamond samples seen byGIA researchers from 1984 to the present and someof our observations: 11 from General Electric, threeproduced by Sumitomo Electric Industries, sixRussian synthetic diamonds (the research facilitywhere they were produced has not been identifiedby the distributor, Chatham Created Gems), 22manufactured by De Beers researchers in SouthAfrica, eight of unidentified manufacture (loanedby Starcorp Inc., of Goleta, California), and the 0.16ct round brilliant mentioned above, which wassubmitted to the GIA Gem Trade Laboratory by aclient in the trade. Most of these 51 samples(specifically, the G.E., Sumitomo, and De Beers


TABLE 1. Properties of near-colorless synthetic diamonds examined by GIA. 1984–1996.

Ref. Discription Color Date Mfr.d Ultraviloet luminescencee Cathodo- Inclusions Attraction Electrical Other featuresno. gradea of Inclusions to a conductivity

study Short-wave Phosphorescence magnet (intensity)UV fluorescence (duration)

10210 0.30 ct round n.t.b 1984 G.E. Weak yellow Yellowish white n.t. n.t n.t. n.t.brilliant

10211 0.31 ct crystal n.a.c 1984 G.E. Yellow Yes (color not reported) n.t. n.t. n.t. n.t.

20121 0.20 ct crystal na. 1986 G.E. Moderate yellowish Strong blue, persistentf n.t. Opaque metallic n.t n.twhite flux

20226 0.50 ct crystal n.a. 1986 Sumitomo Moderate yellow Strong yellow, persistent Strong green-blue; Opaque metallic flux n.t. Yesblue phosphorescence (variable)

20805 0.75 ct crystal n.a. 1988 G.E. Moderate yellow, zoned Strong yellow, persistent Strong blue, zoned; blue Opaque metallic flux n.t. Yes Blue electroluminesphosphorescence cence

20808 0.39 ct round J-K 1988 G.E. Moderate yellow, zoned Moderate yellow, Strong blue, zoned Opague metallic flux n.t. Yes brilliant persistent (weak)

781 0.91 ct crystal n.a. 1991 G.E. Weak orange-yellow Moderate green-yellow, Strong blue, zoned Clouds of gray platelets Yes n.d.g Yellow luminescence persistent to X-rays

782 1.04 ct crystal n.a. 1991 G.E. Weak orange-yellow Moderate green-yellow, Moderate yellow-green, Triangular platelets and n.d. n.d. Yellow luminescencepersistent zoned opaque metallic flux to X-rays

21282 0.78 ct round F-G 1991 G.E. Moderate green-yellow, Moderate green-yellow, n.t. Opaque metallic flux Yes Yes Blue brilliant zoned persistent (variable) electroluminescence

21283 0.29 ct oval H 1991 G.E. Moderate green-yellow, Moderate green-yellow, n.t. Pinpoint Yes Yes Blue brilliant zoned persistent (variable) electroluminescence

21399 0.05 ct round K 1991 De Beers Strong yellow, zoned Strong yellow, persistent Strong yellow, zoned Opaque metallic flux n.d. n.d. Faint yellow and blue brilliant color zones

21620 0.16 ct crystal n.a. 1993 De Beers Moderate green-yellow Moderate yellow, persistent Yellow-green, zoned Triangular platelets and n.d. Yes Faint blue color zoneopaque metallic flux (variable)

21621 0.15 ct crystal n.a. 1993 De Beers Moderate green-yellow Moderate yellow, persistent Yellow-green, zoned Pinpoint Weak n.d.attraction

21622 0.15 ct crystal n.a. 1993 De Beers Moderate green-yellow Moderate yellow, persistent Yellow-green, zoned Opague metallic flux Moderate n.d.attraction

21705 1.25 ct crystal n.a. 1994 Sumitomo Moderate orange-yellow, Orange-yellow Weak blue, zoned Opaque metallic flux Moderate Yes Zoned: blue and red zoned attraction (variable) electroluminescence,

with blue phosphorescence; orange thermoluminescence; weak orange fluorescence to long-wave UV

21706 0.23 ct crystal n.a. 1994 Sumitomo Weak green, zoned Greenish blue, persistent Strong blue Opaque metallic flux n.d. n.d.

21617 0.19 ct crystal n.a. 1994 De Beers Moderate greenish Moderate yellow, persistent Yellow-green, zoned Opaque metallic flux Weak Yes Faint blue and yellow yellow, zoned attraction (variable) color zones

21619 0.17 ct crystal n.a. 1994 De Beers Moderate greenish Moderate yellow, persistent Yellow-green, zoned Opaque metallic flux Weak Yes Faint blue and yellow yellow, zoned attraction (variable) color zones

21618 0.22 ct crystal n.a. 1994 De Beers Moderate greenish Moderate yellow, persistent Yellow-green, zoned Pinpoint n.d. Yes Faint blue and yellow yellow, zoned (variable) color zones

21686 0.42 ct crystal n.a. 1994 Unidentified Very weak yellow Weak yellow, persistent n.t. Opaque metallic flux n.d. Yes Russian co. (variable)(Chatham)

30089 0.11 ct crystal n.a. 1995 Unidentified Weak yellow Moderate yellow, persistent n.t. Opaque metallic flux Yes Yes Russian co. (variable)(Chatham)

30099 0.01 ct round I 1996 Unknown Inert Inert n.t. None visible Too small Too small to testbrilliant (Starcorp) to test

30094 0.09 ct trill iant I 1996 Unknown Weak yellow, zoned Weak yellow, persistent n.t. Opaque metallic flux Yes n.d.(Starcorp)

30091 0.11 ct rectangle K 1996 Unknown Weak yellow, zoned Weak yellow, persistent n.t. Opaque metallic flux Yes Yes Blue electroluminesc(Starcorp) (variable) ence, zoned, withblue

phosphorescence

30096 0.22 ct crystal n.a. 1996 Unknown Moderate yellow Moderate green-blue, n.t. Opaque metallic flux Yes Yes Blue electrolumines(Starcorp) persistent (variable) cence, zoned, with

blue phosphorescence; fluoresced weak yellow to long-wave UV

30092 0.07 rectangle I 1996 Unknown Weak yellow, zoned Strong yellow, persistent n.t. Opaque metallic flux Yes Yes Blue electrolumines(Starcorp) (variable) cence, zoned, with

blue phosphorescence

30095 0.25 ct crystal n.a. 1996 Unknown Weak yellow, zoned Strong yellow, persistent n.t. Opaque metallic flux n.d Yes Blue electrolumines(Starcorp) (variable) cence, zoned, with

blue phosphorescence


Ref. Discription Color Date Mfr.d Ultraviloet luminescencee Cathodo- Inclusions Attraction Electrical Other featuresno. gradea of Inclusions to a conductivity

study Short-wave Phosphorescence magnet (intensity)UV fluorescence (duration)

30098 0.02 ct round H 1996 Unknown Weak yellow Strong yellow, persistent n.t. None visible Too small Too small brilliant (Starcorp) to test to test

30097 0.02 ct round H 1996 Unknown Weak yellow Strong yellow, persistent n.t. None visible Too small Too small brilliant (Starcorp) to test to test

30103 0.41 ct round G 1996 De Beers Weak yellow-green, zoned Strong green-yellow Strong blue Opaque metallic flux n.d. n.d. Blue thermoluminesbrilliant cence

30102 0.52 ct round I 1996 De Beers Weak yellow-green, zoned Strong green-yellow Strong blue Opaque metallic flux, Moderate n.d. Blue thermoluminesbrilliant with a reddish attraction cence

appearance

35018 0.16 ct round F to G 1996 Unknown Weak yellow-green Weak green changing to blue Moderate blue, zoned None visible n.d. n.d.brilliant

35017 0.72 ct round H to J 1996 G.E. Moderate yellow Strong yellow-green n.t. Pinpoint n.d. ndbrilliant

30100 0.61 ct round H 1996 De Beers Weak yellow, zoned Strong green-yellow Strong blue, zoned Blue, triangular n.d. n.d. Blue thermoluminesbrilliant platelets cence

30101 0.91 ct round I 1996 De Beers Weak yellow, zoned Strong green-yellow Strong blue Opaque metallic flux n.d. n.d. Blue thermoluminesbrilliant cence

21966 0.56 ct round H 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Platelet, pinpoint, and Moderate n.d. brilliant zoned opaque metallic flux attraction

21965 0.58 ct round I 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Triangular platelets Weak n.d.brilliant zoned attraction

21962 0.37 ct round H 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Platelet, pinpoint, and n.d. n.d.brilliant zoned opaque metallic flux

21961 0.26 ct round H 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Pinpoint and opaque n.d. n.d.brilliant zoned metallic flux

21960 0.37 ct round H 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Pinpoint and opaque n.d. n.d.brilliant zoned metallic flux


21970 0.42 ct round H 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Pinpoint and opaque Weak n.d.brilliant zoned metallic flux attraction

21969 0.67 ct round I 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Pinpoint and opaque Strong n.d.brilliant zoned metallic flux attraction


21956 0.36 ct round H 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t. Pinpoint and opaque n.d ndbrilliant zoned metallic flux

30105 0.51 ct crystal n.a. 1996 Unidentified Very weak orange Weak yellow n.t. Opaque metallic flux Yes YesRussian co. (Chatham)

30106 0.50 ct crystal n.a. 1996 Unidentified Strong yellow Strong yellow, persistent n.a. Opaque metallic flux Yes YesRussian co.

(Chatham)

30107 0.41 ct crystal n.a. 1996 Unidentified) Strong yellow Strong yellow, persistent n.t. Opaque metallic flux Yes YesRussian co. (Chatham)

30108 0.42 ct crystal n.a. 1996 Unidentified) Very weak orange Weak yellow n.t. Opaque metallic flux Yes YesRussian co. (Chatham)

30109 round brilliant n.a. 1996 G.E. Moderate yellow-green Very strong green n.t. Pinpoint n.t. n.a.

22007 0.27 ct round n.t. 1996 De Beers Moderate yellow-green, Moderate yellow-green n.t Opaque metallic flux n.d. n.d.brilliant zoned

aColor grades are for discussion purposes only. GIA-GTL does not grade synthetic diamonds. No grades are given for unfashioned samples.bn.t. = Not tested.cn.a. = Not applicable.dGE. = General Electric Company: Sumitomo = Sumitomo Electric Industries; De Beers = De Beers Diamond Research Laboratory.eAII but two samples were inert to long-wave UV. The two samples that fluoresced to long-wave UV were ret. nos. 21705 (weak orange, zoned) and 30096 (weak yellow).f Persistent=Phosphorescence that lasted 15 seconds or longer.gn.d. = Tested but no reaction detected.

synthetic diamonds) represent experimental, ratherthan commercial, products. Color-grade equivalentson most of the faceted samples were determined bythe staff of the GIA Gem Trade Laboratory, usingstandard grading procedures, for research purposesonly. (GIA GTL does not provide color or claritygrading services for synthetic diamonds.)

A large portion of this sample consists of the 15faceted near-colorless experimental synthetic dia-monds, produced by De Beers scientists during1995, that were loaned to GIA for examination inthe spring of 1996 (see, again, figure 1 here and alsofigure 1 in Welbourn et al., 1996, which illustratessimilar material). The largest of these samplesweighed 0.91 ct, and the color grades of this groupall fell within the "G" to "I" range.

In May 1994, Thomas Chatham loaned GIAResearch a 0.42 ct near-colorless synthetic diamondcrystal that he reported was of Russian origin (for aphoto and brief description of this crystal, seeKoivula and Kammerling, 1994, pp. 123–124). Twoyears later, in May 1996, Mr. Chatham also madeavailable to us about 100 small (as large as 0.7 ct,but most weighing 0.25 ct or less) near-colorlesssynthetic diamond crystals from Russia, which helater offered for sale at the 1996 JCK jewelry showin Las Vegas. Although these crystals appeared col-orless, for the most part they were distorted inshape, contained numerous metal inclusions, andwere very small; as such, they were not well-suitedfor jewelry. Because we had very little time toexamine these samples, we selected only four of thelarger crystals for testing.

The majority of the samples we examined forthis study would be considered near-colorless on theGIA color-grading scale.

CChhaarraacctteerriizzaattiioonn TTeecchhnniiqquueess.. To observe the color,relative intensity (described as none, very weak,weak, moderate, strong, and very strong), and distri-bution pattern, if any, of the ultraviolet fluorescencein all of the samples, we used a long-wave (366 nm)and short-wave (254 nm) GIA Gem Instruments UVlamp unit in a darkened room with contrast-controlglasses. When the UV lamp was turned off, a nota-tion was made of the color, relative intensity, andduration (by means of an electronic timer) of anyphosphorescence emitted by each sample. Wheneverpossible, we took photos of these UV luminescencereactions to help illustrate the diagnostic value ofthese features. Some of the samples acquired in1996 were examined with the De Beers Diamond

View, which was loaned to GIA in September 1996.Fluorescence images captured with the DiamondView were stored electronically.

Relative transparency to ultraviolet radiationwas observed in all samples examined since 1991 byplacing the sample between the short-wave UVlamp and a piece of synthetic scheelite (a materialthat luminesces blue to short-wave UV). The extentof the UV transparency could then be estimated byjudging the relative intensity of the scheelite's fluo-rescence.

We looked for luminescence to an electron beam(cathodoluminescence) in 21 samples with aNuclide ELM-2B luminoscope (currently manufac-tured by Premier American Technologies,Bellefonte, Pennsylvania), again noting the color anddistribution pattern, if any. Photographs of all thetypes of luminescence were captured whenever pos-sible.

We used a binocular gemological microscope toexamine all but the two earliest G.E. samples forinclusions and other features. (A Nikon SMZ-Uphotomicroscope was used to prepare photomicro-graphs of distinctive visual features for those sam-ples seen since spring 1995.)

Magnetism (caused by the presence of transitionmetal inclusions, such as iron) has been shown to bea valuable test for recognizing some synthetic dia-monds (Koivula and Fryer, 1984; Shigley et al., 1986,1987, 1993a,- Hodgkinson, 1995). We judged mag-netic attraction (none, weak, moderate, or strong)for 41 samples by observing the amount of move-ment or rotation when a rare-earth iron magnet (assuggested by H. Oates and W. Hanneman,-seeHodgkinson, 1995) was brought close to a syntheticdiamond that had been suspended in air by means ofa thin plastic thread. (We did not use this specifictest on the early G.E. samples or those that were toosmall to be suspended from the thread.)

Forty-four samples were tested for electrical con-ductivity with a standard gemological conductome-ter. We also looked for visible luminescence pro-duced by an electrical current (electroluminescence)in these 44 samples. We used a darkened room withthe sample placed between the metal probes of theconductometer.

In five of the samples examined in 1996, wedetected some interesting thermoluminescence byplacing the synthetic diamond in hot water (about140°F/60°C) after exposure to the ultraviolet lamp.Again, a subjective visual judgment was made of theemitted color.



We used either a Beck prism or a Discan digital-scanning diffraction-grating spectroscope to observethe optical absorption spectra of all of our samples.Pye-Unicam 8800 and Hitachi U-4001 spectropho-tometers were also used to record visible-rangeabsorption spectra over the wavelength range250–850 nm (with the sample held at liquid-nitrogentemperatures in an evacuated chamber) for 25 sam-ples. A Nicolet 60SX Fourier-transform infrared spec-trometer was used to record room-temperatureinfrared absorption spectra over the wavelengthrange 400–25000 cm-1 for 24 samples.

RREESSUULLTTSSWe determined that all of these synthetic diamondsamples were type Ha on the basis of a combinationof their infrared absorption spectra and their greatertransparency to ultraviolet radiation. As table 1 indi-cates, some of these were electrically conductive orhad a faint yellow or faint blue component to theircolor. In those samples for which we had infraredspectra, we typically saw evidence of boron- and/ornitrogen-related absorption features.

CChhaarraacctteerriissttiiccss ooff tthhee CCrryyssttaallss.. As with other syn-thetic diamond crystals (Shigley et al., 1986, 1987,1993a), all of the near-colorless crystals in thisstudy had a very distinct shape that consisted of aportion of a cuboctahedron with a flat base.Octahedral {111} and cube {100) faces were mostcommon in both abundance and surface area (size)on these synthetic diamond crystals, but smallerdodecahedral (110) and trapezohedral {113}—and,occasionally, {115}—faces were also seen. In general,these crystals had relatively flat faces, with sharpedges and corners between adjacent faces (figure 2).The flat base was not a true crystal face but agrowth surface, where the tiny seed crystal waslocated (figure 3).

