Diamond Cut: A Foundation for Grading · of blue color in a diamond • Nephrite that mimics serpentine • Pink opal • Rubies clarity enhanced with a lead glass filler • Unusual

THE QUARTERLY JOURNAL OF THE GEMOLOGICAL INSTITUTE OF AMERICA

FA

LL2004

PA

GES

197–282V

OLU

ME

40N

O.3

VOLUME XL FALL 2004

GEM

S & G

EMO

LOG

Y

Diamond Cut:

A Foundation for Grading

pg. 238

VOLUME 40, NO. 3Fall 2004

197

199

230

202

239

240

252

270

273

281

REGULAR FEATURES _____________________2004 Challenge Winners

Lab Notes

Large coral bead necklace • Four blue diamonds from a historic necklace • Irradiated blue diamond crystal • Irradiated type IIb diamond • Unusual causeof blue color in a diamond • Nephrite that mimics serpentine • Pink opal • Rubies clarity enhanced with a lead glass filler • Unusual synthetic ruby tripletdisplaying asterism • Copper-bearing color-change tourmaline from Mozambique

Gem News International

An untreated type Ib diamond exhibiting green transmission luminescence andH2 absorption • Gem amphiboles from Afghanistan, Pakistan, and Myanmar • Recent gem beryl production in Finland • Hessonite from Afghanistan • Interesting abalone pearls • Rhodonite of facet and cabochon quality fromBrazil • Spessartine and almandine-spessartine from Afghanistan • Gem tourmaline from Congo • Dyed horn as an amber imitation • Fake inclusions inquartz • Dyed cultured pearls fading on exposure to heat • Masterpieces ofAmerican Jewelry exhibition

Book Reviews

Gemological Abstracts

Guidelines for Authors

EDITORIAL _____________Unlocking the Secrets of the Fourth CWilliam E. Boyajian

LETTERS ____________

FEATURE ARTICLE _____________A Foundation for Grading the Overall Cut Quality of Round Brilliant Cut Diamonds Thomas M. Moses, Mary L. Johnson, Barak Green, Troy Blodgett, Kim Cino, Ron H. Geurts, Al M. Gilbertson, T. Scott Hemphill, John M. King, Lisa Kornylak, Ilene M. Reinitz, and James E. ShigleyIn the third installment of GIA’s research on diamond cut, the authors describe their use of observation testing to help determine the factors that are important in evaluating the quality of a diamond’s cut. They then intro-duce the new GIA diamond cut grading system, which provides a single overall cut quality grade for standard round brilliants.

NOTES AND NEW TECHNIQUES _________Amethyst from Four Peaks, ArizonaJack Lowell and John I. Koivula

A report on the geology, mining, and gemological properties of amethystfrom the Four Peaks mine, the most important commercial source ofamethyst in the United States.

pg. 249

pg. 203

pg. 261

hen Richard T. Liddicoat created the now de factointernational diamond grading system in the early1950s, he was able to establish standards and

nomenclature for color and clarity that would eventuallybecome part of our everyday diamond lexicon. He was lessconfident in his ability to create a standardized system for cutthat could adequately deal with both the scientific aspects andthe multitude of tastes in the marketplace. Liddicoat thereforedeveloped a system of “corrected weight” from which studentscould arrive at reasonably accurate diamond cut quality (andeven pricing) determinations based on proportion deductionsfrom the so-called American Ideal round brilliant. This systemprovided the basis for GIA’s diamond training in cut for somethree decades. However, when diamond prices skyrocketed inthe late 1970s and subsequently crashed in the early ‘80s, thecorrected weight system became less relevant.

GIA formally revised its diamond courses to reflect thischange in the mid-1980s. For yearsafterward, some in the trade werestill questioning our shift away fromcorrected weight toward a moregeneralized training system that didnot hold the American Ideal cut as astandard to which all other roundbrilliants were compared. And yetanother debate emerged: Some peo-ple wanted a cut grade, while otherswere vehemently against it. Weknew, however, that fundamental tothe concerns of this debate was thequestion of whether cut could beobjectively assessed, and whetherthat assessment could be made sci-entifically, using modern resources.

Our commitment to contemporary research on cut reached anew level in the late 1980s, when computer technology hadadvanced to the point where we could analyze features of dia-mond appearance that heretofore had not been feasible toexplore. As a result, in 1988 we made the critical decision tofund a substantial grant to a young mathematician then ingraduate school at the California Institute of Technology(Caltech) in Pasadena. We had no idea at the time that a pro-ject to develop a highly technical three-dimensional computermodel of a “virtual” diamond using sophisticated ray-tracingsoftware would evolve into a 15-year, multi-million-dollarstudy of every aspect of cut appearance in round brilliant dia-monds. What began as pure research has now resulted in ascientifically based, but eminently practical, grading system forcategorizing cut in standard round brilliants.

The feature article in this issue, authored by Tom Moses and

a host of GIA professionals, is the third in a series of land-mark articles on cut in round brilliants that we are proud tobring you in Gems & Gemology. The first, published in1998, was from a group led by that “young mathematician”from Caltech, Scott Hemphill, and focused on what wecalled “weighted light return,” a metric for reporting bril-liance (which we have since determined is best described asbrightness). The second article, published by Dr. IleneReinitz and co-authors in 2001, focused on dispersed col-ored light return, a metric for fire. These two articles formedthe basis of our assessment of cut in round brilliants, butbrightness and fire alone were not enough.

A third component, scintillation, needed to be examined beforewe could fully understand the factors that contribute to dia-mond cut appearance. Our researchers turned to observationtesting with experienced people from different trade segments,as well as consumers, to explore this element. They found that

although some aspects of scintillation(those related to sparkle) were includ-ed in the brightness and fire metrics,there was also an important underly-ing element that affected the appear-ance of scintillation itself. This ele-ment, called “pattern,” represents thesize and arrangement of bright anddark areas in a diamond, and is a keycomponent of the extent to whichobservers find a diamond attractive.We found from our interaction withthe trade that this phenomenon wasconsidered part of the “life” of a dia-mond—a common term used todescribe desirable stones.

Observation testing and interactionwith the trade also established the importance of other aspectsbeyond cut appearance—design and craftsmanship—in theassessment of a diamond’s overall cut quality. These includedurability and “over-weight” concerns, as well as polish andsymmetry. In addition, we used observation testing to fine-tune our original brightness and fire metrics, so they wouldmore accurately reflect real-world conditions. The current arti-cle discusses all these elements of cut and describes how wevalidated the very sophisticated model we created.

Our authors aren’t the first to use computer modeling to pre-dict the effects of various proportion sets on the cut appear-ance of diamonds. To our knowledge, though, no other orga-nization or research group has validated their models to theextent GIA has with observation testing of actual diamondsby experts in the field, a major part of the research describedin the present article.

EDITORIAL GEMS & GEMOLOGY FALL 2004 197

Wof the FFOOUURRTTHH CC

UUnnlloocckkiinngg the SECRETS

Phot

os b

y H

arol

d &

Eri

ca V

an P

elt

198 EDITORIAL GEMS & GEMOLOGY FALL 2004

While GIA has studied cut in diamonds for decades, ourconcentrated effort—particularly over the past 10 years—hasnow yielded breakthrough research that will undoubtedlyalter historical practices and traditional perceptions of manyin the diamond industry. The main con-clusions are as follows:

• Individual proportions must not beassessed on their own. It is the com-plex interrelationship of individualproportions that matters most in theface-up cut appearance and overallcut quality of a diamond.

• There is no one set of proportions thatyields the most beautiful diamond.Similarly, the long-held view thatexpanding deviations from a fixed,arbitrary set of proportion values pro-duces diamonds with increasinglypoorer appearances is simply not valid.

• Truly consistent and accurate compar-isons of cut in diamond require a stan-dardized viewing and lighting environ-ment that is representative of commonenvironments used in the trade.

• Whereas other systems for assessingcut in round brilliants have from threeto 11 different classification categories,our research found that most individu-als could consistently discern five lev-els of different cut quality.

• For a grading system to be truly unbi-ased and objective, it must allow forpersonal and global preferences in diamond appearance.

Once we determined that a comprehen-sive system for assessing diamond cut inround brilliants could be developed, wewere still left with the fundamental ques-tion of whether a diamond cut grade was useful for the pub-lic and the trade. Although there actually has been a tremen-dous amount of industry interest in the creation of an objec-tive, scientifically based system for assessing diamond cut,ultimately the decisions on this project and the directionsGIA has chosen were derived from its mission—to ensure thepublic trust in gems and jewelry by upholding the higheststandards of integrity, academics, science, and professional-ism through education, research, laboratory services, andinstrument development. GIA firmly believes that the publicinterest is best served by creating such a system, and that itsimpact on the trade will also be positive.

As such, the authors and their colleagues have used thisextensive research to develop a cut grading system for roundbrilliants that will be incorporated into GIA’s diamond train-ing in education and into diamond grading in the laboratory

in mid-2005. For years, many in the tradehave maintained that a wider set of pro-portions could yield an equally beautifuldiamond; to some extent, these people arevindicated by the results of our research.Diamond manufacturers will be able tocut round brilliants to a wider range thanthe current norm and still achieve top-grade, great-looking diamonds. An evenwider range of proportions can producepleasing diamonds in the upper-middle tomiddle grade ranges. Each of these gradeswill, in many cases, allow for greater yieldand weight retention from the rough.

Ultimately, the new GIA diamond cutgrading system will provide answers to thelong-debated questions about the fourth Cin diamond grading. As a result, dealersand retailers will have definitive categoriesfor cut in round brilliants and thus will beable to better serve their clients. And con-sumers will have access to information thatwas heretofore either nonexistent orunavailable as an international standard.GIA will soon propose that global standard.

Over the years, our research objectivesevolved and expanded from first seeking abetter understanding of cut, to establishinga system that would “flag” poorly cut dia-monds, to building one that would alsogive credit to well-made stones. I ampleased to say that our objectives havebeen met with the creation of a truly unbi-ased scientific system to assess the cutgrade of standard round brilliants; further-more, the system has been validated byexpert observers. We believe that this sys-tem will stand the test of time, like thecolor and clarity scales we created morethan 50 years ago. While the process has

been evolutionary, the end result may in fact be revolutionary.Only the future will tell. Research will certainly continue, andit may never stop. Cut is humankind’s unique way of addingvalue to that finely crystallized carbon we have all come toknow and love as the king of gems.

William E. Boyajian, PresidentGemological Institute of America

Phot

os b

y H

arol

d &

Eri

ca V

an P

elt

LETTERS GEMS & GEMOLOGY FALL 2004 199

Following are some of the questions and suggestionswe received in response to this spring’s first-ever Gems& Gemology reader survey. (For a look at some of theresults from the survey, see the editorial in theSummer 2004 G&G, p. 103.) We thank everyone whowrote back with words of praise as well as construc-tive criticism. Since the survey forms were anony-mous, we cannot provide names for those commentsthat were not signed.—Eds.

Issue PackagingI wonder why, when you package the magazine, youconceal the cover photo. If everyone who handledGems & Gemology on the way to delivery could seethe cover, it might help to educate or create an inter-est—who knows? [Anonymous]

Reply: In mailing each issue, we deliberately concealsubject matter such as the cover photo. This is a secu-rity measure that was instituted at the request of sev-eral subscribers, who receive the journal at theirhomes. As a postscript, after we started this practice,we also saw a significant decrease in reports of issuesthat never quite made it into our readers’ mailboxes.

Expressing Magnification in PhotomicrographsThe use of simple numbers such as “30¥” to describethe scale of photomicrographs is unheard of in theworld of credible scientific publication. It is a meaning-less number that is changed at any stage, from the cam-era itself to the digital file and the cropping of thephoto, as well as in the adjustments made by the pub-lisher for the column width, and so on. By the time thisnotation is seen by the reader, all the reader knows isthat the microscope was set on 30¥ for visual observa-tion at the time the photomicrograph was taken.

Respectable scientific journals in all other fieldsrefer to the scale of a photomicrograph with an actualmeasured “field width” of the final image or a simpleline in the image representing millimeters, decimals ofmillimeters, or other appropriate linear measurements.

That being said, I personally think the currentiteration of G&G has done more for the science ofgemology than any other single item or event, and Ilook forward to every issue.

Bob HordLaguna Park, Texas

Reply: This same point was raised by G&G reader Dr.William Hanneman in the Summer 1987 issue (EditorialForum, pp. 111–112). Contributing editor and noted pho-tomicrographer John I. Koivula responded to that letter,and we feel his reply—paraphrased here, with Mr.Koivula’s permission—is just as appropriate today:

The purpose of printing photomicrographs inprofessional gemological publications is to con-vey useful information to the reader. When wesay 45¥, we immediately let the jeweler-gemolo-gist know that this subject would be easilyresolved using a standard gemological microscopewith an upper magnification limit of 45¥ orgreater. If the slide was cropped by the editor andthen enlarged by a factor of 10 to fill the plannedspace, all the editor has done is increase the sizeof the image without gaining or increasing detail.In reality, if the same area was actually viewed at450¥, the scene could be significantly different,with the item in question totally distorted.

From my own experience in the use of precisemeasuring devices in petrology and chemicalmicroscopy, I can understand some of these argu-ments. But each of the sciences has its own set ofstandard practices, and these may or may not applyto others. It is true that we could place a standardscale bar of a specified length in each photomicro-graph as is done in some of the other sciences.However, gemology is more than just a science, itis also an art. As gemologists, we deal with beautyon a daily basis; it is perhaps the greatest appeal ofour profession. And as long as there is no signifi-cant problem generated by the method of magnifi-cation designation as it is currently practiced invirtually all publications in the field, this writersees no reason to detract from the artistic qualityof the image by incorporating a size scale.

LETTERSLETTERS

200 LETTERS GEMS & GEMOLOGY FALL 2004

Use of Technical TermsQuite often, with the exception of prefatory informa-tion and conclusions, the feature articles in G&G aredifficult for me to understand because of the technicalterms. The names of certain advanced instruments,geological terms, mining terms, etc. that are used bythe authors are not found in my GIA course materials.I think it would be very helpful to have a comprehen-sive dictionary or encyclopedia that I could refer towhen I encounter words or concepts that I don’tunderstand. [Anonymous]

Your Summer 2003 issue clearly explained on page 88(of the article by J. L. Emmett et al., “Beryllium diffu-sion of ruby and sapphire”) what SIMS is and how itworks. In my opinion, a publication on non-standardidentification techniques would be of interest. Thissummary should explain how each method works (e.g.,EDXRF, photoluminescence spectroscopy, Raman, etc.)and why it is used. Do you think such a publicationwould be possible?

Heinz KniessRodersdor, Switzerland

Reply: We know that our readers and authors alikecome from many different backgrounds, with educa-tion and experience in a variety of disciplines. Withinthe journal, we make every attempt to define non-standard terms when they are first used. In addition,for some articles on special topics (see, for example,the lead article in this issue, on diamond cut), we willinclude a glossary. GIA is also looking into updatingand republishing the Dictionary of Gems andGemology, which has been out of print for manyyears. In the interim, for geologic and mining terms,we strongly recommend the Glossary of Geology, 4thed. (J. A. Jackson, Ed., American Geological Institute,Alexandria, VA, 1997). For descriptions of many ofthe advanced instruments and techniques currentlyused in gemology, we recommend the article by M. L.Johnson, “Technological developments in the 1990s:Their impact on gemology,” Winter 2000 G&G, pp.380–396. However, these two letters send a clearmessage that, given the rapid advances being made ingemology in the 21st century, we may need to pub-lish such a review article more often than every 10 years.

GIA began its 15-year study of diamond cut by using a computer to model the way light behaveswithin a round brilliant cut diamond. From this model, GIA researchers developed proportion-based metrics to predict how diamonds would perform with regard to brilliance and fire.Continued research revealed several important variables that could not be evaluated effectivelyby computer modeling alone. Thus, the authors asked diamond manufacturers, dealers, retailers,and potential consumers to evaluate brightness (a term selected as more appropriate thanbrilliance), fire, and overall cut appearance of diamonds representing many different proportioncombinations. These observations and discussions confirmed that additional factors, besidesbrightness and fire, contribute to diamond cut appearance, and that factors in addition to face-upappearance are important in assessing the quality of a diamond’s cut. With the trade interactionsas a foundation, the authors (1) tested the brightness and fire metrics to find the best fit withhuman observations, (2) identified and quantified factors in addition to brightness and fire thatcontribute to face-up appearance, (3) developed a standard viewing environment that mimicscommon trade environments, (4) created the foundation for a comprehensive diamond cut grad-ing system, and (5) began development of reference software to predict the overall cut grade of aparticular diamond. The GIA diamond cut grading system described here includes the compo-nents of brightness, fire, scintillation, polish, and symmetry, as well as weight and durability con-cerns, into a single overall grade for cut quality for standard round brilliants.

f the Four Cs (color, clarity, cut, and caratweight), cut is the least understood—andleast agreed upon—aspect of diamond

appearance. Current claims about the superiority ofcertain round brilliant diamond cuts focus mostlyon three approaches:

• The use of specific sets of proportions (e.g., thosefor the AGS 0, the AGA 1A, “Class 1” cuts [aspreviously taught by GIA Education], the HRD“Very Good” grades, “Ideal” cuts, and“Tolkowsky” cuts)

• The use of viewing devices to see specific pat-terns or pattern elements in diamonds (e.g.,FireScope, Symmetriscope, IdealScope, and vari-ous “Hearts-and-Arrows”–style viewers)

• The use of proprietary devices, such as theGemEx BrillianceScope and ISEE2, which mea-sure one or more of the following aspects of dia-mond appearance: brilliance, fire, scintillation,and/or symmetry

For GIA’s research on the evaluation of diamondcut, we started with a different approach, based onthe following questions: What makes a round bril-liant cut (RBC, figure 1) diamond look the way itdoes? To what degree do differences among cutting

202 GRADING OVERALL CUT QUALITY GEMS & GEMOLOGY FALL 2004

O

A FOUNDATION FOR GRADING THEOVERALL CUT QUALITY OF

ROUND BRILLIANT CUT DIAMONDSThomas M. Moses, Mary L. Johnson, Barak Green, Troy Blodgett, Kim Cino, Ron H. Geurts,

Al M. Gilbertson, T. Scott Hemphill, John M. King, Lisa Kornylak, Ilene M. Reinitz, and James E. Shigley

See end of article for About the Authors and Acknowledgments.GEMS & GEMOLOGY, Vol. 40, No. 3, pp. 202–228.© 2004 Gemological Institute of America

proportions create observable distinctions? Whichproportion sets produce results that are deemedattractive by most experienced observers? The firststages of our research—which utilized advancedcomputer modeling—were described briefly byManson (1991), and then in detail by Hemphill et al.(1998) and Reinitz et al. (2001).

Many other groups have used some form of com-puter modeling to predict appearance aspects of dia-mond proportion sets, including: Fey (1975), Dodson(1978, 1979), Hardy et al. (1981), Harding (1986), vanZanten (1987), Long and Steele (1988, 1999),Tognoni (1990), Strickland (1993), Shigetomi (1997),Shannon and Wilson (1999), Inoue (1999), andSivovolenko et al. (1999). To our knowledge, how-ever, few if any of these other studies validated theirmodeling results by using observation tests of actu-al diamonds, a major component of the researchdescribed in the present article. The validation ofcomputer modeling by observations is essential inthe evaluation of diamond cut appearance, as with-out this validation there is a risk of producingresults that are not applicable to the real-worldassessment of diamonds.

In this article, we discuss the key aspects of awell-cut diamond. We describe how we tested ourpreviously published metrics (numerical valuesbased on mathematical models) for brilliance andfire by conducting observations with actual dia-monds in typical trade environments and then

developed new metrics based on our results. Wealso explain how we validated these new metricswith further observation tests, and developed andtested additional methods, including environmentsand procedures, for evaluating other essentialaspects of diamond appearance and cut quality.Finally, on the basis of the information gatheredduring this extensive testing, we constructed a com-prehensive system for assessing the cut appearanceand quality of round brilliant cut diamonds. Thepresent article discusses the framework of this sys-tem, further details of which will be made availablein later publications.

BACKGROUND AND TERMINOLOGYThe face-up appearance of a polished diamond isoften described in terms of its brilliance (or brillian-cy), fire, and scintillation (see, e.g., GIA DiamondDictionary, 1993). Historically, however, diamondappearance has been described using other terms aswell; even the addition of scintillation to this listhas been a relatively recent development (see, e.g.,Shipley, 1948).

Today, while brilliance, fire, and scintillation arewidely used to describe diamond appearance, thedefinitions of these terms found in the gemologicalliterature vary, and there is no single generallyaccepted method for evaluating and/or comparingthese properties in diamonds. Further, experienced

GRADING OVERALL CUT QUALITY GEMS & GEMOLOGY FALL 2004 203

Figure 1. The round bril-liant is the most populardiamond cut. Because of itspopularity, assessment ofthis cut has been the sub-ject of considerableresearch. This image showsa wide range of uses for thisstyle in commercial jewel-ry, as well as loose polisheddiamonds and diamondcrystals. The loose polisheddiamonds weigh 1.05–3.01ct, and the rough crystalsweigh 2.14–2.49 ct. Jewelryand loose polished dia-monds courtesy of BenBridge Jewelers. Compositephoto by Harold & EricaVan Pelt.

members of the diamond trade use additional termswhen they assess the appearance of diamonds. Inthe course of this study, we interviewed dozens ofdiamond manufacturers and dealers, in variousinternational diamond cutting centers and at tradeshows. (We also interviewed retailers and jewelryconsumers, as described below.) We found that, inaddition to brilliance, fire, and scintillation, theytended to use words such as life, pop, lively, dull,bright, or dead to describe a diamond’s cut appear-ance, although they could not always explain pre-cisely what they meant by such terms. In somecases, they would know whether or not they liked adiamond, but were unable to articulate exactly why.

To avoid potential confusion in describing cutappearance, we have refined and expanded the defi-nitions of three essential terms, so that they moreclearly and accurately reflect what experiencedobservers see in actual diamonds in everyday envi-ronments. Throughout this article, we will use thefollowing definitions (see the glossary on p. 223 for alist of key diamond cut terms used in this article):

• Brightness—the appearance, or extent, of internaland external reflections of “white” light seen in apolished diamond when viewed face-up. Notethat although we originally used brilliance todescribe this property (Hemphill et al., 1998;Reinitz et al., 2001), as we proceeded further withour study, we found that many individuals in thetrade and general public include other appear-ance aspects (such as contrast) in their use of thatterm. Hence, we decided to use brightnessinstead.

• Fire—the appearance, or extent, of light dispersedinto spectral colors seen in a polished diamondwhen viewed face-up.

• Scintillation—the appearance, or extent, of spotsof light seen in a polished diamond when viewedface-up that flash as the diamond, observer, orlight source moves (sparkle); and the relative size,arrangement, and contrast of bright and darkareas that result from internal and external reflec-tions seen in a polished diamond when viewedface-up while that diamond is still or moving(pattern).

Note that the definitions for fire and scintillation dif-fer from those currently found for similar terms intheGIA Diamond Dictionary (1993) and those givenin the two earlier G&G articles about this study.They replace those definitions, and brightness

replaces brilliance, for the purposes of this article andthe forthcoming GIA diamond cut grading system.

Our interviews also confirmed that, in additionto brightness, fire, and scintillation, the design andcraftsmanship of the diamond, as evidenced by itsphysical shape (e.g., weight and durability concerns)and its finish (polish and symmetry), are importantindicators of a diamond’s overall cut quality.

MATERIALS AND METHODSThis (third) stage of research evolved from that pre-sented in our previous two articles on diamondappearance (Hemphill et al., 1998; Reinitz et al.,2001). Initially, we focused this stage on exploratorytesting, to compare our computer-modeled predic-tions of brightness and fire with observations byexperienced trade observers of selected actual dia-monds. We found that the observers generallyagreed with each other but, in many cases, not withour predictions. We used these findings to createand test additional brightness and fire metrics, usinga broader group of observers and diamonds.

Extensive observation testing with diamondswas needed to: (1) determine how well the originaland subsequent metric predictions compared toactual observations; (2) establish thresholds atwhich differences defined by the model are nolonger discerned by an experienced observer; (3) seethe broad range of effects that might be statisticallysignificant with a large and varied sample of dia-monds; (4) determine what additional factors mustbe considered when assessing diamond cut appear-ance and quality; and (5) supply enough data foroverall preferences to be revealed amid the widelyvaried tastes of the participants.

Analysis of the observation data did revealwhich metrics best fit our observation results. Italso outlined discernible grade categories for ourmetric results by identifying those category distinc-tions that were consistently seen by observers. Todetermine what additional factors were not beingcaptured by our computer model, we returned tothe trade and asked individuals their opinions ofdiamonds that were ranked with our new bright-ness and fire metrics. Although a majority of thesediamonds were ranked appropriately when metricresults were compared to trade observations, manywere not. By questioning our trade observers, andthrough extensive observations performed by a spe-cialized team (the “Overall observation team”), weexplored additional issues related to face-up appear-



ance (sparkle and pattern) and cut quality (designand craftsmanship) that proved to be essential whenassessing a round brilliant’s cut quality. Theseobservation tests also supplied data that empha-sized the importance of considering personal andglobal preferences when assessing and predictingdiamond cut appearance and quality.

Last, we combined the findings of our observa-tion testing and trade discussions with the predic-tive and assessment capabilities of our brightnessand fire metrics to develop a comprehensive systemcomprised of all the factors identified in this latestphase of research. This became the framework ofour diamond cut grading system.

Methods of Observation Testing. Testing for indi-vidual and market preferences is called hedonicstesting (see, e.g., Ohr, 2001; Lawless et al., 2003) andis often used in the food sciences. Among the typesof tests employed are acceptance tests (to determineif a product is acceptable on its own), preferencetests (comparing products, usually two at a time),difference tests (to see whether observers perceiveproducts as the same or different; that is, which lev-els of difference are perceptible), and descriptiveanalysis (in which observers are asked to describeperceptions and differences, and to what degreeproducts are different). At various times throughoutour research, we used each of these.

The observations focused on individual appear-ance aspects (such as brightness and fire) as well ason the overall cut appearance and quality of pol-ished diamonds. The format and goal of each set ofobservation tests were determined by the questionswe hoped to answer (e.g., Will pairs of diamondsranked in brightness by our brightness metricappear in the same order to observers?), as well asby the findings of previous observation tests. In thisway, as our study evolved, we varied the specificdiamonds used in testing, the environments inwhich the diamonds were viewed, and the ques-tions that we asked.

Our first observation tests for this project wereperformed in February 2001; since then, we havecollected more than 70,000 observations of almost2,300 diamonds, by over 300 individuals. (Approx-imately 200 observers were from all levels of thediamond trade or consumers, and about 100 werefrom the GIA Gem Laboratory and elsewhere atGIA, as described below.)

The trade press has reported on the use of dia-mond observations to test appearance models (e.g.,

Scandinavian Diamond Nomenclature [SCAN DN]in 1967, mentioned by Lenzen, 1983; Nahum Sternat the Weitzmann Institute of Science in Israel,circa 1978 [“Computer used . . .,” 1978]), althoughto the best of our knowledge no results have beenpublished. In addition, we at GIA have used statisti-cal graphics in the past to explain observationalresults (see, e.g., Moses et al., 1997). Thus, this workis an application (and extension) of previouslyapplied techniques.

Diamonds. We purchased and/or had manufactureda set of diamonds of various proportions (somerarely seen in the trade), so that the same set ofsamples would be available for repeated and ongo-ing observation tests. These 45 “ResearchDiamonds” made up our core reference set (seetable 1). Some data on 28 of these diamonds wereprovided by Reinitz et al. (2001).

In our computer model, assumptions were madeabout color (D), clarity (Flawless), fluorescence (none),girdle condition (faceted), and the like. We recognizedthat actual diamonds seen in the trade would differfrom their virtual counterparts in ways that wouldmake the model less applicable. Therefore, to expandour sample universe, we augmented the core refer-ence set with almost 2,300 additional diamonds(summarized in table 2) temporarily made availableby the GIA Gem Laboratory. These diamonds provid-ed a wide range of weights, colors, clarities, and otherquality and cut characteristics. All of these diamondswere graded by the GIA Gem Laboratory and mea-sured using Sarin optical measuring devices. In addi-tion, we developed new methods for measuring criti-cal parameters that previously had not been captured(for a description of the proportion parameters mea-sured and considered, see figure 2).

Observers. Experienced diamond manufacturers andbrokers make purchasing and cutting decisionsbased on aesthetic and economic considerations. Tobegin the verification process for our brightness andfire metrics, we watched these individuals as theyexamined some of our Research Diamonds, both inthe environments where they usually make theirdaily decisions about diamond cut and appearance,and in a variety of controlled environments (detailedbelow). In general, we asked them what we thoughtwere straightforward questions: “Which of these dia-monds do you think is the brightest, the most fiery,and/or the most attractive overall? What differencesdo you see that help youmake these decisions?”

Interactions with trade observers were used intwo ways. First, they provided an initial direction forthis stage of our research project, reinforcing whichaspects of cut quality needed to be considered in addi-tion to brightness and fire. Subsequently, they servedas guidance; throughout our research, we returned totrade observers to compare against the findings wereceived from our internal laboratory teams.

A summary of our observers (including numberand type) is given in table 3. Our core tradeobservers (“Manufacturers and Dealers” and “Re-tailers” in table 3) are experienced individuals fromaround the world who routinely make judgmentson which their livelihoods depend about the quali-ty of diamond manufacture. Many of these menand women have decades of experience in the dia-

TABLE 1. Properties of the core sample group of 45 Research Diamonds.a

Crown Crown Pavilion Table Total Star LowerRD Weight angle height angle size depth length girdle Girdle Girdle Culet Fluores-no. (ct) (º) (%) (º) (%) (%) (%) length (%) thickness condition size Clarity Color cence Polish Symmetry

01 0.61 34.3 15.5 40.6 54 61.2 53.8 81 Thin to Faceted None VS1 E None Very Verymedium good good

02 0.64 33.0 13.0 41.6 59 61.5 55 75 Slightly thick Faceted Very small SI2 E Faint Very Goodto thick good

03 0.55 32.0 11.5 41.0 63 58.6 60 80 Medium to Faceted None VS2 H None Good Goodslightly thick

04 0.70 36.0 15.5 42.0 58 65.4 55 80 Slightly thick Faceted None VVS2 E None Good Veryto thick good

05 0.66 24.0 9.5 42.4 57 58.5 55 85 Medium to Faceted None VS2 F None Very Goodslightly thick good

06 0.59 23.0 9.5 42.0 56 57.2 60 80 Medium to Faceted None VVS2 F Faint Very Veryslightly thick good good

07 0.76 36.5 17.5 41.4 53 64.1 55 90 Thin to Faceted None SI1 F None Very Verymedium good good

08 0.50 33.5 14.0 41.2 57 61.1 55 85 Medium Faceted None VVS1 H None Very Verygood good

09 0.66 23.5 10.0 42.2 55 59.4 60 75 Medium to Faceted None IF F None Very Goodslightly thick good

10 0.68 34.5 16.0 41.0 54 62.1 55 75 Very thin Faceted None VS2 G None Very Goodto medium good

11 0.71 37.0 16.0 42.2 58 64.9 45 85 Medium to Bruted None VS2 D None Good Veryslightly thick good

12 0.71 35.0 15.0 41.0 57 62.6 55 75 Medium to Faceted None SI1 F None Good Veryslightly thick good

13 0.59 33.5 16.0 41.2 52 61.9 60 80 Thin to Faceted None VVS2 E None Very Goodslightly thick good

14 0.71 34.5 14.0 42.0 59 62.4 60 80 Very thin to Faceted None SI1 G None Good Goodslightly thick

15 0.67 25.5 10.0 40.8 59 55.6 55 75 Medium Faceted None VS1 H None Good Good16 0.82 33.5 15.5 40.6 53 61.2 50 75 Thin to Faceted Very small VS1 G None Good Very

medium good17 0.75 26.0 10.0 38.6 59 53.2 50 75 Thin to Faceted None VS2 F None Very Very

medium good good18 0.62 29.0 11.0 41.4 61 57.8 45 75 Medium to Faceted None VVS2 H None Very Very

slightly thick good good19 0.72 29.0 10.5 39.6 62 54.5 50 75 Medium Faceted None VS1 H None Very Very

good good20 0.62 34.5 13.5 40.8 61 59.6 55 80 Medium Faceted None VVS1 I Strong Very Very

blue good good21 0.82 35.5 15.5 41.2 58 62.3 55 75 Thin to Faceted None VVS1 I Strong Very Good

medium blue good22 0.81 35.5 16.5 39.4 54 60.6 55 75 Thin Faceted None VS1 K None Very Very

good good23 0.72 36.5 17.0 40.6 54 63.7 55 80 Medium Faceted None VVS2 I None Very Good

good

a Research Diamonds RD01–RD27 and RD29 were previously reported in Reinitz et al. (2001); variations in proportion values from that article are theresult of recutting, measuring device tolerances, and/or the application of rounding. Verbal descriptions are used here for girdle thickness and culet


mond trade, and most of them routinely handlethousands of polished diamonds per week. (Becauseretailers typically sell diamonds in different envi-ronments from those in which manufacturers anddealers evaluate them [see below], we generallyanalyzed their observations separately.) The resultsof these trade observations were used to define ourinitial quality ranges for brightness, fire, and overall

face-up appearance, as well as to provide usefulinformation on other essential aspects of diamondcut quality.

To expand our population of experienced dia-mond observers, we also established several teamsof individuals from the GIA Gem Laboratory tocarry out the numerous observations that we con-ducted. We developed a team of “Brightness


Crown Crown Pavilion Table Total Star LowerRD Weight angle height angle size depth length girdle Girdle Girdle Culet Fluores-no. (ct) (º) (%) (º) (%) (%) (%) length (%) thickness condition size Clarity Color cence Polish Symmetry

24 0.58 35.5 12.5 39.0 66 56.3 60 75 Thin to Faceted None VVS1 H None Very Goodmedium good

25 0.82 40.0 13.0 42.0 69 60.2 55 75 Thin to Faceted None VVS2 H None Good Verymedium good

26 0.89 38.0 15.0 42.0 61 63.3 55 70 Medium Faceted None VS1 I None Very Verygood good

27 0.44 11.0 15.0 50.8 64 67.8 50 75 Thin to Faceted None VS2 G Strong Very Goodmedium blue good

28b n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a29 0.69 37.5 15.5 42.2 60 62.9 50 75 Thin to Bruted Small SI1 F Faint Excellent Excellent

medium30 0.64 34.5 15.5 40.8 55 60.9 50 75 Medium Bruted None IF I None Very Excellent

good31 0.41 27.0 11.5 40.4 57 58.8 50 75 Slightly thick Faceted Very small VS2 E None Very Good

to thick good32 0.64 35.0 16.5 41.0 53 60.5 45 60 Medium to Faceted Slightly VS2 H Medium Very Good

thick large blue good33 0.64 37.0 16.5 44.0 56 68.0 55 70 Thin to Faceted None VS1 H None Very Very

medium good good34 0.49 41.5 19.5 40.4 56 70.7 55 80 Very thick Faceted None VS1 H None Very Good

good35 0.44 31.0 9.0 43.2 70 58.4 65 80 Thin to Bruted None VS2 D None Good Good

medium36 0.65 37.0 16.5 43.4 57 67.9 55 75 Medium to Faceted None VS2 H None Excellent Very

thick good37 0.50 33.5 9.5 40.2 70 56.9 60 80 Slightly thick Bruted None VS2 F None Good Good

to thick38 0.70 37.0 16.5 41.6 57 69.1 60 85 Very thick Faceted None VS1 H None Very Good

good39 0.70 35.5 15.5 41.2 57 74.0 55 80 Extremely thick Faceted None SI1 F Medium Good Good

blue40 0.70 38.5 14.5 41.0 63 69.3 60 80 Very thick to Faceted None SI1 G None Good Good

extremely thick41 0.71 37.0 17.0 40.2 55 67.3 55 85 Very thick Faceted None VS2 H Medium Good Good

blue42 0.71 37.0 17.0 41.4 54 68.3 55 80 Thick Faceted None VS1 G None Good Very

good43 0.50 38.5 17.5 41.8 57 71.5 55 80 Thick to very Faceted None VVS2 G None Good Good

thick44 0.70 38.0 16.5 41.4 57 68.1 55 80 Medium to Faceted None VVS2 I Faint Good Good

very thick45 0.62 37.0 14.5 45.2 62 69.3 60 85 Medium to Bruted None VS1 F None Good Good

very thick46 0.54 37.0 14.5 37.2 62 54.5 60 85 Extremely thin Bruted None SI2 F None Excellent Good

to thick

size, as they are reported by the GIA Gem Laboratory. Listed properties were determined by the GIA Gem Laboratory.b Not included in sample set for this research because it is a modified round brilliant.

observers” who saw the same differences in bright-ness (within a five-diamond set of our ResearchDiamonds, RD01–RD05; again, see table 1) as ourtrade observers did in a comparable environment.We assembled a different group of specialized indi-viduals to serve as our “Fire observers.” Last, weassembled a team of six individuals from the GIAGem Laboratory (our Overall observation team)who combined had more than 100 years of experi-ence viewing diamonds. This team, whose mem-

bers did not participate in any of the other teams,conducted several sets of tests that focused onjudging diamonds for their overall cut appearanceand quality. The GIA Gem Laboratory observerswere asked to examine larger populations of select-ed diamonds and to answer the same kinds ofquestions as those posed to the trade observers.Early testing showed that the responses of the labgroups were consistent with those of the tradeobservers.

Two other groups who took part in observationswere less experienced GIA personnel and con-sumers. In this way, we met our goal of consideringobservations from people at all levels of the diamondtrade, as well as consumers.

Viewing Environments. To discover how individ-uals in the trade normally evaluate diamonds ona day-to-day basis, we asked them detailed ques-tions about their working environments, and weobserved them while they assessed diamonds inthese environments. This revealed their everydayobservation practices such as colors of clothing,colors of the backgrounds on which they vieweddiamonds, light intensity, lighting and viewinggeometry, light-source specification, and howthey held and moved diamonds when viewingthem.

Our observers examined diamonds in a numberof different environments, some variable and somecontrolled, including:

• Their own offices and workplaces (using desktopfluorescent lamps)

• A conference room at the GIA offices in NewYork (using similar desk lamps and/or the view-ing boxes described below)

• Retail showrooms (usually consisting of a mix offluorescent and spot lighting)

• “Retail-equivalent” environments at GIA inCarlsbad and New York, set up according to rec-ommendations by a halogen light-fixture manu-facturer (Solux)

• Standardized color-grading boxes, including twocommercially available boxes (the GraphicTechnology Inc. “Executive Show-Off” ModelPVS/M—the “GTI” environment—and theMacbeth Judge II Viewing Booth, both with day-light-equivalent D65 fluorescent lamps)


TABLE 2. Ranges of properties and proportions for 2,298other diamonds used for verification testing.a

Parameter Brightness and fire Overall Verificationverification diamonds Diamonds (OVDs)

No. of diamonds 688 1,610Weight range 0.20–1.04 ct 0.25–14.01 ctClarity Internally flawless–I3 Internally flawless–I3Color D–Z D–ZFluorescence None to very strong None to very strongbintensityFluorescence Blue Blue, white, yellowbcolor

Table size 52–72% 46–74%Crown angle 23.0–42.5° 22.5–42.0°Pavilion angle 37.6–45.6° 37.2–44.0°Lower-girdle 60–95% 55–95%facet lengthStar facet length 40–70% 35–70%Depth percent 51.5–71.2 52.8–72.0Crown height 7.0–20.0% 6.5–19.5%Polish Excellent to fair Excellent to fairSymmetry Excellent to fair Excellent to fairCulet size None to very large None to very largeGirdle thickness Very thin to extremely Very thin to extremely

thick thickGirdle condition Faceted, polished, Faceted, polished,

bruted bruted

Total no. observa- 9–29 3–15tions per diamondBrightness 3–11 0c–3observationsper diamondFire observations 5–15 0c–4per diamondOverall appearance 1–3 3–8observations perdiamond

a See figure 2 for a description of diamond proportions mentioned in thistable.b We saw only an extremely small number of fluorescent diamonds in thevery strong range, or in white or yellow; we found the effects of these par-ticular qualities to be insignificant for the diamonds observed.c Brightness and/or fire observations were not conducted for some of theOverall Verification Diamonds.

• At least three versions of a standardized viewingbox of our own design (the common viewingenvironment, or “CVE”)

• A variety of patterned hemisphere environments(to imitate computer-modeled environments)

The same diamond can look quite differentdepending on the type and position of lighting thatis used (figure 3). On the one hand, for cutting dia-monds and for evaluating brightness and the qualityof diamond cutting in general, most manufacturersuse overhead fluorescent lights and/or desk lamps


Figure 2. A round brilliant cut diamond can be described using eight proportion parameters: table size, crownangle, pavilion angle, star facet length, lower-girdle facet length, girdle thickness, culet size, and number of girdlefacets. Other parameters (e.g., crown height) can be calculated from these eight. (A) All linear distances in this pro-file view can be described as a percentage of the girdle diameter, although at the GIA Gem Laboratory girdlethickness and culet size are described verbally based on a visual assessment. (B) In this face-up view of the crown,the star facet length is shown at 50%, so that the star facets extend half the distance from the table to the girdle(indicated here by 0–1). (C) In this table-down view of the pavilion, the lower-girdle facet length is shown at 75%,so that the lower-girdle facets extend three-fourths of the distance from the girdle to the culet center (0–1).Adapted from Reinitz et al. (2001).

TABLE 3. Summary of observers and types of observations.

Trade observers GIA Gem Laboratory observersb

Overallobservationteam

No. ofindividuals 37 159 7 6 6 141 28 384Types of Brightness, fire, Brightness, fire, Brightness Fire Overall Brightness, fire, Brightness, fire,observations overall overall overall overall

a Includes sectors of the trade that work with the public, such as appraisers.b Each of these three teams was composed of members who were not part of other teams.c Includes individuals from the Research department, the GIA Gem Laboratory, and GIA Education.d Includes non-gemological individuals from trade shows and GIA.

Observationgroup

Brightnessteam

Fireteam

Manufacturersand dealers Retailersa Additional GIA

personnelcConsumersd Total

with daylight-equivalent fluorescent bulbs; dealersand brokers generally use similar desk lamps intheir offices (figure 4). However, this type of diffuselighting suppresses the appearance of fire (again, seefigure 3). On the other hand, retail environmentsgenerally provide spot, or point source, lighting(usually with some overall diffuse lighting as well),which accentuates fire (figure 5).

Therefore, when we wanted solely to study theeffects of brightness, we used dealer-equivalent

lighting, which consisted of daylight-equivalentfluorescent lights mounted in fairly deep, neutral-gray viewing boxes (e.g., the Macbeth Judge II, as isused for color grading colored diamonds; see Kinget al., 1994). Similarly, when we wanted to studyonly the effects of fire, we used our retail-equiva-lent lighting, which consisted of a series of threehalogen lamps mounted 18 inches (about 46 cm)apart and six feet (1.8 m) from the surface of thework table, in a room with neutral gray walls thatalso had overhead fluorescent light fixtures.

For observation of overall cut appearance, wedeveloped a GIA “common viewing environment”(CVE [patent pending]), a neutral gray box (shal-lower than the Macbeth Judge II or GTI environ-ment) with a combination of daylight-equivalentfluorescent lamps and overhead white LEDs (light-emitting diodes; figure 6). We established the opti-mum intensity of the fluorescent lamps by observ-ing when a set of reference diamonds showed thesame relative amounts of brightness as theyshowed in the dealer-equivalent lighting. Theintensity of the LEDs was determined by identify-ing a level at which fire was visible in diamondsbut the relative amounts of brightness were stilleasy to observe accurately. In this way, we wereable to observe brightness and fire in a single view-ing environment that preserved the general quali-ties of both dealer and retail lighting.

We also investigated the effects of backgroundcolor (that is, the color in front of which diamondswere observed). Our computer models for bright-ness and fire assumed a black background; yet wefound that most people in the diamond trade usewhite backgrounds of various types (often a foldedwhite business card) to assess diamond appearance.Our observation teams assessed diamonds forbrightness and fire on black, white, and gray trays


Figure 3. A diamond looks different in different lighting and viewing environments. In these images, the samediamond was photographed in diffused lighting (left), mixed lighting (center), and spot lighting only (right).Photos by A. Gilbertson.

Figure 4. Diamond manufacturers and dealers typicallyview and assess diamond appearance and cut qualityin offices with fluorescent desk lamps. Objects in theroom, including the observer, can block or affect light

shining on the crown of a polished diamond.Photo by A. Gilbertson.

to determine if tray color affected brightness andfire results. Additionally, the members of ourOverall observation team observed diamonds onvarious color trays to determine their effect onoverall cut appearance.

For the Brightness and Fire teams, additionalviewing devices were sometimes employed, espe-cially in the early stages of investigation. To testour axially symmetric (that is, hemisphere-like)brightness metrics, we built patterned hemispheres(figure 7; also, see table 1 in theGems &GemologyData Depository at www.gia.edu/gemsandgemology)of various sizes (6, 12, and 16 inches—about 15, 30,and 41 cm—in diameter) in which the diamondswere placed while observers evaluated their rela-tive brightness. The results of these hemisphereobservations were also compared to results fromthe more typical trade environments discussedabove (table 4, “Brightness: verification;” see alsobox A). To be rigorous in our investigation, weexamined a wider range of hemispheres than webelieved were necessary solely to test our bright-ness metrics. In addition, we constructed a “firetraining station,” an environment consisting of alight source and a long tube (figure 8) that enabledFire team observers to grow accustomed to seeingfiner distinctions of dispersed colors in diamonds,and to distinguish among diamonds with differentamounts of fire. Once they were comfortable withthe fire training station, observers made evalua-tions of fire in our retail-equivalent lighting(described above) and, eventually, in our CVE (table4, “Fire: verification”).


Figure 5. Retail environ-ments for diamonds typ-ically use a combinationof spot lighting and dif-fused or fluorescentlighting. Photo courtesyof Dale’s Jewelry, IdahoFalls, Idaho; © Instore .

Figure 6. The GIA common viewing environment(CVE) allows individuals to observe the brightness,fire, and overall cut appearance of a polished dia-mond. This CVE contains daylight-equivalent fluo-rescent lighting (to best display brightness) com-bined with the spot lighting of LEDs (to best dis-play fire) in a neutral gray environment. Photo byMaha Tannous.

http://lgdl.gia.edu/pdfs/cut_table_1.pdf

Early Observation Testing: Brightness and Fire.Our Brightness team examined a set of fiveResearch Diamonds, RD01–RD05 (see table 1), forbrightness differences in the dome environmentsdescribed above. We confirmed that the predictionsof a specific brightness metric (the relative bright-ness order of the five diamonds) matched the obser-vations of the Brightness team in the environmentfor that metric. We then used relative observationsof 990 pairs of Research Diamonds (our core refer-ence set; see table 1 and box A) in dealer-equivalentlighting to select the appropriate brightness metric;that is, we adjusted the modeling conditions (e.g.,lighting conditions or viewing geometry) of ourbrightness metrics until we found one that predict-ed brightness ranking in the same order as theobservation results.

Next, we trained the Fire team to see relativeamounts of fire consistently and asked them tocompare the same 990 pairs of diamonds in a retail-equivalent environment that emphasized thisappearance aspect. Then, as we did with the bright-ness metric, we varied the modeling conditions (in

this case, the threshold levels of discernment) of theReinitz et al. (2001) fire metric to get the best fitwith these observations in this environment.

As part of this early testing process, we alsochose almost 700 diamonds with varying qualitycharacteristics (i.e., with a wide range of clarity,color, symmetry, polish, fluorescence, etc.) and hadboth our Brightness and Fire teams observe them forbrightness and fire in the dealer- and retail-equiva-lent environments. We compared these observationsto brightness and fire metric results to determinewhether any of these characteristics significantlyaffected the correlation between observation resultsandmetric results.

Later Observation Testing: Overall Cut Appearanceand Quality. We used several methodologies forobservation testing of overall cut appearance andquality. One method was to ask observers to look atfive diamonds at a time and rank them from bright-est, most fiery, and/or best looking to least bright,least fiery, and/or worst looking (we also did thisusing three diamonds at a time). We conducted later


Figure 7. Shown here is asmall assortment of the“patterned” hemi-spheres used in the test-ing of the brightnessmetrics. Inset: Using ahemisphere to observediamonds. Photos by A.Gilbertson.

comparisons in a “binary” fashion (that is, compar-ing two diamonds at a time from a set, until eachdiamond had been compared to every other dia-mond in the set). We also conducted observations inwhich diamonds were compared against a smallsuite of Research Diamonds chosen from the corereference set. A fourth methodology consisted ofasking observers to examine larger sets (10 to 24diamonds) and order them by overall appearanceinto as many groups as they wished (for a summaryof all observation tests, again see table 4).

In early sessions, participants were asked toobserve diamonds face-up, without a loupe, whilethe diamonds were in the observation tray.However, we did not restrict their ability to moveor tilt the diamonds, and in most cases participants

tilted or “rocked” them during their examination.Later, when we conducted observations on overallcut quality (as opposed to just face-up appearance),we allowed participants to examine the profiles ofthe diamonds (using a loupe and tweezers) afterthey had provided their first impressions of the dia-monds. This process further helped us recognize theimportance of craftsmanship and other factors inthe assessment of overall cut quality.

In all of these observations, participants wereasked to rate diamonds based solely on face-upappearance or on each diamond’s overall cut quali-ty. Participants were also asked to detail the reasonsfor their decisions (e.g., localized darkness in theface-up appearance or girdles that were “too thick”).These responses along with the participants’ rank-


TABLE 4. Summary of observation tests.

Type of Viewing Type of Diamond Comparison Total no. ofobservation environmenta observerb samples usedc methodd observations

Brightness Manufacturer-equivalent, M&D, GIA personnel, RD01–RD46 Binary, 3x rank, 9,996retail-equivalent, Judge consumers, B-team 5x rank

Brightness: GTI and CVE GIA personnel and Diamonds borrowed Binary with comparison 11,418metric verification B-team from other sourcese “master” diamondsBrightness: Various domes GIA personnel and RD01–RD46 Binary, 3x rank, 17,843metric verification B-team 5x rankBrightness: GTI, Judge B-team Set 1 Binary 280environmentconsistencyFire Manufacturer-equivalent, GIA personnel, RD01–RD46 Binary, 5x rank 688

retail-equivalent B-team, M&DFire: metric Retail-equivalent and F-team Diamonds borrowed Binary with comparison 11,992verification CVE from other sourcese “master” diamondsScintillation Retail-equivalent GIA personnel, Set 1, set 2, diamonds 5x rank 2,122

B-team, F-team borrowed from othersourcese

Overall Retail-equivalent and GIA personnel, RD01–RD46 5x rank, Good/Fair/ 3,608CVE B-team, F-team, Poor rank; dividing

retailers, consumers diamonds into groupsOverall: metric CVE Overall observation Diamonds borrowed Binary with comparison 3,549verification team from other sourcese “master” diamondsOverall: environment CVE with and without Overall observation RD01–RD46 Binary with comparison 396consistency multiple light sources team “master” diamondsBrightness, fire, Retailer environments Retailers Set 1, set 2 5x rank 1,370scintillation, andoverallOverall verification CVE F-team, B-team, Diamonds borrowed Binary with comparison 7,580(brightness, fire, Overall observation from other sourcese “master” diamondsoverall) observations team

a As described in the Materials and Methods section: GTI = Graphic Technology Inc. “Executive Show-Off” Model PVS/M; Judge = Macbeth Judge IIViewing Booth; CVE = the GIA common viewing environment.b Observers are listed as B-team (Brightness team), F-team (Fire team), and M&D (Manufacturers and Dealers). See Materials and Methods section andtable 3 for a description of these teams.c Set 1 consisted of RD01, RD02, RD03, RD04, and RD05; set 2 consisted of RD08, RD11, RD12, RD13, and RD14. See table 1 for properties.d Comparison methods used were binary rank (two diamonds side-by-side), 3x rank (three diamonds side-by-side), and 5x rank (five diamondsside-by-side). “Master” diamonds were chosen from the Research Diamonds.e Summarized in table 2.


ings were then used to develop a methodology foraccurately predicting a diamond’s overall cutappearance and quality.

Computer Modeling and Calculations. Our compu-tational methods for the modeling of brightness andfire were essentially the same as those given in ourtwo previous papers (Hemphill et al., 1998; Reinitzet al., 2001). Although our modeling software iscustom and proprietary, it can be used on any com-puter that can run programs written in the C lan-guage; to calculate the metric results for almost onemillion proportion combinations, we ran them onsixteen 500 MHz Pentium III processors (laterupdated to sixteen 2.5 GHz Pentium IV processors)and two 2.4 GHz Pentium IV processors.

Metrics. We generated more than 75 different, yetrelated, brightness and fire metrics to comparewith our ongoing observations (see table 2 in

the Gems & Gemology Data Depository atwww.gia.edu/gemsandgemology). To define anappearance metric, assumptions must be madeabout the modeled diamond, the modeled observer(position and angular spread of observation), themodeled environment (including illumination), andthe property being quantified.

In the metrics for this work (compared to thosepresented in our two previous G&G articles), wevaried:

• The position of the observer and the angular spreadof observation for brightness.

• The distribution of dark and light in the environ-ment for brightness.

• The absence or presence of front-surface reflec-tions (specular reflection, or “glare”) for brightness.

• The visual threshold for fire. (This was an explic-itly variable factor in our fire metric; again, seeReinitz et al., 2001.)

We collected relative brightness and fire observationson diamonds in many environments, and we exam-ined a number of possible brightness and fire metrics.To compare metric values with observation results,we had to convert both into rank orders.

Members of the Brightness and Fire teams com-pared each of the Research Diamonds to each otherin pairs for brightness or fire, respectively; this gave990 binary comparisons under each condition. As istypical with observation data, not all observersagreed on every result (although some results wereunanimous). This makes sense if the relative rankingof two diamonds is not considered simply as a mea-surement, but as a measurement with some accom-panying uncertainty; that is, a distribution of values.(For example, 4 is always a larger number than 3which is a larger number than 2; but a number mea-sured as 3 ± 1.2 could in fact be greater than 4 or lessthan 2.) We therefore assumed that the observedbrightness (or fire) rank for each diamond could berepresented by a probability distribution, and thenfound the relative order that maximized the probabil-ity of obtaining the observational data we had.

Sometimes, the data showed that all observerssaw one diamond to be better (or worse) than all theothers. In such a case, all the pair-wise comparisons tothat diamond were set aside from the rest of the data

set; this process was repeated, if necessary, to deter-mine the relative order of the remaining diamonds,fromwhich overall rankings could then bemade.

For both observed ranks (described above) andmetric ranks (based on their metric values), we usedscaled rank orders (i.e., the orders did not have to bean integer value, but the highest-ranking diamondcame in first, and the lowest-ranking diamond camein 45th).

The scaled-rank data sets were compared usingthe Pearson Product Moment Correlation. Thismethod produces the “r”-value seen in linear correla-tions (see, e.g., Kiess, 1996; Lane, 2003). The metricwith the highest r-value to the observed data wasselected as the best fitting metric.

We then used Cronbach’s alpha (see, e.g.,Cronbach, 1951; Nunnally, 1994; Yu, 1998, 2001) totest the reliability of the metric predictions relativeto our observers. Cronbach alpha values rangebetween 0 and 1, with near-zero values representingnoncorrelated sets of data. Values of 0.70 and higherare considered acceptable correlations for reliability.More importantly, if results from a predictive systemare added to a dataset as an additional observer andthe alpha coefficient remains about the same, thenthat system is strongly correlated to (i.e., is equallyreliable as) the observers.

BOX A: STATISTICAL EVALUATION OF BRIGHTNESS AND FIRE METRICS

http://lgdl.gia.edu/pdfs/cut_table_2.pdf


As before, the proportions of the modeled dia-monds were the input parameters that determinedthe metric values, so the proportion sets could varywithout changing the fundamental nature of themetrics. Also as in our earlier articles, the comput-er-modeled diamonds were colorless, nonfluores-cent, inclusion-free, and perfectly polished.Although at first we assumed the diamonds werecompletely symmetrical, later we measured all thefacets on certain diamonds to input their exactshapes into metric calculations.

Comparison of the observation results with themetrics proved to be quite challenging, and detailsof some of the statistical methods we used are givenin box A. These tools enabled us to decide which ofour metrics were the most appropriate to predictlevels of brightness and fire (i.e., the calculatedappearance values that best matched results fromobservers looking at actual diamonds in realisticenvironments).

Our new metrics were based on the previouslypublished WLR and DCLR metrics and then furtherdeveloped by varying observer and environmentalconditions, and the effect of glare, until we foundsets of conditions that best fit the observation datain dealer- and retail-equivalent environments. TheHemphill et al. (1998) WLR (weighted light return)metric for brilliance and the Reinitz et al. (2001)DCLR (dispersed colored light return) metric for fireboth assume a distributed observer who is posi-tioned over the entire hemisphere, above the dia-mond, infinitely far away. The weighting for eachpossible angle of observation is determined by anangular relationship to the zenith of the hemi-sphere. (The zenith, looking straight down on thetable of the diamond, is weighted the strongest inthe final result; this is like someone who rocks thediamond, but allows the table-up view to create thestrongest impression.)

To obtain stronger correlations with our dia-mond observation results, this time we also mod-eled a localized observer. This virtual observer onlydetected light from the diamond from a face-upposition and within a narrow—3° angular spread—area (like a person who looks at a diamond from amostly fixed position and from a reasonably closedistance, in this case about 14–20 inches—roughly35–50 cm—as we noted in most trade observations).Although the published WLR observer did notdetect light reflected directly from the upper sur-faces (that is, glare, or luster), for this work we con-sidered brightness metrics both with and without

glare. As for previous metrics, we assumed ourobserver had normal color vision.