The evidence of extensive mechanical abrasionor chemical etching that is seen on many natural(octahedral) diamond crystals was absent on our syn-thetic samples. Nevertheless, many of these syn-thetic diamond crystals had interesting surfacemarkings that, if retained after faceting, might be ofdiagnostic value. Such diagnostic markings includedendritic or striated patterns (figures 2 and 4).However, features that appear identical to the well-known trigons seen on natural diamond crystalsmay also be found on some synthetic diamonds (fig-ure 5), so it is important not to conclude that a dia-mond is natural based on the presence of these trian-gular markings.

Figure 2. The cuboctahedral form exhibited by thesetwo experimental Sumitomo synthetic diamond crys-tals (0.23 and 1.25 ct; reference nos. 21706 and21705, respectively) is typical of synthetic diamondsfrom all known manufacturers. The 1.25 ct crystalshows the larger octahedral faces and smaller cube,dodecahedral, and trapezohedral faces that are char-acteristic of this material. Note also the striationscovering the surface, which are not seen in naturaldiamonds. Photo © GIA and Tino Hammid.

UUllttrraavviioolleett FFlluuoorreesscceennccee aanndd PPhhoosspphhoorreesscceennccee.. Allbut two samples showed no fluorescence to long-wave UV radiation. No response to either long- orshort-wave UV was observed in sample no. 30099.However, all other samples did fluoresce yellow toorange-yellow or yellow-green to short-wave UV(see, e.g., figure 6). The intensity of this reactionvaried from very weak to moderate. Because thevery weak short-wave fluorescence observed insome of these synthetic diamonds could be mistak-en for no fluorescence reaction, care must be taken


Figure 3. The flat base of this 0.22 ct Starcorpsynthetic diamond crystal (reference no. 30096)shows the square imprint of the seed from whichthe crystal grew. Photomicrograph by ShaneElen; magnified 5×.

Figure 4. Magnification (12×) reveals the distinc-tive striation pattern on this Sumitomo synthet-ic diamond crystal (reference no. 21706).Photomicrograph by John I. Koivula.

De Beers DiamondView instrument is shown in fig-ure 7 (see also Welbourn et al., 1996). Note that thecolor of the UV fluorescence as seen with theDiamondView is different from that seen with a stan-dard UV lamp unit. In either case, it is the fluores-cence pattern that is of greatest diagnostic impor-tance.

In addition, when the UV lamp was turned off,most of our study samples continued to phosphorescea weak to strong yellow, yellow-green, or blue for15–60 seconds or longer. In some cases, this phospho-rescence was so intense that the glowing syntheticdiamond could be seen from a distance of several feetin a darkened room.

CCaatthhooddoolluummiinneesscceennccee.. Of the samples tested, 19(representing all sources of production) exhibited blue(figure 8), yellow, or green-yellowCathodoluminescence. This was unevenly distribut-ed in a pattern (again, often cross-shaped) that differsfrom those patterns seen in a natural diamond.

MMaaggnniiffiiccaattiioonn.. When viewed with a binocular gemo-logical microscope, the study samples revealed a fewdistinctive features. Most prominent were fluxmetalinclusions, which usually appeared opaque in trans-mitted light and metallic in reflected light. The fluxinclusions were usually elongate with rounded edges,and could be seen singly or in small groups,-some hada dendritic appearance (figure 9). The elongate metal-lic inclusions were present in most of the samples(see table 1). In general, however, they

Figure 5. The triangular, pyramid-like markings onan octahedral crystal face of this near-colorless DeBeers synthetic diamond (reference no. 21618)strongly resemble the trigons seen on natural dia-monds. Photomicrograph by John Koivula; magnified10 ×.

in making these observations. This is why fluo-rescence should be tested in a completely dark-ened room, and only when one's eyes have hadtime to adjust to those viewing conditions (atleast several minutes).

In many of our study samples, the short-waveUV fluorescence was unevenly distributed, withcertain internal growth sectors (sometimes inthe form of a black, cross-shaped pattern; again,see figure 6) exhibiting no fluorescence. Anexample of the UV fluorescence pattern obtainedwith the new


Figure 6. This 0.91 ct De Beers synthetic diamond(reference no. 30101) displays the zoned fluores-cence to short-wave UV that is commonly seen insynthetic diamonds. The black, cross-shapedareas where there is no fluorescence correspond tointernal growth sectors that lack the impuritiesresponsible for the UV fluorescence emitted bythe other growth sectors. Photo by Shane Elen.

Figure 7. The uneven pattern of fluorescence typi-cal of synthetic diamonds-here, in a cross shape-isreadily apparent in this reference photo takenwith the De Beers DiamondView verificationinstrument. Note that because two very differentexcitation sources are used, the color of fluores-cence as seen with the DiamondView is very dif-ferent from that seen with a standard UV unit(e.g., figure 6). Photo by Shane Elen.

colored sample (no. 30095) exhibited no obviousinternal color zoning. (In several instances,though, a very faint color zone was seen with mag-nification; table 1.)

seemed to be more numerous in the Russian syn-thetic diamonds. In some cases, the synthetic dia-mond was attracted by a magnet because of thesemetallic inclusions. In approximately onethird ofour samples, we saw tiny pinpoint inclusions thatprobably also were metallic flux.

In eight of the De Beers synthetic diamonds,we saw groups of thin, translucent, oriented, trian-gular inclusions of uncertain identity (figure 10).These inclusions, often associated with weakstrain (anomalous birefringence), were translucentand blue in transmitted or polarized light (figure11, left), and less translucent and reddish brown inreflected light (figure 11, right). These triangularinclusions often were apparent only in certain ori-entations of the sample (with respect to the direc-tion of the light source); otherwise, they werenearly transparent and could easily go unnoticedduring observation with a gemological microscope.Close inspection of these features revealed surfacedetails similar to the large tetrahedral stackingfaults observed using X-ray topography (as illus-trated in Field, 1979, p. 442).

We did not see intersecting graining patterns inany of the samples; these patterns are an impor-tant identification feature in colored synthetic dia-monds (Shigley et al., 1995). Even the most strong-

Figure 8. Like fluorescence, cathodoluminescencein near-colorless synthetic diamonds is also typi-cally uneven and very different from the patternsseen in natural diamonds. A faint cross shape isvisible in this 0.61 ct near-colorless De Beers syn-thetic diamond (reference no. 30100). Photo byShane Elen.


Figure 9. An unusual dendritic inclusion accom-panies the elongate inclusions in this 1.25 ctSumitomo synthetic diamond crystal (referenceno. 21705). Photomicrograph by John I. Koivula;transmitted light, magnified 20×.

BBiirreeffrriinnggeennccee.. When observed between crossedpolarizing filters with the microscope, the syn-thetic diamonds typically displayed weak anoma-lous birefringence ("strain"; see, e.g., figure 6 inShigley et al., 1993b, p. 194), as was the case formany of the colored synthetic diamonds weexamined. This weak "strain" is indicated by low-order interference colors (typically just black,gray, or white, and frequently in a cross-shapedpattern).

MMaaggnneettiissmm.. About one-half of the study samplesexhibited some attraction to a magnet. Sampleswith more metallic inclusions seemed to displaya stronger attraction.

EElleeccttrriiccaall CCoonndduuccttiivviittyy.. In 20 of the 45 samplestested, we noted some electrical conductivitywith a strength that varied depending on thepoint on the sample that we tested with the con-ductometer probes. For the crystals, this electri-cal conductivity could be seen to vary dependingon which pair of crystal faces were selected fortesting with the probes (although it was not pos-sible to determine exactly which faces were con-ductive because of their small size and the diffi-culty of ensuring that we were touching just oneface with the probe).

EElleeccttrroolluummiinneesscceennccee.. Of the samples thatshowed some electrical conductivity, six also dis

played aninteresting blue luminescence (usually spo-radic and weak) when the sample was touched withthe conductometer probes (figure 12). To ourobservers, this luminescence did not appear to beemitted by the entire sample. Rather, it appeared tobe localized, and to extend between the points onthe sample touched by the two probes (somewhatlike a flash of lightning). Because this reaction issimilar to that seen in many natural diamonds, thistest usually is not useful to identify the syntheticstones. However, one Sumitomo crystal displayedboth blue and red electroluminescence, dependingon where it was touched with the probes; when theprobes were removed, the diamond continued tophosphoresce blue for several seconds.

As was the case with electrical conductivity, wecould not relate the visible electroluminescence inthe unfashioned samples to the particular crystalfaces being touched by the conductometer probes(although the impurities causing the conductivitywere presumed to be distributed unevenly betweenthe internal growth sectors of the synthetic dia-mond).

TThheerrmmoolluummiinneesscceennccee.. When immersed in hot waterafter exposure to UV radiation, all five samples test-ed were seen to emit luminescence (which contin-ued briefly when the still-warm sample was removedfrom the water). Four emitted blue and one (refer-ence no. 21705) emitted orange. We have notobserved this reaction in natural diamonds.

Figure 10. Groups of tiny, translucent, oriented, tri-angular inclusions of uncertain identity can be seenin this 0.61 ct De Beers synthetic diamond (researchno. 30100). Photomicrograph by Shane Eleri; unpo-larized reflected light, magnified 15×.


Figure 11. In polarizedtransmitted light (left), thetriangular growth featuresappeared translucent blueand often had localizedstrain. In unpolarizedreflected light (right), theyappeared less translucentand reddish brown. Sampleno. 21967 (De Beers); pho-tomicrographs by ShaneElen, magnified 12.5×.

AAbbssoorrppttiioonn SSppeeccttrruumm.. The vast majority of the syn-thetic diamonds in this study did not reveal anysignificant absorption bands with either the spec-trophotometer or the handheld spectroscope.(Although a well-defined band at 270 nm has beenidentified in near-colorless synthetic diamonds [C.Welbourn, pers. comm., 1997], it was not resolv-able with our instruments.) However, the two G.E.synthetic near-colorless diamonds we examined in1984 displayed a very weak band at about 732 nmin the spectra recorded with the Pye-Unicam spec-trophotometer (a feature discussed by Lawson andKanda, 1993). Since we have not seen this band insynthetic diamonds produced more recently, we donot consider it to be significant.

Although the infrared spectra allowed us toestablish that all of our samples were type Ila dia-monds (and, as mentioned earlier, that some had

weak boron- and/or nitrogen-related absorption fea-tures), they contained no features by which wecould separate a natural type Ila diamond from asynthetic type IIa stone.

DDIISSCCUUSSSSIIOONNNatural near-colorless diamonds are usually type Ia;those that contain little or no nitrogen are referredto as type IIa (for a brief discussion of diamondtypes, see Box A in Fritsch and Scarratt, 1992, pp.38–39). The synthetic diamonds described here areall type IIa, although some have a small type lib ortype IaB component. To date, we have not seen atype la near-colorless gem synthetic diamond; nor,to our knowledge, has one ever been reported. Near-colorless type IIa diamonds have a greater degree oftransparency from the blue end of the visible spec-trum into the near-ultraviolet region than do near-colorless type la diamonds.

In addition, type IIa diamonds do not display thenitrogen-related sharp absorption peaks in theirabsorption spectra ("Cape lines," with the strongestpeak at 415 nm and additional peaks between 415and 478 nm) that are seen in type la diamonds. Thisis the principal behind the new De BeersDiamondSure instrument (Welbourn et al., 1996).Once it is established that the diamond in questionis a type IIa, other tests such as UV fluorescence orcathodoluminescence can be used to determinewhether it is natural or synthetic.

Natural diamond crystals are typically in theshape of an octahedron or dodecahedron, wheregrowth has taken place outward in all directionsfrom a central core to give an equant crystal (see fig-ure 7 in Welbourn et al., 1996 p. 163). Many naturaldiamond crystals have rounded surfaces that are dueto chemical dissolution (etching) of the diamondswhile they were still in the Earth, or to mechanicalabrasion during transport from the host rock in astream or river. Synthetic diamonds have a very dif-ferent crystal morphology (again, see figure 5 in the

Figuie 12. Blue electroluminescence was emittedby some of the samples when they were touchedwith the conductometer probes. Sample no. 30096(Starcorp); photomicrograph by Shane Elen, magni-fied 2.5×.


article by Welbourn et al., 1996, p. 162). Unlike nat-ural diamonds, growth only takes place outward andupward from the seed location at the flat base.

In his description of the early G.E. synthetic dia-monds, Crowningshield (1971) reported that the twonear-colorless samples had an unusual reaction toultraviolet radiation. They did not fluoresce to long-wave UV, but did fluoresce to short-wave UV, witha yellow color that continued, in his words, "for along time" even after the UV lamp had been turnedoff (phosphorescence). He mentioned how differentthis reaction was to that of natural near-colorlessdiamonds, and how observation of both fluorescenceand phosphorescence to short-wave UV would beimperative for synthetic diamond verification.Thus, at this early date, he established what hasbecome a practical means by which synthetic dia-monds can be recognized by the gemologist.

Observations of the fluorescent and phosphores-cent reactions to UV radiation among our studysamples confirm Crowningshield's results.Specifically, strong short-wave UV fluorescence (rel-ative to long-wave UV) and, in some cases, phospho-rescence are very distinctive of synthetic diamonds(Shigley et al., 1995). Almost all the near-colorlesssynthetic diamonds we have examined to dateexhibited these phenomena. Weak or absent long-wave UV fluorescence, and the presence of strongershort-wave UV fluorescence and phosphorescence,do not immediately prove that a "colorless" dia-mond is a synthetic. However, any diamond thatdisplays these reactions should be considered sus-pect and should be examined by other gem-testingmethods. The typical form of zoned fluorescenceseen in synthetic diamonds is not found in naturaldiamonds. When present, fluorescence in a naturalnear-colorless diamond is typically blue (rarely, it isyellow) to both long- and short-wave UV radiation,with the reaction almost always being more intenseto long-wave UV.

Cathodoluminescence is an important additionalmeans of identifying a diamond as being synthetic,since the pattern of luminescence from the differentinternal growth sectors is even more visible withthis technique than with conventional observationof UV fluorescence (Ponahlo, 1992; Shigley et al.,1995). This equipment has become more standard atgem-testing laboratories, although the availability ofthe new De Beers DiamondView may make theneed for cathodoluminescence instrumentation less-critical.

Strong magnetism in a diamond suggests that itis synthetic. However, because some synthetic dia-monds (including a number examined during thisstudy) contain few if any metallic inclusions, mag-netism is not an effective test for all stones.Our latest findings, reported here, support previousindications that the most diagnostic gemologicalfeatures for distinguishing near-colorless type Hasynthetic diamonds (so far, from any source) are stillthose summarized by Shigley et al. (1995):

The cuboctahedral crystal shape, and the possibility of striations and other unusual patterns onthe surfaces of the crystal faces. Metallic inclusions, which seem to be relativelyabundant in the near-colorless Russian-grownsynthetic diamonds we have examined so far.Attraction of the synthetic diamond to a magnet.The short-wave ultraviolet fluorescence andphosphorescence.The greater UV transparency, which is indicativeof a type IIa diamond (most natural near-colordiamonds are type la).Weak or absent anomalous birefringence("strain").Electrical conductivity.In contrast to the situation with colored synthet-

ic diamonds, our near-colorless synthetic sampleslacked color zoning, graining patterns, and, in mostcases, distinctive absorption bands in the visible andinfrared spectra. Thus, these features are of little ifany diagnostic value for the separation of naturalfrom synthetic near-colorless diamonds.

CCOONNCCLLUUSSIIOONNA few polished synthetic diamonds have been seenin the jewelry industry, but we know of only oneconfirmed instance of a near-colorless sample ofunknown origin appearing in the trade withoutbeing represented as synthetic (the 0.16 ct roundbrilliant mentioned earlier). Because of technicalchallenges and the high cost of production, we ques-tion the likelihood that fashioned near-colorlesssynthetic diamonds over 25 points will enter thejewelry industry in commercial quantities. In ouropinion, the greatest possibility is the availability ofnear-colorless melee, because at these small sizessynthetic diamonds can be grown relatively fast andefficiently.


The group of near-colorless type IIa syntheticdiamonds described in this article is the largestknown to have been examined by standard andadvanced gem-testing techniques for the expresspurpose of establishing practical identification cri-teria. The most useful identification clues for syn-thetic diamond crystals are the crystal shape andsurface features. For a polished stone, key identifi-cation features include metallic inclusions andthe related possibility of attraction to a magnet,the possibility of electrical conductivity, and thecharacteristic short-wave UV fluorescence andphosphorescence. These gemological propertiesappear to be diagnostic of near-colorless type IIasynthetic diamonds, and are very different fromnear-colorless natural type la and type IIa dia-monds.