Another factor to consider when modeling anobserver for fire is the visual threshold at which anindividual can readily detect colored light. In ourprevious research (Reinitz et al., 2001), we deter-mined visual thresholds by using a hemisphere onwhich chromatic flares from the crown of a pol-ished diamond were reflected. With this hemi-sphere, we concluded that about 3,000 levels ofintensity of the colored light could be observed. Inthe course of our observation tests for firediscernment, we found that an individual could

Figure 8. This configuration was used to train ob-servers to see differences in fire in polished diamonds.Inset: In this schematic diagram, spot incandescentlighting from above is blocked and channeled into along tube that shines a narrow beam of directed lightonto a polished diamond. Photo by A. Gilbertson.

observe more levels of intensity with this hemi-sphere than when observing fire directly from thecrown of a polished diamond. Thus, for the presentwork we varied this threshold in our metric untilwe found the best fit with observation results.

The environment for the WLR metric wasassumed to be a hemisphere of uniform (that is,fully diffused) illumination above the diamond’sgirdle (everything below the diamond’s girdle isdark). By contrast, for the present work we weretrying to model environments and lighting condi-tions used in the trade to buy or sell diamonds.Real-life environments for observing brightness areconsiderably more complicated. For example, lightaround a diamond often is disrupted by objects inthe room, and much of the light directly over a dia-mond’s table is reflected off the observer (again, seefigure 4). We modeled hemispheres with variouspatterns of light and dark (again, see figure 7) untilwe found a modeled environment that closely cor-related with the brightness results from typicaltrade environments.

The environment for the DCLR metric was auniformly dark hemisphere (again, above the dia-mond’s girdle, with all space below the girdle planealso dark) with parallel rays of illumination comingfrom a point light source, centered over the table.This is a reasonable approximation of a single spotlight (for an observer who is not blocking the lightsource, and who is rocking the diamond a lot) or ofmany, arbitrarily placed spot lights, including oneabove the diamond, for an observer who rocks thediamond only a little. For our current research, we

adjusted the visual discernment thresholds withinthe metric to improve correlation with actual obser-vations of fire in retail-equivalent lighting and view-ing environments. This change in metric thresholdswas the only one needed to create a new fire metricthat correlated well with fire observations.

Finally, the property being quantified by WLR(and our new brightness metric, discussed below)was the total amount of white light returned to theobserver from the crown of the diamond (in the caseof the new brightness metric, this includes glare);for DCLR, it was the amount of dispersed coloredlight (i.e., fire) returned to the observer (see table 5for a summary of these model conditions).

Calculations Derived from Standard ProportionParameters. From the eight proportion parametersdescribing a perfectly symmetrical round brilliantcut diamond with a faceted girdle (i.e., table size,crown angle, pavilion angle, star facet length, lower-girdle facet length, girdle thickness, culet size, andnumber of girdle facets; again, see figure 2), it is pos-sible to calculate other proportions and interrela-tionships. These include not only commonly quot-ed proportions such as crown height, paviliondepth, and total depth, but also, for example:

• Facet geometry (e.g., facet surface areas and inter-facet angles)

• Extent of girdle reflection in the table when thediamond is viewed face-up (i.e., if too extensive, a“fisheye” effect)

• Extent of table reflection when viewed face-up

• Several parameters related to localized darknessin the crown when viewed face-up

• Weight-to-diameter ratio

We ran such calculations for all the ResearchDiamonds and for most of the diamonds in table 2;these were used to explore scintillation aspects (seebelow) and other factors related to the physicalshape (e.g., weight concerns) of the diamonds.

Evaluation of Overall (Face-Up) Cut Appearance.Our initial observation tests revealed that, as weexpected, our best brightness and fire metrics wereable to predict specific observation results (i.e.,brightness and fire), but they were not adequate topredict and evaluate a diamond’s overall cut appear-ance and quality. An example of this can be seen infigure 9, which displays brightness and fire metricresults for 165 randomly selected diamonds evaluat-


TABLE 5. Comparison of old and new model conditionsfor calculating brightness and fire.

Modeled Modeled OtherProperty Metric observer environment factors

Brightness Old Spread over 180° White hemi- No glareabove diamond sphereand “weighted”

New Localized 3° Dark circle Glareangular spread with radius included

of 23° aroundzenith

Fire Old Spread over 180° Dark Largeabove diamond hemisphere threshold—and “weighted” 3,000 bright-

ness levelsNew Spread over 180° Dark hemi- Small

above diamond sphere threshold—and “weighted” 18 bright-

ness levels


ed by our Overall observation team for their overallface-up cut appearance. The boundaries on this plotdelineate five discernible appearance categories,which were based on observation results for bright-ness and fire previously obtained for the ResearchDiamond set. Of these 165 diamonds, the overallcut appearance for 95 (58%) was accurately predict-ed using brightness and fire metrics alone. In addi-tion, all the diamonds were within one category ofthe predicted result based only on a combination ofcalculated brightness and fire results.

Obviously, additional factors played a significantrole in the observation results for the remaining42% of these diamonds. Hence, the next stage ofour investigation concerned how to identify andcorrectly evaluate those diamonds for which thebrightness and fire metric results alone did notaccurately predict overall cut appearance, withoutaffecting the results for diamonds already adequate-ly “predicted.”

With this in mind, we looked at comments pro-vided by trade observers and the Overall observa-tion team on the visual appearance of every dia-mond they examined. In many cases, these com-ments supported the metric results (for example,that a diamond was dark overall). In other cases, theobservers’ comments described appearance effects

that caused the diamond to look worse than expect-ed on the basis of brightness and fire alone. Whenwe studied these additional appearance factors, werecognized them as various aspects of scintillation(see box B).

We used specific comments provided by theOverall observation team and by members of thediamond trade to develop methods of capturingscintillation aspects of overall (face-up) appearancethat were not being addressed already by our bright-ness and fire metrics (again, see box B). We used sev-eral rounds of observation tests (listed together intable 4) to create and test a methodology for identi-fying, quantifying, and categorizing the variouseffects that indicate deficiencies in scintillation.

Members of our Overall observation team com-pared “Overall Verification Diamonds” (OVDs;again, see table 2), one at a time, to a suite of appear-ance comparison diamonds assembled from ourResearch Diamonds. (Some OVDs were looked atmore than once, and some were also observed bythe Brightness and Fire teams.) Observations weredone in the CVE environment on gray trays (which,at this point, we had determined were most appro-priate for assessing cut appearance; see Results).These observers were asked to rank the overall cutappearance of diamonds on a scale of 1 to 5, and to

Figure 9. This plot shows165 of the diamonds usedfor our overall observa-tion tests plotted againsttheir brightness and firemetric results. The fivegrading categories delin-eated are based on bright-ness and fire observationresults obtained for the45 Research Diamonds.Although many of these165 diamonds werefound to be predicted cor-rectly by our brightnessand fire metrics (whencompared to overallobservation results),many were not. Thisnecessitated furtherresearch to determinewhat other factors mightbe influencing observers’assessments of face-upcut appearance.

provide specific reasons for the rankings they gave.We used these reasons (which were in the form ofdescriptions about each diamond’s appearance) tofind ways to predict specific pattern-related scintil-lation aspects that caused a diamond to appear less

attractive than expected from our brightness andfire metrics.

This developed into a system for addressingthose diamond proportion sets that led to lower-than-expected appearance rankings (due to pattern-


In recent history, scintillation has been defined asthe “flashes of white light reflected from a polisheddiamond, seen when either the diamond, the lightsource, or the observer moves” (see, e.g., GIADiamond Dictionary, 1993, p. 200). This was wide-ly recognized as the third essential appearanceaspect that worked with brightness and fire to cre-ate the overall face-up appearance of a diamond.

However, we found through our interactionwith members of the diamond trade and our overallobservation tests that scintillation encompassesmore than just this flashing of light. When askedabout the face-up appearance of the diamonds theywere observing, many trade members also men-tioned the importance of the distribution of brightand dark areas seen in the crown of a diamond.Differences in this distribution, especially changesbrought on when the diamond moves, were seen tounderlie and influence the flashes of light describedin the above definition of scintillation.

Thus, given the interdependence of flashinglight and distribution, we decided to use two termsto represent these different aspects of scintillation.Sparkle describes the spots of light seen in a pol-ished diamond when viewed face-up that flash asthe diamond, observer, or light source moves.Pattern is the relative size, arrangement, and con-trast of bright and dark areas that result from inter-nal and external reflections seen in a polished dia-

mond when viewed face-up while that diamond isstill or moving. As such, patterns can be seen aspositive (balanced and cohesive patterns; see figureB-1) or negative (e.g., fisheyes, dark centers, or irreg-ular patterns; see figure B-2).

Many of these pattern-related aspects of scintilla-tion are already taken into consideration by experi-enced individuals in the diamond trade. Often theywere included in the general assessments of dia-monds we recorded during observation tests, usuallydescribed with terms such as dark spots or deadcenters, in addition to fisheyes. Our main findingwas that pattern-related effects were often used todescribe why a diamond did not perform as well as itotherwise should based on its brightness and fire.

Many sparkle-related aspects of scintillation arealready included in our brightness and fire metrics.These consist of specular reflections from facet sur-faces (now included in the brightness metric) andthe dispersed light that exits the crown but has notyet fully separated, so is not seen as separate colorsat a realistic observer distance (included in the firemetric). We also found that sparkle was stronglytied to our fire metric, in that those diamonds thatdisplayed high or low fire were found to displayhigh or low sparkle, respectively. Therefore, weconcluded that we did not need to address sparkleany further. However, we developed proportion-based limits and pattern calculations to specificallypredict and assess the pattern-related aspects ofscintillation.

BOX B: SCINTILLATION

Figure B-1. These diamonds are, in general,viewed positively by experienced members of thediamond trade, due to the overall balance oftheir patterns and the lack of any negative pat-tern-related traits. Photos by A. Gilbertson andB. Green.

Figure B-2. These diamonds are viewed negativelyby experienced members of the diamond trade,due to a variety of unattractive pattern-relatedtraits such as a fisheye (left), dark upper-girdlefacets (center), and a busy, broken overall pattern(right). Photos by A. Gilbertson and B. Green.

related scintillation). We used proportion-range lim-its along with proportion-derived calculations topredict specific pattern-related effects.

As we completed each set of observations, wedeveloped and refined our pattern-related method-ology, so we could test its efficacy during the nextset of observation tests. In this way, we refined pro-portion-range borders as appropriate, adding newpredictive calculations as needed. Thus, we wereable to use early test results to address the addition-al aspects that observers considered (either con-sciously or unconsciously) while assessing overallcut appearance in later tests. In addition, the tens ofthousands of observations we conducted duringthis process have provided a real-world confirma-tion of our predictive system, allowing us to feelconfident in predicted results, even in cases wherewe may not have seen a diamond with that specificset of proportions.

RESULTSBrightness. In early observation experiments, wefound that the WLR (weighted light return) metricof Hemphill et al. (1998), although an accurate pre-dictor of a diamond’s brightness when tested in anenvironment similar to the model, was not as effec-tive at predicting the brightness observations bymanufacturers and experienced trade observers intheir own environments. Consequently, we devel-oped a new brightness metric that included a moreappropriate lighting condition, a more limitedobserver placement, and an additional observationfactor (i.e., glare, the direct reflections off the facetsurfaces).

We first confirmed that observations with hemi-spheres agreed with our predictions of the relativeorder of the diamonds based on the correspondingbrightness metrics. We then used the statisticaltechniques described in box A to determine whichof these metrics gave the best fit to observations ofbrightness in dealer-equivalent environments (e.g.,the GTI, Judge, and CVE). Cronbach alpha valuesfor our brightness testing were determined to be0.74 for observers alone, and 0.79 for observers plusour brightness metric; the closeness of the two val-ues shows that the brightness metric is at least asreliable as the average observer.

Our final brightness metric assumes a diffused,white hemisphere of light above the girdle plane ofthe diamond, with a dark circle located at thezenith of this hemisphere (see figure 10). The area

below the girdle plane is dark. The total angularspread of observation is 3°, located directly over thecenter of the diamond’s table. In addition, glare isincluded in the final metric results.

Fire. Also as described above, the DCLR (dispersedcolored light return) metric of Reinitz et al. (2001)did not correlate well with the collected fire obser-vations in standard lighting and viewing condi-tions. This is probably because it assumed agreater ability to discern fire than observersdemonstrated when they looked at diamondsinstead of projected dispersed-light patterns (seeMaterials and Methods). Therefore, we varied thethreshold for readily observable fire to find the bestfit. Again using statistical methods mentioned inbox A, we found that the best match to the obser-vation data was for a threshold of 101.25, whichgives about 18 distinct levels of light intensity forobserved fire.

Cronbach alpha values for our fire observationswere determined to be 0.72 for observers alone, and0.75 for observers plus our fire metric; again, thecloseness of the two values shows that the fire met-ric is at least as reliable as the average observer.Since the final fire metric correlated well with thefire observation data, we did not vary any of theother model assumptions.


Figure 10. This diagram shows the environmentand viewing conditions for our brightness metric. Itassumes a diffused, white hemisphere of lightabove the girdle plane of the diamond, with a darkcircle located at the zenith of this hemisphere thathas a radius formed by a 23° angle from the cen-tered normal of the diamond’s table. The areabelow the girdle plane is dark, and the angularspread of observation is 3°, located directly over thecenter of the diamond’s table.

The Effect of Other Diamond Properties andConditions on Brightness and Fire. Our Brightnessand Fire teams evaluated the brightness and fire of688 diamonds with a range of color, clarity, polishand symmetry grades, girdle condition (bruted, pol-ished, or faceted), and blue fluorescence1 intensity(from none to very strong), as given in the first col-umn of table 2. From these evaluations, we assessedthe interaction of these properties or conditionswith apparent brightness and fire (by comparing thepredicted metric values of these diamonds). Wefound, as would be expected, that apparent bright-ness decreases as the color of the diamond becomesmore saturated in the GIA D-to-Z range (includingbrowns). Grade-determining clouds in the SI2 and Iclarity grades diminish the appearance of fire. Fairor Poor polish causes both apparent brightness andfire to diminish; and Fair or Poor symmetry nega-tively affects apparent brightness. Neither fluores-cence nor girdle condition showed any effect onapparent brightness or fire.

Addressing Overall Cut Appearance. The next stepwas to compare brightness and fire metric resultswith observer assessments of overall appearance. Forthis exercise, we used the experienced observers whocomprised our Overall observation team and a set of937 diamonds borrowed from various sources (a sub-set of the 1,610 Overall Verification Diamonds). Wealso conducted observation tests with tradeobservers using the core reference set of ResearchDiamonds. Based on tests that placed diamonds intogroups, these two observer populations distinguishedfive overall appearance levels. A number of addition-al results emerged:

1. Differences in body color did not influence theability of observers to assess overall cut appear-ance.

2. To be ranked highest by the observers, a diamondhad to have both high brightness and high firemetric values.

3. Not all diamonds with high values for either orboth metrics achieved the highest rank.

For the subset of 937 Overall Verification Diamondsfor which we had measurements, quality informa-tion, system predictions, and a detailed set of obser-vations, the observer ranks for about 73% corre-

sponded to the ranks that would be anticipatedbased on brightness and fire alone; most of the restwere ranked one level lower than would be expect-ed solely based on those two metrics. An additionalfactor—perhaps more than one—was contributingto overall face-up appearance.

Scintillation. At this point, we did not believe thatdeveloping a specific “scintillation metric” was theright approach. (Recall that most of the sparkleaspect of scintillation was already being captured inour metrics for brightness and fire; again, see box B.)Instead, we needed to find a methodology for cap-turing and predicting the pattern-related effects ofscintillation. We accomplished this using a dualsystem of proportion-based deductions and calcula-tions for specific negative pattern-based featuressuch as fisheyes. (For example, we downgraded dia-monds with pavilion angles that were very shallowor very deep because these proportions generallychanged the face-up appearance of the diamond inways that made it less desirable to experiencedtrade observers.)

Based on the results of the OVD examinations,we found that some overall cut appearance cate-gories were limited to broad, yet well-defined,ranges of proportions. Changes in table size, crownangle, crown height, pavilion angle, star facetlength, lower-girdle facet length, culet size, girdlethickness, or total depth could lead to less desirableappearances. Therefore, based on our observationtesting, we determined limits for each of these pro-portions for each of our overall cut quality cate-gories. We also developed calculations to predictpattern-related effects of scintillation (based on pro-portion combinations) that included the fisheyeeffect, table reflection size, and localized dark areasin the crown when the diamond is viewed face-up(see examples at the end of the Discussion section).A diamond has to score well on each of these pat-tern-related factors to achieve a high grade.

Design and Craftsmanship. After speaking with dia-mond manufacturers and retailers, we verified anumber of additional aspects of a diamond’s physi-cal attributes as important: A diamond should notweigh more than its appearance warrants (i.e., dia-monds that contain “hidden” weight in their girdlesor look significantly smaller when viewed face-upthan their carat weights would indicate; figure 11);its proportions should not increase the risk of dam-age caused by its incorporation into jewelry and


1 Less than 2% of all diamonds that fluoresce do so in colors otherthan blue.


everyday wear (i.e., it should not have an extremelythin girdle); and it should demonstrate the caretaken in its crafting, as shown by details of its finish(polish and symmetry). Diamonds that displayedlower qualities in these areas would receive a loweroverall cut quality grade.

Putting It All Together. Each of these factors(brightness, fire, scintillation, weight ratio, durabil-ity, polish, and symmetry) individually can limitthe overall cut quality grade, since the lowest gradefrom any one of them determines the highest over-all cut quality grade possible. When taken togeth-er, these factors yield a better than 92% agreementbetween our grading system and Overall observa-tion team results (for comparison, observers in ourOverall observation team averaged a 93% agree-ment). Similar to our brightness and fire metrics,these results confirm that our grading system is asreliable as an average observer. Such high agree-ment percentages are considered a reliable measureof correlation in the human sciences; this is espe-cially true in those studies influenced by preference(Keren, 1982). We found that many diamonds inthe remaining percentage were often “borderline”cases in which they could be observed by our teamas a certain grade one day, and as the adjacent gradethe next. The difficulties inherent in the assess-

ment of cut for “borderline” samples are similar tothose faced in the assessment of other quality char-acteristics. Observation testing with members ofthe retail trade and consumers confirmed thesefindings as well.

Grading Environment. When diamonds are beingassessed for overall cut appearance, a standardizedenvironment is essential. Therefore, we developedthe GIA common viewing environment, whichincludes the diffused lighting used by manufactur-ers and dealers to assess the quality of a diamond’scut, and the directed lighting used by many retail-ers, within an enclosed neutral gray viewing booth.Our CVE contains a mix of fluorescent daylight-equivalent lamps (to best display brightness) andLEDs (to best display fire). Observation tests andtrade interaction confirmed that this environmentis useful for consistently discerning differences inoverall cut appearance.

After testing with laboratory observers whowore either white or black tops, we determined thatobservers provided more consistent results forassessing brightness (that is, independent observerswere more likely to reach the same results) whenthey wore a white shirt. Shirt color did not influ-ence fire and overall appearance observations.

During our observation testing with trade

Figure 11. These twodiamonds are fairly sim-ilar in diameter andface-up appearance.However, the diamondon the right containsextra, or “hidden,”weight located in thethickness of the girdle.The diamond on the leftweighs 0.61 ct, while thediamond on the rightweighs 0.71 ct. Photosby A. Gilbertson andMaha Tannous.

members and our Overall observation team, wealso found that in many cases background colorcould affect the ease with which observers distin-guished the face-up appearance of one diamondfrom another. We determined that white trays(which mimic the white folded cards and whitedisplay pads often used in the trade) can some-times cause a diamond to look brighter by hidingor masking areas of light leakage (areas where lightis not returned from the diamond because it exitsout of the pavilion rather than back to the observ-er). Alternately, black trays were shown to demon-strate possible areas of light leakage, but in manycases they overemphasized them so the diamondlooked too dark. We found that a neutral gray tray(similar in color to the walls of our CVE) was themost appropriate choice for assessing a round bril-liant’s overall face-up appearance.

DISCUSSIONThrough our research (computer modeling, observa-tion testing, and trade interaction) we found that tobe attractive, a diamond should be bright, fiery,sparkling, and have a pleasing overall appearance,especially as can be seen in the pattern of bright anddark areas when viewed face-up.

Aspects of overall face-up appearance seen aspositive features include facet reflections of even,balanced size, with sufficient contrast betweenbright and dark areas of various sizes so thatsome minimal level of crispness (or sharpness) ofthe faceting is displayed in the face-up pattern.There are also appearance aspects that are consid-ered negative traits: For example, a diamondshould not display a fisheye or large dark areas inits pattern.

In the same manner, we recognized that morethan just face-up attractiveness should be incor-porated into evaluating overall diamond cut qual-ity. Quality in design and craftsmanship (as evi-denced by a diamond’s weight ratio, durability,polish, and symmetry), even if face-up appearanceis barely affected, also should be evident in a dia-mond’s fashioning.

Overall Cut Grade: Components of the GIADiamond Cut Grading System. Seven components(brightness, fire, scintillation, weight ratio, dura-bility, polish, and symmetry) are consideredtogether to arrive at an overall cut grade in oursystem. These seven components are considered

equally in the system, as the lowest result fromany one component determines the final overallcut grade (e.g., a diamond that scores in the high-est category for all components except durability,in which it scores in the second highest category,would only receive the second highest overall cutgrade; see the pull-out chart for examples). Usingthis approach ensures that each diamond’s overallcut grade reflects all critical factors, includingaspects of face-up appearance, design, and crafts-manship.

In practice, the GIA diamond cut grading sys-tem [patent pending] operates by first establishingthe diamond’s light-performance potential throughmetric calculations of brightness and fire (i.e., thebest grade possible considering the combination ofaverage proportions and how well they worktogether to return white and colored light to theobserver). That potential is then limited by pattern-,design-, and craftsmanship-related determinationsbased on calculations, proportion-range limits, andpolish and symmetry, so that the grade takes intoaccount any detrimental effects. These determina-tions work together with the brightness and firemetrics as a system of checks and balances; the cutgrade of a diamond cannot be predicted by eitherthe metric calculations or any of the other compo-nents alone.

We found through our observation tests thatmost experienced individuals can consistently dis-cern five levels of overall cut appearance and quali-ty. Thus, the GIA diamond cut grading system iscomposed of five overall grade categories.

Design and Craftsmanship. “Over-weight” dia-monds are those with proportions that cause thediamond, when viewed face-up, to appear muchsmaller in diameter than its carat weight wouldindicate. Consider, for example, a 1 ct diamondthat has proportions such that its diameter isroughly 6.5–6.6 mm; this diamond will have theface-up appearance of a relatively typical 1 ctround brilliant. A comparable 1 ct diamond with adiameter of, for example, only 5.7 mm should sellfor less. A person who contemplates buying one ofthese diamonds might believe that the latter was a“bargain” (since both diamonds weigh 1 ct, but thelatter costs less). However, that person would endup with a diamond that appeared smaller whenviewed face-up because much of the weight wouldbe “hidden” in the overall depth of the diamond.Such diamonds are described in the trade as


“thick” or “heavy.” A similar difference in valuewould apply if two diamonds had roughly thesame diameter but one weighed significantly more(again, see figure 11).

Often, an assessment of a diamond as over-weight can be deduced from the combination of itscrown height, pavilion depth, total depth, and/orgirdle thickness. We developed a calculation that


GLOSSARYBrightness the appearance, or extent, of internal and

external reflections of “white” light seen in a pol-ished diamond when viewed face-up.

Brightness team the team of individuals used in obser-vation testing to validate the brightnessmetric.

Common viewing environment (CVE) for this study,a neutral gray box with a combination of daylight-equivalent fluorescent lamps and overhead whiteLEDs (light-emitting diodes), used to view theoverall cut appearance and quality of diamonds.

Computer model a computer program that re-cre-ates the properties and characteristics of an object,along with the key factors in its interaction withspecified aspects of its environment.

Craftsmanship a description of the care that wentinto the crafting of a polished diamond, as seen inthe finish (polish and symmetry) of a diamond.

Design decisions made during the fashioning pro-cess that determine a diamond’s physical shape,as seen in a diamond’s proportions, weight ratio,and durability.

Durability the characteristic of a polished diamondthat accounts for the risk of damage inherent inits proportions (i.e., the risk of chipping in a dia-mond with an extremely thin girdle).

Face-up appearance the sum appearance (brightness,fire, and scintillation) of a polished diamond whenit is viewed in the table-up position. This appear-ance includes what is seen when the diamond is“rocked” or “tilted.”

Fire the appearance, or extent, of light dispersed intospectral colors seen in a polished diamond whenviewed face-up.

Fire team the team of individuals used during obser-vation testing to validate the fire metric.

Metric a calculated numerical result obtainedthrough computer modeling; for the GIA diamondcut research project, metrics were calculated forbrightness and fire for both hypothetical and actu-al diamonds.

Overall cut appearance and quality a description of apolished diamond that includes the face-upappearance, design, and craftsmanship of that dia-mond.

Overall observation team the team of six individuals(who combined had over 100 years of diamondexperience) used during observation testing to dis-cover additional aspects related to face-up appear-ance, as well as to validate the predictions of theGIA cut grading system.

Overall Verification Diamonds diamonds used inthis study to validate the predictive accuracy ofthe GIA diamond cut grading system. Each ofthese diamonds was observed for its overall cutappearance and quality by the members of theOverall observation team.

Research (reference) Diamonds (RD) the core set of45 polished diamonds (which represented a widerange of proportion combinations) that were pur-chased and/or manufactured to be used as a con-sistently available sample group during the courseof the diamond cut study.

Scintillation the appearance, or extent, of spots oflight seen in a polished diamond when it isviewed face-up that flash as the diamond, observ-er, or light source moves (sparkle); and the rela-tive size, arrangement, and contrast of bright anddark areas that result from internal and externalreflections seen in a polished diamond whenviewed face-up while that diamond is still ormoving (pattern).

Weight ratio a description of a diamond’s overallweight in relation to its diameter.

combines the effects of all these factors into onevalue (the weight ratio of a diamond). This ratiocompares the weight and diameter of a round bril-liant to a reference diamond of 1 ct with a 6.55 mmdiameter, which would have a fairly standard set ofproportions (see the pull-out chart for examples).

Durability is another trait of overall diamondcut quality that was emphasized throughout ourinteraction with members of the diamond trade.Diamonds fashioned in such a way that they are atgreater risk of damage (i.e., those with extremelythin girdles) receive a lower grade in the GIA dia-mond cut grading system.

Finish (that is, the polish and physical symme-try of a diamond) also affects cut appearance andquality. Much like weight ratio and durability, pol-ish and symmetry were highlighted by tradeobservers as important indicators of the care andcraftsmanship that went into the fashioning of adiamond, and therefore had to be considered in anycomprehensive grading system. These are assessedbased on standard GIA Gem Laboratory gradingmethodology, and lower qualities of either canbring the grade of the diamond down (again, see thepull-out chart for examples). Note, however, thatunlike other traits, there is not a direct correlationbetween a finish grade and an overall cut grade(e.g., a diamond with a “Very Good” finish mayreceive a top cut grade).

Other Diamond Quality Factors. Our observer testsenabled us to examine the effects of other diamondquality factors (e.g., color, clarity, fluorescence, andgirdle condition) on overall cut appearance.Although in cases of very low color or clarity, wefound some impact on overall appearance, in gener-al observers were able to separate these factors outof their assessments. Therefore, we determined thatthe GIA diamond cut grading system does not needto take these factors into consideration in its finaloverall cut quality grades; it applies to all standardround brilliant cut diamonds, with all clarities, andacross the D-to-Z color range as graded by the GIAGem Laboratory.

Optical Symmetry. One aspect of pattern-relatedscintillation that has gained more attention inrecent years is often called “optical symmetry” (see,e.g., Cowing, 2002; Holloway, 2004). Many peoplein the trade use this term for “branded” diamondsthat show near-perfect eight-fold symmetry by dis-playing eight “arrows” in the face-up position (and

typically eight “hearts” table-down) when observedwith specially designed viewers. To investigate thepossible benefits of optical symmetry, we includedseveral such diamonds in our observation testing.We found that although many (but not all) dia-monds with distinct optical symmetry were ratedhighly by our observers, other diamonds (with verydifferent proportions and, in many cases, no dis-cernible optical symmetry) were ranked just as high.Therefore, both types of diamonds can receive highgrades in our system.

Examples from the GIA Diamond Cut GradingSystem. On the enclosed pull-out chart, we haveprovided three examples (from our core set ofResearch Diamonds) for each of five categories inthe GIA diamond cut grading system, includingtheir proportions and other grade-determining fac-tors. For the purposes of this article only, categoriesare listed as “first” through “fifth,” with “first” rep-resenting the best; this nomenclature does not inany way reflect future nomenclature of the GIAdiamond cut grading system.

In the first category, we see a relatively widerange of proportions. For these three examples,brightness and fire metric values indicated thatthey could belong in the top category. Also, noneof these diamonds were subject to downgradingbased on proportion values or calculated pattern-related scintillation problems. Finally, these dia-monds all had polish and symmetry grades thatwere Very Good. These factors combined to creatediamonds that would receive the highest grade.

Our research found that the top grade includedeven broader proportion ranges than are shown inthe chart. For example, we have established thatdiamonds in this category could have crown anglesranging from roughly 32.0° to 36.0° and pavilionangles ranging from 40.6° to 41.8°. It is important tonote, however, that not all proportions within theseranges guarantee a diamond that would rate a topgrade. As we have previously stated, it is not anyone proportion, but rather the interrelationship ofall proportions, that determines whether a particu-lar diamond will perform well enough to receive atop grade.

By further studying the data in the pull-outchart, one can see various reasons why particulardiamonds would receive a lower cut grade in theGIA system. For example, RD07 falls in the sec-ond category based on its fire and scintillation, itstotal depth of 64.1% and crown height of 17.5%,


and its weight ratio. This is a good example of adiamond where the proportion values cause lowerlight performance and a less-than-optimal face-upappearance.

We have found through our research that pro-portion ranges for the second category are muchwider than those considered by other cut gradingsystems. Likewise, our trade observers were oftensurprised when they learned the proportions of dia-monds they had ranked in this near-top-level cate-gory, although they supported our findings. Here,crown angles can range from roughly 27.0° to 38.0°,and pavilion angles can range from roughly 39.8° to42.4°. Tables also can range from roughly 51% to65% for this grade category. Once again, it isimportant to note that not all individual propor-tions within these ranges guarantee a diamond thatwould fall into the second category.

RD06 on the pull-out chart falls into the thirdcategory in the GIA diamond cut grading systemfor at least two reasons: It has a crown height of9.5% and a crown angle of 23.0°. These factorscombine in this diamond to produce a shallowcrown, which negatively affects overall appearance.In addition, this diamond is downgraded for a lackof contrast in its scintillation and a localized dark-ness in the crown area (especially in the table),which results from the interaction of the shallowcrown with this particular pavilion angle. There-fore, this is a good example of a diamond thatscores high on our brightness and fire metrics, yetis down-graded based on individual proportion val-ues that cause undesirable pattern-related scintilla-tion effects.

It is interesting to note that many in the tradewould not consider cutting a diamond with acrown angle this shallow. Yet our research hasshown that diamonds with these proportions scorein the middle category overall, and might be a veryuseful alternative for diamond cutters in some cir-cumstances. Typical ranges for this grade categoryare roughly 23.0° to 39.0° for crown angles, 38.8°to 43.0° for pavilion angles, and 48% to 68% fortable sizes.

An example of a diamond that would fall in thefourth category is RD37, which has low brightnessand fire metric scores, a table size of 70%, anddowngrading for a fisheye that becomes moreprominent when the diamond is tilted slightly.Here is another example of a “shallow” diamond,but this one is less attractive because of the fish-eye produced by the combination of a large table

and a shallow crown height (9.5%) with a pavilionangle of 40.2%.

RD39 is an interesting example of a diamondthat would receive the lowest grade. Its brightnessand fire metric results—and polish and symmetrygrades (each was assessed as Good)—would placeit in the second category, and a calculated predic-tion for localized darkness would place it in thethird category. However, it falls into the fifth cate-gory in the GIA diamond cut grading systembased on its total depth (74.0%) and its weightratio (which was calculated to be 1.52—that is,52% more “hidden” weight than a diamond withthis diameter should have). Although these pro-portions may seem extreme, this diamond waspurchased in the marketplace. This diamondmight be considered better in a less comprehen-sive system that only accounted for brightness,fire, and finish; however, we believe that this dia-mond’s overall cut quality (which includes itsexcess weight) is properly accounted for andappropriately graded in our system.

Please examine the pull-out chart for additionalexamples of diamonds in the various grade categories.

Personal Preferences and Their Effect on DiamondGrading. Although a diamond’s performance isquantifiable, “beauty” remains subjective. (Thatis, metrics are not subjective but individual tasteis.) No cut system can guarantee that everyonewill prefer one set of proportions over another;instead, as you move down the cut grade scale,the diamonds in the grade categories change fromthose that almost everyone likes, to those thatonly some people might like, to those that no oneprefers. A grading system that fails to acknowl-edge differences in taste is neither practical norhonest in terms of human individuality and pref-erence.

We have found through our research and exten-sive interaction with the trade that even for dia-monds within the same grade, some individualswill prefer one face-up appearance over another(figure 12). Individual preferences have evengreater impact in the lower categories. The inher-ent role of personal preference in diamond assess-ment will often lead to a situation in which someobservers will not agree with the majority; thus,no cut grading system should expect to assess per-ceived diamond cut quality perfectly for everyone.Instead, what we have tried to accomplish withour grading system is to “capture” within each


grade category those diamonds that, in general,most individuals would consider better in appear-ance and cut quality than diamonds in the nextlower category.

CONCLUSIONSDuring the 15 years of our research into the rela-tionship of proportions and overall cut quality, wehave accomplished a great deal: developed a com-puter model and created metrics to predict bright-ness and fire; developed a methodology to validatethose metrics and assess other aspects of cutappearance and quality using observation testing;created a common “standardized” viewing envi-ronment; and, finally, combined all of these ele-ments to create a comprehensive system for grad-ing the cut appearance and quality of round bril-liant diamonds.

In the course of this research (including researchdescribed in our earlier articles by Hemphill et al.,1998, and Reinitz et al., 2001), we arrived at manyconclusions. Among them:

• Proportions need to be considered in an interre-lated manner. The combination of proportions ismore important than any individual proportionvalue.

• Attractive diamonds can be manufactured in awider range of proportions than would be sug-gested by historical practice or traditional tradeperception.

• For consistent comparisons between diamonds,cut grading requires a standardized viewing envi-ronment that is representative of common envi-ronments used by the trade.

• Personal preference still matters. Diamonds withdifferent appearances can be found within eachcut grade, so individuals need to look at the dia-mond itself, not just its grade, to choose the onethey like the best.

Our research and trade interaction also necessi-tated the further refinement of the terms we use todescribe the appearance of a polished diamondwhen it is viewed face-up. Our definitions of theseterms are:

• Brightness—the appearance, or extent, of internaland external reflections of “white” light

• Fire—the appearance, or extent, of light dispersedinto spectral colors

• Scintillation—the appearance, or extent, of spotsof light that flash as the diamond, observer, orlight source moves (sparkle); and the relative size,arrangement, and contrast of bright and darkareas that result from internal and external reflec-tions seen while that diamond is still or moving(pattern)

The GIA Diamond Cut Grading System. Webelieve that to best serve the public and the trade,an effective diamond cut grading system shouldensure that well-made diamonds receive therecognition they deserve for their design, crafts-manship, and execution. Conversely, it shouldensure that diamonds that are not pleasing inappearance, or that warrant a discount for weightor durability reasons, are rated appropriately. Inaddition, the individual categories in this systemshould allow for personal and global differencesin taste.


Figure 12. Both of these dia-monds would score in thetop category of the GIA dia-mond cut grading system,yet we found that differentobservers prefer one or theother based on face-upappearance. Since personalpreference plays such animportant role in perceivedcut quality, it is essentialthat the purchaser examinethe actual diamond (and notrely solely on its proportionsor its cut grade). Photos byA. Gilbertson and B. Green.

Extensive observation testing and trade inter-action made it very clear that for a diamond cutgrading system to be useful and comprehensive, ithad to consider more than just brightness, fire,and scintillation (i.e., more than only face-upappearance). For these reasons, we decided thatour system should also include elements of designand craftsmanship (which can be seen in a dia-mond’s physical shape and finish respectively).Therefore, the GIA diamond cut grading system,which applies to standard round brilliant dia-monds on the GIA D-to-Z color scale, encompass-es the following seven components: brightness,fire, scintillation, weight ratio, durability, polish,and symmetry.

Brightness and fire, including aspects of sparkle-related scintillation, are assessed using computer-modeled calculations that have been refined andvalidated by human observations. Pattern-relatedaspects of scintillation are assessed using a combi-nation of determinations based on proportion rangesand calculations developed to predict specific detri-mental patterns (both derived from observation test-ing). Weight ratio (which is used to determinewhether a diamond is so deep that its face-up diam-eter is smaller than its carat weight would usuallyindicate) and durability (in the form of extremelythin girdles that put the diamond at a greater risk ofdamage) are calculated from the proportions of eachdiamond. Polish and symmetry are assessed usingstandard GIA Gem Laboratory methodology. Thegrading scale for each of these components was vali-dated through human observations; these individualgrades are considered equally when determining anoverall cut grade.

In summary, our research has led us to concludethat there are many different proportion sets thatprovide top-grade diamonds, and even wider rangesof proportions that are capable of providing pleasingupper-middle to middle-grade diamonds. Althoughit is important to consider many components whenassessing the overall cut appearance and quality ofa round brilliant diamond, an individual’s personalpreference cannot be ignored. The GIA cut gradingsystem provides a useful assessment of a diamond’soverall cut quality, but only individuals can saywhich particular appearance they prefer. With thissystem of cut grading, the diamond industry andconsumers can now use cut along with color, clari-ty, and carat weight to help them make balancedand informed decisions when assessing and pur-chasing round brilliant diamonds.

GIA Diamond Cut Grading Reference System.During our research and trade interaction, itbecame clear that for our grading system to be use-ful to all levels of the diamond trade (includingmanufacturers, dealers, retailers, and appraisers), aswell as consumers, we needed to provide a methodfor individuals to predict the cut grade of a polisheddiamond (even if that diamond was only in the“planning” stage of fashioning) from that diamond’sproportion parameters. To this end, we began theprocess of developing reference software.

This software [patent pending] will provide a pre-dicted overall cut grade from proportion values inputby the user, with different versions allowing varia-tion of some or all relevant proportions. Final resultswill be in the form of an estimated overall cut gradeby itself (in the basic version of the application) orthe estimated overall cut grade presented within alarger grid that would allow a user to explore possi-ble alternative proportion sets that might provide animproved final result. GIA plans to release severalversions of this software (as well as a printed version)concurrently with the release of the new cut system.

Next Steps. We plan to incorporate the findings fromthis research, as well as the foundations and frame-work of our cut grading system, into future GIAEducation courses, GIA Alumni and Research pre-sentations, and Institute informational brochures. Inaddition, we plan to incorporate some of this infor-mation (e.g., expanded proportion data and an overallcut grade) into future GIA Diamond GradingReports and the GIA Diamond Dossier®. To thisend, we are also planning to publish future articleson other aspects of the cut grading system, referencesoftware, and changes to GIA Gem Laboratory grad-ing reports.

Although a primary goal of this research projecthas been to develop a cut grading system for roundbrilliant diamonds, there are other benefits that wehave gained from this work. Most importantly, thisresearch project has allowed us to create and validatea method of modeling the behavior of light in a pol-ished diamond along with a methodology to verifythe findings from that modeling using observationtesting by experts in the field. We can now applythese technologies andmethods to other shapes, cut-ting styles, and colors of diamond to determinewhether similar grading systems can be developed.We will continue to identify new goals and ques-tions related to diamond cut as we move forward inour research, beyond the standard round brilliant.



REFERENCESComputer used to set standard of gem beauty (1978) Retail

Jeweller, January 19, p. 40.Cowing M., Yantzer P., Tivol T. (2002) Hypothesis or practicali-

ty: The quest for the ideal cut. New York Diamonds, Vol. 71,July, pp. 40, 42, 44.

Cronbach, L.J. (1951) Coefficient alpha and the internal structureof the tests. Psychometrika, Vol. 16, pp. 297–334.

Dodson J.S.(1978) A statistical assessment of brilliance and firefor polished gem diamond on the basis of geometrical optics.Optica Acta, Vol. 25, No. 8, pp. 681–692.

Dodson J.S. (1979) The statistical brilliance, sparkliness, and fireof the round brilliant-cut diamond. Diamond Research 1979,pp. 13–17.

Fey E. (1975) Unpublished research completed for GIA.GIA Diamond Dictionary, 3rd ed. (1993) Gemological Institute

of America, Santa Monica, CA, 275 pp.Harding B. (1986) GemRay [unpublished computer program].Hardy A., Shtrikman S., Stern N. (1981) A ray tracing study of

gem quality.Optica Acta, Vol. 28, No. 6, pp. 801–809.Hemphill T.S., Reinitz I.M., Johnson M.L., Shigley J.E. (1998)

Modeling the appearance of the round brilliant cut diamond:An analysis of brilliance. Gems & Gemology, Vol. 34, No. 3,pp. 158–183.

Holloway G. (2004) The Ideal-Scope. Precious Metals,http://www.preciousmetals.com.au/ideal-scope.asp [accessedJune 22, 2004].

Inoue K. (1999) Quantification and visualization of diamond bril-liancy. Journal of the Gemmological Society of Japan, Vol. 20,No. 1–4, pp. 153–167.

Keren G., Ed. (1982) Statistical and Methodological Issues inPsychology and Social Sciences Research. L. ErlbaumAssociates, Hillsdale, NJ, 390 pp.

Kiess H.O. (1996) Statistical Concepts for the BehavioralSciences. Allyn and Bacon, Boston, 604 pp.

King J.M., Moses T.M., Shigley J.E., Liu Y. (1994) Color grading ofcolored diamonds in the GIA Gem Trade Laboratory. Gems&Gemology, Vol. 30, No. 4, pp. 220–242.

Lane D.M. (2003) Pearson’s correlation (1 of 3).Hyperstat Online,http://davidmlane.com/hyperstat/A34739.html [accessedAug. 23, 2004].

Lawless H., Chapman K., Lubran M., Yang H. (2003) FS410:Sensory evaluation of food. Cornell University, Fall,http://zingerone.foodsci.cornell.edu/fs410 [accessed Sept. 11,2003].

Lenzen G. (1983) Diamonds and Diamond Grading. Butter-worths, London, p. 164.

Long R., Steele N. (1988) Ray tracing experiment with round bril-

liant. Seattle Faceter Design Notes, January.Long R., Steele N. (1999) United States Faceters Guild

Newsletter, Vol. 9, No. 4.Manson D.V. (1991) Proportion considerations in round brilliant

diamonds. In A. S. Keller, Ed., Proceedings of the Inter-national Gemological Symposium 1991, GemologicalInstitute of America, Santa Monica, CA, p. 60.

Moses T.M., Reinitz I.M., Johnson M.L., King J.M., Shigley J.E.(1997) A contribution to understanding the effect of blue fluo-rescence on the appearance of diamonds. Gems & Gemology,Vol. 33, No. 4, pp. 244–259.

Nunnally J.C. (1994) Psychometric Theory, 3rd ed. McGraw-Hill,New York, 752 pp.

Ohr L.M. (2001). Formulatingwith sense. Prepared Foods, February;http://www.preparedfoods.com/CDA/ArticleInformation/features/BNP__Features__Item/0,1231,114431,00.html[accessed Aug. 23, 2004]

Reinitz I.M., Johnson M.L., Hemphill T.S., Gilbertson A.M.,Geurts R.H., Green B.D., Shigley J.E. (2001) Modeling theappearance of the round brilliant cut diamond: An analysis offire, and more about brilliance. Gems & Gemology, Vol. 37,No. 3, pp. 174–197.

Shannon P., Wilson S. (1999) The great cut debate rages on.Rapaport Diamond Report, Vol. 22, No. 5 (February 5), pp.89–90, 95.

Shigetomi G.T. (1997) Diamond industry breakthrough. BangkokGems& Jewellery, Vol. 10, No. 8, pp. 97, 98, 100.

Shipley R.M. (1948, date approximate) Assignment 2-30: Thecritical angle and comparative brilliancy. Diamonds Course,Gemological Institute of America, Los Angeles, p. 1.

Sivovolenko et al. (1999) http://www.gemology.ru/cut/english/document4.htm [accessed Dec. 15, 2003].

Strickland R. (1993) GEMCAD [computer program]. Austin,Texas.

Tognoni C. (1990) An automatic procedure for computing theoptimum cut proportions of gems. La Gemmologia, Vol. 25,No. 3–4, pp. 23–32.

van Zanten P.G. (1987) Finding angles for optimal brilliance bycalculation. Seattle Faceter Design Notes, November.

Yu C.H. (1998) Using SAS for item analysis and test construc-tion, http://seamonkey.ed.asu.edu/~alex/teaching/assess-ment/alpha.html [accessed Aug. 23, 2004].

Yu C.H. (2001) An introduction to computing and interpretingCronbach Coefficient Alpha in SAS. In Proceedings of the 26thSAS Users Group International, April 22–25, 2001, Long Beach,California, Paper 246-26, http://www2.sas.com/proceedings/sugi26/p246-26.pdf [accessed Aug. 23, 2004].

ABOUT THE AUTHORSMr. Moses is vice president of Identification and ResearchServices, Mr. King is Laboratory Projects officer, and Dr.Reinitz is manager of Research and Development at the GIAGem Laboratory in New York. Dr. Johnson is manager ofResearch and Development, Mr. Green is TechnicalCommunications specialist, Mr. Gilbertson is a researchassociate, Dr. Shigley is director of GIA Research, and Ms.Cino is director of Administration for the GIA Gem Laboratoryin Carlsbad, California. Dr. Blodgett is a research scientist forGIA Research, and is located in Flagstaff, Arizona. Mr. Geurtsis Research and Development manager at GIA in Antwerp,Belgium. Mr. Hemphill is a research associate for the GIAGem Laboratory and is located in Boston, Massachusetts.Ms. Kornylak is a Research Laboratory technician for GIAResearch, and is located in Tyler, Texas.

ACKNOWLEDGMENTS: The authors wish to thank the othermembers of the GIA Diamond Cut Team, past and present, in

addition to those listed as authors: Kelly A. Yantzer, Phillip M.Yantzer, Russell Shor, Brooke Goedert, and Jim Enos. JohnMcCann and Dr. Elliot Entin provided useful perspectives fromother sciences; Peter De Jong also provided assistance inarranging and conducting observation testing in Antwerp.

The authors are very appreciative of those who gave theirtime and expertise while taking part in diamond observationsand sharing their insights on various issues related to thisproject. Many diamond manufacturers, brokers, dealers,retailers, and other trade-related individuals helped during thecourse of this research. For a full list of acknowledgments,please see the Gems & Gemology Data Depository atwww.gia.edu/gemsandgemology.

Invaluable assistance was also received from many dedicatedindividuals within various departments at GIA. The authorsthank all of these individuals for their help in bringing this pro-ject to fruition.

http://lgdl.gia.edu/pdfs/acknowledgments_cut.pdf

For more than a century, the Four Peaks mine inMaricopa County, Arizona, has produced gem-quality amethyst from crystal-lined or crystal-filledcavities and fractures in a brecciated quartzite hostrock. Crystals from the deposit exhibit a range ofpurple colors, uneven color zoning, and variabletransparency, which present challenges for obtain-ing a steady supply of material suitable for faceting.Faceted material may display fluid inclusions andtiny reddish brown hematite flakes, growth zoning,and Brazil-law twinning, all of which provide visualclues to separating the Four Peaks material fromsynthetic amethyst. Recovery of amethyst at thisdeposit continues at this time on a limited basis.

he most important commercial source of gem-quality amethyst in the United States is theFour Peaks mine in Maricopa County,

Arizona (figure 1). Discovered by accident in theearly 1900s by a gold prospector, this deposit hasbeen worked intermittently on a small scale eversince. Good-quality amethyst from a U.S. occur-rence has special value in the marketplace, but thechallenge at this deposit has been to produce suffi-cient quantities of commercial-grade material on acontinuing basis to satisfy market demand.

The mine’s restricted accessibility and remotelocation in the rugged Mazatzal Mountains of cen-tral Arizona, combined with the low market valueof amethyst from all sources, has limited produc-tion in the past. In the late 1990s, the mine wasreopened with the prospect of providing a regularsupply of calibrated material in a range of sizes

and shapes (Lurie, 1998, 1999; Johnson andKoivula, 1998). Currently, it is owned and operat-ed by Four Peaks Mining Co. LLC of OceanGrove, New Jersey. The deposit is again producinggood-quality amethyst, with some of the cutstones exceeding 20 ct.

This article provides a brief description of thegeologic occurrence of amethyst at the deposit andsummarizes the gemological properties of thismaterial. Although previous descriptions of thedeposit and local geology can be found inSinkankas (1957, 1976, pp. 373–374), Lowell andRybicki (1976), Estrada (1987), Chronic (1989, pp.174–175), and Lieber (1994, p. 122), this article pro-vides the first gemological characterization of FourPeaks amethyst.

LOCATION AND ACCESSThe Four Peaks mine is named for the mountain areawhere it is located. The Four Peaks are four promi-nent, steep-sided mountains aligned north-south nearthe southern end of the 80-km-long Mazatzal moun-tain range (figure 2). This area is approximately 75km (46 miles) east-northeast of Phoenix (and is visi-ble from the city) within the Four Peaks Wildernessarea of the Tonto National Forest (figure 3). Theamethyst deposit is located on a 20-acre patentedmining claim that lies below the second peak fromthe south (again, see figure 2) at an elevation of 1,980

230 AMETHYST FROM FOUR PEAKS, ARIZONA GEMS & GEMOLOGY FALL 2004

T

AMETHYST FROMFOUR PEAKS, ARIZONA

Jack Lowell and John I. Koivula

See end of article for About the Authors and Acknowledgments.GEMS & GEMOLOGY, Vol. 40, No. 3, pp. 230–238.© 2004 Gemological Institute of America

NOTES & NEW TECHNIQUES

m (6,500 ft) above sea level. The area has sparse vege-tation and an arid, high-desert climate.

The workings consist of an elongate open cutand a tunnel that penetrates about 10 m into themountainside (figures 4 and 5). Because of the veryrugged terrain and location in a wilderness area,access is limited to foot travel or helicopter. Exceptduring the winter months, when the peaks aresometimes covered with snow, the area can beapproached by vehicle on U.S. Forest Service roadsfrom State Route 188 or Highway 87. A narrow trailthen climbs approximately 760 m over a distance of

7.2 km to the mine. Entrance to the tunnel is closedwhen the mine is not in operation, and at all timesprior permission from the mine owners is requiredto enter the property.

GEOLOGYThe Four Peaks are eroded remnants (what geolo-gists call “roof pendants”) of Precambrian-agemetasedimentary rocks that were intruded frombelow by a granitic batholith. The amethyst miner-alization occurs within one stratigraphic unit of

AMETHYST FROM FOUR PEAKS, ARIZONA GEMS & GEMOLOGY FALL 2004 231

Figure 1. These two cutstones (19.25 and 17.02ct) are excellent exam-ples of the fine-qualityamethyst obtained fromthe Four Peaks mine inMaricopa County,Arizona. Courtesy ofJack Lowell; photo © Jeff Scovil.

Figure 2. This aerialphotograph, taken in

1994 toward the north-east, shows the Four

Peaks mountains at thesouthern end of Mazat-zal Range. The tallest,

on the far left, is knownas Browns Peak. The

trail and mine are visi-ble in the right fore-

ground. Photo by ToddPhotographic Services;

courtesy of CommercialMineral Co.

these rocks: the Mazatzal Formation, a light-coloredquartzite consisting of a tough, closely packedaggregate of cemented, angular quartz fragments.This quartzite and other sedimentary units wereuplifted during the intrusion of the batholith in themiddle Proterozoic; the uplifting was accompaniedby faulting and brecciation of zones in the quartzite.

Erosion since Precambrian times has exposed por-tions of both the granite batholith and the overlyingMazatzal Formation.

Quartz deposition (colorless and smoky quartz, aswell as amethyst) is thought to have occurred in sev-eral stages along fractures and cavities in the brec-ciated quartzite. These crystal-lined spaces are irreg-ular in form and can vary from several centimetersto a few meters in maximum dimension. Smalleropenings are completely filled with interlockingamethyst crystals (figure 6), whereas larger cavitiesoften contain either crystal druses attached to cavitywalls, or loose crystals suspended in a vug-fillingalteration material. The amethyst appears to haveformed along fractures by crystallization of silica-containing hydrothermal solutions thought to havebeen derived from the cooling granite batholith.Fractured zones within the quartzite exhibit evi-


Figure 3. The amethyst deposit is located within theFour Peaks Wilderness area, approximately 75 kmeast-northeast of Phoenix.

Figure 4. An elongate open cut (in the center andupper right) can be seen from the trail approaching the

Four Peaks mine. Rugged outcrops are formed by theMazatzal quartzite. Photo by J. E. Shigley.

Figure 5. A short tunnel penetrates about 10 m fromthis entrance. All the Four Peaks amethyst pro-duced in the past seven years has come from theunderground workings. Courtesy of CommercialMineral Co.


dence of hydrothermal alteration from these or othersolutions. This evidence includes not only thequartz mineralization, but also the occurrence offine-grained apatite, hematite, and unidentified clayminerals within open spaces (Sinkankas, 1957).

MININGRecovery of amethyst for both mineral specimensand gemstones has taken place on a limited scale andintermittent basis. Initial work was carried out fromsurface outcrops of the amethyst-mineralized zones.Later, mining by open-pit methods employed handtools and (unsuccessfully) a bulldozer. More recently,a tunnel was driven by hand to access productiveareas of the deposit. Several measures have recentlybeen implemented so that the operation is legallycompliant with mine safety regulations. A crew ofthree miners is working the deposit within the hori-zontal tunnel. Explosives are rarely used, and only tobreak up boulders when needed. Otherwise, the min-ers use a hand-held pneumatic chisel for digging.Challenging conditions are caused by the remotenessof the location as well as hot summer temperaturesand the lack of water and power at the mine. Gemmaterial is taken out on foot or by helicopter.

Since 1998, Commercial Mineral Co. ofScottsdale, Arizona, has had an exclusive arrange-ment with the mine owner to acquire the mined

material. However, a certain amount of illegal min-ing has occurred over the years, with the materialperiodically sold to local gem cutters.

DESCRIPTION OF THE AMETHYSTThe amethyst crystals from this locality display amorphology that is typical of gem amethyst world-wide. Rhombohedral faces are predominant, whilethe prism faces are either poorly developed or absent(figure 7). This habit is common for quartz crystals

Figure 6. Amethyst mineralization occurs along frac-tures and cavities in the brecciated quartzite, asshown in this May 2004 image. Courtesy ofCommercial Mineral Co.

Figure 7. Amethyst crys-tals from the Four Peaks

mine display a typicalmorphology for ame-

thyst, consisting mainlyof rhombohedral faces.

Facetable materialvaries from light to dark

purple, and generallyonly small portions of

the crystals are suitablefor faceting. The cut

stones shown here rangefrom 3.83 to 13.00 ct.

Courtesy of Commer-cial Mineral Co.


that grew simultaneously on the walls of open cavi-ties. The best cutting material shows dissolutionbasal pinacoid (c) faces, which are rare in naturalamethyst. Euhedral crystals of good form and withlustrous faces are seen rarely at this deposit; whenfound, they range from a few centimeters to about20 cm in maximum dimension. More typically,crystals (or crystal fragments) are etched and corrod-ed due to attack from late-stage hydrothermal solu-tions. These solutions appear to have preferentiallyattacked the joints between adjacent crystals, there-by loosening the crystals and perhaps causing themto break away from a small point of attachment atthe base. Crystal faces are normally frosted or coatedby apatite or hematite. Some display internal areasthat have a hazy or translucent fog-like appearance.

Four Peaks amethyst crystals show great variabil-ity in the distribution and quality of color, whichcan range from light to dark purple and includessome purplish red material. Most crystals display anuneven color distribution (figure 8), with darker pur-ple areas separated by sharp boundaries from areasthat are lighter or near-colorless. Color banding isoriented parallel to the rhombohedral crystal faces,with most crystals showing the strongest colorzones under the larger rhombohedral faces.

MANUFACTURING AMETHYST FOR JEWELRY PURPOSESFour Peaks amethyst crystals also vary in size andquality. The rough is cobbed or trim sawn to pro-duce relatively clean, darker pieces suitable forfaceting. Because of the color zoning, even largecrystals often contain only small portions that are ofsuitable color and transparency for faceting. Thecuttable areas are usually near the pyramidal crystalterminations. These areas can sometimes be bestexamined after acid cleaning and application ofmineral oil. More commonly, however, because ofthe corroded or encrusted nature of the crystal faces,the gem material is judged after the crystal has beensawn or broken into pieces. Only a small percentageof the total production exhibits intense, even col-oration and good clarity. Polished stones of goodcolor can also be created by positioning the zones ofbest color in the culet. A significant amount of FourPeaks material shows red overtones and/or a deepreddish purple bodycolor when faceted; these stonesare known in the gem trade as “Siberian” quality.

All of the material cut by Commercial MineralCo. has come from the underground workings andis faceted overseas. Material not suitable for facetingis stockpiled for future carving or cabbing.According to Mike Romanella (pers. comm., 2004),vice president of Commercial Mineral Co., theamethyst is faceted in both calibrated and free sizes.The non-calibrated amethyst ranges from 4 to 15 ct;

Figure 9. This 68.39 ct amethyst from Four Peaks isthe largest faceted stone from the deposit known tothe authors. Photo by J. Lowell.