Diamond dealers, jewelers, and gemologistsmust be prepared to handle the identification ofjewelry-quality near-colorless synthetic diamondsif they become more widely available. Testing willrequire a more careful gemological examination of

near-colorless diamonds. The synthetic nature of adiamond must be disclosed at the time of sale orappraisal. The GIA Gem Trade Laboratory will con-tinue its policy of issuing only identification reports,reports, on synthetic diamonds.

Acknowledgments: The following individuals and orga-nizations provided synthetic diamonds and, in somecases, information for this study: Rick D 'Angela andDr. William Banholzer of General ElectricSuperabrasives, Worthington, Ohio; Dr. ThomasAnthony of General Electric Research &Development,Schenectady, New York; Dr. Shuji Yazu and his col-leagues at Sumitomo Electric Industries, Itami, Japan;Thomas Chatham of Chatham Created Gems, SanFrancisco, California; and Marion Matthews ofStarcorp, Goleta, California. The largest group of sam-ples in this study were produced by Dr. Robert Burnsand his colleagues at the De Beers Diamond ResearchLaboratory in Johannesburg, South Africa; they weremade available through Martin Cooper of the DiamondTrading Company (DTC) Research Centre inMaidenhead, United Kingdom.

RREEFFEERREENNCCEESSAnthony T.R., Banholzer W.F., Fleischer J.F., Wei L, Kuo P.K.,

Thomas R.L., Prycr R.W. (1990) Thermal diffusivity of iso-topically enriched 12C diamond. Physical Review B(Condensed Matter), Vol. 42, Third Series, No. 2, pp. 1104-1111.

Chatham to sell "created" diamonds (1993). New YorkDiamonds, No. 22, Autumn, pp. 44, 46.

Crowningshield R. (1971) General Electric's cuttable syntheticdiamonds. Gems & Gemology, Vol. 13, No. 10, pp. 302-314.

Field J.E. (1979) Properties of Diamond. Academic Press, NewYork.

Fritsch E., Scarratt K. (1992) Natural-color nonconductivegray-to-blue diamonds. Gems & Gemology, Vol. 28, No. 1,pp. 35–42.

Hobbs J. (1981) A simple approach to detecting diamond simu-lants. Gems & Gemology, Vol. 17, No. 1, pp. 20–33.

Hodgkinson A. (1995) Magnetic wand-Synthetic gem diamonddetector. Rapaport Diamond Report, May 5, pp. 34-35.

Koivula J.I., Fryer C.W. (1984) Identifying gem-quality synthet-ic diamonds: An update. Gems & Gemology, Vol. 20, No. 3,pp. 146–158.

Koivula J.I., Kammerling R.C. (1994) Gem news: Near-color-less Russian synthetic diamond examined. Gems &Gemology, Vol. 30, No. 2, pp. 123–124.

Lawson S.C., Kanda H. (1993) An annealing study of nickelpoint defects in high-pressure synthetic diamond. Journal ofApplied Physics, Vol. 73, No. 8, pp. 3967–3973.

Nassau K. (1980) Gems Made by Man, Chilton Book Co.,Radnor, PA.

Nassau K. (1993) Are synthetic diamonds from Russia athreat? Rapaport Repoit, November 5, pp. 29–32.

Ponahlo J. (1992) Cathodoluminescence (CL) and CL spectra of

De Beers' experimental synthetic diamonds. Journal ofGemmology, Vol. 23, No. 1, pp. 3–17. Rooney M-L.T., Welboum CM., Shigley J.E., Fritsch E., Reinitz

I. (1993) De Beers near colorless-to-blue experimental gem-quality synthetic diamonds. Gems & Gemology, Vol. 29,No. 1, pp. 38–45.

Scarratt K., Du Toit G., Sersen W. (1994) Russian syntheticsexamined. Diamond International, No. 28 (March/April), pp.45–52.

Shigley J.E., Fritsch E., Stockton CM., Koivula J.I., Fryer C.W.,Kane R.E. (1986) The gemological properties of theSumitomo gem-quality synthetic diamonds. Gems &Gemology, Vol. 22, No. 4, pp. 192–208.

Shigley J.E., Fritsch E., Stockton CM., Koivula J.I., Fryer C.W.,Kane R.E., Hargett D.R., Welch C.W. (1987) The gemologicalproperties of the De Beers gem-quality synthetic diamonds.Gems & Gemology, Vol. 23, No. 4, pp. 187–206.

Shigley J.E., Fritsch E., Reinitz I., Moon M. (1992) An updateon Sumitomo gem-quality synthetic diamonds. Gems &Gemology, Vol. 28, No. 2, pp. 116–122.

Shigley J.E., Fritsch E., Koivula J.I., Sobolev N.V., MalinovskyI.Y., Pal'yanov Y.N. (1993a) The gemological properties ofRussian gem-quality synthetic yellow diamonds. Gems &Gemology, Vol. 29, No. 4, pp. 228–248.

Shigley J.E., Fritsch E., Reinitz I. (1993b) Two near-colorlessGeneral Electric type IIa synthetic diamond crystals. Gems& Gemology, Vol. 29, No. 3, pp. 191–197.

Shigley I.E., Fritsch E., Reinitz I, Moses T.M. (1995) A chart forthe separation of natural and synthetic diamonds. Gems &Gemology, Vol. 31, No. 4, pp. 256–264.

Welboum CM., Cooper M., Spear P.M. (1996) De Beers naturalversus synthetic diamond verification instruments. Gems &Gemology, Vol. 32, No. 3, pp. 156–169.

54 Gem Trade Lab Notes GEMS & GEMOLOGY Spring 1997

EEddiittoorrC. W. Fryer, GIA Gem TradeLaboratory

CCoonnttrriibbuuttiinngg EEddiittoorrssGIA Gem Trade Laboratory, EastCoast G. Robert CrowningshieldThomas Moses llene ReinitzGIA Gem Trade Laboratory, WestCoast Karin Hurwit Mary L.Johnson Shane F. McClure

Cheryl Y. Wentzell

DDIIAAMMOONNDD

TTwwoo NNootteewwoorrtthhyy SSttoonneessffrroomm tthhee AAmmeerriiccaassThe October 11, 1996, RapaportDiamond Report included anarticle on the purchase of a 28.3ct rough diamond, recentlyrecovered from the Kelsey Lakemine in Colorado (see alsoWinter 1996 Gem News, pp.282–283). The highly etchedbrownish yellow crystal yielded a5.39 ct pear-shaped brilliant (fig-ure 1). A rather deep, etchedfeather was seen in the pavilion

of the finished stone (figure 2).Althoughthis stone was submit-ted to the laboratory for anIdentification and Origin of ColorReport only, we would have grad-ed the clarity in the SI range; thecolor was graded as Fancy DeepBrownish Yellow.

Fluorescence to long-waveultraviolet radiation was strongin intensity and predominantlyblue with small zones of strongyellow. The diamond fluoresceda weaker yellow to shortwaveUV. The absorption spectrum,seen at low temperature using

Figure 2. An etched feather can beseen in the pavilion of the KelseyLake diamond shown in figure 1.Magnified 57×.

Figure 1. This 5.39 ct Fancy Deep Brownish Yellow diamond was cutfrom a 28.3 ct piece of rough recovered from the Kelsey Lake mine inColorado.

a desk-model spectroscope, showed astrong "Cape series" and weak sharpbands at about 545 and 563 nm.These properties are consistent withhydrogen-rich diamonds (see E.Fritsch and K. Scarratt, Journal ofGemmology, "GemmologicalProperties of Type la Diamonds withan Unusually High HydrogenContent," Vol. 23, No. 8, 1993, pp.451–60). Although the distinctiveproperties of this class of diamondwere observed decades ago, hydro-gen's role in causing the color wasonly recognized in the last five years.These hydrogen-rich diamonds havealso occurred in other recently

Editor's note: The initials at the end of each itemidentify the contributing editor(s) who providedthat item.Gems & Gemology, Vol. 33, No. 1, pp. 54-59© 1997 Gemological Institute of America

Gem Trade Lab Notes GEMS & GEMOLOGY Spring 1997 55

developed sources, such as inAustralia, Russia, and the Jwanengmine in Botswana.

Only a few days after we sawthe Kelsey Lake stone, a long-timeGTL client submitted a 2.15 ct old-minecut diamond (figure 3) thatwas reportedly from Brazil. GradedFancy Intense Greenish Yellow,this stone revealed gemologicalproperties common to stones withstrong "green transmission" lumi-nescence. The H3 optical centerthat is associated with this effectgives rise to absorption bands atabout 494 and 503 nm. Diamondswith this optical center not onlyroutinely fluoresce strongly to UVradiation, but they also are excitedby visible light—often appearingquite green when exposed to astrong incandescent light source(figure 4). This strong lumines-cence may influence the colorgrade of such diamonds, as it didwith this one. In our experience,some diamonds with these proper-ties do originate from Brazil andVenezuela, thus lending support tothis stone's reported provenance(see also J. F. Cottrant and G.Calas, "Étude de la coloration dequelques diamants du MuséumNational d'Histoire Naturelle,"Revue de Gemmologie, No. 67,1981, pp. 2–8).

According to our client, thestone was reportedly given byPedro II (emperor of Brazil from1831 to 1889) to his niece. A mem-ber of the Braganza royal family,

which ruled Portugal from 1640until 1910, Pedro II was purported-ly greatly interested in diamondsand mineral specimens.

The submission of diamonds ofknown or well substantiated prove-nance provides the laboratory withthe opportunity not only to docu-ment the properties of such stones,but also to help answer ongoingquestions related to the origin ofcolor in diamonds. TM

DDaammaaggeedd ffrroomm WWeeaarrWhen appropriate, GIA GTL dia-mond grading reports on roundbrilliants contain a comment thatthe average crown angles are lessthan 30° or exceed 35°. These com-ments are intended to alert thereader of the report that the crownangles deviate seriously from

industry norms and, in the case ofshallow angles, warn of a threat tothe stone's durability.

The East Coast lab recentlyissued a report on just such astone—a 2.66 ct yellow diamondthat had been removed from afour-prong ring setting. As can beseen in figure 5, the unusuallythin girdle was chipped in manyplaces. When the stone wasviewed from the side (figure 6),the crown angles appeared to bevery shallow. Just how shallowwas seen when the stone wasviewed in a GIA ProportionScope (figure 7): The angles aver-aged less than 25°. In addition,the pavilion anglewere less than

Figure 5. Many chips can be seenon the girdle of this 2.66 ctround-brilliant-cut diamond,which was originally set in afour-prong lady's ring.

Figure 3. This 2.15 ct old-mine-cut diamond was graded FancyIntense Greenish Yellow.

Figure 6. Damage to the dia-mond in figure 5 is related to thecombination of a thin girdle andshallow crown angles, as can beseen in this profile view.

Figure 4. The green transmissionluminescence in the diamond infigure 3 is best seen when thestone is excited with a strongincandescent light source, suchas fiber-optic illumination.

Figure 7. A GIA ProportionScope revealed that the dia-mond in figures 5 and 6 hadcrown angles of about 25°.


41°. This combination of cuttingfaults usually leads to damage dur-ing normal wear. The potential fordamage to such a stone can be min-imized by choosing mountings,such as gypsy or bezel settings, thatprotect the girdle area.

GRC and TM

IIddeennttiiffyyiinngg FFiilllleedd FFrraaccttuurreess::NNeeww CChhaalllleennggeess

Over the last six months, the EastCoast lab examined four diamondswith unusual-appearing fractures.These stones gave us an advancedlesson in distinguishing filled frac-tures from unusual-appearing frac-tures that are unfilled, a lesson thatrequired checking for all the fea-tures summarized by McClure andKammerling in "A Visual Guide tothe Identification of FilledDiamonds" (Summer 1995 Gems &Gemology, pp.114–119, plus chart).

A 1.05 ct oval modified brilliantshowed muted, subtle flash-effectcolors (figure 8), similar to thoseobserved in filled diamonds thathad been heated (see Kammerlinget al, "An Update on FilledDiamonds: Identification andDurability," Fall 1994 Gems &Gemology, pp. 142–177, especiallyfigure 27). This prompted us toexamine the fracture closely athigher magnification, whichrevealed a dendritic pattern typicalof filled fractures, together withflow structure and cloudiness inthe fracture plane (figure 9).

A contrasting unfilled example

was provided by a cleavage in a 1.23ct round brilliant (figure 10).Although large in area, this breakwas nearly invisible from mostangles. Higher magnification, witha variety of lighting techniques,revealed neither flash colors norany internal structure—all of whichindicated that the cleavage was notfilled. An X-radiograph confirmedthis conclusion. The entire stone,including the fracture, was trans-parent to X-rays, whereas fracture-filling material would have beenopaque.

An unusual inclusion that

proved to be fracture-filled wasseen in the 1.01 ct round brilliantshown in figure 11. A laser drillhole led to a crystal inclusion cen-tered in a large fracture. The wallsof the drill hole showed a roughtexture, as did the fracture at the point of drill-hole contact. At thepoint of contact, the surface of thefracture appeared burned. The restof the fracture was smoother andshowed flow structures. Weobserved flash colors only in theareas of the fracture that surround-ed the crystal; these ranged fromorange to pink and blue.

The complicated group of frac-tures in figure 12 illustrate theimportance of viewing angle inidentifying filled fractures. Frommost sides, this 1.48 ct marquiseshowed both the characteristicflash colors of filled fractures andthe natural iridescence of unfilledfractures. The profile view of thisdiamond (figure 13) showed oneunfilled fracture that displayed awhole rainbow of interference col-ors, and another fracture, nearlyperpendicular to the first, with ablue flash color. This diamondillustrates the importance of nothalting the examination of a stonebefore it is viewed from all sides, asnot all fractures in a filled stonecan or will absorb the filling. The

Figure 8. Only a muted pink flashcolor could be seen in this filledfracture in a 1.05 ct oval dia-mond. Magnified 30 ×.

Figure 10. This large cleavage ina 1.23 ct round brilliant dia-mond stretches from the pavil-ion to the table. Although notfilled, the cleavage was only vis-ible from a few viewing angles.Magnified 15×.

Figure 11. Flash colors and flowstructure characteristic of frac-ture filling are visiable in partsof this complex inclusion in a1.01 ct round brilliant diamond.Magnified 45×.

Figure 9. At higher (63 ×) magni-fication, the dendritic pattern ofthe flow structure and cloudi-ness of the fracture shown in fig-ure 8 con firmed that it wasfilled.


many fractures and their reflec-tions in this diamond made thepresence of fracture filling unusu-ally difficult to identify.

IR and Vincent CraccoIImmiittaattiioonn From time to time, mineral speci-mens—loose crystals or pieces ofrough—are submitted to GIA GTLfor identification. Transparent-to-translucent, near-colorless speci-mens arc usually submitted in thehope that they will turn out to bediamond. Sometimes these speci-mens have been doctored to resem-ble diamond crystals. Althoughoccasionally minerals other thandiamond are used for this purpose,the most common imitation dia-mond rough that we see is fash-ioned from cubic zirconia (see, forinstance, Lab Notes, Winter 1988,pp. 241–242, and Fall 1996, p. 205;Gem News, Spring 1994, p. 47).

Standard gemological testingalone was sufficient to identify aspecimen of near-colorless roughthat was submitted to the EastCoast lab last winter. The roughhad triangular engravings, resem-bling trigons, on part of its surface(figure 14). The specific gravity wasalso approximately that of dia-mond—and very different fromthat of CZ. However, one flat

(cleaved or polished) surface gaverefractive indices of about1.61–1.63. The fact that the materi-al was doubly refractive meant thatit clearly could not be either dia-mond or CZ. With magnification,we observed fluid inclusions,which also typically are notencountered in diamond.

This combination of propertiesidentified the rough as topaz. Themarkings were probably added to

simulate "trigons," to better con-vince the unwitting buyer that thiswas a diamond. Although the termis a misnomer and rarely usedtoday, colorless topaz was oncecalled "slave's diamond" because ofits resemblance to diamond (see M.Bauer, Precious Stones, Charles E.Tuttle & Co., Rutland, Vermont,1969, p. 330).

Nicholas DelRe

HHOORRNNBBIILLLL ""IIVVOORRYY""

An elaborate yellow metal brooch(figure 15) was sent to our EastCoast lab for identification of itslarge center carving. Accented bytwo large pearls, this semi-translu-cent red and yellow portrait ofBuddha measured about 70 × 47 ×12.5 mm.

Only very limited gemologicaltests could be performed becauseof the mounting. However, we didget a vague spot refractive indexreading, in the low 1.50s, from asmall, fairly well-polished area onone side. The carving fluoresced afaint bluish white to short-waveultraviolet radiation, which is sim-ilar to the response of some den-tine ivories as well as of hornbill"ivory," a nondentine material thatis among the rarest of gem rriateri-als.

Observation with a microscopeat about 10× magnificationrevealed numerous thin, parallelfibers in the yellow part of thecarving, with a few small dark-pig-mented granules scatteredthroughout. These structural char-acteristics are typical of an organicmaterial such as ivory. The thinlayer around the main carving wasuniformly red, with no internalfeatures apparent. The transitionfrom the red to yellow areas wasgradual, with no obvious demarca-tion separating the different colors.These characteristics proved thatthe carving was a single piece andnot assembled. We concluded fromthese properties, especially the dis-tinctive red and yellow coloringand the structure seen with magni-fication, that this carving was fash-ioned from hornbill "ivory."

Figure 12. Some of the fracturesin this 1.48 ct marquise arefilled and others are unfilled.

Figure 13. In profile, the stone infigure 12 can be seen to containnaturally iridescent unfilledfractures (to the right of thetweezer) as well as a blue flasheffect (visible throughout theregion to the left) from reflec-tions of filled fractures.Magnified 20×.

Figure 14. Trigons were engravedon this 17.56 ct colorless topazrough to make it look like a dia-mond.

Hornbill "ivory" is obtainedfrom the casque (the large protu-berance covering a portion of thehead and bill) of the helmetedhornbill, an exotic bird (Rhinoplaxvigil) native to southeast Asia andBorneo. (Because this material isnot fashioned from teeth or toothmodifications, such as tusks, it isnot considered a true

ivory.) Comprehensive articleshave appeared in Gems &Gemology (A. S. Munsuri,"Hornbill Ho-Ting," Fall 1973, pp.208–211; R. E. Kane, "HornbillIvory," Summer 1981, pp. 96–97)and in the Journal of Gemmology(G. Brown and A. J. Moule,"Hornbill Ivory," January 1982, pp.8-19). The distinctive bright redborder of the yellow helmet makesthis material easy to identifydespite its rarity. KH

MMOOGGUULL TTAALLIISSMMAANNSS

An interesting set of three largebaroque beads, which reportedlyonce belonged to the Mogul emper-or Akbar Shah (1542–1605), were

submitted to the East Coast labora-tory for testing. The set (figure 16)consisted of a natural pink spinel, anatural blue sapphire, and an aqua-marine. Standard gemological test-ing readily identified the materialsfrom which each bead was fash-ioned.

The spinel's gemological prop-erties were within normal pub-lished values. Mineral inclusionsestablished that the stone was nat-ural. It fluoresced strong red tolong-wave UV radiation and mod-erate red to short-wave UV. Asexpected from the fluorescence, thestone showed several sharp "organpipe" lines beyond 650 nm in thered end of the spectrum, which aredue to the presence of chromium.The sapphire fluoresced weak redto long-wave UV. It had a moder-ately strong 450 nm line with weak"chrome lines" in its absorptionspectrum. Long, fine needles, pris-tine crystals, and fluid-filled "fin-gerprint" inclusions were visiblewith magnification, indicating thatthe stone had not been heat treat-ed.

The aquamarine had a "spot"refractive index of 1.57 and a ratherstrong iron line at 427 nm. Themost distinctive feature was itscolor. Before the now-common


Figure 15. This portrait ofBuddha is carved from rarehornbill "ivory."

Figure 16. A set of beads, which reportedly once belonged to a Mogulemperor, consists of a 114.38 ct spinel, a 106.57 ct sapphire, and an87.72 ct aquamarine

Figure 17. The spinel bead infigure 16 is inscribed with thenames "Akbar Shah" andJehangir," as well as with the(Moslem calendar) date "971."

Editor's Note: The helmetedhornbill has been declaredendangered throughout theregion where it lives. Commercein recent hornbill "ivory" is nowillegal in most countries,including the United States.However, hornbill-" ivory"pieces 100 years old or oldermay be legal to own andimport, provided they have notbeen repaired recently.

practice of heat treating aqua-marines to make them blue, thegreen color was actually consideredmore desirable (see, e.g., K. Nassau'sGemstone Enhancement, 1st ed.,Butterworth, Oxford, England, 1984,p. 98). The aquamarine bead is thisgreen color.

Because it is often difficult tovalidate historical information, wewere unable to verify the age(s) ofthe beads. However, careful observa-tion can sometimes provide indica-tions of authenticity. The greencolor of the aquamarine bead is onesuch indication. In addition, all thebeads had crude conical drill holes,which indicates that they weredrilled with hand tools and not bymachine. Further, the names AkbarShah and Jehangir (his son), togetherwith the date 971 (on the Moslemcalendar, which is 1565 on theGregorian calendar) are inscribed onthe spinel (figure 17). These inscrip-tions are in keeping with a traditionof engraving gemstones with thenames of rulers in India. Aninscribed Mogul spinel from thetime of Jehangir is in the collectionof the Victoria & Albert Museum inLondon (Benjamin Zucker, pers.comm., 1996).

TM and GRC

OOPPAALL,, wwiitthh TTrruuee CChhaattooyyaannccyy

In 1983, a piece of greenish yellowbanded rough—identified as opal—was donated to the GIA collection.Later that year, we were shown a 1.5ct brownish orange chatoyant stonethat was purportedly cut from

one of the bands of this type ofmaterial (see Spring 1983 LabNotes, pp. 45–46). Although thiscabochon was also identified asopal , i t was very d i f fe rent incolor from the rough specimen inour col lect ion. The Fal l 1990Gem News section (pp. 232–233)reported on a 0.76 ct opal thathad a sharp "eye" and showed anexce l lent "mi lk and honey"effect. Because such chatoyancyis relatively rare in opal, the lab-oratory was not able at that timeto determine the reason for thisphenomenon. (A different type ofchatoyancy—and even asterism—is seen in opals with chatoyantplay-of -color , including somestones from Jalisco, Mexico [see,e.g., Winter 1990 Gem News, p.304], and especially Idaho.)

Earlier this year, staff mem-bers a t the East Coast l abremembered the greenish bandedrough while identifying a 4.86 ctcat's-eye opal. When this cabo-chon and the rough piece wereexamined side by side (figure 18),the similarity in color of the twospecimens was evident; now itwas easier to believe that theprevious examples had been cutfrom similar rough. This materi-al is unusual in that it showschatoyancy caused by l inearinclusions—that is, "true" cha-toyancy. Such inclus ions areuncommon in opal, which is pri-mar i ly composed o f smal lspheres of silica.

TM and GRC

QQUUAARRTTZZ,,CCaatt''ss--eeyyee EEffffeecctt CCaauusseedd bbyy LLaarrggeeRRuuttiillee NNeeeeddlleess

The cat's-eye phenomenon in gem-stones is usually caused by either oftwo forms of fine parallel inclusions.One form involves hollow tubes, asin cat's-eye emeralds and tourma-lines (see, for instance, Winter 1990Gem News, pp. 306–307). The otherinvolves needle-like inclusions, as inthe cat's-eye apatite illustrated inthe Fall 1995 Gem News (pp.205–206). In many samples of cat's-eye and star quartz, the inclusionsthat cause this phenomenon are veryfine rutile needles. This was not thecase, however, with a 2.92 ct cabo-chon submitted to the West Coastlaboratory in summer 1996: Therutile inclusions responsible for thecat's-eye effect were large andprominent (figure 19). In fact, hadthis stone not demonstrated chatoy-ancy, we would have simply calledit rutilated quartz. However,because this stone fit the criteria forboth cat's-eye quartz and rutilatedquartz, our conclusion stated that itwas cat's-eye rutilated quartz.

MLJ and SFM


Figure 19. Chatoyancy in this 2.92 ct quartz cabo-chon is caused by reflection of light from the par-allel array of coarse rutile needles.

PPHHOOTTOO CCRREEDDIITTSSNicholas DelRe supplied the pictures used infigures 1,2, 14, 15, and 18. Figures 3 4, and19 were taken by Maha DeMaggio. The pho-tos in figures 5–13 were taken by VincentCracco. Figures 16 and 17 are courtesy ofDr. Gary Hansen, St. Louis, Missouri.

Figure 18. This 4.86 ct cat's-eye opal cabochonwas cut from a band similar to those in theaccompanying 8.59 ct piece of rough

60 Gem News GEMS & GEMOLOGY Spring 1997

EEddiittoorrss Mary L. Johnson and John I. Koivula

CCoonnttrriibbuuttiinngg EEddiittoorrssDino DeGhionno and Shane F. McClure,

GIA GTL, Carlsbad, CaliforniaEmmanuel Fritsch, University of Nantes, France Henry A. Hänni, SSEF, Basel, Switzerland

TUCSON ‘97

For the colored stone market, the new year begins noton January 1, but in late January to early February, atthe many shows in Tucson, Arizona. The Gem Newsstaff visited 17 of the 23 "official" shows—of gems, min-erals, beads, and finished goods. We also combed theblocks-long bazaar of open-air booths and tents that par-alleled the Interstate 10 highway.

This year, a one-week gap separated two of thebiggest shows, the American Gem Trade Association(AGTA) show and the retail segment of the TucsonGem and Mineral Society (TGMS) show, both held atthe Tucson Convention Center, so the overall "Tucsonexperience" lasted just over three weeks (from January26 through February 16). Several of the more interestingmaterials and individual stones that we uncovered arediscussed here, and more will be profiled throughoutthe year. Special thanks go to GIA Gem TradeLaboratory gemologists Maha DeMaggio (who also pho-tographed many of the specimens), Cheryl Wentzell,Nick DelRe, and Phillip Owens for their tireless quest-ing after the new, the unusual, and the downright odd.

DDIIAAMMOONNDDSS

""OOppaalleesscceenntt"" aanndd ootthheerr uunnuussuuaall ddiiaammoonnddss.. Althoughthe Tucson shows primarily showcase colored stones,some interesting diamonds were also available, includ-ing black diamonds, treated-color green and blue dia-monds, and naturally "colored" (by cloud-like inclu-sions) "white" diamonds. These "white" diamonds aresometimes called "opalescent" because flashes of spec-tral colors, caused by dispersion from the back facets,resemble play-of-color when viewed through theirmilky white body color. One such diamond is shown infigure 1: In the table-down position, it has a "J" colorgrade and is faintly brown. Similar diamonds were alsodiscussed in the Spring 1992 Gem News section (p. 58).

Diamond crystals were prominent at the TGMSshow. David New, of Anacortes, Washington, showeddiamond crystals from many localities, including five

from the new Kelsey Lake mine in Colorado and a"Star-of-David" twinned made from South Africa. (Asimilar crystal is described and illustrated on pp.117–118 of the Summer 1991 Gem News section).

DDiiaammoonndd ""ppeeaarrllss.."" For some time, one editor (MLJ) hasbeen intrigued by the idea of a cabochon diamond. Thisyear, at the booth of Crystal Classics, Okehampton,Devon, England, she found the next best thing: dia-mond "pearls" (figure 2). Dr. Heinz Malzahn, of Berlin,Germany, has adapted the technique for making roundpearlescent diamonds from one used by De Beersaround 1970 to process industrial diamonds. Accordingto Dr. Malzahn, rough diamonds with roundish shapesare ground to rough spheres and then "'cooked" in sodi-um carbonate, in an inert atmosphere at about 800°C,to produce the pearly surface. This process is some-times referred to as chemical polishing by selective

Figure 1. This 0.03 ct "opalescent" diamond wasamong several marketed by Malhotra Inc., NewYork City. Photo by Maha DeMaggio.

GEM N.E.W.S

Gem News GEMS & GEMOLOGY Spring 1997 61

dissolution: Comer sites on the surface of the diamondsetch faster than edge sites, and edges dissolve fasterthan flat or curved surfaces, resulting in a smoothly pol-ished round stone. Because the process is destructive,we have only seen the results of treating "industrial"(non-gem) diamond crystals.

Custodiam, a Brussels firm, is also marketing "dia-mond pearls," according to the December 1996 issue ofAntwerp Facets (pp. 37, 39). They also grind rough dia-monds into spheres, which they then "polish" by chemi-cal etching. Belgian designers—such as Christiguey andJan Pycke—are using these "pearls" in jewelry.

CCOOLLOORREEDD SSTTOONNEESS AANNDD OORRGGAANNIICC MMAATTEERRIIAALLSS

AAnnddrraaddiittee ffrroomm AArriizzoonnaa.. Mineral collectors are familiarwith the clusters of brownish green andradite crystals,some with iridescent surfaces, that have been found formany years at Stanley Butte in Graham County,Arizona. Although these specimens typically havebright surfaces and sharp crystal faces, they are nottransparent enough to facet. This year at Tucson, how-ever, Charles Vargas of Apache Gems, San Carlos,Arizona, was offering gem andradites (figure 3) from anew locality near Apache Camp on the San CarlosApache Reservation,- this is about 10 miles (17 km)from the Stanley Butte locality.

These garnets come from a contact zone betweencarbonate-rich basement rocks and basalt lava flows,according to Mr. Vargas. The garnet deposit was discov-ered only recently, in June 1996, and there was no evi-dence that it had been worked earlier. The garnets occuras dodecahedral crystals, with calcite, in pockets in mul-tiple decomposing dikes within a 15 square-mile (about40 km2) area; up to 18 inches (about 45 cm) of topsoilcovers the dikes. As of early April 1997, Mr. Vargasreports, the largest fashioned andradite weighed about 4ct; production was a few kilos per month of facetedstones, ranging in size from melee to just over 1 ct.

We examined 10 andradites (see, e.g., figure 3),which weighed 0.10 to 1.51 ct. All had properties typicalfor andradite garnet: brownish greenish yellow to yel-

low- brown color, singly refractive optic character, R.I.greater than 1.81 (over the limit of a standard refrac-tometer), and no reaction (inert) to both long- and short-wave ultraviolet radiation. Specific gravity values—determined hydro—statically-averaged 3.92 (18 readingson nine stones). All 10 samples had a line at 440 nm in adesk-model spectroscope. Visible through the micro-scope were healed fractures, "fingerprints," growth band-ing, a few dark crystals (in two stones), and two-phaseinclusions (in four stones). The growth banding was vis-ible as linear or roiled features in plane-polarized lightand showed first- to low-second-order interference col-ors between crossed polarizers. No "horsetail" inclu-sions were seen.

CCaarrvveedd aaqquuaammaarriinnee wwiitthh nnaattuurraall ccrryyssttaall ffaacceess.. In theWinter 1996 Gem News section (p. 283), we reported onfaceted stones that had incorporated naturally etchedfaces of the original crystal. The stones illustrated weregem varieties of beryl; this mineral, occurring as it doesin late-stage pegmatites, often shows shattered or dis-solved surfaces that have healed into well-defined crys-tal faces. Mark Herschede Jr., of Turmali and Herschede,Sanibel, Florida, showed one of the editors another vari-ation on this same idea using aquamarine from theUkraine. Not only did the approximately 45 ct stoneinclude part of the natural crystal, but carver GlennLehrer (Lehrer Designs, San Raphael, California) hadalso fashioned the bail as part of the pendant (figure 4).

CCaatt''ss--eeyyee cchhrryyssoobbeerryyll.. One of the notable stones at theAGTA show was a large, fine cat's-eye chrysoberyl (fig-ure 5), shown by Evan Caplan of Los Angeles,California. The 72.68 ct cabochon is reportedly from theFaisca area in northeastern Minas Gerais, Brazil, whichhas been producing relatively large quantities of finechrysoberyl since 1939 (see, e.g., K. Proctor,"Chrysoberyl and Alexandrite from the PegmatiteDistricts of Minas Gerais, Brazil," Gems & Gemology,Spring 1988, pp.16–32). Two flat sur faces on the back

Figure 2. These 1.66 and 0.28 ct diamond "pearls"were polished by selective dissolution in hot sodi-um carbonate. The larger "pearl" measures about5.76 ×5.58 mm. Photo by Maha DeMaggio.

Figure 3. These nine faceted andradites (0.10–1.51ct) are from a newly discovered contact metamor-phic deposit on the San Carlos ApacheReservation in Arizona. Stones courtesy of ApacheGems-, photo by Maha DeMaggio.


of the stone marked where two pieces had beenremoved to cut two smaller cabochons. Mr. Caplan the also had a large (317.7 ct) piece of rough chatoyantchrysoberyl that also reportedly originated from theFaísca area.