Figure 8. Distinctive color zoning is evident in thisamethyst crystal fragment (5 cm in diameter).Virtually all the material from the Four Peaks mineshows color zoning in certain orientations. Photo byJ. Lowell.

fine-quality stones in the 15–20 ct range are quiterare. The calibrated goods range from 4 mm (0.25 ct)to 8–10 mm (~3 ct). The material is classified intothree quality grades, and at press time the companyhad thousands of carats of the high-quality facetedstones in its inventory. The largest faceted stoneknown to the authors weighs 68.39 ct (figure 9). Afew pieces have been carved to utilize the color zon-ing to good effect.

Like amethyst from other localities (e.g., Brazil),some material from Four Peaks is heated to lightenthe color. According to Mr. Romanella (pers.comm., 2004), 20–30% of the amethyst he hasfaceted is over-dark, so this material is heated to350–450°C. The treatment has about a 50% successrate; the remainder becomes unsalable due to frac-turing. Although Sinkankas (1957) reported thatheating of some Four Peaks amethyst can result in apale green color, the present authors have beenunable to verify or reproduce this behavior.

MATERIALS AND METHODSGemological properties were measured on fivefaceted amethyst samples using standard gem-test-ing instruments. These samples, ranging from 1.25to 3.71 ct (figure 10), are representative of the mate-rial currently being produced from the Four Peaksmine.

Refractive indices were obtained with a DuplexII refractometer and a near-monochromatic lightsource. Birefringence reactions were observed witha polariscope and a calcite dichroscope. Reactionsto ultraviolet radiation were checked in a darkened

room with conventional four-watt long-wave (366nm) and short-wave (254 nm) Ultra-Violet Productslamps. Observation of absorption spectra was madewith a Beck prism spectroscope. The visual fea-tures of the study samples were observed with agemological microscope. Photomicrographs weretaken with a Nikon SMZ-10 photomicroscopeunder various lighting conditions. A polishedamethyst plate oriented perpendicular to the opticaxis was prepared to allow for better observation oftwinning patterns.

A solid mineral inclusion was identified using aRaman Renishaw 1000 microspectrometer. Mid-infrared absorption spectra for two of the sampleswere recorded at room temperature with a NicoletMagna-IR Fourier-transform infrared (FTIR) spec-trometer over the range 400–6000 cm-1, with a reso-lution of 1.0 cm-1. Qualitative chemical analyses ofthe same two samples were obtained with a Kevex(now Thermo-Electron) Omicron X-ray fluorescence(EDXRF) system operating at accelerating voltagesof 10, 25, and 35 kV, and beam currents of 1.70,2.15, and 3.30 mA.

RESULTS AND DISCUSSIONThe refractive indices of each of the seven samplesstudied were 1.543 (w) and 1.551 (e). The birefrin-gence calculated from these values is 0.008. Withthe polariscope, a uniaxial “bull’s-eye” optic figurewas seen in all samples; slight distortions of thisoptic figure were seen near the edges of twinningboundaries (a feature noted before in amethyst).Purple and bluish purple dichroic colors were seen


Figure 10. These faceted samples of Four Peaks amethyst (1.25–3.71 ct) were examined for this study. Theamethyst is typically faceted to minimize the appearance of color zoning in the face-up position (left). Whenviewed table-down (right), the color zoning of these samples becomes apparent. Photos by Maha Tannous.

through a calcite dichroscope. All the samples wereinert to both long- and short-wave UV radiation,and they displayed no absorption spectra whenexamined with a Beck spectroscope.

Both primary and secondary fluid inclusionswere observed in the samples studied. The formerwere relatively small, the largest being 0.7 mm inlength (figure 11), and they showed minimal nega-tive crystal form. While some of the primary fluidinclusions appeared to be two-phase with liquidand gas components, many others appeared to con-

tain tiny anhedral solid phases (figure 12). No bire-fringence was evident in these solid phases whenthey were examined in transmitted light betweencrossed polarizers. The secondary fluid inclusionsexhibited typical veil-like patterns, and they wereoften seen in association with the larger primaryfluid inclusions (figure 13).

Other than the tiny anhedral solid phases influid inclusions, the only mineral inclusionsobserved within the amethysts were small reddish


Figure 11. Primary fluid inclusions, such as thoseshown here, were common in the Four Peaksamethyst examined. The largest inclusion measures0.7 mm long. Photomicrograph by J. I. Koivula.

Figure 12. Many of the primary fluid inclusions inthe amethyst samples examined also containedtiny anhedral solid phases, as shown in the lowerright of this image. Photomicrograph by J. I.Koivula; magnified 25¥.

Figure 13. Veil-like patterns of secondary fluid inclu-sions were commonly observed in the amethystsamples studied. Photomicrograph by J. I. Koivula;magnified 20¥.

Figure 14. Where dense accumulations of reddishbrown hematite flakes were seen, the host amethystmaterial tended to be a lighter purple. Photo-micrograph by J. I. Koivula; magnified 15¥.

brown grains of hematite (identified by Ramananalysis). These either occurred in dense accumula-tions (figure 14) or were sparsely scattered. To thebest of our knowledge, such dense accumulationsof hematite inclusions have not been reported inamethyst from other localities. Where hematitewas present in considerable amounts, the color ofthe amethyst tended to be lighter purple, whereasdarker-color material had fewer of these inclusions.

In transmitted light between crossed polarizers,Brazil-law twinning was obvious in all the samples

studied (see, e.g., figure 15). In many of the Brazil-law-twinned crystals and faceted stones, therewere sufficiently large untwinned areas to suggestthat it would be possible to cut Four Peaksamethysts that would not show this form of opti-cally active twinning.

The most unusual visual feature was a white,wedge-shaped zone that was seen oriented parallelto the rhombohedral direction in a 27.13 ct heart-shaped faceted stone (figure 16). While only slightlyvisible in darkfield illumination, this zone became


Figure 15. This pho-tomicrograph, taken intransmitted lightbetween crossed polar-izers, shows the Brazil-law twinning patternobserved in a polishedplate oriented perpen-dicular to the optic axisof the amethyst. Theopaque areas are denseaccumulations ofhematite inclusions.Photomicrograph by J. I.Koivula; magnified 2¥.

Figure 16. This 27.13 ctcut stone (inset; photoby J. Lowell) displayed

an unusual white,wedge-shaped growth

zoning pattern that wasespecially visible in

reflected light when thesample was illuminated

with a fiber-optic lightsource. Photomicro-

graph by J. I. Koivula;magnified 5¥.

much more apparent in reflected light from a fiber-optic light source. The cause of this unusual type ofdecorated growth zoning is unknown, and such fog-like inclusions have not been reported in amethystfrom other localities.

EDXRF analysis of two faceted amethyst samplesdetected silicon (as expected for quartz) and a weakfluorescence peak due to iron. IR spectra of these

same two samples did not show the 3543 cm-1 peakthat is characteristic of most synthetic amethyst(although also seen in some natural amethyst, suchas that from Caxarai, Brazil; Kitawaki, 2002).

The Four Peaks amethyst tested exhibitedinclusions, growth zoning, and Brazil-law twin-ning that should enable gemologists to separatethis material from its synthetic counterpart. Ifneeded, infrared spectroscopy can provide addi-tional evidence that the material is not laboratorygrown (Balitsky et al., 2004). When present, denseaccumulations of reddish brown hematite plateletsand wedge-shaped fog-like zones (again see figures14 and 16) not only prove that an amethyst is nat-ural, but they also strongly suggest that it origi-nates from the Four Peaks deposit.

CONCLUSIONThe Four Peaks mine in Maricopa County, Arizona,has sporadically produced fine-quality amethystsince the early 1900s, and it continues to supplycommercial quantities of amethyst for the jewelryindustry (figure 17). Usually the color is typical ofmaterial from most other localities, but a small per-centage of Four Peaks amethyst exhibits red over-tones and/or a deep reddish purple body color, acolor sometime described in the trade as “Siberian.”Future production will likely remain limited due tothe small size of the deposit and its remote location.


REFERENCESBalitsky V.S., Balitsky D.V., Bondarenko G.V., Balitskaya O.V.

(2004) The 3543 cm-1 infrared absorption band in natural andsynthetic amethyst and its value in identification. Gems &Gemology, Vol. 40, No. 2, pp. 146–161.

Chronic H. (1989) Roadside Geology of Arizona. Mountain PressPublishing Co., Missoula, MT, 322 pp.

Estrada J.J. (1987) Geology of the Four Peaks area, Arizona. M.S.thesis, Arizona State University, 76 pp.

Johnson M.L., Koivula J.I., Eds. (1998) Gem News: Amethystfrom Arizona. Gems & Gemology, Vol. 34, No. 3, pp.219–220.

Kitawaki H. (2002) Natural amethyst from the Caxarai mine,Brazil, with a spectrum containing an absorption peak at 3543

cm-1. Journal of Gemmology, Vol. 28, No. 2, pp. 101–108.Lieber W. (1994) Amethyst–Geschichte, Eigenschaften,

Fundorte. Christian Weise Verlag, Munich, 188 pp.Lowell J., Rybicki T. (1976) Mineralization of the Four Peaks

amethyst deposit, Maricopa County, Arizona. MineralogicalRecord, Vol. 7, No. 2, pp. 72–77.

Lurie M. (1998) Amethyst venture breathes new life into oldmine. Colored Stone, Vol. 11, No. 3, pp. 72–73.

Lurie M. (1999) Uphill battle. Lapidary Journal, Vol. 53, No. 2,pp. 290–293, 328.

Sinkankas J. (1957) “Green” amethyst from Four Peaks, Arizona.Gems & Gemology, Vol. 9, No. 3, pp. 88–95.

Sinkankas J. (1976) Gemstones of North America, Vol. 1. VanNostrand Reinhold, New York.

ABOUT THE AUTHORSMr. Lowell is a geologist and gem cutter living in Tempe,Arizona (www.coloradogem.com), and Mr. Koivula is chiefresearch gemologist at the GIA Gem Trade Laboratory inCarlsbad, California.

ACKNOWLEDGMENTS: The authors thank Joseph Hyman,former owner of the Four Peaks mine, for providing study

material and historical documents on the property. They alsothank Dr. Donald Burt of the Geology Department of ArizonaState University, as well as Dr. Emmanuel Fritsch of theInstitut des Matériaux de Nantes, University of Nantes,France, for useful information. Dr. James Shigley of GIAResearch assisted with the preparation of the article. Mikeand Jerry Romanella (Commercial Mineral Co., Scottsdale,Arizona) kindly provided photos and helpful information.

Figure 17. The Four Peaks amethyst mine has sup-plied attractive material for the gem trade. Thiswhite gold jewelry features a 6.89 ct Four Peaksamethyst in the ring. Courtesy of CommercialMineral Co.

ARGENTINA Buenos Aires: Ricardo Angel Tuccillo • AUSTRALIA Slacks Creek, Queensland: Ken Hunter • BELGIUM Diegem: Guy Lalous. Diksmwide: Honoré Loeters. Ghent: Jan Loyens. Hemiksem: Daniel DeMaeght. Koksijde: Christine Loeters. Overijse: Margrethe Gram-Jensen. Ruiselede: Lucette Nols. Tervuren:Vibeke Thur • CANADA Calgary, Alberta: Diane Koke. Bobcaygeon, Ontario: David Lindsay. Kingston,Ontario: Brian Randolph Smith. St. Catherines, Ontario: Alice Christianson. Montreal, Quebec: Tricia Anderson,Marie-France Rosiak. • ENGLAND Tenterden, Kent: Linda Anne Bately • FINLAND Oulu: Petri Tuovinen • FRANCE Paris: Pierre-Henri Boissavy • GREECE Athens: Gihan Zohdy • MALAYSIA Pasir Puteh, Kelantan:Kiam Kui Yang • THE NETHERLANDS Rotterdam: E. Van Velzen • PORTUGAL Figueira: Johanne Jack • SCOTLAND Edinburgh: James Heatlie • SPAIN Vitoria, Alava: Ignacio Borras Torra • SWITZERLANDRodersdorf: Heinz Kniess. Zurich: Doris C. Gerber, Eva Mettler • THAILAND Bangrak, Bangkok: SomapanAsavasanti, Siriya Siripanich, Pattarat Termpaisit, Anchalee Udomkhunatham • UNITED ARAB EMIRATESDubai: Iamze Salukvadze • UNITED STATES Arizona Chandler: Gary Dutton, LaVerne Larson. Gilbert:Margaret Hodson. Sun City: Anita Wilde. Tucson: Dave Arens. California Burlingame: Sandra MacKenzie-Graham. Burney: Willard C. Brown. Carlsbad: Cindi Doyle, Michael Evans, Mark Johnson, Thorsten Strom, RicTaylor, Jim Viall, Lynn Viall, Michael Wobby, Marisa Zachovay. Marina Del Rey: Veronika Riedel. Orange: AlexTourubaroff. San Rafael: Robert Seltzer. Ukiah: Charles "Mike" Morgan. Watsonville: Janet Mayou. ColoradoAurora: Ronda Gunnett. Denver: Mary Shore, Alan Winterscheidt. Florida Clearwater: Tim Schuler. SatelliteBeach: Consuelo Schnaderbeck. Tampa: R. Fred Ingram. Illinois Bethany: Richard Gallagher. Chicago: LindaNerad. Joliet: Christopher Shumard. Northbrook: Frank Pintz. Indiana Carmel: Mark Ferreira. Indianapolis:Wendy Wright Feng. Maine Braintree: Arthur Spellissy, Jr. Maryland Baltimore: Alissa Ann Leverock. PatuxentRiver: Pamela Dee Stair. Massachusetts Braintree: Alan Howarth. Brookline: Martin Haske. Lynnfield: JohnCaruso. Uxbridge: Bernard Stachura. Minnesota Paw Paw: Ellen Fillingham. Missouri Chesterfield: LeslieFaerber. Perry: Bruce Elmer. New Jersey Fort Lee: Wendi Mayerson. New York City Island: Marjorie Kos. CliftonPark: Sarah A. Horst. New York: Anna Schumate. North Carolina Advance: Donna Cranfill. Asheville: ChristianRichart. Creedmoor: Jennifer Jeffreys-Chen. Kernersville: Jean Marr. Ohio Cuyahoga Falls: Catherine Lee.Holland: Mary Jensen. Toledo: Nicholas Licata. Pennsylvania Schuylkill Haven: Janet Steinmetz. Yardley: PeterStadelmeier. South Carolina Sumter: James Markides. Tennessee Clarksville: Kyle Hain. Texas Corpus Christi:Warren Rees. Katy: Christine Schnaderbeck. Virginia Herndon: Lisa Marsh-Vetter. Newport News: ShannonSmith. Reston: Beth Carter. Vienna: Eugene May, Jr. Washington Ferndale: Candi Gerard. Seattle: Janet SuzanneHolmes, A. Samsavar. Sumner: Lois Henning. Wisconsin Milwaukee: William Bailey.

This year, 253 dedicated readers participated in the 2004 GEMS & GEMOLOGY Challenge. Entries arrivedfrom all corners of the world, as readers tested their knowledge on the questions listed in the Spring 2004issue. Those who earned a score of 75% or better received a GIA Continuing Education Certificate recog-nizing their achievement. The participants who scored a perfect 100% are listed below. Congratulations!

Answers (see pp. 89–90 of the Spring 2004 issue for the questions): 1 (d), 2 (a), 3 (c), 4 (b), 5 (c), 6 (c), 7 (a), 8 (b), 9(b), 10 (a), 11 (d), 12 (c), 13 (b), 14 (a), 15 (b), 16 (c), 17 (b), 18 (d), 19 (c), 20 (d), 21 (a), 22 (c), 23 (d), 24 (a), 25 (a)

CHALLENGE WINNERS GEMS & GEMOLOGY FALL 2004 239

Large CORAL Bead NecklaceSought after as a gem material formore than 2,000 years, today coral isone of the most popular organic gemsin the marketplace, next to pearls(“Coral: More than seasonal,” TheGuide, May/June 2003, pp. 7–8, 13).Coral’s enduring appeal may be due toits broad range of color saturation,

from vivid “ox-blood” red, throughthe softer pinkish orange of “salmon,”further to the pale pink of “angelskin,” and beyond even that to white.Combined with a Moh’s hardness of3.5, it is not surprising that coral is afavorite carving material for use asobjets d’art and in jewelry, as well asfashioned into cabochons and beads.

The coral used as a gem material isactually the accumulated calcium car-bonate secreted by colonies of tinyanemone-like sea animals. Coralpolyps are fairly delicate sea creatures,sensitive to changes in water depth,temperature, and clarity, as well as tomodern hazards such as pollution andoverfishing (see, e.g., R. Webster,Gems, 5th ed., rev. by P. G. Read,Butterworth-Heinemann, Oxford,1994, pp. 559–564). These modernhazards and quotas imposed to protectendangered corals have led to short-ages, particularly of the finer material.

The East Coast laboratory wastherefore very fortunate to have theopportunity to identify a necklacecontaining 11 large round pinkishorange beads that were graduated insize from 34.45 to 24.45 mm (figure1). All had a high polish, and only afew contained minor blemishes,which were almost hidden close totheir drill holes. Each bead displayedcoral’s classic wavy parallel fibrousstructure. Standard gemological test-ing identified the beads as coral, andswabbing with acetone in tiny incon-spicuous places did not reveal anydye. Fourier-transform infrared (FTIR)

240 LAB NOTES GEMS & GEMOLOGY FALL 2004

© 2004 Gemological Institute of America

GEMS & GEMOLOGY, Vol. 40, No. 3, pp. 240–251

Editor’s note: The initials at the end of each itemidentify the editor(s) or contributing editor(s)who provided that item. Full names are given forother GIA Gem Laboratory contributors.

EDITORSThomas M. Moses, Ilene Reinitz, Shane F. McClure, and Mary L. JohnsonGIA Gem Laboratory

CONTRIBUTING EDITORSG. Robert CrowningshieldGIA Gem Laboratory, East Coast

Cheryl Y. WentzellGIA Gem Laboratory, West Coast

Figure 1. The 11 coral beads in this necklace, which graduated in sizefrom 34.45 to 24.45 mm, were found to be untreated.

spectroscopy found no evidence ofpolymer impregnation. Thus, we con-cluded that these beads were, in fact,of natural color and not impregnat-ed—an impressive result for such awell-matched, highly polished set ofbeads this size. Alternating with yel-low metal spheres pavé set withnumerous transparent yellow roundbrilliants, these coral beads were thehighlight of a stunning necklace.

Wendi M. Mayerson

DIAMONDFour Blue Diamonds from aHistoric Necklace

The Cullinan blue diamond necklace(figure 2) holds a special place in thehistory of African diamond mining.While its classic style, workmanship,and rare blue diamonds have made itfamous, it started as a personalmemento—a gift to celebrate a specialevent. This Edwardian gold back, sil-ver top, festoon necklace was present-ed by Thomas Cullinan, then chair-man of the Premier mine, to his wifeAnnie in 1905 to commemorate thegift of the 3,106 ct Cullinan diamondto England’s King Edward VII andCullinan’s subsequent knighthood.Mr. Cullinan could not have knownat the time that blue diamonds suchas those in the necklace would oneday become as synonymous with thePremier (recently renamed Cullinan)mine as the famous large colorlesscrystals it has produced. Nor wouldhe have known that blue diamondswould remain rare and among themost highly valued of all diamonds.

The exhibition of the Cullinanblue diamond necklace at GIA’smuseum in Carlsbad this fall present-ed a welcome opportunity to examinethe necklace for the first time in manyyears, and to do so in detail for thefirst time ever. For this occasion, thecurrent owner requested that detailedgemological information accompanythe exhibition of the piece so as to bet-ter inform the public about thisunique necklace. For our examination,

the four principal blue diamonds (anoval shape weighing 2.60 ct—referredto as the Cullinan blue diamond—andthree Old European cut brilliantsweighing 0.75 ct, 0.73 ct, and 0.42 ct)were removed from the mounting forgrading (figure 3), and the necklaceitself was given a thorough inspection.Three of the four blue diamonds (allexcept the 0.42 ct) had been previous-ly graded by GIA, most recently in1993. That grading took place prior tothe 1995 enhancements to GIA’s col-ored diamond color grading system.Additionally, the 0.75 ct and 0.73 ctOld European brilliants had been grad-ed in the mounting, which did notallow a precise assessment. Thus, thiswas an important opportunity toreview the grading of these diamonds.

The 2.60 ct oval-shaped centerstone was described as Fancy Intenseblue. The three Old European bril-liants were each color graded Fancygrayish blue. The range of color inwhich blue diamonds occur is rela-

tively compressed in saturation, andvaries more widely in tone. Therefore,appearance differences are often dif-ferences in lightness to darknessrather than the strength or purity ofthe color. The four blue diamonds arerelatively similar in tone, whichexplains their selection for the neck-lace. It is the stronger saturation ofthe 2.60 ct oval that accounts for itsFancy Intense grade.

Other gemological propertieswere consistent with those of typicaltype IIb natural-color blue diamonds(see J. M. King et al., “Characterizingnatural-color type IIb blue dia-monds,” Winter 1998 Gems &Gemology, pp. 246–268). None of thediamonds showed a noticeable reac-tion to long- or short-wave ultravioletradiation, and all of them were elec-trically conductive. Electrical con-ductivity is measured by placing thestone on a metal base plate andtouching a probe carrying an electri-cal current to various surfaces of the

LAB NOTES GEMS & GEMOLOGY FALL 2004 241

Figure 2. The Cullinan blue diamond necklace dates back to 1905 andcontains several rare type IIb blue diamonds, the largest of which is2.60 ct. Courtesy of S. H. Silver Co., Menlo Park, California.


diamond. The conductivity value canvary with direction, so a number ofmeasurements are made and thehighest value is recorded. As noted inprevious studies (again, see King etal., 1998), the highest electrical con-ductivity value can vary widely with-in and between color grades. Wefound this to be the case with thesediamonds, as each showed a range ofvalues. Interestingly, the FancyIntense blue oval—the most stronglycolored—displayed the lowest value.

It is common for type IIb blue dia-monds to phosphoresce after expo-sure to short-wave UV radiation, andthis characteristic was noted for allfour stones. Although the best-known phosphorescent reaction inblue diamonds is that of the famousHope diamond, the Hope’s long-last-ing strong red phosphorescence isactually quite rare. It is much morecommon for phosphorescence to bevery weak to weak blue or yellowand of short duration; this more com-mon reaction was exhibited by thediamonds from the necklace.

Microscopic examination revealedcharacteristics consistent with other

type IIb blues. Such diamonds fre-quently have an uneven color distri-bution, which was observed here. It isalso typical for them to have relative-ly few solid inclusions, with fracturesand indented naturals being morecommon. This was also consistentwith our findings. As would be

expected, the blue diamonds, as wellas other diamonds in the necklace,exhibited minor chips and abrasionsconsistent with the necklace’s nearly100-year history (figure 4).

Examination of these historicstones also gave us an importantopportunity to study spectroscopicfeatures of known natural IIb dia-monds. Infrared spectroscopy showedabsorptions in all four stones at 2801and 2454 cm-1, which are typical fea-tures of type IIb diamonds. Photo-luminescence spectra collected usingan argon laser (at 488 nm excitation)revealed emission features character-istic of natural diamonds. A relativelystrong 3H emission (503.5 nm) wasdetected in all four stones.

Luminescence imaging is a usefulway to study diamond growth, andwas performed using the De BeersDiamondView. As shown in figure 5,blue luminescence and networks ofpolygonized dislocations are evident.Similar dislocation features wereobserved in all four stones. This typeof dislocation network is a specificfeature of natural diamond, and hasnever been observed in synthetic dia-mond.

John M. King, TMM, and Wuyi Wang

Figure 3. The four main blue diamonds are shown here removed fromthe necklace. The Fancy Intense blue oval at top weighs 2.60 ct, whilethe three Fancy grayish blue Old European brilliants weigh, from left toright, 0.75 ct, 0.42 ct, and 0.73 ct.

Figure 4. Minor chips and abra-sions such as those seen here onthe 0.42 ct Old European bril-liant are often encountered ondiamonds that have been wornover a long period of time.

Figure 5. This luminescenceimage of the 2.60 ct oval-shapeddiamond was collected usingthe De Beers DiamondView. Ablue luminescence and net-works of polygonized disloca-tions are evident; such disloca-tion networks are characteristicof natural diamond.


Irradiated Blue Diamond CrystalIrradiation with or without annealingis a common technique used toenhance the color of diamonds, andthose suspected of being treated inthis manner are routinely submittedto the laboratory for origin-of-colortesting. Most such stones are treatedafter faceting, and depending on thetype of treatment, diagnostic colordistribution features are sometimesseen. Among the few laboratory-irra-diated rough diamonds we haveexamined was a 2.95 ct well-formedbluish green octahedron (Winter 1989Lab Notes, pp. 238–239).

Recently, a large crystal thatresembled the sample in the 1989 LabNote was submitted to the East Coastlaboratory for origin-of-color determi-nation. The experienced client whobrought the stone to our attention wassuspicious of its origin even though itwas represented as coming directlyfrom the mine in central Africa.

The 11.60 ct crystal (figure 6),which measured 13.11 ¥ 12.85 ¥ 9.09mm, exhibited typical octahedralcrystal morphology and showed obvi-ous resorption features on its surface.It also showed a distinct greenish bluecoloration. In contrast to similarlycolored natural diamonds, no green orbrown radiation stains were observedwith magnification. However, a slightcolor concentration was evident atthe edges of the crystal faces (again,see figure 6). We observed a strongblue fluorescence to long-wave UVradiation and a weak green-yellowreaction to short-wave UV.

Infrared spectroscopy revealed fea-tures typical of a type Ia diamond withvery high nitrogen content and a weakabsorption due to hydrogen impurities.In rare cases, a high concentration ofstructurally bonded hydrogen in dia-mond could produce blue coloration,but that definitely was not the case forthis crystal. A weak H1a absorptionwas present (at 1450 cm-1), but therewere no H1b, H1c, or H2 absorptions.In the UV-visible spectrum collectedwhen the diamond was cooled by liq-uid nitrogen (figure 7), strong N3 (415

nm), moderate N2 (478 nm), weak H3(503 nm), and weak 595 nm absorp-tions were detected; a strong and broadGR1 (741 nm) absorption also wasapparent.

These gemological and spectro-scopic features led to the conclusionthat this crystal was artificially irradi-

ated without subsequent annealing.Considering the large size of the crys-tal, it is possible that the radiation-related color does not penetrate evenlythroughout, as is suggested by theconcentration of color along the edges.

Wuyi Wang and TMM

Irradiated Type IIb DiamondType IIb diamonds are often blue, dueto small amounts of boron impuri-ties. Depending on the occurrence ofother defects (e.g., plastic deforma-tion), some type IIb diamonds exhibita gray or, more rarely, a brown color(see, e.g., King et al., 1998, cited inearlier entry). Blue also can be pro-duced in an otherwise colorless dia-mond by exposure to radiation (eithernaturally or in a laboratory) to createa vacancy defect. Although it is tech-nically possible to enhance the bluecolor of a type IIb diamond by irradia-tion, mixing of two different color-causing mechanisms in the same dia-mond may not necessarily producean attractive color. Brown natural

Figure 6. The strong blue col-oration in this 11.60 ct diamondoctahedron proved to be theresult of laboratory irradiation.

Figure 7. A strong and broad GR1 absorption (741 nm), together with aweak absorption at 595 nm, indicated that the 11.60 ct greenish bluediamond was irradiated in a laboratory. The pre-existence of relativelystrong Cape absorptions (e.g., N3 and N2) resulted in the green modifier.

radiation stains were reported on anunusual type IIb blue diamond (Fall1991 Lab Note, pp. 174–175), but theradiation did not appear to have hadany effect on the color of this facetedstone. An artificially irradiated typeIIb diamond recently submitted to

the East Coast laboratory gave us anextremely rare opportunity to exam-ine a sample of this nature.

The 2.00 ct oval brilliant cut(10.52 ¥ 7.14 ¥ 4.21 mm) in figure 8was color graded Fancy Dark green-gray. Although this hue was not out-side the range of hues we have seen innatural-color type IIb diamonds, thecause of the color was not immediate-ly obvious. No green or brown naturalradiation stains were observed on thesurface with magnification and dark-field illumination. Nor did the dia-mond have any notable internal char-acteristics such as solid inclusions orfractures, or any distinct colored grain-ing. We did not see fluorescence toeither long- or short-wave UV radia-tion, but very weak yellow phospho-rescence was detected after exposureto short-wave UV. Infrared absorptionspectroscopy only showed featurestypical of a type IIb diamond (e.g.,strong and clear absorptions at 2801and 2454 cm-1 due to substitutionalboron impurities). However, when the

diamond was examined more careful-ly with low-power binocular magnifi-cation and diffused light, a distinctblue color concentration was notednear the culet (figure 9). This type ofcolor zoning is typical for diamondsthat have been artificially irradiatedwith a low-energy source, as is usuallydone today with an electron beam.

Absorption spectroscopy collectedat liquid nitrogen temperature in theultraviolet-visible range (figure 10)showed a strong GR1 band (vacancy,741 nm) and several other lines of theGR series (GR2 through GR8). A clearTR12 absorption at 469.9 nm alsowas detected. It is very rare for a natu-ral-color type IIb diamond to showany detectable GR1 or TR12 absorp-tion with UV-Vis absorption spec-troscopy. In addition, in the photolu-minescence spectrum seen with alaser Raman microspectrometer, theintensity of the 3H defect with a zero-phonon line at 503.5 nm (another typ-ical radiation-related defect in dia-mond) was significantly stronger than


Figure 8. This 2.00 ct FancyDark green-gray type IIb dia-mond was found to have beenlaboratory irradiated.

Figure 9. When examined with amicroscope and diffused light,the diamond in figure 8 revealeda strong concentration of bluenear the culet, which proved thatit had been artificially irradiated.Field of view is 6.0 mm high.

Figure 10. This UV-Vis absorption spectrum of the green-gray diamondshows a strong GR1 band and some additional lines from the GR series.A clear TR12 band also was detected. A gradual increase in absorptionat higher wavelengths is attributed to the presence of boron impurities.


that in any natural-color type IIb dia-monds we have examined. All theseobservations led to the conclusionthat this type IIb diamond had beentreated by irradiation in a laboratory.

Little is known about the interac-tion between vacancies and boron indiamond and the possible impact oncolor. The color of this specific sam-ple is not very attractive. A possiblereason is that prior to irradiation thestone may have had a strong browncomponent. This also would explainthe green coloration after treatment.

Wuyi Wang, TMM, and Thomas Gelb

Unusual Cause of Blue Color in a DiamondIt is unusual for the laboratory toencounter natural blue diamondswith a cause of color other than boronimpurities, natural irradiation, or thepresence of hydrogen (see, e.g., E.Fritsch and K. Scarratt, “Natural-colornonconductive gray-to-blue dia-monds,” Spring 1992 Gems &Gemology, pp. 35–42; and King et al.,1998, cited above). Thus, it was quitea surprise when the East Coast labrecently tested an oval diamond over12 ct (figure 11, center) with a bluecolor typical of that produced byboron but none of the other propertiesone would expect for such diamonds.

When color graded, the oval wasclassified as Very Light blue. On test-ing for electrical conductivity, the

diamond was found to be nonconduc-tive. We observed a moderate bluereaction to both long- and short-waveUV radiation. Infrared spectroscopyproved that the diamond was type IaBand did not show any elevated levelsof hydrogen. Spectroscopic analysisfurther ruled out any possibility thatthis stone was irradiated in a labora-tory. However, microscopic examina-tion revealed clouds of pinpoints andinternal whitish graining.

In this case, we believe the colorwas caused by a scattering of light inthe diamond from the clouds of pin-point inclusions. This effect is similarto the one that causes cigarette smoketo appear blue, a phenomenon oftenreferred to as the Tyndall effect.Named after its discoverer, 19th-cen-tury British physicist John Tyndall,

this effect is caused by reflectionand/or scattering of light by very smallparticles in suspension in a transparentmedium. It is often seen from the dustin the air when sunlight enters a roomthrough a window or comes downthrough holes in clouds. Most dia-monds that have dense clouds and alsoshow scattering have a predominantlywhite or gray appearance (e.g., figure12). The difference in color is probablyrelated to the size of the particles andthe density of the cloud.

We have documented diamondswith these properties from a few dif-ferent geographic locations, includingIndia and Australia. This unusualstone indicates that micro-inclusionsin diamond can generate colors otherthan black, gray, and white. Thenature of the micro-inclusions (chem-istry, particle size, density, and distri-bution) and the size of the stone,along with the light source and itsdistribution of output, are all factorsin determining the final color appear-ance of the diamond.

John M. King, Wuyi Wang, TMM

NEPHRITE that Mimics Serpentine Because of their similar appearances,some nephrite and serpentine can beimpossible to distinguish by visualobservation alone. Although the twomineral species usually can be readilyseparated on the basis of their refrac-tive index and specific gravity values,in some cases even these physicalproperties can be misleading. TheWest Coast laboratory recentlyreceived a 142.27 ct translucent mot-tled green-gray carving of a water buf-falo that exhibited both an artificialwater-soluble reddish stain outliningdetails of the carving, and a lightbrownish yellow coating that partiallyconcealed the underlying host materi-al (figure 13).

Microscopic examination revealedthat the host was a fine-grained aggre-gate. When tested in an inconspicu-ous area, the coating dented andscratched easily with a metal probe,similar to the response of a wax.

Figure 11. The color appearance of the 12+ ct Very Light blue type IaBoval modified brilliant in the center is due to scattering of light and notto any of the typical causes of natural blue color. The 6+ ct pear-shapeddiamond on the left is colorless (D-color); the 5.5+ ct type IIb on theright is Fancy Light blue.

Figure 12. More typically, thescattering of light from microinclusions produces a predomi-nantly white (as shown in this7.00 ct marquise) or gray colorin diamond.


However, it did not melt readilywhen exposed to a thermal reactiontester; instead it merely yielded moreeasily and turned brown under themicroscope with prolonged contact. Italso did not react to a dilute HCl solu-tion. This coating may have been anattempt at “antiquing” the carving;however, it is not known whether thecoating was natural or an artificiallyapplied substance. In addition, a verythin coating, which may have beenseparate from—or an extension of—the built-up areas, created an invisiblefilm over the host material. It toocould be scratched off, possiblydeceiving an unwary observer intobelieving that the host material wassoft. However, only the thin filmyielded to the tip of the probe, where-as the material beneath was harderthan the metal point.

The R.I. of the greenish areas wasapproximately 1.58, varying from 1.57to 1.59. The higher value wasobtained on a small area where thethin coating had been removed withsolvent; areas with the thicker brown-ish yellow coating had lower R.I. val-ues than areas where the green-grayhost material was clearly visible.

The reaction to long-wave UV radi-ation was a combination of an inert

background mottled with mediumchalky green-yellow fluorescence thatappeared weak to medium yellow onthe areas with the thicker brownishyellow coating. The reaction to short-wave UV was also a combination of aninert background with a mottled veryweak to weak yellow fluorescence onthe coating. There was no significantvisible-light spectrum.

With only this initial examina-tion—the appearance of the carving,the illusory “soft” nature of the mate-rial, and the low R.I. values rarelyattained by nephrite—a hasty evalua-tion could lead a gemologist tomisidentify this material as serpen-tine. However, further testing revealedits true composition. The specificgravity, measured hydrostatically,was approximately 2.96; the FTIRspectrum was consistent with thoseof other nephrites; and the Ramanspectrum of the greenish areas wasconsistent with the nephrite referencein the database. To characterize thematerial further, EDXRF analysis wasperformed by GIA senior researchassociate Sam Muhlmeister. Themajor elements present were Si, Ca,and Mg, with a trace of Mn and Fe.The presence of Ca in the structureexcluded serpentine, while the chem-

istry, S.G., and FTIR and Raman spec-tra were all consistent with nephrite.

In an attempt to confirm orexplain the uncharacteristically lowR.I. values, the thin film that coatedthe specimen was polished off in asmall inconspicuous area to betterexpose the host material. A highervague R.I. reading between 1.60 and1.61 was then obtained, indicatingthat the coating was responsible forlowering the apparent R.I. This piecewas a valuable reminder of the careand diligence that should be exer-cised when obtaining physical prop-erties for gem identification.

CYW

Pink OPAL This past summer, the East Coast lab-oratory received for identification twopairs of partially drilled translucent tosemi-translucent variegated pinkdrops (figure 14). The larger pair mea-sured 35.90 ¥ 13.95 ¥ 13.90 mm and35.60 ¥ 13.95 ¥ 13.90 mm and had atotal weight of 64.73 ct; the smallerpair had a total weight of 29.28 carats.

At first glance, the drops resem-bled conch pearls in their coloration.However, they did not exhibit anyflame-like structure; nor were theythe standard shape of conch pearls. Infact, their symmetrical shape indicat-ed they were not pearls at all. Theyalso resembled massive rhodo-chrosite, but they lacked the botry-oidal structure and distinctive colorzoning commonly associated withthat mineral. Standard gemologicaltesting revealed spot refractive indicesof 1.45, weak-to-medium whitish flu-orescence to long-wave UV, veryweak yellow to no reaction to short-wave UV, and specific gravities rang-ing from 2.16 to 2.23. With a desk-model spectroscope, all four dropsshowed a weak band at approximately490 nm, a weak line at approximately550 nm, and a cutoff from 430 nm.These properties suggested opal, butglass could not be ruled out.

A Fall 1982 Lab Note (pp. 172–173)described a variegated light pink and

Figure 13. This nephrite carving (43.95 ¥ 30.35 ¥ 9.98 mm) could be con-fused with serpentine due to the effects of an unidentified coating.


gray carving of a bird that was identi-fied as opal by the West Coast laborato-ry. Not until the Winter 1991 issue (pp.259–260) did information becomeavailable in G&G regarding the loca-tion of the mines that reportedly wereproducing this translucent-to-opaquepink material, along with blue opal: theAcari copper mining area near Are-quipa, Peru. It was said that some of thepink material even “exhibits a colorreminiscent of rhodochrosite” (p. 259).

Our next step was to use energy-dispersive X-ray fluorescence (EDXRF)to investigate the chemistry of thefour drops and compare the findings toa sample of known pink opal fromPeru. EDXRF revealed the presence ofMg, Al, Si, K, Ca, Mn, and Fe. As opalis amorphous silica, the other ele-ments can be attributed to mineralimpurities, such as palygorskite, thatare commonly associated with thistype of opal (see J. Hyrsl, “Gemstonesof Peru,” Journal of Gemmology, Vol.27, No. 6, 2001, pp. 328–334, and theFall 1982 Lab Note). The spectra werenearly identical, clearly identifyingthese drops as pink opal, possibly from Peru.

Wendi M. Mayerson and David Kondo

RUBIES, Clarity Enhanced with a Lead Glass FillerHeat treatment of natural rubies fre-quently leaves a glassy residue withinsurface-reaching fractures and cavities.This filling is produced during the heat-ing process and can facilitate the partialhealing of fractures within the rubies. Itis often detectable with magnificationby observing the difference in lustercompared to the surrounding corun-dum (see, e.g., Fall 2000 Lab Notes, pp.257–259). This type of glass residue hasa relatively low R.I. compared to corun-dum and is usually not considered to

be a clarity enhancement on its own.Earlier this year, however, the

Gemmological Association of AllJapan (GAAJ) issued a lab alert describ-ing rubies that had not been heattreated, but that showed a flash effectin their fractures caused by a clarityenhancement similar to that tradi-tionally used in diamonds (“Leadglass impregnated ruby,” GAAJ LabAlert, 2004, GAAJ Research Laborato-ry, http://www.gaaj-zenhokyo.co.jp/researchroom/kanbetu/2004/gaaj_alert-040315en.html). Their EDXRFchemical analyses revealed the pres-ence of elevated levels of lead (Pb) inthe material filling the fractures inthese stones. The results were laterconfirmed by the AGTA (“New rubytreatment arrives in the UnitedStates,” AGTA Gemstone Update,http://www.agta.org/consumer/news/20040702rubytreatment.htm).

Recently, the West Coast laborato-ry had the opportunity to examine twopurplish red mixed-cut oval stones(3.19 and 2.76 ct) that containednumerous large surface-reaching frac-tures and cavities filled with a glassysubstance (figure 15). Both gems wereidentified as natural ruby by theirinclusions, R.I., and visible absorptionspectra in a desk-model spectroscope.Unlike the stones described by GAAJ,the presence of thermally alteredinclusions showed that these rubieshad been heated. The glassy fillingscontained numerous flattened gas bub-bles or voids (figure 16), and showed a

Figure 14. These four variegated pink drops, which range from 14.01 to32.68 ct, were identified as pink opal.

Figure 15. These two rubies contain numerous fractures filled with ahigh-lead-content glass that undoubtedly significantly improved theirapparent clarity.


distinctly higher surface luster than thecorundum (figure 17). The filled frac-tures were very low relief in all viewingdirections and displayed weak to mod-erate blue-to-violet and orange flasheffects. In some directions, the frac-tures had a slightly hazy appearance(figure 18). The 3.19 ct ruby, in particu-lar, had several large filled cavities.

The filling material in the largestcavity was translucent, yellow in color,and contained many spherical gas bub-bles (figure 19). In another large filledcavity, the filling material was also yel-

low, but transparent with no visible gasbubbles. The filling material in thefractures did not have a visible bodycolor, but this was undoubtedlybecause of the small amount of materi-al present. All the filled cavitiesshowed poor polish at the surface, indi-cating a much lower hardness than thesurrounding corundum (again, see fig-ure 17). Unlike the residues we haveseen previously in heat-treated rubies,the nature and abundance of the fillingsin these two stones suggested an intentto hide the fractures and improve theapparent clarity of the gems.

To learn more about the fillingmaterial, we employed EDXRF analy-sis, Raman spectroscopy, and fluores-cence imaging. EDXRF, performed byGIA senior research associate SamMuhlmeister, showed elements typi-cal of ruby (Al, Cr, and Fe) and elevatedlevels of Pb similar to those reported

by GAAJ. None of the other elementsthat have been reported in glass-likeruby fillings (Si, P, Ca, and Ti) weredetected (again, see the Fall 2000 LabNote). It is important to note thatlighter elements such as boron cannotbe measured using EDXRF.

Raman spectroscopy was per-formed to compare the glass filler inthese two rubies with five differentglass samples known to contain sig-nificant Pb. Although the spectrumfor this filler was not a match for anyof the glasses in our collection, it hadmany of the same luminescence fea-tures as a sample of lead borate glass.

Last, we observed reactions tohigh-energy short-wave UV radiationusing the De Beers DiamondViewinstrument. The filling material re-sponded very strongly, fluorescingbright blue in contrast to the redreaction (caused by chromium) of thesurrounding corundum. For compari-son, traditional glass-filled cavities inrubies show an inert to dull gray reac-tion (figure 20, left and right). Noneof the glass samples from our collec-tion showed any reaction in theDiamondView.

Key gemological identification fea-tures of this new filler include a dis-tinct luster, flash effects, haziness, gasbubbles and voids, together with verylow relief of the fractures in all viewingdirections. In addition, elevated Pb andthe absence of other significant ele-

Figure 16. The glassy fillings in these clarity-enhanced rubies all con-tained numerous flattened gas bubbles (left) and elongated and irregu-lar voids (right). Magnified 27¥.

Figure 17. The luster of this fillermaterial was actually higherthan the ruby, which is the oppo-site of what is typically seenwith glassy residues. Surface cav-ities filled with this lead glassalso showed a poor polish com-pared to the surrounding corun-dum, indicating a much lowerhardness. Magnified 40¥. Figure 18. In some directions,

the filled fractures showed aslightly hazy appearance.Magnified 36¥.

Figure 19. In one large filled cavi-ty, the filler was translucent andyellow in color. One of the manyspherical gas bubbles that werepresent can be seen in the centerof this image. Magnified 27¥.


ments attributable to the filler in theEDXRF analysis are characteristic ofthe samples we have seen thus far. Bluefluorescence in the DiamondView mayalso be useful for identification.

Kimberly M. Rockwell andChristopher M. Breeding

GIA Gem Laboratory, Carlsbad

Unusual SYNTHETIC RUBY Triplet Displaying AsterismThe identification of loose assembledstones is relatively straightforward.However, when they are mounted injewelry (often bezel set), almost all

signs of assembly can be hidden.Frequently only the top portion of theassemblage can be identified, withfew indications as to the identity ofthe bottom part possible because ofthe restrictions the mounting placeson testing procedures (as was the casewith the assembled stone reported inthe Fall 1993 Lab Notes, p. 205).

Recently, the New York laboratorywas asked to identify a large red doublecabochon that displayed a weak six-rayed star; it was set in a yellow metalring with 22 transparent near-colorlessbrilliants (figure 21). A 1.76 spot R.I.reading, characteristic absorption spec-trum, and the presence of curved striaeand gas bubbles seen with magnifica-tion identified the top of the cabochon

as synthetic ruby manufactured by amelt process. Further examinationthrough the top of this mounted cabo-chon revealed a deeper layer of color-less cement with numerous oval gasbubbles, followed by another layer thatexhibited the hexagonal growth struc-ture typical of natural corundum.

Although at this point the cabo-chon appeared to be a simple assem-bled stone consisting of a syntheticruby top and a natural ruby bottom,held together by a near-colorlesscement layer, closer inspection of theconvex bottom with a fiber-optic lightand magnification revealed the pres-ence of curved striae. This indicatedthat the bottom piece was also amelt-grown synthetic ruby. This wasvery unusual, but to fully identify theparts of this assemblage it would haveto be removed from the mounting.Fortunately, the cabochon was held inplace by only four small prongs, sowith the client’s permission we had itunmounted for further analysis.

The bottom piece exhibited diag-nostic properties identical to those ofthe synthetic ruby top. Immersion ofthe cabochon in methylene iodiderevealed a thin grayish blue layer com-pletely encapsulated within a thicklayer of colorless cement (figure 22).Due to its location within the glue, itwas not possible to definitively identi-fy this layer. However, the hexagonalgrowth structure seen in figure 23

Figure 20. These DiamondView images show the inert to dull gray reac-tion of a traditional glass filler (left) compared to the bright blue reac-tion of the new lead glass filler (right).

Figure 21. The red asteriatedcabochon in this ring (13.00 ¥11.80 ¥ 10.55 mm) was identi-fied as a triplet consisting of asynthetic ruby top and bottomjoined by a near-colorless gluelayer that surrounded anunidentified center section.

Figure 22. When the cabochon infigure 21 is viewed in profile,unmounted and immersed inmethylene iodide, the distinc-tive layers that make up thisassemblage are clearly visible,as are numerous gas bubbles inthe near-colorless cement.

Figure 23. Seen through the bot-tom of the cabochon while it wasimmersed in methylene iodide,the hexagonal growth structureof the unidentified section isapparent within the cement.


indicated that it most likely was athin slice of natural corundum thatcontained the three directions of rutileneedles, or “silk,” necessary to createasterism. The conclusion thatappeared on the gem identificationreport read as follows: “Triplet con-sisting of a synthetic ruby top and asynthetic ruby bottom, joined togetherwith a near colorless cement contain-ing an unidentified center section.”

Siau Fung Yeung and Wendi M. Mayerson

Copper-bearing Color-ChangeTOURMALINE from MozambiqueColor-change tourmaline has beenreported in G&G on only three prioroccasions: dravite from East Africa(Fall 1991 Gem News, pp. 184–185),uvite possibly from East Africa (Fall2000 Gem News, pp. 270–271), andcopper-manganese-bearing elbaitefrom Nigeria (Fall 2001 Gem NewsInternational, pp. 239–240).

Elsewhere in the literature, thereare limited reports of color-changetourmalines, which typically provedto be chromium- and vanadium-bear-ing specimens in the dravite-uviteseries (H. Bank and U. Henn, “Colour-changing chromiferous tourmalinesfrom East Africa,” Journal of Gem-mology, Vol. 21, No. 2, 1988, pp.102–103; A. Halvorsen and B. B.Jensen, “A new colour-change effect,”Journal of Gemmology, Vol. 25, No.5, 1997, pp. 325–330).

Noel Rowe of Rough to Cut inSan Jose, California, recently submit-ted to the West Coast laboratory acolor-change tourmaline that differed

significantly from those previouslydescribed. Viewed table-up, this 5.68ct transparent emerald-cut stone(11.63 ¥ 8.76 ¥ 7.48 mm) exhibited adistinct color change from purple influorescent light to gray–bluish greenin incandescent light (figure 24).Particularly notable is the fact thatthe colors exhibited were the reverseof the usual alexandrite effect: coolcolors such as green, blue, or gray inday or fluorescent light and warmcolors such as purple, red, or brownin incandescent light. All other previ-ously reported color-change tourma-lines followed the usual convention.

This stone was eye-clean, and evenwith magnification it was exceptional-ly free of inclusions except for a tinycrystal near the surface of the pavilionand three miniscule needles near thegirdle. Although optical properties andRaman analysis proved that this stonewas indeed tourmaline (the Ramanspectrum was consistent with rubellitein the database), the lack of inclusionsand the very unusual phenomenonwarranted further investigation.

To facilitate the investigation ofthis intriguing material, Mr. Roweand the owner of the stone, Bruce Fryof Mars, Pennsylvania, supplied twoadditional tourmalines that exhibitedsimilar properties. One was a 5.47 ctpear shape pre-form that had roughlythe same color change as the emer-ald-cut stone. The other was a 5.37 ctoval that had a weaker color changefrom grayish purple in fluorescentlight to gray in incandescent light(again, see figure 24). According toMr. Fry, these three stones are theonly tourmalines of this type that heor his South African suppliers have

ever seen. He stated that the emer-ald-cut and pear pre-form stones werefrom Moiane in northern Mozam-bique, approximately 200 km northof Nampula. The origin of the oval isonly known as northern Mozam-bique. The specimens were alluvialin origin, and at least one of thestones was reportedly mistaken for alow-value iolite after being discov-ered by a local gold miner. Both theemerald cut and the pear shape pre-form were thought to be from thesame lot; although cut recently, theywere probably recovered several yearsago. The oval was cut from a crystalwith a color distribution that includ-ed a thin “rind” of pink skin andother small pink zones. Mr. Fryretained a flame-shaped zone of pinkwhen he faceted the stone approxi-mately four years ago.

The pear pre-form had a frostedsurface with small areas of the origi-nal skin. Although the view into thestone was obscured by the frostedsurface, magnification did revealsome sparse “trichites” (intercon-nected secondary two-phase inclu-sions) and a couple of fracturesextending in from the surface thatcontained an orangy yellow residue,possibly iron-oxide staining. Thestone was vaguely bicolored with apink zone encompassing approxi-mately half of it toward the tip. Theoval also contained trichites, in addi-tion to the pink flame-shaped zonethat extended inward from the side.All the stones displayed the strongdoubling at facet junctions typical oftourmaline.

The R.I. values of the emerald-cutand oval stones were w = 1.639–1.640

Figure 24. These three cuprian tourmalines from Mozambique (5.47, 5.68, and 5.37 ct) exhibit a strong“reverse” color change from fluorescent light (left) to incandescent light (right).


and e = 1.621 (values for the pear pre-form were unattainable due to therough surface). The hydrostatic spe-cific gravities ranged from approxi-mately 3.05 to 3.06 for the three sam-ples. All three stones were inert toboth long- and short-wave UV radia-tion. Dichroism was similar in allthree—grayish violet to violet andgrayish green to pale green—varyingmostly in intensity.

To characterize these stones fur-ther, GIA senior research associateSam Muhlmeister performed EDXRFqualitative chemical analysis. Al andSi were the major elements detected;the oval had a minor amount of Mn,while traces of Mn were seen in theother two samples. All the stonesalso showed traces of Ca, Cu, Ga,and Bi, with the oval showing Znand Sr, and both the emerald cut andthe oval showing some Pb (the pres-ence of which was inconclusive inthe pre-form).

The copper content is especiallynoteworthy. The only other commer-cially available tourmalines that con-tain copper are the copper-manganese-bearing elbaites, well-known for theirexquisite blue colors, from São José daBatalha, Paraíba, Brazil (and sur-rounds), and western Nigeria (see, e.g.,E. Fritsch et al., “Gem-quality cuprian-elbaite tourmalines from São José daBatalha, Paraíba, Brazil,” Fall 1990Gems & Gemology, pp. 189–205; andB. Laurs et al., “More on cuprianelbaite tourmaline from Nigeria,”Spring 2002 Gem News International,pp. 99–100). One Nigerian cuprianelbaite that was examined by theGübelin Gem Lab—a 22.98 ct violetgemstone—also exhibited a colorchange (C. P. Smith et al., “Nigeria asa new source of copper-manganese-bearing tourmaline,” Fall 2001 Gem

News International, pp. 239–240).However, both the dichroism (purple-violet and slightly grayish violet-blue)and the color change (violet to purple)were different from what was seen inthe three stones from Mozambique.

It is likely that Cu and Mn, atleast in part, are the cause of color inthese three specimens (Fritsch et al.,1990). Trace amounts of Pb in thesesamples could possibly be the resultof residue from the polishing processand may not be reliable (D. Dirlam etal., “Liddicoatite tourmaline fromAnjanabonoina, Madagascar,” Spring2002 Gems & Gemology, pp. 28–53).

The EDXRF analyses helped nar-row the list of possible tourmalinespecies to which these stones belong.The lack of significant amounts ofCa, Fe, and Mg eliminates the calcictourmalines and several other endmembers (e.g., dravite). Solely basedon the elements detected, the remain-ing possibilities include the alkalitourmalines elbaite and olenite andthe X-site-vacant tourmaline, ross-manite (see, e.g., F. C. Hawthorne andD. J. Henry, “Classification of theminerals of the tourmaline group,”European Journal of Mineralogy,1999, Vol. 11, pp. 201–215). However,quantitative chemistry, using amethod that would detect lighter ele-ments such as Na and Li, would benecessary to fully characterize thesestones and the tourmaline species towhich they belong.

Polarized UV-visible spectroscopy(also performed by Sam Muhlmeister)and oriented FTIR spectroscopy wereconducted for comparison to othertourmalines. The UV-Vis spectrawere very similar to those reportedfor the cuprian elbaites from theParaíba mines in Brazil, with broadabsorption peaks centered around

485–520 nm (which was attributed toMn3+ in unheated Paraíba tourma-lines) and Cu2+ absorptions centeredaround 690 and 895–905 nm in theE^c direction and 710–720 and900–920 nm in the E || c direction(again, see Fritsch et al., 1990). Al-though the FTIR spectra were consis-tent with other tourmalines, deter-mining whether or not subtle differ-ences exist to distinguish thesestones would require further study.

Several aspects of these threetourmaline specimens are veryintriguing: their strong and distinctcolor change, the fact that the colorchange relative to the type of lightsource is opposite the normal alexan-drite effect, the copper content, andthe source location in East Africa.

We have no current explanationfor the reverse nature of the colorchange. We are continuing researchinto this seemingly new variety oftourmaline, with a focus on obtain-ing quantitative chemistry to morefully characterize this material anddetermine the species to which itbelongs. Quantitative chemical anal-ysis will also provide greater insightinto the cause of color in these stonesand possibly even clues to the unusu-al nature of this phenomenon.Perhaps East Africa will become anew source of copper-manganese-bearing change-of-color tourmalines.

CYW

PHOTO CREDITSElizabeth Schrader—1, 3, 6, 8, 11, 12, 14,and 21; Harold and Erica Van Pelt—2; Wuyi Wang—4, 5, and 9; C. D. Mengason—13 and 24; Maha Tannous—15; Shane F.McClure—16–19; Christopher M. Breeding—20; Wendi M. Mayerson—22 and 23.

estimated from the spectrum to be 15 ppm by comparisonwith samples of known nitrogen content.

A strong “amber center” with its main peak at 4165cm−1 was visible in the near-infrared region of the spectrum(again, see figure 2), indicating a deformation-related col-oration. This also was unusual, since the amber center istypical for type Ia brown diamonds colored by deformationand related defects (L. DuPreez, “Paramagnetic Resonanceand Optical Investigation of Defect Centres in Diamond,”Ph.D. dissertation, University of Witwatersrand, Johannes-burg, 1965). In “olive” and brown type Ib diamonds withdeformation-related coloration, the main absorption of thisdefect has been found at 4115 cm−1 or as a doublet at 4165and 4065 cm−1 (T. Hainschwang, “Classification and colororigin of brown diamonds,” Diplôme d’Université deGemmologie, University of Nantes, France, 2003). A “sin-gle” amber-center peak at 4165 cm−1 in a type Ib diamondhas not previously been described.