UUnnuussuuaall ttrraappiicchhee eemmeerraalldd.. Ron Ringsrud ofConstellation Colombian Emeralds, Saratoga,California, displayed a variety of trapiche emeraldsfrom Colombia at his booth in the AGTA show. Onecabochon was particularly unusual, in that it had twocentral columns, each with the black "spokes" typicalof trapiche emeralds (figure 6). The 3.50 ct oval cabo-chon measured 10.24 × 8.95 × 5.97 mm. The double-column structure extended completely through thestone and appeared to taper only slightly from thedome to the base of the cabochon. Perhaps the doublecenter resulted from parallel growth of two individualcrystals that were later incorporated into one piece asgrowth continued. This double-center trapiche emeraldwas the first that Mr. Ringsrud or any of the GemNews editors had encountered (although K. Nassau andK. A. Jackson showed a trapiche emerald slice withtwo inter-grown crystals in the April 1970 LapidaryJournal, p. 82).

TThhee rreettuurrnn ooff jjeett.. The natural hydrocarbon jet becamepopular in mourning jewelry in the late Victorian era.With the recent interest in black opaque materials (see,for instance, "Some Gemological Challenges inIdentifying Black Opaque Gem Materials" by M. L.Johnson et al, Winter 1996 Gems & Gemology, pp.252–261), we noticed more jet at Tucson this year,including bead necklaces from Czechoslovakia. At thePueblo Inn, Russian dealer Serguei Semikhatov ofLithos, Irkutsk, had blocks of massive jet and a fewcarvings, including the bear shown in figure 7. Thematerial is being quarried near the village of Matagan,on a river leading into Lake Baikal north of the city ofIrkutsk, according to Mr. Semikhatov. The Russianname for this material is "gagate."

""BBiiccoolloorr"" llaabbrraaddoorriittee.. So-called "spectrolitc"—

Figure 4. This approximately 45 ct aquamarine pen-dant, shown here as viewed from the side, wascarved both to include faces from the original crys-tal and to provide its own bail. Carving by GlennLehrer, courtesy of Turmali and Herschede-, photoby Maha DeMaggio.

Figure 5. Some particularly fine cat's-eyechrysoberyl from Brazil was shown in Tucson thisyear, including this 72.68 ct cabochon and theaccompanying 317.7 ct piece of rough. Courtesy ofEvan Caplan; photo by Robert Weldon.

Figure 6. The double center in this 3.50 ct trapicheemerald is most unusual. Photomicrograph by JohnI. Koivula; magnified 5×


labradorite feldspar with iridescent schiller—is by nomeans a new material, and the locality at Ylämaa, Finland, is also well known. However, this year, G.P.G.Trading of New York City, in conjunction with JoganOy, Mikkeli, Finland, was marketing large quantitites ofcabochons—in a variety of shapes—that had been cutfrom this material. Subsequently, one of the editors sawa nicely patterned 9.51 ct rectangular cabochon with adistinct delineation between two colors of sheen (figure8) at the booth of another firm, the D&R Collection, LosAngeles.

According to Don Pier, of G.P.G. Trading, the lease-holder recently extracted 650 tons of material from thislocality, using earth-moving equipment. The Ylämaadeposit is located close to the Russian border and theArctic circle,- the nearest town is Lappeenranta. Mr. Pierestimates that the deposit is about 5 m wide by 40 mlong by 80 m deep; because it is in a permafrost region,

mining is possible only three months a year. Thelabradorite is being fashioned into calibrated goods bycutters in Sri Lanka.

""RRaaiinnbbooww"" oobbssiiddiiaann hheeaarrttss.. We have reported in the paston "rainbow" obsidian from Mexico, which shows lay-ered schiller in interference colors (see, e.g., Summer1993 Gem News, p. 133). In 1996, many dealers wereshowing this material fashioned in a new way: A groovecut into the widest part of an oval or pear-shaped cabo-chon the size of a paperweight produced a pattern ofconcentric hearts. This year, Zee's of Tucson, Arizona,added yet another twist: Jewelry-size cabochons withthe heart pattern were drilled through their tops, so thata chain or cord could be passed directly through thepiece (figure 9).

BBoottrryyooiiddaall wwhhiittee ooppaall ffrroomm MMiillffoorrdd,, UUttaahh.. The botry-oidal opal from Milford, Utah, has been known for atleast 20 years. Since 1996, William Cox, a gem cutterfrom Provo, Utah, has been cutting this material intoflat-topped free-form cabochons with curved schiller—aresult of the botryoidal internal layering. Mr. Cox ismarketing these cabochons as Satin Flash opal.

We examined the cabochon shown in figure 10. Withmagnification, we saw a structure of curved layersthroughout this material; some layers showed crazing inreticulated patterns, but cracks were confined to individ-ual layers for the most part and did not extend throughthe stone. Some small rounded white crystals were alsovisible,- these were clumped in small patches ratherthan distributed evenly throughout the material. Whenexamined with crossed polarizers, the stone did not

Figure 7. This 71.1 × 42.1 × 41.1 mm bear wascarved from the hydrocarbon jet, which is beingquarried near Matagan, Russia. Photo by MahaDeMaggio.

Figure 9. This 20.58 ct "rainbow" obsidian cabo-chon (19.20 × 16.39 × 10.91 mm) has been carvedat its broadest point to produce a heart-shapedpattern of iridescence. Note the hole-visible on theright, where the cabochon has been drilled for easystringing on a chain or cord. Photo by MahaDeMaggio.

Figure 8. This 9.51 ct labradorite cabochon fromFinland (17.65 × 13.75 × 4.01 mm) has blue andorange schiller in a "bicolor" pattern; the centerwhite stripe probably marks the boundary betweentwo adjacent crystals. Photo by Maha DeMaggio.

show the extremely high strain characteristics of hyaliteopals from most localities. The cabochon had a typicalopal R.I. of 1.45, and a specific gravity—measuredhydrostatically— of 2.14; it was inert to both long- andshort-wave UV.

Similar opal from Utah—banded in translucentbrown and translucent-to-transparent white—was seenin the booth of Charles R. Richmond, Blountville,Tennessee. Marketed as Candy-Stripe opal, this materialcould be carved into cameos or striped cabochons,depending on the orientation.

PPeettrriiffiieedd ppaallmm wwoooodd aass aann oorrnnaammeennttaall mmaatteerriiaall..Women's fashions sometimes affect the popularity ofcertain gemmaterials. For instance, last year's emphasis

on bright orange, yellow, and lime green clothing helpedspark interest in "Mandarin" spessartine garnets, peridotfrom Pakistan, and grossular-andradite garnets fromMali. This year, one fashion trend is a growing interestin "animal prints"—fabrics designed to resemble thehide patterns of leopards, zebras, and other animals. JoeJelks of the Horizon Mineral Co., Houston, Texas, mar-kets cabochons of certain naturally patterned stones as"animal print" cabochons, including Australian "zebrarock" (sec, e.g., Gem News, Summer 1994, pp. 128–129).Another material that is enjoying increased popularity—in part because of its resemblance to animal patterns—is petrified palm wood from Louisiana, Texas, andArkansas. Although this is not a new material, customjewelry makers are beginning to explore its design possi-bilities. The three pieces of petrified palm wood in fig-ure 11 show some of the colors and patterns in whichthe material is available. The 45 × 42 mm brooch, byNature/Man, Lyons, Colorado, contains tan wood witha leopard pattern in black; while the rough samplesfrom Ancient Earth, Hurricane, Utah, are reddish brownwith tan patterns and tan with brown patterns.

DDrruussyy iirriiddeesscceenntt ppyyrriittee ffrroomm RRuussssiiaa.. The interest indrusy materials continues unabated. One new materialthis year was drusy pyrite, found as crystals lining con-cretions in the bed of Russia's Volga River. We bor-rowed samples of rough and fashioned material (figure12) from Joe Jelks of the Horizon Mineral Co. The con-cretions are gathered from one section of the Volgawhen the water is shallow. The pyrite is naturally iri-descent, according to Mr. Jelks, and has not been artifi-cially tarnished. The concretion material itself is suffi-ciently strong that a separate backing material is notrequired.

""WWaasspp--ttaaiill"" cciittrriinnee//ssmmookkyy qquuaarrttzz.. Klaus Schafer, ayoung gem carver from Idar-Oberstein, Germany, cutsexacting stones from humble materials. This year, heshowed several bar-shaped faceted stones cut from a


Figure 10. This 37.18 ct opal fmm Milford, Utahdemonstrates schiller from its curved internal sur-faces. Stone cut by William Cox, provided cour-tesy of Richard Shall, Out of Our Mines, Arcata,California; photo by Maha DeMaggio.

Figure 12. This rough section of a concretion fromthe Volga River, and the cabs cut from similarmaterial, have surfaces of naturally iridescentdrusy pyrite. The largest cab measures 21.73 ×13.02 × 4.61 mm. Stones courtesy of foe Jelks; photoby Maha DeMaggio.

Figure 11. Petrified palm wood shows a variety ofcolors and animal patterns. Photo by MahaDeMaggio.

single heat-treated piece of banded citrine and smokyquartz (figure 13). The bars had been carefully cut—per-pendicular to the growth zoning that paralleled a rhom-bohedral face—so that the color bands were distinct. Mr.Schäfer calls these pieces "wasp-tail" quartz, but heagreed that they also looked like bar codes. He showedthese at the booth of Richard Bernhard and Company,Idar-Oberstein.

QQuuaarrttzz wwiitthh ""rraaiinnbbooww"" hheemmaattiittee iinncclluussiioonnss .. .. .. We haveseen quartz included with red iron oxides fromMadagascar (Fall 1995 Gem News, pp. 209–210) andKazakhstan ("strawberry quartz," Spring 1995 GemNews, pp. 63–64). This year at Tucson, VladimirChernavtsev of Charmex, Moscow, was showing quartzincluded with very thin hexagonal to irregular plates ofhematite, showing an iridescent "rainbow" tarnish; thismaterial came from Aldan Mountain in Yakutia. A25.59 ct polished free-form hexagonal tablet (22.20×18.47 × 6.50 mm) is shown in figure 14. The thickerhematite plates look black and opaque, but those thatare thin appear transparent red and some plates showregions of both colors.

...... AAnndd wwiitthh ppyyrriittee iinncclluussiioonnss.. Aventurescence in quartzis generally thought to be caused by the presence ofmany tiny platelets of the chromium-colored variety ofmica known as fuchsite. In such quartzes, the inclusionsare randomly oriented. Thus, in almost any cutting ori-entation, a sufficient number of the platelets showessentially the same reflective direction, so that aven-turescence occurs in reflected light without a specialneed to orient the rough before cutting.

However, Gem News editor John Koivula acquiredanother form of aventurescent quartz from theGemological Center, Belo Horizonte, Brazil. The hostmaterial was a 70.48 ct oval faceted rock crystal quartzthat measured 40.01 × 21.44 × 12.67 mm (figure 15).Here, the mineral inclusion causing the aventurescencewas pyrite.

The pyrite inclusions creating the distinctivereflectance did not display the isometric crystal habitstypical for this mineral. Instead, these crystals grew asdisks in a fracture plane in the quartz host. As seen infigure 16, these inclusions strongly resemble the so-


Figuie 13. These quartz "wasp-tails" (4.4 to 9.61 ct)were cut from a chunk of heat-treatedcitrine/smoky quartz from Brazil. Stones courtesyof Klaus Shäfer; photo by Maha DeMaggio.

Figure 15. Aventurescence in this 70.48 ct rockcrystal quartz is caused by light reflecting off amultitude of pyrite inclusions. Photo by MahaDeMaggio.

Figure 14. Thin plates of bright red-to-blackhematite, with an iridescent "rainbow" tarnish, arevisible in this 25.59 ct polished free-form hexago-nal tablet of quartz from Yakutia. Photo by MahaDeMaggio.

Figure 16. The disk-shaped aventurescence-causingpyrite inclusions in the stone shown in figure 15formed along a single fracture plane.Photomicrograph by John I. Koivula; magnified l0×.


called pyrite "sand dollars" or "suns" that are found asconcretions between bedding planes in black and grayshale and slate.

To create the unusual aventurescence, the lapidaryhad to position the layer of pyrite disks at an angle justoff-parallel to the table facet. This was done so that theaventurescent effect could be seen without the distrac-tion of light reflections off the table facet surface. Thisneed for specific orientation of the inclusion plane is indirect contrast to the random orientation of cuttingquartz in which the aventurescence is caused by fuch-site.

OOrraannggee ssaapppphhiirree aanndd ootthheerr ggeemmss ffrroomm tthhee TTuunndduurraarreeggiioonn.. East Africa continues to produce astoundinggems. One example is shown in figure 17: a 2.28 ctbright slightly reddish orange natural sapphire from the

Tunduru region of Tanzania, shown by Malhotra Inc.,New York City. Mr. Malhotra said that he had not seena stone of such color in his 40 years in the trade. Thestone had been cut in Bangkok and may have been heattreated.Other unusual materials from Tanzania (seen at thebooths of New Era Gems, Grass Valley, California)included: peridot and "mint" (light green) and "honey"(brownish yellow) grossular from Mcrelani; and light-pink and light-purple to "rose pink" spinel from Pokot,near Tunduru. According to Bill Vance of New EraGems, some blue-to-purple spinels from Tunduru lookpink when viewed through the Chelsea filter, so hesuspects that they contain cobalt.

TTaannzzaanniittee iinn aabbuunnddaannccee:: GGeemm ccaarrvviinnggss aanndd ""eennhhyyddrroo""ccrryyssttaall sseeccttiioonn.. Tanzanite was readily available inTucson this year. One possible reason was that-in addi-tion to the "traditional" tanzanite mines (see, e.g.,Summer 1996 Gem News, pp. 135–136)—the gem isnow being recovered as a byproduct of large-scalegraphite mining in the Merelani region, as reported inthe Mineralogical Record (T. Moore, "What's New inMinerals: Denver Show 1996," Vol. 28, No. 1, 1997,especially p. 60). In addition to plentiful supplies ofcalibrated goods, we saw some notable individualpieces. New Era Gems showed several examples,including a 255.75 ct terminated blue crystal that con-tained an obvious fluid inclusion, with a movable gasbubble (figure 18). They also showed contributing edi-tor Shane McClure a trio of tanzanite gem carvings: arabbit, a frog, and a horse head (figure 19).

Another firm, Affro Gems of Dallas, Texas, hadsome tanzanites with unusual colors, including a medi-um-toned grayish blue stone that was similar in colorto some blue zircons.

TToouurrmmaalliinnee ffrroomm tthhee NNeeuu SScchhwwaabbeenn rreeggiioonn,, NNaammiibbiiaa::AA mmaajjoorr nneeww ppllaayyeerr.. After an initial offering at theHong Kong show last September, tourmaline fromIndigo Sky Gems of Windhoek, Namibia, made its U.S.debut at Tucson. The Neu Schwaben deposit is alluvialin nature, according to the company's managing director

Figure 17. The color in this 2.28 ct Tundura sap-phire is exceptionally vivid. Courtesy of MalhotraInc.; photo by Maha DeMaggio.

Figure 18. This 255.75 cttanzanite crystal containsa fluid inclusion with alarge gas bubble that, asis evident in these twophotos, moves freely with-in the liquid. Courtesy ofNew Era Gems; photos byMaha DeMaggio.


Chris Johnston, it is expected to produce fine green,blue, and blue-green tourmalines of all sizes, from cali-brated goods to major single gems, for many years tocome. Several articles about this deposit have alreadybeen published in the trade press, including theNovember 1996 issue of Modem Jeweler ("NeuSchwaben Tourmaline," by David Federman, pp. 17-18);the December 1996 issue of Asia Precious ("NamibianTourmaline Touches the Sky," by Robin Bower, pp.96–98); and the January 1997 issue of Colored Stone("Blue Tourmaline Makes Name for Namibia," by GregBartolos, pp. 539, 558–560). Mr. Johnston co nfirmedthat the production information in these articles wasessentially correct.

The Neu Schwaben deposit has been known tosources in Idar-Oberstein since 1925, but it has beenmined by the present company only for the last year orso. Current estimates of the reserves are 12.5 millioncarats of gem-quality tourmaline (from 500,000 to750,000 tons of alluvial gravels). The deposit covers atourmaline-bearing dike—a potential source of addition-al material—that is about 1 km long. The Indigo SkyCompany will only market cut stones, not rough; cur-rent production is 10,000 calibrated stones per month,about 10% of which are 1 ct or larger. Several largestones (see, e.g., figure 20) have been cut, with thelargest to date approximately 65 ct.