The deformation-related color was confirmed by obser-vation in transmitted light between crossed polarizers. Thediamond showed very distinct parallel gray to black extinc-

DIAMONDS An untreated type Ib diamond exhibiting green transmis-sion luminescence and H2 absorption. These contributorsrecently analyzed a small greenish brownish yellow(“olive yellow”) diamond that exhibited green transmis-sion luminescence (figure 1) as well as an unusual combi-nation of absorption features. At first sight, the 0.12 ct dia-mond did not appear particularly remarkable, except for itsUV luminescence, which was green to long-wave andgreenish yellow to short-wave UV radiation. However, aninfrared spectrum showed that the diamond was a low-nitrogen type Ib/IaA, with the single nitrogen clearly dom-inating the A-aggregates (figure 2; see specifically the insetshowing the 1358–1000 cm−1 region). This was surprising,since green luminescence caused by the H3 center (thecombination of paired nitrogen [A aggregate] with a vacan-cy) is commonly observed in type Ia diamonds but seenonly very rarely in type Ib diamonds. Furthermore, theextremely low A aggregate concentration would not nor-mally indicate the formation of distinct H3 luminescence.The total amount of nitrogen in the diamond was roughly

252 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY FALL 2004

EDITORBrendan M. Laurs ([email protected])

CONTRIBUTING EDITORSEmmanuel Fritsch, IMN, University of Nantes, France ([email protected])

Henry A. Hänni, SSEF, Basel, Switzerland([email protected])

Kenneth V. G. Scarratt, AGTA Gemological TestingCenter, New York ([email protected])

Karl Schmetzer, Petershausen, Germany ([email protected])

James E. Shigley, GIA Research,Carlsbad, California ([email protected])

Christopher P. Smith, GIA Gem Laboratory,New York ([email protected])

Figure 1. This 0.12 ctdiamond, shown on theleft in daylight, exhibitsgreen transmissionluminescence in dark-field illumination(right), as well as anunusual combination ofabsorption features.Photos by T.Hainschwang.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY FALL 2004 253

tion in two directions along octahedral growth planes, fol-lowing “olive”-colored graining visible in the stone (figure2, inset). This colored graining and extinction along thegraining are very common features in “olive” diamonds.The order of extinction indicates that the stone is severelydeformed and thus not optically isotropic. This is explainedby the fact that the dislocations (broken bonds) caused bythe deformation interfere with the direct passage of lightthat would be expected in truly isotropic materials. Eventhough strain associated with linear extinction and/orinterference colors can be found to some degree in practi-cally all diamonds, strong parallel extinction in a coloreddiamond provides a good indication for deformation-relatedcoloration. Regardless of whether their green coloration ishydrogen- or radiation-related, “olive” diamonds seldom

exhibit the strong strain pattern described here. These observations prompted further analysis of the

sample. A low-temperature spectrum was recorded in thevisible/near-infrared range, which added to the unusualassemblage of absorption centers found in this diamond.The spectrum exhibited a combination of weak H3 (503nm), NV− (637 nm), and H2 (986 nm) absorptions withassociated structures, plus moderate broad bands at about550 and 800 nm (figure 3). The H3, NV−, and H2 absorp-tion centers are typical of the visible-NIR spectra ofHPHT-treated type Ia diamonds (A. T. Collins et al.,“Colour changes produced in natural brown diamonds byhigh-pressure, high-temperature treatment,” Diamondand Related Materials, Vol. 9, 2000, pp. 113–122), and alsocan be created in type Ia or Ib diamonds through irradia-tion followed by annealing (A. M. Zaitsev, OpticalProperties of Diamond: A Data Handbook, Springer-Verlag, Berlin, 2001, pp. 136–137), although these featurestypically are moderate to strong in treated-color diamonds.

Despite this correlation, other features indicatedunambiguously that the diamond had not been treated byeither method, and was indeed natural color. HPHT treat-ment of a brown-to-olive type Ib diamond, even at verymoderate temperatures, would aggregate much of the sin-gle nitrogen and destroy the amber center (T.Hainschwang et al., “HPHT treatment of different classesof type I brown diamonds,” Journal of Gemmology, Vol.29, 2004 [in press]; A. N. Katrusha et al., “Application ofhigh-pressure high-temperature treatment to manipulatethe defect-impurity content of natural diamond singlecrystals,” Journal of Superhard Materials, No. 3, 2004, pp.47–54). Besides the temperature/pressure conditions, theaggregation is influenced by the total nitrogen content andthe types of defects present. The combination of factors in

Editor’s note: The initials at the end of each item identify the editor or contributing editor who provided it. Full names and affiliations are given for other contributors.

Interested contributors should send information and illustrations to Brendan Laurs at [email protected] (e-mail), 760-603-4595 (fax), or GIA, 5345 Armada Drive, Carlsbad, CA 92008. Original photos will be returned after considera-tion or publication.GEMS & GEMOLOGY, Vol. 40, No. 3, pp. 252–269© 2004 Gemological Institute of America

Figure 2. The FTIR spectrum of the diamond in figure1 reveals that it is a type Ib/IaA, with isolated nitro-gen clearly dominating the A-aggregates. The regionbetween at least 5000 and 4165 cm−1 comprises the“amber center,” of which the 4165 cm−1 peak is themain feature. The amber center is deformation relat-ed; the deformation is apparent in the extinction pat-terns visible with crossed polarizers (see inset; photoby T. Hainschwang).

Figure 3. The low-temperature Vis-NIR spectrum ofthe diamond in figure 1 shows weak absorptions forthe H3, NV−, and H2 centers, which is an unusualcombination for an untreated natural diamond.


this diamond would enhance nitrogen aggregation undercommonly used HPHT conditions. In addition, irradiationof a type Ib diamond followed by annealing would create avery distinct NV− absorption, resulting in pink to purplecoloration (E. Bienemann-Küespert et al., GmelinsHandbuch der Anorganischen Chemie, Verlag Chemie,Weinheim, Germany, 1967, p. 237). Radiation treatmentwould also leave other traces, such as the 595 nm and theH1a and possibly H1b absorptions, which were not detect-ed in this diamond.

The authors have recently seen the H2 center in a suiteof very rare type Ib diamonds containing large concentra-tions of single-nitrogen that may exceed 400 ppm. Thesestones showed no deformation-related features and weredistinctly different from the stone reported here. In contrastto these high-nitrogen type Ib diamonds, this is the firsttype Ib “H2 diamond” we have seen that shows a combina-tion of H3, NV−, and H2 centers with classic strain patternsbetween crossed polarizers and a very low nitrogen concen-tration. The properties observed for this diamond are, atthis point, difficult to explain. The deformation pattern andcolor distribution indicate octahedral growth and dynamicpost-formation conditions. Besides the strong post-growthdeformation and associated defects (dislocations, vacancies,and interstitials), the observed features suggest prolongednatural annealing at a low enough temperature to avoidaggregation of the single nitrogen, but nevertheless result-ing in the combination of defects noted.

Thomas Hainschwang ([email protected])Gemlab Gemological Laboratory

Vaduz, Liechtenstein

Franck NotariGemTechLab Laboratory

Geneva, Switzerland

COLORED STONES ANDORGANIC MATERIALSGem amphiboles from Afghanistan, Pakistan, andMyanmar. The amphibole group consists of several com-mon rock-forming minerals, as well as many unusualspecies. Examples that are best known to gemologists aretremolite and actinolite, which as fine-grained aggregatesform nephrite. Like most amphiboles, these are opaque, ortranslucent at best. In the past few years, however, someunusual transparent amphiboles from three localities inAsia have been faceted. These include light yellow rich-terite from Afghanistan, green pargasite from Pakistan,and brown pargasite and near-colorless edenite fromMyanmar. One of these contributors (DB) has obtainedfacet-quality examples of all these amphiboles from localdealers in Peshawar (Pakistan) and Mogok (Myanmar), andalso recently visited one of the deposits.

The Afghanistan richterite was first seen in thePeshawar mineral market in October 2001. The materialwas sold with sodalite and hackmanite crystals, often

associated on the same specimen. The source was reportedto be in the vicinity (i.e., an approximately six- to eight-hours’ walk) of the Sar-e-Sang lapis lazuli deposits, whichare located in the Kokcha Valley, Badakhshan Province.During 2002, DB saw at least 5 kg of rough material,including some attractive crystals (figure 4); about 20%was facet grade. However, due to the mineral’s perfectcleavage, very few stones have been cut. The largest rich-terite cut by DB weighed 1.86 ct; attempts to cut largerstones have been unsuccessful.

The Pakistan pargasite appeared on the mineral marketin the mid-1990s, typically as broken crystals embedded ina marble matrix. Similar material from China wasdescribed in the Spring 2002 Gem News International sec-tion (p. 97). Due to its attractive green color (figure 5), thePakistan pargasite is sometimes referred to as “Hunzaemerald” by local dealers. DB visited the mining area inNovember 2003. It is located about 3 km east of theKarakoram Highway bridge that crosses the Hunza Rivernear Ganesh in the Hunza region. The pargasite is foundwithin marble boulders that contain small seams of phlo-gopite. In addition to the green pargasite, the area hasyielded translucent-to-opaque red, pink, and purple-bluecorundum (to 7.5 cm) and “maroon,” blue, dark brown,and black spinel (to 2.5 cm). Pargasite crystals up to almost3 cm have been found, sometimes associated with the

Figure 4. Beginning in late 2001, facet-quality richteritehas been recovered from the vicinity of Afghanistan’slapis lazuli deposits. The crystal shown here is 1.9 cmtall, and the oval brilliant weighs 1.72 ct. Courtesy ofDudley Blauwet Gems; photo © Jeff Scovil.


dark brown spinel, but transparent material is very rare,yielding faceted stones of less than 1 ct.

The two transparent amphibole species fromMyanmar(figures 6 and 7) were purchased by DB in Mandalay inJune 2002. The vendor, a geology student from a universi-ty in Yangon, reported that both the pargasite and edenitewere mined from the well-known Mogok deposit of OhnBin. Only a few pieces were available, although severalmore samples of both minerals turned up in 2003. Thelargest pargasite and edenite cut by DB weighed 1.78 ctand 0.29 ct, respectively (again, see figures 6 and 7).

To positively identify the specific amphibole speciespresent, one of us (FCH) analyzed the faceted stones pic-tured in figures 4–7 by electron microprobe. Also analyzedwas one additional sample of brown pargasite fromMyanmar. Approximately 10 point analyses were obtainedfrom each sample, and their averages were used to calcu-late the formulas listed in table 1. The mineralogical clas-sification within the amphibole group of each stone wasthen established using the conventions published by B. E.Leake et al. (“Nomenclature of amphiboles: Report of thesubcommittee on amphiboles of the International Miner-alogical Association, Commission on New Minerals andMineral Names,” Canadian Mineralogist, Vol. 35, 1997,pp. 219–246). All the amphiboles contained relatively highamounts of fluorine. The analyses also revealed that thegreen pargasite from Pakistan contains traces of vanadium;up to 1.3 wt.% V2O3 was reported in similar material byV. M. F. Hammer et al. (“Neu: Grüner Pargasit ausPakistan,” Lapis, Vol. 24, No. 10, 1999, p. 41).

Gemological properties were obtained on the samefour samples (by SM and EPQ) using standard testingequipment and a gemological microscope; the data aresummarized in table 1. The properties of each stone fellwithin the ranges reported for these amphibole varieties inmineralogical textbooks (see, e.g., W. D. Nesse, Intro-duction to Mineralogy, Oxford University Press, NewYork, 1991, pp. 277–290). The relatively low R.I. values ofthe richterite are consistent with its Mg-rich composition(i.e., lacking iron); its properties may overlap those of col-orless tremolite. The properties of the pargasites and theedenite also are consistent with the literature—that is,with values reported for hornblende. Material referred toas “hornblende” also includes other closely related speciesof the amphibole group; because their physical propertiesmay overlap, conclusive identification of these speciesrequires quantitative chemical analysis.

Dudley Blauwet ([email protected])Dudley Blauwet Gems

Louisville, Colorado

Frank C. HawthorneUniversity of Manitoba

Winnipeg, Manitoba, Canada

Sam Muhlmeister and Elizabeth P. QuinnGIA Gem Laboratory, Carlsbad

Figure 6. Transparent yellowish brown pargasite wasrecently found in Myanmar. The crystal is 1.5 cm tall,and the faceted stone weighs 1.78 ct. Courtesy ofDudley Blauwet Gems; photo © Jeff Scovil.

Figure 5. This vivid green pargasite from Pakistan iscolored by vanadium. The specimen is 2.1 cm tall,and the oval brilliant weighs 0.59 ct. Courtesy ofDudley Blauwet Gems; photo © Jeff Scovil.


Recent gem beryl production in Finland. In 1982, a shardof transparent colorless topaz was found at a road con-struction site near Luumäki in southern Finland.Subsequently, local mineral enthusiasts and gem cuttersstaked a claim to the source of this stone, a granitic peg-matite running parallel to the road, and initial miningresulted in the discovery of at least one gem beryl pocket.

This deposit subsequently produced some significantcrystals and gem material (see S. I. Lahti and K. A.Kinnunen, “A new gem beryl locality: Luumäki, Finland,”

Spring 1993 Gems & Gemology, pp. 30–37). The peg-matite was mined until 1995, during the summer seasons.Although several pockets were found in areas adjacent toquartz core zones, none of them contained any beryl min-eralization. Indeed, the distribution of the beryl in the peg-matite proved to be quite sporadic.

In the past three years, renewed work at this deposit(now known as the Karelia Beryl mine; figure 8) by anewly formed mining company has yielded some addition-al production, which recently included what many believeare some of the finest and largest green gem beryls everfound in western Europe (figure 9). This contributor had anopportunity to witness the removal of some of this materi-al during a May 2004 visit to the deposit, which is situatedon a small island in one of the 70,000 Finnish lakes. Theexact location is Kännätsalo (Finnish for drunken forest),Kivijärvi (Stone Lake), Luumäki, Karelia, Finland.

The new gem beryl pocket had a vertical orientationand measured approximately 2 ∞1.5 ∞4 m. The location ofthis large cavity in the pegmatite was unusual, in that itwas found only 20–40 cm from the hanging wall. Thepocket contained two layers of large gem beryls, one nearthe top and the other about 30 cm from the bottom.Between these layers were found broken translucent berylcrystals (up to 10–15 cm in diameter) of cabochon andcarving quality; some areas showed chatoyancy. The berylin this pocket ranged from light yellow to deep “golden”yellow, and from green-yellow to green.

Most of the gem beryl was found within 10–25 cm ofthe cavity walls, together with crystals of mica (typically5–15 cm) and albite (1–8 cm), as well as pocket rubble con-sisting of broken shards of microcline and, rarely, quartz.A total of more than 110 kg of beryl was produced, withabout 30 kg being suitable for faceting. Most of the fac-

TABLE 1. Characteristics of gem amphiboles from Afghanistan, Pakistan, and Myanmar.a

Fluorescence

Long-wave UV Short-wave UV

Richterite Afghanistan 1.72 Light brownish 1.599– 0.023 3.11 Weak orange Weak yellowish Cavities, fractures, two-phase yellow 1.622 orange inclusions

Pargasite Pakistan 0.59 Yellowish green 1.625– 0.015 3.03 Inert Very weak yellowish Numerous fractures (some along1.640 green cleavage directions), needles,

two-phase inclusionsPargasite Myanmar 1.78 Yellowish brown 1.620– 0.020 3.17 Inert Weak yellow Numerous three-phase inclu-

1.640 sions (some with tension frac-tures), two-phase inclusions, needles, fractures

Edenite Myanmar 0.29 Near colorless 1.612– 0.019 3.14 Inert Weak yellow Needles, fractures, three-phase (very light yellow) 1.631 inclusions (some with tension

fractures), two-phase inclusions

a None of the samples showed any phosphorescence or any absorption features with a desk-model spectroscope. b Microscopy also revealed evidence of clarity enhancement in the 0.59, 1.72, and 1.78 ct samples.c Electron microprobe analysis of an additional sample (a grain mounted in epoxy) of yellowish greenish brown potassian fluorian pargasite from Mogokyielded the formula (Na0.58K0.39 )Ca2.00(Mg3.90 Fe0.05 Cr0.01Ti0.13 Al0.80 )(Si5.84 Al2.16 )O22(OH)1.13 F0.87.

Name Locality Color R.I. Birefringence S.G. Internal featuresbWeight(ct)

Figure 7. Myanmar is also the source of facetableedenite, as shown by this 1.4-cm-wide crystal and

0.29 ct round brilliant. Courtesy of Dudley Blauwet Gems; photo © Jeff Scovil.


etable material was found as small (1–3 cm) fragments (fig-ure 10), but some pieces reached 5–8 cm. About 10 kgcomprised “flawless” material with no eye-visible inclu-sions, and in exceptional cases these crystals exceeded 1kg each. These large stones are of a fine green to yellowishgreen color, and are rather distinct from the better-knownUkrainian material in terms of both color and morpholo-gy. The Finnish crystals are typically etched and striated(but have retained their hexagonal shape), and their termi-nations are rounded (as pictured by Lahti and Kinnunen,

(Na0.60K0.32)(Ca1.14Na0.85Mg0.01)(Mg4.90Al0.10) Potassian fluorian richterite(Si7.83Al0.17)O22(OH)1.59F0.41

(Na0.85K0.07)Ca2.00(Mg4.18V0.21Al0.61)(Si6.25Al1.75) Fluorian vanadian pargasiteO22(OH)1.57F0.43

(Na0.52K0.43)(Ca1.95Na0.03Fe0.02) Potassian fluorian pargasite(Mg3.97Fe0.07Ti0.08Al0.88)(Si6.05Al1.95)O22(OH)1.29F0.71

(Na0.87K0.09)(Ca1.80Na0.15Fe0.03Mg0.02) Fluorian edenite(Mg4.45Ti0.01Al0.54)(Si6.62Al1.38)O22(OH)1.54F0.46

Chemical compositionc Classification

Figure 9. In May 2004, several world-class crystals ofgreen gem beryl—some exceeding 1 kg, as shownhere—were recovered at the Luumäki pegmatite.Photo by P. Lyckberg.

Figure 8. Mining activi-ties have resumed atthe gem beryl pegmatitenear Luumäki, Finland.Heavy machinery isused to reach the gem-bearing zones, whichare then carefullymined by hand to avoiddamaging the crystals.Photo by P. Lyckberg.


1993). Although it would be possible to cut stones weigh-ing several thousand carats from the largest fine gem crys-tals, these crystals have been sold as specimens to collec-tors to preserve their natural beauty.

When present, inclusions in the beryl are similar to

those described by Lahti and Kinnunen (1993). For exam-ple, a 3 cm golden yellow heliodor studied by this contrib-utor had dozens of parallel channels/tubes along the c-axis,and these were filled by rust-colored clay minerals. Noneof the beryl was found to contain the clouds of minuteinclusions that are typical of Ukrainian beryl.

Approximately 100 large stones (25–50 ct) from therecent production have been cut by Finnish masterfaceter Reimo Armas Römkä. Most of them are light yel-low and some are yellowish green. In addition, about11,000 round brilliants (10 mm in diameter) have beencut in China. Most of the cut stones are sold into thedomestic Finland market.

The 2004 mining season will continue until the severewinter halts activities for the following six months.

Peter Lyckberg ([email protected])Luxembourg

Hessonite from Afghanistan. Over the past few years,Afghanistan has become a significant source of gem-quali-ty hessonite (grossular garnet), and some of the materialhas been found mixed into parcels with other orange tored stones from this region (see Summer 2001 Gem NewsInternational, p. 144). To our knowledge, the gemologicalproperties of the Afghan hessonite have not been pub-lished, so we were interested to examine several facetedsamples and mineral specimens that were recently loaned(and, in some cases, donated) to GIA by Sir-Faraz(“Farooq”) Hashmi of Intimate Gems, Jamaica, NewYork. In addition, Peter Lyckberg loaned six facetedexamples for our examination. Mr. Hashmi reported thatthe hessonite comes from eastern Afghanistan; twoknown deposits are Munjagal in Kunar Province (produc-ing roughly 1,500–2,000 kg/year of mixed grade) andKantiwow, Nuristan Province (up to 5,000 kg/year). Mostof the clean rough weighs 0.5–1 gram, in colors rangingfrom yellowish orange to red-orange. Although thousandsof kilograms of this garnet have been produced, mininghas waned in recent months due to lack of demand in thelocal market (i.e., in Peshawar, Pakistan) and the migra-tion of miners to the kunzite deposits in the same regionof Afghanistan.

The specimens we examined consisted of euhedralgarnets that were commonly intergrown with anhedralmassive quartz. This assemblage formed within massivegarnet that was intergrown with quartz and, less com-monly, a green mineral (probably epidote) and anotherwhite mineral (possibly wollastonite). This mineralassociation is typical of a skarn-type deposit formed bycontact metamorphism of carbonate rocks by a graniticintrusion. The euhedral grossular crystals typically mea-sured up to 1 cm in diameter, although some partialcrystals of larger dimension (up to 4 cm) also were pres-ent. The crystals contained abundant fractures, butsome had small areas that were transparent enough forfaceting.

Figure 10. Most of the Luumäki gem beryl from the May2004 pocket consisted of relatively small fragments. Theinset shows a green beryl weighing approximately 15 ctthat was cut in Finland. Photos by P. Lyckberg.

Figure 11. Afghanistan has produced attractive hes-sonite over the past few years. These stones, weigh-ing 2.64, 7.39, and 3.51 ct (from left to right), show

the range of color that is commonly encountered inthis material. Courtesy of Intimate Gems;

photo by Maha Tannous.


Three representative faceted stones (2.64–7.39 ct; figure11) were selected for examination by one of us (EPQ), andthe following properties were obtained: color—yellowishorange, orange, and red-orange; diaphaneity—transparent,R.I.—1.739 or 1.740; S.G.—3.63 or 3.64; weak to moderateADR in the polariscope; and inert to both long- and short-wave UV radiation. Weak absorption bands at 430 and 490nm were observed with a desk-model spectroscope.Microscopic examination revealed transparent near-color-less crystals (one of which was identified as apatite byRaman analysis), needles, “fingerprints,” stringers of parti-cles, fractures, and straight and angular growth lines. Onestone showed evidence of clarity enhancement. The R.I.values of these samples are slightly lower than thosereported in the literature for hessonite (see R. Webster,Gems, 5th ed., rev. by P. G. Read, Butterworth-Heinemann, Oxford, 1994, pp. 201–202). Notably, thethree Afghan samples did not show the roiled or oilyappearance that is commonly seen in hessonite; nor wasthis feature noted upon further examination of severaladditional faceted stones.

BML

Elizabeth P. QuinnGIA Gem Trade Laboratory, Carlsbad

Interesting abalone pearls. In the Winter 2003 Gem NewsInternational section (pp. 332–334), this contributorreported on some interesting pearls that had been loanedby Jeremy Norris of Oasis Pearl in Albion, California.Two of those were unusual specimens from the greenabalone Haliotis fulgens and the red abalone H. rufescens.Mr. Norris recently loaned GIA two additional abalonepearls from the waters off Baja California, Mexico.

One of these pearls was an exceptional example of a

pearl from the pink abalone H. corrugata. This 4.90 ctlight-toned pearl (11.4 ∞ 8.9 ∞ 6.9 mm) displayed a stun-ning array of colors (figure 12).

The other pearl was a 41.03 ct horn-shaped specimen(36.6 ∞ 23.0 ∞ 14.5 mm) from H. fulgens. What was sounusual about this particular pearl was its remarkableresemblance to an eagle’s head, complete with eye, brow,and beak structures (figure 13). Such horn shapes—a formtypically exhibited by abalone pearls—may be solid, butoften they are hollow. This particular specimen was fun-nel shaped, with the narrow end of the hole forming theapparent “eye.” The nacre displayed vibrant hues of blue,

Figure 12. Showing a beautiful array of colors, this4.90 ct abalone pearl is an exceptional example fromthe pink abalone H. corrugata. Courtesy of JeremyNorris; photo by C. D. Mengason.

Figure 13. This unusual 41.03 ct pearl from the green abalone H. fulgens has a remarkable resemblance to aneagle’s head (left). The funnel-shaped hole, which starts at the wide opening at the back of the “head,” and exitsout the eagle’s “eye,” exhibits the same smooth vibrant nacre as the outside of the pearl (right, looking down thewide opening). Courtesy of Jeremy Norris; photos by C. D. Mengason.


green, and purple-pink, even in the interior (again, see fig-ure 13). For the nacre to have formed so evenly on theinside surfaces of the pearl, Mr. Norris stated that theabalone’s nacre-secreting tissue must have passed all theway through the pearl, so that the pearl completely encir-cled part of the gastropod’s anatomy, and in essenceentrapped its host. This pearl has wonderful potential for ajewelry designer who could incorporate the bird-like imageinto a one-of-a-kind creation.

Cheryl Y. Wentzell ([email protected])GIA Gem Laboratory, Carlsbad

Rhodonite of facet and cabochon quality from Brazil.Rhodonite typically occurs as a translucent-to-opaqueornamental stone with an attractive pink color. Trans-

parent facetable material has so far been very rare andhas been reported only from Broken Hill, New SouthWales, Australia, where it is frequently associated withthe closely related mineral pyroxmangite (see H. Bank etal., “Durchsichtiger rötlicher Pyroxmangit aus BrokenHill/Australien und die Möglichkeiten seiner Unter-scheidung von Rhodonit [Transparent reddish pyroxman-gite from Broken Hill/Australia and the criteria for dis-tinguishing it from rhodonite],” Zeitschrift derDeutschen Gemmologischen Gesellschaft, Vol. 22, No.3, 1973, pp. 104-110). Recently, however, transparent-to-semi-transparent rhodonite also has been found in MinasGerais, Brazil.

During the February 2004 Tucson gem shows, JosephRott (of Tropical Imports, formerly based in New York andnow located in Grand Island, Nebraska) showed one ofthese contributors (BML) some attractive rough and cutexamples of Brazilian rhodonite. He said that the materialcomes from the same region in Minas Gerais that histori-cally has been mined for manganese (see, e.g., F. R. M.Pires, “Manganese mineral paragenesis at the LafaieteDistrict, Minas Gerais,” Anais da Academia Brasileira deCiencias, Vol. 54, No. 2, 1982, p. 463). He also indicatedthat the deposit contains both rhodonite and pyroxman-gite. Although these minerals can be visually indistin-guishable and may be intergrown within the same piece,he reported that only minor amounts (if any) of pyroxman-gite are known to be present in the gem material from thisdeposit. Most of the faceted rhodonites are between 1 and5 ct, although larger gems have been cut; cabochons typi-cally range from 5 to 10 ct (figure 14).

Mr. Rott loaned several rough and cut samples to GIAfor examination. Two polished stones were examined byone of us (EPQ), a transparent oval modified brilliant (1.77ct) and a semi-transparent oval double cabochon (8.55 ct),as illustrated in figure 15. The following properties wererecorded: color—orangy red, with moderate pinkish orangeand purplish pink pleochroism; R.I.—1.720–1.733 (facetedstone) and 1.73 (spot reading on the cabochon); birefrin-gence—0.013; S.G.—3.69 and 3.66, respectively; and fluo-

Figure 14. A new find of rhodonite from MinasGerais, Brazil, has yielded facet- and cabochon-quali-

ty material. The faceted stones weigh 1.84 and 8.22ct, and the cabochons are 9.98 and 32.0 ct. Courtesy

of Alan Freidman, Beverly Hills, California; photo © Harold & Erica Van Pelt.

Figure 15. The Brazilian rhodonite samples examinedfor this study consisted of an oval modified brilliant(1.77 ct) and an oval cabochon (8.55 ct). Courtesy ofJoseph Rott; photo by Maha Tannous.


rescence—inert to very weak red to both long- and short-wave UV radiation. The following absorption bands wereobserved with the desk-model spectroscope: a strong bandat 410 nm, weak lines at 430 and 460 nm, a moderate lineat 500 nm, and a wide band at 520–560 nm. Microscopicexamination revealed numerous randomly oriented curvedneedles, cleavage cracks, “fingerprints,” and two-phaseinclusions in both samples, as well as some transparentbrown-yellow euhedral crystals in the cabochon.

The gemological properties are comparable to thosereported for rhodonite in the literature (see e.g., R.Webster, Gems, 5th ed., rev. by P. G. Read, Butterworth-Heinemann, Oxford, 1994, p. 365). Although some of theproperties of rhodonite and pyroxmangite overlap, the rel-atively low R.I.’s and birefringence of the two samples weexamined are indicative of rhodonite. Also, Raman analy-sis of two spots on the cabochon and one spot on thefaceted stone yielded spectra that more closely matchedrhodonite than pyroxmangite. EDXRF analyses of the twosamples by GIA Gem Laboratory senior research associateSam Muhlmeister showed major amounts of Si and Mn,as well as traces of Fe and Ca (and possibly Zn in thefaceted stone).

As with the Australian material, the main challengewith cutting the Brazilian rhodonite is its perfect cleavagein two directions. This, combined with a Mohs hardnessof 51/2–61/2 and the fact that limited transparent materialis available for faceting, means that it will remain a collec-tor’s stone. Nevertheless, the availability of even a smallamount of faceting-quality rhodonite from the Braziliansource has created interesting opportunities for setting thematerial into jewelry (figure 16).

Elizabeth P. Quinn ([email protected])GIA Gem Laboratory, Carlsbad

BML

Spessartine and almandine-spessartine from Afghanistan.Beginning in mid-2002, these contributors received occa-sional reports of new spessartine discoveries inAfghanistan, and a few faceted stones stated to be from thisproduction were seen at the Tucson gem shows in 2003and 2004. Recently, a multitude of rough and cut samplesof this material were loaned (and, in some cases, donated)to GIA by Sir-Faraz (“Farooq”) Hashmi of Intimate Gems.Most of these samples were purchased in late 2003, in themineral bazaar at Peshawar, Pakistan. The dealers reportedthe garnets were mined from pegmatites in the Darre Pecharea of Kunar Province, where they were apparently recov-ered as a byproduct of mining for kunzite and tourmaline.

The rough material we examined (see, e.g., figure 17)consisted of a 385-gram parcel of loose pieces and two

Figure 17. Since mid-2002, increasing amounts ofspessartine (and some almandine-spessartine) haveemerged from Afghanistan. The spessartine in matrixmeasures at least 3 cm in diameter, and the loosecrystals weigh 3.59–15.47 ct. Courtesy of IntimateGems; photo by Maha Tannous.