The Indigo Sky Company is operating underNamibia's Export Processing Zone regulations, whichare intended to encourage employment of local work-ers,- the company hopes to employ 150 local miners,with some of the profits being invested in providing reg-ular medical care for workers and their families.According to Mr. Johnston, attempts are being made tomine the deposit in an "ecologically responsible fash-ion," so that sustainable long-term development overthe mine's projected 20-year life will be accompanied orfollowed by reclamation of the mine site.

Gemological information on this material willappear in a forthcoming issue of Gems &Gemology.

TTRREEAATTMMEENNTTSS GGoolldd--eelleeccttrrooppllaatteedd jjaaddee aanndd ssiillvveerr oorree.. Bill Heher of theRare Earth Mining Company, Trumbull, Connecticut,always brings unusual material to Tucson. New thisyear were nephrite jade cabochons that had been electro-plated with gold (figure 21). According to Mr. Heher,only the green and black nephrite from an old depositnear Victorville, California, responds to this treatment.This material contains magnetite inclusions, to whichthe gold attaches. Sometimes, only some of the blackinclusions accept the gold plate (again, see figure 21).Mr. Heher first saw such treated material (not fashionedinto cabochons) many years ago in Idar-Oberstein, andhe had been trying to obtain some of this nephrite formany years. Finally, letters he wrote to several WestCoast lapidary clubs resulted in the discovery of a fewpounds of rough.

SSYYNNTTHHEETTIICCSS AANNDD SSIIMMUULLAANNTTSSAAsssseemmbblleedd ssttoonneess uussiinngg iinnsseett ggeemm mmaatteerriiaall.. It is wellknown that color concentrated in the culet of a stonewill appear to be spread evenly throughout the gem

Figure 19. These are the first carvings made fromgem tanzanite that the editors have seen. This 55 ×22 × 33 mm frog, 62 × 37 × 33 mm rabbit, and 44 ×26 × 12 mm horse head were at the booth of NewEra Gems. Photo by Maha DeMaggio.

Figure 20. This 46.48 ct (23 × 18 mm) emerald-cuttourmaline from Neu Schwaben, Namibia, is rep-resentative of the fine material that this regionproduces. Stone faceted by Martin Key, courtesyof Indigo Sky Gems; photo by Maha DeMaggio.


when that stone is viewed face up. David Brackna, agem cutter from Germantown, Maryland, told staffgemolo-gist Cheryl Wentzell that he has found anotheruse for this effect. For example, he set an approximate-ly 3 mm cabochon of Australian opal within the culetarea of a 14 × 12 mm faceted green beryl from the UralMountains in Russia (figure 22, left). Face up, the opal'splay-of-color is reflected throughout the larger stone(figure 22, right). Mr. Brackna calls this technique"optical inlay."

To create this effect, Mr. Brackna first fashionedthe rough beryl into an antique cushion mixed cutwith a step-cut crown ("Harlequin cut"). The culet was

cut off, and a concave "divot" was fashioned into theculet area of the main piece. A small opal cabochonwas glued into the divot with UV-curing epoxy; theUV content of daylight was sufficient to set thisglue in about 15 seconds. A replacement culet wasthen fashioned from a piece of beryl similar to themain piece. To increase the dispersive effects of theopal's play-of-color, Mr. Brackna carves high crownson his pieces. The concave facet on the inside ofeach piece magnifies the effect of the inset material.

In the two months before the Tucson shows, Mr.Brackna cut and assembled 25 such stones. Othershapes are possible, including round and fancy cuts. Forthe bodies of these assembled stones, Mr. Brackna hasused amethyst and citrine, as well as aquamarine andgolden beryl.

RReedd aanndd ppuurrppllee hhyyddrrootthheerrmmaall ssyynntthheettiicc bbeerryyll.. Twomysterious emerald cuts were turned over to GIA bythe Tucson police in 1996. One, 0.81 ct, was red andthe other, 0.90 ct, was red-purple (figure 23).Apparently, they had been found at a show that yearand were never claimed by their owner. The red sam-ple (measuring 7.06 × 5.10 × 3.30 mm) was stronglypleochroic—with orange (ordinary ray) and purple-red(extraordinary ray) colors; the pleochroism was sostrong that the color of the sample varied from orangyred to purplish red depending on its position and orien-tation relative to the viewer and the light source. Thered-purple emerald cut (7.05 × 5.16 × 3.64 mm) wasalso strongly pleochroic, with the ordinary ray orangyred and the extraordinary ray red-purple. The two sam-ples were cut in different orientations: The optic axisran down the length of the 0.81 ct piece, but throughthe width of the 0.90 ct piece.

Gem properties in common for both samplesincluded: specific gravity (2.68, measured hydrostati-cally), lack of reaction (inert) to both long- and short-wave ultraviolet radiation, lack of visible lumines-cence, and an orangy red color when viewed through a"Chelsea" color filter. Features visible with magnifica-tion were "chevron" growth zoning (typical forhydrothermal synthetic beryl), pinpoint inclusions,

Figure 21. Magnetite inclusions in some nephrite,such as this 56.62 ct crescent-shaped cabochon(65.8 × 17.75 × 4.38 mm), form artistic patternswhen electroplated with gold. Photo by MahaDeMaggio.

Figure 22. Left, a small opalcabochon was inserted intothe culet area of this 11.82 ctassembled stone, and thencovered with a replacementculet of green beryl similar tothat used for the body. Right,when the stone is viewed faceup, the play-of-color from theopal appears throughout.Courtesy of David Brackna;photos by Maha DeMaggio.

and stringers. The samples differed in some properties,however: The 0.81 ct red had refractive indices of1.568-1.572 (birefringence 0.004) and a complicatedabsorption spectrum when viewed through the hand-held spectroscope—with a diffuse band at 420 nm, abroad 435-465 nm band, a faint band at 500 nm, a dif-fuse weak band at 530 nm, a strong 545 nm band, andlines at 560 and 585 nm. In contrast, the red-purple0.90 ct emerald cut had refractive indices of1.575–1.581 (birefringence of 0.006) and a simplerabsorption spectrum—a diffuse 435 nm band, a diffuseband at 540–580 nm, and weak lines at 650, 660, and670 nm. EDXRF revealed major amounts of Al and Siin both samples, as well as trace amounts of Mn, Fe,Ni, Cu, Zn, Ga, and Rb; the red-purple 0.90 ct examplealso contained Ti and Cr. All these properties wereconsistent with hydrothermal synthetic beryl; thesamples' complicated absorption spectra are not sur-prising, given the many chromophores present.

DDiicchhrrooiicc ggllaassss aass aann ooppaall iimmiittaattiioonn.. NancyGoodenough is an artist from San Francisco,California, who works in glass. She creates dichroicglass cabochons (figure 24) that could be mistaken forwhite opal, nonphenomenal white opal, and the blueopal from Peru. Another artisan, Susan Slayton ofEarthly Enchantments, Covina, California, has beencreating free-form cabochons of dichroic glass that shethen etches to create a matte surface (again see, figure24). Neither artist creates these items specifically toimitate opal. However, we saw a display at one show inwhich such cabochons were scattered among free-form

cabochons of boulder opal. They could easily have beenmistaken for opals by the unsuspecting.

AAvveennttuurreesscceenntt ssyynntthheettiicc qquuaarrttzz.. In the quest to devel-op more varieties of synthetic quartz, not all experi-ments are successful. One exhibitor from the stateenterprise VNIISIMS (the Russian Research Institutefor the Synthesis of Minerals and Pilot Plant),Alexandrov, Vladimir Region, Russia, showed con-tributing editor Shane F. McClure two examples of anexperiment gone awry. The scientists at VNIISIMSwere attempting to synthesize aventurescent quartz byincluding flakes of copper in their hydrothermal syn-thesis; however, the copper flakes attached only to themiddle seed plane of the synthetic quartz (figure 25).

SSyynntthheettiicc ooppaall iinn ggllaassss.. Opal is in some respects anatypical gem material because of its durability prob-lems,- however, the beauty of opal gives it consider-able value despite the risks of damage inherent inwearing it. Gilson synthetic opal is actually moredurable than most natural opal, in our experience, as itis more heat resistant and does not craze. Using thisdurability advantage, Gerry Manning of ManningInternational, New York City, was offering briolettes,hearts, and spheres (figure 26) of glass with Gilson syn-thetic opal chips in the center, similar to the faceted"Gemulets" mentioned in the 1992 Tucson Gem Newsreport (Spring 1992, p. 65). When examined betweencrossed polarizers, two spheres showed strain colors upto first-order blue, as well as flow lines and sphericalgas bubbles. This implies that the glass was in amolten state when it was poured around the syntheticopal chips. A natural opal probably would not have sur-vived this treatment.


Figure 24. These cabochons of dichroic glass couldbe mistaken for opals. The large free-form piece, bySusan Slayton, measures 34.38 × 24.24 × 10.61 mmand weighs 64.82 ct; the round cabochons, byNancy Goodenough, range from 1.30 to 4.40 ct (thelargest is 10.4 mm in diameter). Photo by MahaDeMaggio.

Figure 23. These samples (0.81 and 0.90 ct), leftunclaimed after the 1996 Tucson show, turned outto be hydrothermal synthetic beryls. Photo byMaha DeMaggio.


MMIISSCCEELLLLAANNEEOOUUSS

""AAmmpphhoorraaggeemmss.."" Many brightly colored gemmy mate-rials are primarily available as small crystals or crystalfragments; they only appear as larger pieces on rareoccasions, or at prohibitive prices. While thinkingabout the design possibilities inherent in a smallParaiba tourmaline crystal, Brian Charles Cook ofNature's Geometry, Graton, California, had the idea ofamplifying the effect of its color by enclosing it in amagnifying setting of optical-grade quartz. This idea isthe basis of the assemblages that he calls"Amphoragems," in which small euhedral crystals withsaturated colors are placed in liquid-filled wells incurved quartz bottles (figure 27). Crystals of more thanone material may be used in the same piece, and thecap of the "amphora" may be quartz or another materi-al. The "bottles" may stand alone or be worn as pen-

dants. Among the materials that have been placed inthe centers of these "Amphoragems" are spessartineand chrome pyrope garnets, Paraiba tourmalines, redberyls, gold nuggets, diamond crystals, and abalonepearls; materials chosen for the caps include chryso-prase, chrysocolla, rutilated quartz, amethyst, aquama-rine, and black jade.

AANNNNOOUUNNCCEEMMEENNTTSS

GGeemmss && GGeemmoollooggyy wwiinnss tthhrreeee oouutt ooff ffiivvee.. For the thirdtime in the past five years, Gems & Gemology wonfirst place as best journal in the prestigious AmericanSociety of Association Executives (ASAE) Gold Circlecompetition. Gems & Gemology has won awards inthe competition for five consecutive years, includingtwice winning first-place trophies for best scientific-educational feature articles. Gems & GemologyTechnical Editor Carol M. Stockton, of Alexandria,Virginia, accepted the award in early December onbehalf of the journal.

Figure 25. These two crystals of synthetic quartz(44.28 and 15.28 ct, respectively) were produced inan attempt to simulate aventurescent quartz. Thespangled inclusions are metallic copper. Samplescourtesy of VNIISIMS; photo by Maha DeMaggio.

Figure 26. This 26.55-mm-diameter sphere has acenter of Gilson synthetic opal. Photo by MahaDeMaggio.

Figure 27. The "Amphoragem"assemblage in this necklace containscrystals of (top to bottom): greengahnite, spessartine from Brazil, andred beryl from Utah. The body of thepiece is quartz and the cap isrhodochrosite from Tadjikistan.Courtesy of Nature's Geometry;photo by Maha DeMaggio.

71 Gems & Gemology Challenge GEMS & GEMOLOGY Spring 1997

Anyone who missed even one issue of Gems & Gemology in 1996missed a great deal. Spring started the year with African diamondsources and Russian hydrothermal synthetic emeralds. Summerbrought information on Madagascar sapphires, Russian demantoid gar-nets, and Ethiopian opals. Fall highlighted the new De Beers diamondverification instruments, growth-structure analysis in gem identifica-tion, and Russian flux-grown synthetic alexandrites. The Winter issuecovered such diverse topics as imitation tanzanite, trapiche rubies,gems from Sri Lanka, black opaque gem materials, and the CapãoImperial topaz mine. Now, we invite you to test your knowledge ofthese important topics by taking the 1 lth annual Gems & GemologyChallenge.

The following 25 questions are based on information from the four1996 issues of Gems & Gemology. Refer to the feature articles andNotes and New Techniques in these issues to find the single bestanswer for each question,- then mark your choice with the corre-sponding letter on the response card provided in this issue (sorry, nophotocopies or facsimiles will be accepted; contact the SubscriptionsDepartment if you wish to purchase additional copies of the issue).Mail the card so that we receive it no later than Monday, August 18.Be sure to include your name and address (print legibly, please) Allentries will be acknowledged with a letter and an answer key.

Score 75% or better, and you will receive a GIA Continuing Education Certificate. Earn a perfect score,and your name will also be featured in the Fall 1997 issue of Gems & Gemology.

1. The rarest color of topaz from 3. Of the top 10 diamond-produc 5. Demantoid's usefulness inthe Capao mine is ing countries from antiquity to jewelry is dictated by its hardA. purple. 1994, Africa is home to ness, which is just underB. "sherry" red. A. three. A. 8.C. orange-yellow. B. five. B. 7.D. reddish orange. C. seven. C. 6.

D. nine. D. 5.

2. The De Beers DiamondSure™ 4. Before its recent fame as a sourceinstrument has been designed of fine blue sapphire, Madagascar 6. Some brownish yellow andfor the rapid examination of was best known for producing orange topazes from Ouro Pretolarge numbers of beryls, tourmalines, and are altered to pink byA. natural diamonds. A. rubies. A. dyeing.B. rough diamonds. B. emeralds. B. heat treatment.C. polished diamonds. C. garnets. C. diffusion treatment.D. synthetic diamonds. D. diamonds. D. irradiation.

72 Gems & Gemology Challenge GEMS & GEMOLOGY Spring 1997

7. One reason that the black 13. For lapidaries and gem cutters, 20. Scapolite from Embilipitiyamembers of the spinel mineral the most familiar durable black can be distinguished easilygroup represent an identifica- opaque material is probably the from similar-color citrine andtion challenge is that the group variety of quartz called feldspar on the basis ofcontains

A. "black jade." A. fluorescence and birefringence.A. species with several B. "black onyx." B. inclusions and transparency.

morphologies. C. "burnt topaz." C. optic character and immerB. species with different D. "black spinel." sion microscopy

specific gravities. D. growth zoning and low-C. 21 species 14. The negative d plane, as a domi relief inclusions.D. Shades of gray. nant feature, is most common in

A. natural rubies.8. In addition to advanced testing B. flux-grown synthetic rubies. 21. Tanzanite generally can be easi

techniques, "Tairus" C. flame-fusion synthetic rubies ly separated from its many imihydrothermal synthetic emer D. none of the above. tators byalds can be separated from nat 15. Trapiche rubies are composed of A. growth-structure analysis.ural emeralds on the basis of triangular or trapezoidal sectors B. R.I. andS.G.their formed by the fixed arms of C. chemical analysis

D. ultraviolet fluorescence.

A. internal features. A. a six-rayed star.B. polariscope reactions. B. a four-rayed star.C. reaction to ultraviolet radiation. C. an eight-rayed star. 22. Fine demantoids in commercialD. refractive indices. D. all of the above. size are found almost exclusively

in the

9. To date, the negative d plane has 16. A new, commercial source in A. Ukraine.been observed in natural rubies Madagascar for fine blue sap B. Volga River.thought to be from phires is C. Ural Mountains.

A. Sri Lanka and Tanzania. A. Iankaroka. D. Kamchatka Peninsula.B. Vietnam and Myanmar. B. Amboasary.C. Myanmar and Thailand. C. Antananarivo.D. Thailand and Cambodia. D. Andranondambo. 23. One indication that an emerald

may be synthetic is the presence of17. Much of the material that enters

10. The surest means of separating the gem trade as "tanzanite" is A. a dominant growth patternnatural alexandrite from actually parallel to the s plane.Russian synthetic alexandrite is B. Cr lines seen with a handA. optic character and specific A. irradiated blue iolite. spectroscope.

gravity. B. heat-treated colorless spinel. C. third-order interference colors.B. UV-visible absorption spectra. C. heat-treated brown zoisite. D. a reaction to long-wave ultra-C. characteristic growth patterns. D. synthetic corundum. violet radiation.D. X-ray fluorescence and 18. Before the introduction of pipe

infrared spectroscopy. mining in South Africa in 1869, 24. The De Beers DiamondView™what percentage of the world's enables the separation of natural

11. Opal mining at Yita Ridge, annual diamond production from synthetic diamonds on theEthiopia, is presently came from Brazil? basis of

A. highly mechanized. A. 35%. A. cathodoluminescence.B. diminishing. B. 50%. B. the presence of the 415 nm line.C. near capacity. C. 75%. C. the fluorescence image.D. small in scale. D. 90%. D. the colors of fluorescence.

19. The first country outside South12. The most prestigious source of Africa to have an economic 25. The arms of the star in a trapiche

natural alexandrites is kimberlite pipe was ruby are created primarily by

A. Mexico. A. Ghana. A. oriented rutile needles.B. Brazil. B. Rhodesia. B. chatoyancy.C. Russia. C. Tanzania. C. aventurescence.D. Tanzania. D. Ivory Coast. D. concentrations of inclusions.

Spring 1997 Gemological Abstracts GEMS & GEMOLOGY 73

DDIIAAMMOONNDDSS

CCoonnsseeqquueenncceess ooff rreeccyycclleedd ccaarrbboonn iinn ccaarrbboonnaattiitteess.. DD.. SS..Barker, Canadian Mineralogist, Vol. 34, 1996, pp.373–387.

The question of the cycling of carbon through the Earth'scrust and mantle is one of great interest with regard tothe origin and evolution of diamonds. Carbon-bearingphases in the mantle include calcite and dolomite, aswell as diamond, graphite, silicon carbide, dissolvedmethane, carbon monoxide, and carbon dioxide in fluids,and "carbonate components" in magma. The phases thatform are a function of the availability of oxygen, as wellas carbon, in the mantle.

Evidence has been accumulating that carbon fromthe crust is recycled deep into the mantle by subduc-tion,- an important destination for this carbon—otherthan in diamonds—is the igneous rock carbonatite. Thecarbon isotopes in carbonatites do not match those ofeither marine carbonates (e.g., from coral reefs) or organ-ic carbon (e.g., from plant debris); however, these iso-topes are similar to those in peridotitic diamonds. Fromthe study of trace-element distributions, especially ofultramafic xenoliths in carbonatites, the author con-cludes that these rocks probably result from repeatedinteractions of carbonate-rich mag-matic fluids withrestricted volumes of metasomatically altered mantlerocks. Several influxes of carbon-rich magma have beenimplicated in the formation of diamonds at the Premiermine, and of coated diamonds from Africa, Siberia, and

Australia. Carbonatites are uncommon in the Earth'scrust, but this is probably because the mantle "soaks up"most batches of carbonate-rich magma, and only a fewmanage to reach the surface. MLJ

SSoouutthheerrnn AAffrriiccaa''ss ooffffsshhoorree ddiiaammoonnddss.. Mining Journal,London, January 26, 1996, p. 63.An increasing share of South Africa's diamond produc-tion comes from offshore deposits; for instance, in 1994,31% of the country's total output (407,000 carats out of1.31 million carats [Mct]) came from offshore, and 1995production probably exceeded 500,000 carats. "Hugequantities" of gem-quality diamonds have been deposited

GEMOLOGICALA B S T R A C T S

C. W. FRYER, EDITOR

RREEVVIIEEWW BBOOAARRDDCharles E.Ashbaugh III Emmanuel Fritsch Mary L. Johnson Himiko Naka Isotope Products Labortories University of Nantes, France GIA Gem Trade Lab, Carlsbad Pacific Palisades, CaliforniaBurbank, California A. A. Levinson Gary A. Roskin

Anne M. Blumer Michael Gray University of Calgary European Gemological Bloomington, Illinios Missoula, Montana Calgary, Alberta, Canada Laboratory Los

Angeles, California

Andrew Christie Patricia A. S. Gray Loretta B. Loeb James E. Shigley GIA, Santa Monica Missoula, Montana Visalia, California GIA, Carlsbad

Jo Ellen Cole Professdor R. A. Howie Elise B. Misiorowski Carol M. Stockton GIA, Carlsbad Royal Holloway GIA, Santa Monica Alexandria, Virginia

Maha DeMaggio University of London Jana E. Miyahira Duisburg UniversityGIA Gem Trade Lab, Carlsbad United Kingdom GIA, Carlsbad Duisburg, Germany

This section is designed to provide as complete a record as practical ofthe recent literature on gems and gemology. Articles are selected forabstracting solely at the discretion of the section editor and his review-ers, and space limitations may require that we include only those arti-cles that we feel will be of greatest interest to our readership.Inquiries for reprints of articles abstracted must be addressed to theauthor or publisher of the original material.The reviewer of each article is identified by his or her initials at the endof each abstract. Guest reviewers are identified by their full names.Opinions expressed in an abstract belong to the abstracter and in noway reflect the position of Gems & Gemology or GIA.© 1997 Gemological Institute of America

74 Gemological Abstracts GEMS & GEMOLOGY Spring 1997

along a 1,000 km stretch of Namibia's and South Africa'swest coast; the total resource is estimated at 3,000 Met.Most of the deposits are along the Namibian coastbecause of prevailingly northward currents and waves.To date, about 80 Met have come from these deposits,but on-shore production from marine terraces and beach-es in Namibia is expected to decline in the future.

Most offshore diamonds are presently recovered byDe Beers Marine, from Namdeb concessions at 100 mdepth, about 40 km from the coast. However, NamibianMinerals Corp. (Namco) has the largest exploration pro-gram in southern Africa, with concessions at LuderitzHarbor and Hottentots Bay in Namibia and in two areasoff South Africa's west coast; its reserves are estimatedat 80 Met. Namco is using a converted Russian spy shipto mine in water as deep as 100 m.

Diamond Field Resources (DFR) also has a concession(with BHP/Benguela) in the Luderitz area, between DiazPoint at the south and Hottentots Bay in the north,extend ing from the shoreline to 12 km offshore (waterdepths to 125 m). On the basis of modest amounts ofexploration, a resource of over 1 Met, with an averagevalue of $165/ct, has been confirmed. In 1995, DFR wasgranted two South African concessions, including theCape Canyon area (200-500 m depth), which is thoughtto be a major repository of diamonds carried along a for-mer trajectory of the Orange River. Last, Capetown-based Ocean Diamond Mining is expanding its small-scale mining around the 12 Penguin islands, offNamibia's southern coast. MLJ

TThhee ttyyppiiccaall ggeemmmmoollooggiiccaall cchhaarraacctteerriissttiiccss ooff AArrggyyllee ddiiaammoonnddss.. JJ.. Chapman, G. Brown, and B. Sechos,Australian Gemmologist, Vol. 19, No. 8, 1996, pp.339-346.

The Argyle mine produces more diamonds (about 40 mil-lion carats annually) than any other single source in theworld. Yet there is a paucity of published data on thegemo-logical properties and characteristics of these dia-monds.

Most Argyle diamonds are classified as industrial(45%) or near-gem (50%); only 5% are considered gemquality. Of this small percentage of gem-quality dia-monds, the main colors are brown (80%), yellow (16%),colorless (2%), gray (2%), and pink and green (each <1%). This article provides a valuable compilation of theproperties of three color groups of gem-quality material:colorless, brown, and pink-red-mauve.

The following characteristics of these three cate-gories are clearly tabulated and compared: classification(based on nitrogen aggregation), nitrogen content, habit,colors, color zoning, UV fluorescence and phosphores-cence, absorption spectra (infrared, visible, and ultravio-let), and characteristic inclusions. Thirteen color photosillustrate surface and internal features, such as etch pits,color zoning, graining, and inclusions.

The authors found that more than 75% of the inclu-sions were of eclogitic origin, with orange garnet particu-larly abundant. However, epigenetic graphite liningcleavages and fractures was the most common inclusion

in the Argyle diamonds studied. Also summarized is thediscovery and geology of the Argyle pipe. Availablereserves (if underground mining goes below the 300 mplanned for the open pit) total about 100 million tonnes;a diamond con tent of 3.7 ct per tonne is estimated.

AAL

GGEEMM LLOOCCAALLIITTIIEESS

CCaalliiffoorrnniiaa jjaaddee,, aa ggeeoollooggiiccaall hheerriittaaggee.. A. Pashin,California Geology, Vol. 48, No. 6, 1995, pp. 147–154.

Nephrite, and to a lesser extent jadeite, has been found atseveral localities in California. California jade rangesfrom white to almost black, but the most common col-ors are shades of green.

Primary deposits are found throughout the state'sextensive serpentinite zones, with significant localitiesin the counties of Monterey, San Benito, Mendocino,Mariposa, Tulare, Marin, and Siskiyou. However, mostof the material recovered comes from secondary alluvialdeposits of rivers draining the Klamath Mountains,Coast Ranges, and Sierra Nevada. Jade Cove nephrite canbe found in the rough waters of the Pacific Ocean.

Commonly, jade is found in or adjacent to serpenti-nite bodies. An exception is in Riverside County, wherea dark-green-to-black nephrite is found in a contactmeta-morphic zone between quartz monzonite anddolomitic limestone. A Mendocino deposit is unique inthat jadeite and nephrite are found together, often in thesame rock.

The most common California jade simulants are massive vesuvianite (idocrase or californite) and serpentine.The vesuvianite varieties have been marketed as "HappyCamp Jade" and "Pulga Jade." A California serpentinecalled bowenite is marketed as "New Jade." LBL

CChhaarrooiittee:: AA uunniiqquuee mmiinneerraall ffrroomm aa uunniiqquuee ooccccuurrrreennccee..M. D. Evdokimov, World of Stones, No. 7, 1995, pp.3-11.

Charoite was first discovered in 1949, but it was not con-firmed as a new mineral species until 1978. Unlike someother decorative stones, charoite occurs in several differ-ent colors-such as lilac, brown, and gray-often in thesame specimen. Contrary to prior beliefs, charoite is notnamed after the Chara River (the closest bend of which is70 km from the occurrence). "Instead, it was named forthe impression that it gives: chary, a Russian wordmeaning charms or magic."

Charoite occu rs in about 25 areas along the Ditmarand Davan streams. The structure and subsequent min-eralization of the occurrence are quite complicated. Thestages of formation, which began in the early Archean(about 4.6 billion years ago), are described in detail.Charoite formed in breccia zones, the result of a mixtureof solutions that was either deposited between the brec-cia fragments or metasomatically replaced the fragmentsin the zone. Discussions of the origination of the MurunAlkaline Complex and metasomatic rocks within theComplex complete a complicated and highly technicalexplanation of the formation of charoite.

Gemological Abstracts GEMS & GEMOLOGY Spring 1997 75

Eight structurally diverse varieties of charoite reflectthe order in which they were formed. The earliest toform was massive, structureless charoite. This was fol-lowed by parallel fibrous, undulatory-fibrous, felted, radi-al aggregates, mosaic-fibrous, slate-like, and, finally,sheaf-like varieties. Each of these is illustrated by intri-cate pen-and-ink sketches (which, unfortunately, are notclearly identified with figure numbers). Crystallinecharoite has not been seen, so it is impossible to definethe crystal structure and exact chemical formula of thismineral. There are at least nine different variations of theproposed formula. LBL

CCoommbbaarrbbaalliittáá:: AAnn oorrnnaammeennttaall kkaaoolliinniittee ffrroommCCoommbbaarrbbaalláá,, CChhiillee.. R. R. Coenraads, AustralianGemmologist, Vol. 19, No. 8, 1996, pp. 325–334.

Combarbalitá, a beautifully colored and patterned orna-mental clay material, formed through the deep kaoliniticweathering of a metamorphosed volcanic and sedimenta-ry rock sequence. It is found only in a region near thetown of Combarbala, Chile—hence, the material's name.The small town is about 900 m above sea level, 90 kmsouth of Ovalle and 80 km from the Pacific Coast.Mining this ornamental material (by hand) and craftingobjects from it are important activities in this smalltown.

Long before the arrival of the Spanish, the originalinhabitants of the area crafted beautiful pieces from thissoft, easily worked and polished material. Examples ofthis somewhat primitive early art include pendants,necklaces, bowls, plates, arrowheads, and figurines. Theyhave been found mainly in the local indigenous cemeter-ies, or as abandoned domestic items.

Hundreds of small quarries in the hills aroundCombarbalá produce combarbalita, with each quarryyielding specific colors or patterns. The mines are allclose to the surface, the only area from which materialcan be recovered because weathering has made the origi-nal rock soft enough to be worked. The mines areworked by hand with basic tools—crowbars, hammers,and chisels. Pieces larger than 10 cm3 are collected andtested with a hacksaw blade. If the blade easily saws thematerial, then it is soft enough to be turned on a lathe orworked by hand.

Most highly favored is a blue combarbalita that variesfrom a "blue-white" to a "strong sky-blue," very muchresembling turquoise. These scarce varieties occur nearcopper mineralization, which points to copper as the possible coloring element. MD

EEtthhiiooppiiaa:: AA nneeww ssoouurrccee ffoorr pprreecciioouuss ooppaall.. D. B. Hoover,T. Z. Yohannes, and D. S. Collins, AustralianGemmologist, Vol. 19, No. 7, 1996, pp. 303–07.

A new source for precious opal has been discovered nearthe village of Mezezo, Shewa Province, Ethiopia. Thedeposit is located about 200 km northeast of AddisAbaba, on the Ethiopian plateau in central Ethiopia.Although Africa is not known for precious opal, artifactsshow that ancient native cultures used it. Perhaps thesource of these artifacts was the North Shewa deposit

near Mezezo,- the Kushitic people, a group of stone agefarmers, inhabited the region about this time.

Opal-filled lithophysae (hollow, bubble-like struc-tures composed of concentric shells of finely crystallinealkaline feldspar, quartz, and other minerals found insilicic volcanic rock) weather out of the host rhyolite andare found scattered on the ground. The smaller nodulesare remarkably spherical, while the larger ones tend to beslightly elliptical. Careful sampling at one site indicated80 nodules per square meter at the face of the weldedtuft host unit. About 10% of the nodules are solidlyfilled, the remainder have either partial or no opal filling.

The common opal that fills the nodules occurs inmany colors- colorless, white, brownish red, orange, yel-low, greenish, pale lavender, and an unusual chocolate-brown. The opal can be opaque, or it may be semi-translucent to transparent. Much of the material is trans-parent enough to facet; some is reminiscent of redMexican "fire" opal. Precious opal is estimated to be pre-sent in about 1% of the nodules, but much of this has aweak play-of-color. Body color is similar to that found inthe common opal, but no black opal has yet been found.

More than 50% of the precious opal is of the crystalto semi-crystal type. Typically, the play-of-color coversthe spectrum in each piece. Green and blue or only blueplay-of-color is rare; red is common. Pinfire patterns arerare; most examples show "flashfire." In better-qualitymaterial, the play-of-color is evident in many lightingdirections. Some of the opal, both common and precious,has been found to be of the hydrophane type, whichbecomes more transparent when immersed in liquid.Some specimens of translucent gem hydrophane opalbecome completely transparent after they have beensoaked in water for an hour. At that point, these opalslose most or all of their play-of-color, but they regaintheir fire once they are dry.

The long-term stability of this new opal is notknown. Only a limited amount of material from near thesurface has been collected and worked to date.Weathering has already taken its toll. Thus far, theauthors of this arti cle have not experienced any majorproblems, such as frac turing or crazing, during cutting.Government regulations limit opal production to acqui-sition of representative sam ples for testing and evalua-tion, in advance of detailed mapping of the deposit andfiling of a formal development plan. MD

Editor's note: A detailed article on this locality was alsopublished by Johnson et al. in the Fall 1995 issue ofGems & Gemology, Summer 1996, pp. 112–120.

EEuuccllaassee ffrroomm CCoolloommbbiiaa sshhoowwiinngg tthhrreeee--pphhaassee iinncclluussiioonnss.. J.M. Duroc Danner, Journal of Gemmology, Vol. 25,No. 3, 1996, pp. 175-176.

This note documents the properties of a greenish blueeuclase from Colombia. The specimen was notable forits wealth of "well-shaped" three-phase inclusions, asillus trated by a single photomicrograph. Other proper-ties were characteristic for euclase. CMS


GGooooddlleettiittee--aa bbeeaauuttiiffuull oorrnnaammeennttaall mmaatteerriiaall ffrroomm NNeewwZZeeaallaanndd.. G. Brown and H. Bracewell, Journal ofGemmology, Vol. 25, No. 3, 1996, pp. 211–217.

The ornamental gem material from New Zealand called"goodletite" has been known for more than 100 years.Although mentioned as early as 1962, it has not beendetailed in the gemological literature. This is partlybecause of its rarity. It is a coarsely granular rock consisting of red to blue corundum, green Cr-rich margarite-mus-covite mica, and green Cr-tourmaline. Examination oftwo specimens revealed, not surprisingly, variable properties consistent with the composition of this rock. Thebest means of identification is magnification, as illustrat-ed by two photomicrographs. No in situ deposits havebeen located yet, and the availability of this material isstill extremely limited. CMS

IInnvveessttiiggaattiioonnss oonn ssaapppphhiirreess ffrroomm aann aallkkaallii bbaassaalltt,, SSoouutthhWWeesstt RRwwaannddaa.. M. S. Krzemnicki, H. A. Hänni, R.Guggenheim, and D. Mathys, Journal of Gemmology,Vol. 25, No. 2, 1996, pp. 90–106.

Yet another new locality for natural blue and treatablesapphire surfaces in this description of material from theCyangugu district of South West Rwanda. Following sum-mary descriptions of the sapphires' geologic setting andapparent origin from alkali basalts, the results of testing20 samples is reported. Absorption spectra (visible andinfrared), quantitative chemistry (microprobe), and micro-thermometry were performed on two specially preparedsamples. Ten prepared specimens were examined by SEM-EDS for inclusions, and eight rough samples were alsostudied by SEM for detailed surface analysis. The apparently meticulous methodology yielded high-qualitydetailed information that is a welcome addition to thegemological literature, despite the small sample size.Numerous illustrations also indicate that this materialyields fine-quality gems of brilliant blue color. However,no mention is made of its commercial availability orpotential supply. CMS

NNoorrtthh CCaarroolliinnaa''ss lliinnkk ttoo tthhee WWaasshhiinnggttoonn MMoonnuummeenntt.. K.H. Roll, Rock & Gem, Vol. 27, No. 2, February 1997,pp. 32–34.

The author relates the little-known story of the five-pound solid aluminum "pyramidion"—made by process-ing rubies and sapphires mined in North Carolina—thatwas placed atop the Washington Monument in 1885. Thisleads to an account of corundum mining in NorthCarolina and the history of the Corundum Hill mine,once the world's largest producer of commercial corun-dum.

In the 19th century, aluminum was $1 an ounce andthe North Carolina corundum was prized as a source ofnearly pure aluminum oxide for use in the manufacture ofthe newly discovered metal. The Western world's primarysource of corundum-based abrasives and jewel bearingsbefore the first World War, the Carolina deposits wereplayed out by the early 1900s. Still, recreational panning

for the small gem-quality rubies and sapphires, notable forthe distinctive asterism that many reveal, continues to bea tourist attraction in privately owned streambeds aroundthe Corundum Hill area. AC

OOllddeerr aanndd wwiisseerr.. G. Poinar, Lapidary Journal, Vol. 49,No. 10, January 1996, pp. 52–56.

The differences between copal (dried resin) and amber (fossilized resin)—a median of about 200 years versus 2–4million years—is discussed, along with a simple hot-needletest for separating the two similar-appearing mate-rials. Copal is popular as counterfeit amber. Large, color-ful insects or lizards can be pressed into resin, which thenhardens into copal in about a month. The author exam-ined copal from deposits in New Zealand and Colombia(the latter is the basis of a bustling "amber" trade). Carbondating showed that several samples of Colombian copalwere between 10 and 500 years old. Nuclear magneticspectrometry analysis confirmed that the samples did notdiffer significantly from resin from modern trees. Theauthor concludes that most Colombian "amber" is reallycopal, found in fields and alluvial soil. Theoretically,amber could be found there, but it would most likely bein sedimentary formations. AC

JJEEWWEELLRRYY HHIISSTTOORRYY

GGlleeaanniinngg ttrreeaassuurree ffrroomm tthhee SSiillvveerr BBaannkk.. T. Bowden,National Geographic, Vol. 190, No. 1, July 1996, pp. 90–105.

On October 31, 1641, the Spanish galleon Nuestra Señorade la Pura y Limpia Concepción struck a reef north of the(now) Dominican Republic and sank with most of hercrew. The Concepcion was carrying treasure from thePhilippines and Mexico; the reef where she sank wasnamed the "Silver Bank" because the sailors piled treasurefrom the sinking boat on it. They did this not only to savethe treasure but also to provide a higher place to stand, toliterally keep their heads above water.

The first successful salvager was William Phips, whohired native pearl divers to recover goods in 1687; theBurt Webber expedition visited the site in 1978. Theauthor of this article returned to the site recently, under acontract with the Dominican Republic's Commission forUnderwater Archaeological Recovery, and used ship-mounted vacuum-pumps to expose more artifacts. Shownin the article are gold chains, three gold-and-diamond pen-dants, a set of diamond-centered gold studs, and anamethyst ring. It is interesting that some of the porcelainsrecovered from the site show false markings, in an appar-ent attempt to indicate manufacture in an older (and moredesirable) era than the one in which they were actuallymade. MLJ

JJEEWWEELLRRYY MMAANNUUFFAACCTTUURRIINNGG AARRTTSS

EElluussiivvee ttaalleenntt.. Cathleen McCarthy, Lapidary Journal, Vol.50, No. 10, February 1997, pp. 14-17, 66, 68, 70, 71.

This article gives insight into one of America's most cre-

ative carvers, Steve Walters, a true craftsman who canpick out the right piece of material, look at it, and "see"the design. This rare in-depth interview, well worthreading, covers Mr. Walters' beginnings, influences, andcurrent thoughts on his creative talents and their con-tinuing evolution. (He has no formal art training.) Theunpretentious and prolific Mr. Walters insists that he isa craftsman and not an artist. Still, his work is so muchin demand that his booth routinely sells out in the firstfour hours of the week-long Gem and Lapidary DealersAssociation show in Tucson each February. How doeshe come up with his popular and unusual designs? "Idon't know why I do it, and I don't know why it works .. . I've lucked out," he says. AMB

MMaakkiinngg sseemmii--pprreecciioouuss pprreecciioouuss.. C. Frankel, Art &Antiques, Vol. 19, No. 9, October 1996, pp.

90–94.

Andrew Grima has long been a force in jewelry design.His use of unusual materials, from the common to theunique, has marked his flamboyant style for the past 50years. Grima believes that good design need not alwaysuse the biggest stones and most costly materials.Originally a draftsman in a civil-engineering firm, he isthankful that he never formally went to school to learndesign; he would have ended up turning out the samekinds of pieces as everyone else, he believes. His jewel-ry, though based on organic forms, tends toward theabstract, with emphasis on textured gold and exoticgems.

Grima's London gallery was opened in 1966. Hemoved his shop to Gstaad, Switzerland, in late 1993.His clients have included Queen Elizabeth, JacquelineOnassis, and Elizabeth Taylor. JEC

SSttuucckk oonn gglluuee.. A. Oriel, Lapidary Journal, Vol. 50, No.10, February 1997, pp. 34–38, 39, 41, 44.

Andy Oriel talks about a product with which most inthe jewelry and gem trade are unfamiliar—adhesives.These adhesives are very different from the gloppy"white" glue of our childhood. According to five jewelersand lapidaries who use high-tech adhesives in theirwork, the new adhesives stick with a vengeance and lastnot forever, perhaps, but certainly for a very long time.Thomas Harth Ames, Sid Berman, Henry Hunt, MartinKey, and Eugene Millerexplain the different adhesivesthat each uses and their properties, although they do notdetail how the adhesives work chemically. All intervie-wees agree that adhesive has a place in working withgem materials, and they urge that we rethink the stigmaof impermanence sometimes associated with "glue."The use of these new adhesives will allow more jewelryand carving possibilities than ever before. AMB

JJEEWWEELLRRYY RREETTAAIILLIINNGGRReettaaiill ppeeaarrll ssaalleess iinnccrreeaassee.. National Jeweler, February

16, 1997, p. 16.U.S. retail pearl sales increased an average of 18% in

1996 over 1995, according to a Cultured PearlInformation Center (CPIC) poll. Almost 89% of retailerssurveyed reported an increase in their annual culturedpearl sales; 5% reported no change, and 6% reported adecrease. In terms of pearl sales, 41% of the retailersinterviewed said they had an excellent year, 32% saidthey had a good year, 23% said they had a fair year, and4% rated the year poor. The CPIC cited several reasonsfor the large increase in 1996, including more moneyspent on advertising, larger pearl inventories, better-quality pearl jewelry stocked in stores, and greaterawareness of Tahitian and South Sea pearls. In addition,pearls appear to have become a more important fashionaccessory. Retailers expect an average 14% increase in1997 over 1996. MD

TThhee rreevviisseedd FFTTCC gguuiiddeess--wwhhaatt''ss ppeerrmmiissssiibbllee,, wwhhaatt''ss nnoott..Jewelers' Circular Keystone, Vol. 167, No. 8,August 1996, pp. 152–154.

This article focuses on the many changes present in thenewly revised guidelines issued by the Federal TradeCommission for the jewelry industry. Included arereports on some of the commission's internal discussionof the issues involved. Sections on disclosing artificialcoloring, misrepresentation of weight and total weight,and misuse of words such as ruby, sapphire, emerald,and birthstone are examined. This article is particularlyhelpful for anyone involved in manufacturing, sales, orappraisals. JEC

TTRREEAATTMMEENNTTSS

DDyyeedd ooppaalliisseedd ssaannddssttoonnee aanndd ccoonngglloommeerraattee:: AA nneewwpprroodduucctt ffrroomm AAnnddaammooookkaa.. 1. L. Keeling and I. J.Townsend, Australian Gemmologist, Vol. 19, No.5, 1996, pp. 226–231.

Opal-cemented sandstone and conglomerate from theAndamooka opal fields in south Australia is being treated bya process that is said to differ from the traditional sugar-acidmethod. The finished material is marketed for use in carv-ings, inlay work, cabochons, and assembled stones. Becauseof the high porosity of the sandstone and conglomeratestarting material, it was very difficult to remove all traces ofacid from the finished product when the conventional sugar-acid treatment was used. The new treatment process, theexact method for which (a "closely guarded secret") is notgiven in this article, purportedly has solved this problem.

The article did reveal that an organic solution is converted to carbon in a nonoxidizing environment at temperatures above 500°C. This produces an overall darkerbody color in the material, making any play-of-color moreapparent. It also results in the production of the typicalblack microscopic specks of carbon that are associated withthe well-known sugar-acid treatment method. These carbonspecks made the sandstone and conglomerate treated by this"new" method very easy for this gemologist to identify astreated. However, it may not be possible to distinguishbetween this "new" treatment method and sugar-acidtreatment. JIK

Gemological Abstracts GEMS & GEMOLOGY Spring 1997 77

OOnn tthhee iiddeennttiiffiiccaattiioonn aanndd ffaaddee tteessttiinngg ooff MMaaxxiixxee bbeerryyll,,ggoollddeenn bbeerryyll aanndd ggrreeeenn aaqquuaammaarriinnee.. K. Nassau,Journal of Gemmology, Vol. 25, No. 2, 1996, pp.108-115.

Dr. Nassau discusses his latest findings on the distinguishing characteristics and color stability of Maxixe,golden, and green beryl. Having originally made the dis-tinction between Maxixe and "Maxixe-type" beryl in the1970s, he now proposes that both be referred to by theterm Maxixe, on the premise that the distinction "servesno useful purpose in gemmology." Maxixe samples testedin less-than-ideal conditions revealed spectra (680 nmabsorption) typical for Maxixe-type beryl. Theyalso responded to color-stability testing with thefading typical of this material.

Similar testing of four "yellow to greenish" beryls pro-duced no fading, but exposure to 100° ± 10°C heat (indarkness) for a week visibly reduced the yellow colorcomponent of two out of the four. Eight light-green togreenish yellow beryls proved to be aquamarine-type (pre-treat-ment) beryls with no Maxixe component. All of thecolors studied could be natural or the result of laboratoryirradiation, with no possibility of distinguishing betweenthese origins of color. CMS

MMIISSCCEELLLLAANNEEOOUUSS

AA ccrriittiiccaall mmaatttteerr.. M. Heckenberg, Australian Gold Gem& Treasure, Vol. 11, No. 2, February 1996, p. 26.

The working definition of a gemstone's critical angle isthe minimum angle at which the gem can be cut before itwill have a "fish-eye" effect when viewed through thetable. Light generally travels in a straight line, but it trav-els at different speeds in materials of different optical den-sities. Thus, light bends at the interface between media ofdifferent optical densities (such as air and water—a stickappears bent where it goes through the interface betweenthe two). Light passing at an angle into a different medi-um is split into two parts: Some bends into the medium,and some reflects off the surface. At an angle greater thanthe critical angle, no light escapes from the denser medi-um (gemstone) into the less-dense air; all of it is reflectedback into the gemstone.

In cutting a stone, one wants the light coming inthrough the table to be completely internally reflected bythe pavilion facets,- none should "leak out" the back. Thehigher the optical density of the gem material is, the high-er the refractive index—and the smaller the criticalangle—will be. If you know the refractive index of a par-ticular gem material, you can look up the critical angle ina book of mathematical tables. It can also be calculatedfrom the nat ural cosecant of the refractive index. MLJ

PPrrooggrraamm ttoo ssttuuddyy ccrruusstt aanndd mmaannttllee ooff tthhee AArrcchheeaann ccrraattoonniinn ssoouutthheerrnn AAffrriiccaa.. R. W. Carlson, T. L. Grove, M. J.de Wit, and J. J. Gurney, EOS, Transactions, American

Geophysical Union, July 16, 1996, pp. 273, 277.A research program has begun that will systematicallystudy the Kaapvaal craton, which covers 1.2 million km2

(extending across South Africa, Lesotho, Swaziland, andBotswana, and grading into related rocks in Mozambiqueand Zimbabwe as well). The Kaapvaal craton containsrocks between 3.7 and 2.6 billion years old, and was lastdisturbed about 2.5 billion years ago; it is cut by kimber-lite pipes, including the Jagersfontein, Finsch, Premier,Northern Lesotho, Farm Louwrcnsia, and Ramapfeliso.

Cratons are mysterious because of their long-termstability, which is thought to be caused by different typesof sub-basement rocks in the mantle (for instance, therocks under cratons are colder than other mantle regionsat the same depths). Most of the information about theserocks comes from xenoliths in kimberlite pipes; theupcoming study will examine these xenoliths systemati-cally in terms of their chemical and isotopic composi-tions, petrologies, and petrofabrics. Also, a portable seis-mic array will be deployed to look at seismic velocities(and their anisotropies) under the Kaapvaal craton.Experimental petrology of lower crustal rock types (thatis, growing examples of these rocks at high pressures andtemperatures), and studies of the sedimentary history ofthe Kaapvaal craton, are also planned. MLJ

AA ssaallttyy hheerriittaaggee.. J. Raloff, Science News, April 27, 1996,pp. 264–265.

Chandeliers and life-size statues, altars, and "terrazzo"floors-all have been carved from rock salt (halite, NaCl),underground in Poland's Wieliczka mine. The dry air inthe mine has preserved even the earliest carvings, whichdate to the late 17th century; recently, however, high pol-lution levels and variable humidity (exacerbated byhordes of tourists) have eroded fine features of the oldestand most valuable works.

An international team of scientists has studied themoisture-related deterioration. They found that, althoughpure rock salt does not start to pick up water until the rel-ative humidity reaches 75%, pollutant-related salts(suchas ammonium nitrate) that collect on the rock-saltcarvings can absorb water at relative humidities as low as20%. A special dehumidifier has been designed to lowerthe moisture content of the mine's air—and, it is hoped,protect Wieliczka's unusual art. MLJ

WWhhaatt eevveerryyoonnee sshhoouulldd kknnooww aabboouutt tthhee hhaarrddnneessss ooff ddiiaa--mmoonnddss aanndd ootthheerr mmiinneerraallss.. B. Cordua, Rocks Digest,1995, Vol. 8, No. 2, p. 20.The durability and hardness of minerals can be explainedin many different ways. Tests include Mohs and Knoop.Important terms are brittleness, tenacity, and strength.This article not only describes these tests and terms, butit also provides some insight into why various rocks canbe made up of either hard or soft minerals, yet have verydifferent "strengths" due to internal structure. CEA


Documents

Spring 1997 Gems & Gemology - GIA · Geophysics in Gem Exploration GEMS & GEMOLOGY Spring 1997 5 Gemology and geophysics are similar in many ways. The nondestructive techniques that