Figure 16. The 2.67 ct Brazilian rhodonite in this pendant is set with 111 yellow and 25 pink dia-monds (with a total weight of 0.48 and 0.12 carats,respectively). Courtesy of Alan Freidman; photo © Harold and Erica Van Pelt.


matrix specimens. One of the specimens was a smallcrystal of kunzite (2.6 cm long) associated with spessar-tine and feldspar, while the other consisted of spessartinein a matrix of albite (cleavelandite variety) and K-feldsparthat was covered by a thin layer of a porcelaneous clay-like material (again, see figure 17). The spessartine crystalin the latter specimen measured at least 3 cm in diameter,with some areas suitable for faceting. The rough parcelconsisted of broken fragments and a few well-formed crys-tals, as well as pieces that were moderately to heavily cor-roded (as is typical of spessartine from some pegmatites).

The faceted examples we examined comprised two dis-tinct color groups. Each group was cut from rough pur-chased at different times, but represented as being from thesame mining area. One group ranged from yellow-orange toorange (figure 18), and the other ranged from orange-red todark red (figure 19). Two stones from each color group werechosen by one of us (EPQ) for examination. The followingproperties were obtained: R.I.—1.799 (yellow-orange), 1.802(orange), and both red stones were above the limits of astandard refractometer; S.G.—4.26 (yellow-orange), 4.28(orange), and 4.22 (orange-red), and 4.24 (red); fluores-cence—all were inert to both long- and short-wave UV radi-ation; and all had similar absorption spectra when viewedwith a desk-model spectroscope. The absorption featuresconsisted of strong bands at 410 and 430 (although thesetwo bands converged in the red stones, creating a cutoff at440 nm), with weaker bands at 460, 480, 505, 520, and 570nm. In the two red stones, the 505 and 570 nm bands weremore pronounced than they were in the orange stones.This is consistent with their greater inferred iron content,as indicated by their darker and redder color. Based on theseproperties, the orange-red to dark red garnets are probably amixture of spessartine and almandine.

Microscopic examination revealed “fingerprints,” two-phase inclusions, and needles in all four of the samples.The properties of the yellow-orange to orange stones arecomparable to those reported for spessartine from otherdeposits (see compilation in B. M. Laurs and K. Knox,“Spessartine garnet from Ramona, San Diego County,California,” Winter 2001 Gems & Gemology, pp.278–295), except for the higher S.G. values obtained for theAfghan samples in this study.

According to Mr. Hashmi, most of the facetable roughseen thus far in the Peshawar market has weighed lessthan 2 g, although some 3–5 g pieces were available andthe largest clean rough known to him weighed 15 g (a

Figure 20. At 12.58 ct, this oval brilliant provides afine example of a relatively large spessartine fromAfghanistan. Courtesy of Mark Kaufman ofKaufman Enterprises, San Diego, California; photoby Maha Tannous.

Figure 18. This group of yellow-orange to orange spessar-tines from Afghanistan ranges from 0.78 to 1.68 ct.

Courtesy of Intimate Gems; photo by Maha Tannous.

Figure 19. This group of orange-red to dark redalmandine-spessartines (0.41–1.28 ct) is reportedlyfrom the same mining area as the orange spes-sartines in figure 18. Courtesy of Intimate Gems;photo by Maha Tannous.


well-formed crystal). The faceted material has typicallyranged up to 2 ct, although a 12.58 ct oval brilliant report-edly from this locality was seen at the 2003 Tucson gemshows (figure 20).

Curiously, faceted examples of this spessartine look verysimilar to the hessonite that also has come from easternAfghanistan in recent years (see entry on pp. 258–259 of thissection). In fact, Mr. Hashmi cautioned that some roughparcels he has examined contained both types of garnets.

Elizabeth P. QuinnGIA Gem Laboratory, Carlsbad

BML

Gem tourmaline from Congo. Africa has long been animportant source of gem tourmaline; of particular interestare the countries of Nigeria, Namibia, Zambia, andMozambique. In recent years, however, the DemocraticRepublic of the Congo (DRC, formerly known as Zaire)has occasionally yielded attractive gem rough and collec-tor-quality crystals. While information on the exactsources in the DRC was not available to these contribu-tors, the appearance and composition of these tourma-lines indicates they are derived from granitic pegmatites.

The Central African pegmatite province includesnumerous pegmatite fields in a broad region encompass-ing Uganda, Rwanda, Burundi, the eastern DRC, north-ern Zambia, western Kenya, and western Tanzania. Mostof these pegmatites are categorized in the rare-metalclass and are associated with Early Proterozoic granites(800–1,000 million years old; V. Ye. Zagorsky et al.,Miarolitic Pegmatites, in Vol. 3 of B. M. Shmakin and V. M. Makagon, Eds., Granitic Pegmatites, Nauka,Siberian Publishing Firm RAS, Novosibirsk, Russia, 1999[in Russian]). In the eastern DRC, the pegmatites arelocated in the Nord-Kivu, Sud-Kivu, and Katangaprovinces (N. Varalamoff, “Central and West Africanrare-metal granitic pegmatites, related aplites, quartzveins and mineral deposits,” Mineralium Deposita, Vol.7, 1972, pp. 202–216). These deposits have been minedfor decades for cassiterite (Sn) and industrial beryl (Be),but miarolitic pegmatites that host gem-quality tourma-line, beryl, and other minerals are apparently uncommonin the DRC.

In mid-2000, gem dealer John Patrick (El Sobrante,California) obtained about 200 g of variously coloredDRC tourmaline through an African supplier. The sup-plier indicated the material came from the Virungaregion north of Goma (Nord-Kivu Province). Mr. Patrickdonated three green/pink crystals (2.5–4.0 cm long) andfive multicolored slabs (1.5–2.9 cm wide) to GIA forresearch purposes. The slabs had irregular outlines andwere concentrically zoned around the c-axis in pink andgreen; one sample had a blue core. All of these sampleswere semitransparent due to fluid inclusions and fis-sures, as are typically seen in tourmaline.

More recently, an undisclosed locality in the DRChas yielded transparent prisms of mostly green-to-bluetourmaline (figure 21). Steve Ulatowski (New EraGems, Grass Valley, California) first obtained thismaterial in mid-2003, and he estimates that about20–30 kg/month were produced in early 2004. Sincethen, however, production appears to have diminished.Mr. Ulatowski purchased about 3 kg of the rough, ofwhich 30% was facetable (see, e.g., figure 22), 60% wascabochon grade, and 10% was bead quality. The largestclean piece of gem rough he acquired weighed approxi-mately 30 grams, although stones faceted from the

Figure 22. These tourmalines (5.78 and 15.94 ct) werecut from material similar to that in figure 21. Thesestones were faceted by Thomas Trozzo (Culpeper,Virginia). Photo by C. D. Mengason.

Figure 21. Transparent tourmalines, for the most partgreen to blue, recently have been recovered from theDemocratic Republic of the Congo. These crystals, upto 4 cm long, are courtesy of New Era Gems; photo byMaha Tannous.


prismatic crystals typically weighed 3–5 ct. A smallproportion of the crystals had a slightly waterwornappearance, but most showed no evidence of alluvialtransport. Most of the tourmaline was green, althoughsome blue-green, yellow-green, and rare bright pinkmaterial also was produced.

Figure 23. Heating experiments were conducted onthe bottom portions of these four crystals of DRCtourmaline, while the top untreated portions wereretained for comparison. All samples were heatedin air as follows (from left to right, with weight oforiginal crystal in parentheses): 427°C over a periodof 5 hours (31.61 ct), 482°C over 6 hours (34.84 ct),566°C over 8 hours (19.11 ct), and 621°C over 9hours (19.89 ct). No appreciable change in thegreen-to-blue colors was seen in any of the heatedportions, but abundant microcracks significantlyreduced their transparency. Donated to GIA collec-tion (nos. 30834–30837) by New Era Gems; photoby Maha Tannous.

Figure 25. The attractive color zoning shown by theseDRC tourmalines (up to 3.6 cm long) ranges frompink to greenish yellow to yellowish green at thepyramidal terminations. Courtesy of William Larson;photo by Maha Tannous.

Figure 24. Electron-microprobeanalyses of five multicoloredslabs and the four green-to-bluetourmalines (before heating) fromthe DRC showed that they werepredominantly elbaite and liddi-coatite. In addition, rossmanitewas found in two analyses (over-lapping) of a yellowish green area,and one each of a pink and agreen area of the slabs. Thesepoints fall very close to theboundaries with elbaite and lid-dicoatite and only barely crossinto the rossmanite field. Thecolor of each data point roughlyapproximates the color of thearea analyzed. There is no corre-lation between X-site occupancyand color.


Mr. Ulatowski reported that the overdark gem roughdoes not respond well to heat treatment. To demonstratethis, he heated portions of four green-to-blue crystals in airat temperatures ranging from 427 to 621°C for 5–9 hours.A comparison to the unheated portions of these crystalsrevealed no appreciable change in color, as well as dimin-ished transparency of all samples due to abundant micro-cracks (figure 23).

To investigate the composition of the DRC tourma-line, the five multicolored slabs and eight green-to-bluesamples (i.e., the four clipped crystals in figure 23, beforeand after heating) described above were analyzed by elec-tron microprobe at the University of New Orleans. Theslabs were analyzed in core-to-rim traverses of 9–13points each, and the green-to-blue samples had eithercore-to-rim or along-rim traverses of 5 points each. Figure24 shows that most compositions ranged from elbaite toliddicoatite. Of note, however, are four analyses (fromgreen, yellowish green (2), and pink portions of the slabs)that fell just slightly into the rossmanite field. All of thegreen-to-blue samples consisted of elbaite, whereas theslabs were elbaite ± liddicoatite (with the latter found intwo of the five slabs). As expected from our previous anal-yses of gem tourmaline (see, e.g., Winter 2002 Gem NewsInternational, pp. 356–357), there was no relationshipbetween the color and the tourmaline species.

However, significant variations were seen among thechromophoric elements Fe and Mn in the multicoloredslabs (see the G&G Data Depository at www.gia.edu/gemsandgemology). Traces of vanadium (up to 0.10 wt.%V2O3) also were commonly found in the slabs, but Ti wasrarely found (up to 0.06 wt.% TiO2) and no Cr or Bi wasdetected. The green-to-blue samples likewise had appre-ciable Fe and Mn, but they had slightly higher traces of Tiand no detectable V, Cr, or Bi.

Another example of attractive Congolese tourmaline,but with much different coloration, is shown in figure 25.These gem-quality prisms are mostly pink, grading intogreenish yellow and then yellowish green at their pyrami-dal terminations. According to William Larson (PalaInternational, Fallbrook, California), these crystals alsowere produced in recent years from the DRC. Clearly,they show the potential for the DRC to become a moreimportant source of gem tourmaline in the future.

BML

William “Skip” Simmons and Alexander FalsterUniversity of New Orleans, Louisiana

SYNTHETICS AND SIMULANTSDyed horn as an amber imitation. A necklace with 46rounded beads of a translucent yellow-brown material wassent to the SSEF Swiss Gemmological Institute for identi-fication (figure 26). The 50-cm-long necklace was pur-chased in Africa, where it was represented as amber.

However, despite its general resemblance to amber, someof its details were suspicious. A closer look at the surfacerevealed a peculiar structure and yellow color concentra-tions (figure 27). Due to the number of beads on the strandand their rough surfaces, it was not possible initially toobtain S.G. or R.I. values. An infrared spectrum takenfrom a small powder sample was consistent with horn,although our reference spectrum did not indicate the

Figure 27. A closer look at one of the beads in figure26 shows features inconsistent with amber, althoughthe rough surface made gemological testing difficult.Width of view is approximately 15 mm; photomicro-graph by H. A. Hänni, © SSEF.

Figure 26. This necklace of dyed horn beads (20–22mm in diameter) was sold as amber in Africa. Photoby H. A. Hänni, © SSEF.

http://lgdl.gia.edu/pdfs/congo_slices.pdf

http://lgdl.gia.edu/pdfs/congo_crystals.pdf

http://lgdl.gia.edu/pdfs/congo_crystals.pdf


species origin (e.g., cow, deer, etc.). The spectrum also pro-vided evidence of a yellow dye.

With the client’s permission, we removed one of thebeads and polished several areas on the surface to studythe growth structure. The S.G. of this bead was measuredhydrostatically as 1.27, which is consistent with hornbut not amber. R.I. measurements on the polished areaswere not possible because of the porosity of the material.A Raman spectrum of the original surface showed clearpeaks related to the yellow dye (at 1402, 1434, 1144,1188, 1609 cm−1, in order of decreasing intensity). Themost prominent FTIR peaks were at 1660 and 1540 cm−1.The polished areas showed that the structure of thematerial was characteristic of horn and the yellow pigmentation was restricted to a thin outer layer (figure 28).

Amber imitations such as this have been described inthe past. R. Webster reported that dyed bull horn beadshave been marketed as “red amber” in various parts ofAfrica (Gems, 5th ed., rev. by P. G. Read, 1994, p. 597).

HAH

Fake inclusions in quartz, “Made in Brazil.” Quartz iswell known for its attractive and varied inclusions, andspecimens with distinctive inclusions can fetch highprices from collectors. However, this demand alsoencourages the manufacture of fakes and imitations.Brazil is one of the largest producers of quartz speci-mens with unusual inclusions, and it is also where thiscontributor encountered some interesting new faked“inclusions” in July 2004.

Figure 28. Polished areas of this imitation amber bead show that the structure (both parallel and perpendicularto the “grain”) is consistent with horn. The yellow color is clearly restricted to a thin surface layer. In each,width of view is approximately 1 cm; photomicrographs by H. A. Hänni, © SSEF.

Figure 29. These quartzcabochons containedunusual “inclusions”

that proved to be manu-factured fakes. The

largest cabochon is 6 cmlong. Photo by J. Hyrsl.


The first examples were seen in Governador Vala-dares, where a local dealer had a parcel of 19 specimensthat he had bought for a high price in Teófilo Otoni. All ofthe specimens were cut as cabochons and contained natu-rally occurring inclusions of mica or chlorite near theirbase (figure 29). These cabochons also contained veryunusual coral-like features, which appeared to grow fromthe micaceous inclusions (figure 30). The branches wereup to ~30 mm long and 3 mm thick, occurring both singlyand in groups. Their color varied, with green, brown,pink, and yellow examples seen; a few even showed mul-tiple colors. In two specimens, it was apparent that smallportions of the “branches” were not completely filled bythe colored material.

Two days later, when this contributor visited TeófiloOtoni, the mystery surrounding these cabochons wassolved. Several colleagues had numerous examples of thismaterial. They were familiar with the person who was fab-ricating them, and reported that the “branches” weredrilled into the cabochons and then filled with a mineralpowder, probably mixed with glue, via a syringe. The filledholes were then covered with small pieces of feldspar andquartz mixed with glue, and sometimes “sealed” with alarger piece of feldspar (figure 31).

The glue was soft when poked with a sharp needle, andshowed weak yellowish green luminescence to UV radia-tion (stronger with long-wave UV). The unique shape ofthe inclusions, sometimes incomplete filling, and especial-ly the presence of glue on the base of the cabochons pro-vide evidence for the artificial origin of these inclusions.

Jaroslav Hyrsl ([email protected])Kolin, Czech Republic

TREATMENTSDyed cultured pearls fading on exposure to heat. Due tothe popularity (and expense) of natural-color dark culturedpearls, dyed-color cultured pearls (figure 32) are quitecommon in the marketplace. Although the risks of dam-

Figure 31. On the base of the cabochon in figure 30,the filled holes were seen to be covered with pieces offeldspar. The glue used to attach the feldspar couldbe indented by a needle; it fluoresced yellowish greenwhen exposed to UV radiation. Photo by J. Hyrsl.

Figure 32. After the pendant in this suite of culturedpearl jewelry was accidentally left on the dashboardof a car during several days of hot weather, it showednoticeable fading when compared to its original color(as represented by the earrings here). The fading ofthe treated color was likely due to dehydration of theconchiolin and possibly delamination of the nacre.Natural-color black pearls may also fade on exposureto prolonged heat. Photo by Maha Tannous.

Figure 30. The faked inclusions form interesting pat-terns, and appear to emanate from areas of naturallyoccurring mica or chlorite within the quartz. Photoby J. Hyrsl.


aging pearls through harsh cleaning methods are fairlywell known (see, e.g., D. D. Martin, “Gemstone durabili-ty: Design to display,” Summer 1987 Gems & Gemology,pp. 63–77), excessive heat can also cause fading.

The cultured pearl jewelry suite in figure 32 wasacquired by a GIA staff member during the 2003 Tucsongem shows (“Cultured pearls with diamond insets,”Spring 2003 Gem News International, p. 56). During arecent stretch of hot weather in southern California, thependant was accidentally left on the dashboard of a car forseveral days. When recovered, the cultured pearl’s colordisplayed noticeable fading.

The pendant was examined by one of these contribu-tors (SE) using EDXRF spectroscopy, which detected thepresence of silver in the cultured pearl. This indicates thatit had been treated with silver nitrate to create the darkcolor.

Although K. Nassau (Gemstone Enhancement, 2nd ed.,Butterworth-Heinemann, London, 1994, pp. 171–172) report-ed that the color induced through silver nitrate dyeing is“non-fading,” exposing any pearl—whether natural or treat-ed color—to prolonged heat (such as that generated inside aclosed car) is likely to be detrimental to the color. It wouldnot be surprising for the color to fade due to dehydration ofthe conchiolin and possibly delamination of the nacre.

This example serves to underscore the fact that allpearls must be treated with care.

Thomas W. Overton ([email protected])GIA, Carlsbad

Shane ElenGIA Research, Carlsbad

MISCELLANEOUSMasterpieces of American Jewelry Exhibition. “Jewelry,like all true art, can be a remarkable expression of anentire culture,” said Judith Price, president of the NationalJewelry Institute (NJI). Her words are beautifully reflectedin the NJI’s inaugural exhibition, “Masterpieces ofAmerican Jewelry,” which will be hosted by the AmericanFolk Art Museum in New York City from August 20,2004 to January 23, 2005. During a tour of the exhibition,this contributor was impressed by the artful displays andoverall presentation of the show.

This exhibit, the first museum show to focus exclu-sively on the jewelry history of the U.S., emphasizes thecreativity and artistry of American jewelry from the late1700s to the 1980s. More than 120 dazzling pieces are dis-played, most of which were selected by Ralph Esmerian,curator of this exhibit, vice-chairman of NJI, and chair-man of the Folk Art Museum. Included are unique cre-ations that were crafted by American designers, manufac-tured in American workshops, or retailed by Americanfirms such as Tiffany & Co. and Harry Winston. Also ondisplay are pieces from foreign jewelers such as Cartier,

Van Cleef, and Bulgari, that were designed, manufactured,and distributed in the United States.

The show focuses on five major themes commonthroughout America’s jewelry history: Americana, nature,humor, pastimes, and high style. The Americana sectionpays tribute to the American spirit, with pieces thatdepict victories in the War of 1812 or a single Americanflag. The Nature section showcases 19th-century nature-inspired pieces such as the exotic Tiffany orchid brooches.Disney charm bracelets designed by Cartier, a New YorkYankees watch by Hamilton, a sailing ship pin by Marcus& Co., and ballerina brooches by Van Cleef & Arpels (fig-ure 33) are featured in the Humor and Pastimes sections.The High-style exhibit includes photographs of womenwho distinguished themselves by their personal styles.

Figure 33. Among the highlights of the “Masterpiecesof American Jewelry” exhibition is this miniatureballerina brooch (7 ∞4 cm) designed in 1940 byMaurice Duvalet for Van Cleef & Arpels, New York.It is set with rubies, emeralds, and diamonds in plat-inum. Courtesy of Masterpieces of American Jewelry(by J. Price, Running Press, Philadelphia, 2004).


On display are Jacqueline Kennedy Onassis’ gold cuffs byVan Cleef & Arpels and Joan Crawford’s diamond braceletby Raymond C. Yard.

“Masterpieces of American Jewelry” is truly a master-piece worth seeing, and the companion book by Judith Pricewill enthrall readers. For more information, visitwww.folkartmuseum.org or e-mail [email protected].

Siau Fung Yeung ([email protected])GIA Gem Laboratory, New York

ANNOUNCEMENTSG&G beryllium diffusion article wins AGS Liddicoat jour-nalism award. “Beryllium diffusion of ruby and sapphire,”published by John Emmett and co-authors in the Summer2003 Gems & Gemology, has received the American GemSociety’s Richard T. Liddicoat Journalism Award in theJewelry Industry/Trade Reporting class. This award wasdeveloped in remembrance of GIA Chairman Richard T.Liddicoat to honor journalists who have made exceptionalcontributions to the understanding of gemology, as well asthe ideals of ethics, education, and consumer protection.Gems & Gemology previously won the inaugural 2003Liddicoat Journalism Award in the same category for“Photomicrography for Gemologists” by John I. Koivula inthe Spring 2003 issue (see Fall 2003 Gem News Inter-national, p. 248).

Visit Gems & Gemology in Tucson. Meet the editors andtake advantage of special offers on subscriptions and backissues at the G&G booth in the Galleria section (middlefloor) of the Tucson Convention Center during the AGTAshow, February 2–7, 2005.

GIA Education’s traveling Extension classes will offerhands-on training in Tucson with “Gem Identification”(January 31–February 4) and “Advanced Gemology”(February 5). To enroll, call 800-421-7250, ext. 4001.Outside the U.S. and Canada, call 760-603-4001.

The GIA Alumni Association will host a Dance Partyin Tucson on February 4, featuring a silent auction, anindustry awards presentation, and a live auction. To reservetickets, call 760-603-4204 or e-mail [email protected].

ExhibitsPearls at the Royal Ontario Museum. “Pearls: A NaturalHistory,” a traveling exhibition tracing the natural andcultural history of pearls organized by the AmericanMuseum of Natural History (New York) in collaborationwith the Field Museum (Chicago), will be on display at theRoyal Ontario Museum in Toronto from September 18,2004 to January 9, 2005. Among the many exhibits will bedisplays on pearl formation and culturing, as well as his-torical pearl jewelry that once belonged to Great Britain’sQueen Victoria and Marie Antoinette of France. Visitwww.rom.on.ca.

Carnegie Gem & Mineral Show. Held November 19–21,2004, at the Carnegie Museum of Natural History,Pittsburgh, Pennsylvania, this show will feature sapphiresin special exhibits and invited museum displays. Visitwww.carnegiemuseums.org/cmnh/minerals/gemshow.

Mineralien Hamburg. The International Show for Minerals,Fossils, Precious Stones, and Jewellery will take place inHamburg, Germany, on December 3–5, 2004. Special exhi-bitions will feature pearls and carved mineral and gem mate-rials from China. Visit www.hamburg-messe.de/mineralien.

ConferencesRapaport International Diamond Conference 2004. HeldOctober 12, this conference will take place in New Yorkand feature an insider’s look at the international dia-mond and jewelry industry. Visit www.diamonds.net/conference.

CGA Gemmology Conference 2004. The CanadianGemmological Association is holding its annual confer-ence at the Terminal City Club in Vancouver on October22–24. Contact Donna Hawrelko at 604-926-2599 or [email protected].

Pegmatites at GSA. A topical session titled “GraniticPegmatites: Recent Advances in Mineralogy, Petrology,and Understanding” will be held at the annual meeting ofthe Geological Society of America in Denver, Colorado,November 7–10, 2004. The meeting will also feature a ses-sion covering advanced mineral characterization tech-niques. Visit www.geosociety.org/meetings/2004.

Antwerp Diamond Conference. The 3rd AntwerpDiamond Conference, presented by the Antwerp DiamondHigh Council (HRD), will take place in Antwerp onNovember 15–16, 2004. The conference will focus on syn-thetic diamonds as well as on strategies to promote con-sumer confidence in natural diamond. Visit www.hrd.be.

Diamond Synthesis and History. To commemorate the50th anniversary of the successful repeatable synthesis ofdiamond, the H. Tracy Hall Foundation is organizing aone-day symposium on “Diamond Synthesis and History”on December 16, 2004, in Provo, Utah. The conferencewill focus on high-pressure research and equipment devel-opment. Visit www.htracyhall.org/Symposium.htm.

GemmoBasel 2005. The first open gemological conferencein Switzerland will be presented by the SSEF SwissGemmological Institute at the University of Basel April29–May 2. Among the events scheduled is a field trip to aSwiss manufacturer of synthetic corundum and cubic zir-conia. Visit http://www.gemmobasel2005.org or [email protected].

270 BOOK REVIEWS GEMS & GEMOLOGY FALL 2004

Blood from Stones: The SecretFinancial Network of TerrorBy Douglas Farah, 225 pp., publ. byBroadway Books, New York, 2004.US$24.95

In 1999, the U.S. government attempt-ed to cut the purse strings of theTaliban and al Qaeda by freezing some$240 million in assets. Forced to lookoutside the conventional banking sys-tem, al Qaeda leaders turned tountraceable commodities they couldeasily transport and exchange to fundterrorist operations. Were diamondspart of their financial network?

In Blood from Stones, investigativereporter Douglas Farah claims that inthe late 1990s, al Qaeda began buyingup millions of dollars of better-qualityrough diamonds in Sierra Leone. Inexchange for these “blood diamonds,”al Qaeda operatives allegedly paid cashto warlords from the RevolutionaryUnited Front (RUF), a group then inthe midst of waging a decade-long civilwar in Sierra Leone. Known for itsmany atrocities, the RUF operatedwith the backing of former Liberianleader and U.N.-indicted war criminalCharles Taylor. Farah maintains thatthe diamonds were then smuggled toAntwerp, where they entered the sup-ply chain. The arrangement armed theRUF and added to Taylor’s personalfortune, and in return al Qaeda wasgiven a secret source of funding.According to Farah, the Lebanon-basedHezbollah organization had made sim-ilar inroads into Sierra Leone’s dia-mond trade in the early 1980s.

Farah, the former West Africanbureau chief for the Washington Post,says he “stumbled on” the diamondtrail shortly after the September 11attacks. His investigation began overlunch when one of his sources, a

Taylor associate named CindorReeves, casually asked, “What is Hez-bollah?” Reeves explained that a groupof Arab diamond clients had invitedhim to watch videos of Hezbollah sui-cide bombings. When Farah latershowed him photos of known alQaeda operatives published in News-week, Reeves immediately recognizedthem as men he had taken to the dia-mond fields of Sierra Leone. Farah helda secret meeting with a group of RUFcommanders, who confirmed the iden-tification. In November 2001, he brokethe story on the front page of the Post.

In addition to the decades of cor-rupt rule in Sierra Leone and neighbor-ing Liberia, the author lays blame onthe U.S. Central Intelligence Agencyfor overlooking compelling evidence ofa diamonds-terrorism link. Farah alsocriticizes the diamond industry for fail-ing to take “serious measures” toeliminate conflict diamonds, suggest-ing that it was more interested in lob-bying to water down the CleanDiamond Trade Act passed in 2001.He charges the World Diamond Coun-cil (WDC) with doing “little of sub-stance” and downplays the effective-ness of its Kimberley Process Certifi-cation Scheme in policing the dia-mond supply chain.

Diamonds do not figure in the lat-ter half of the book, which is a primeron al Qaeda’s funding tactics. Theauthor shows how al Qaeda has usedhawala, an informal exchange systembased on trust between brokers, tolaunder funds raised through Islamiccharities and a host of petty scams.The book concludes with an unset-tling picture of the barriers to combat-ing terrorism, most notably shrinkingcounterterrorism budgets and bureau-cratic infighting among U.S. govern-ment agencies.

The author’s fast-paced, straight-forward reporting is at its strongest inportraying the West African night-mare created by a network of corruptpolitical leaders, arms smugglers, anddrugged-out child soldiers. While thesection on al Qaeda fundraising opera-tions is thorough and well document-ed, the alleged diamonds-terrorismconnection that inspired the book’stitle is less so. Farah’s story relies tooheavily on the accounts of charactersassociated with Taylor and the RUF,and he abruptly drops the argumenthalfway through the book.

Indeed, the question of a possiblelink between the diamond trade and alQaeda is still in dispute. Since the pub-lication of Blood from Stones, the finalreport from the National Commissionon Terrorist Attacks, known as the9/11 Commission, cleared the dia-mond trade of links to al Qaeda. Thisfinding was welcomed by the WDCbut disputed by United Nations war-crimes prosecutors and the humanrights group Global Witness. Yet withthe end of Sierra Leone’s civil war inJanuary 2002, Charles Taylor’s exileunder international pressure in August2003, and the implementation of theKimberley Process the followingmonth, few could deny that prospectsfor peace and stability in the WestAfrican diamond trade are brightertoday.

STUART OVERLINGemological Institute of America

Carlsbad, California

2004Book REVIEWS

EDITORSSusan B. JohnsonJana E. Miyahira-SmithStuart Overlin

*This book is available for purchasethrough the GIA Bookstore, 5345Armada Drive, Carlsbad, CA 92008.Telephone: (800) 421-7250, ext.4200; outside the U.S. (760) 603-4200. Fax: (760) 603-4266.

BOOK REVIEWS GEMS & GEMOLOGY FALL 2004 271

Starting to Collect Antique JewelleryBy John Benjamin, 191 pp., illus.,publ. by Antique Collectors’ Club,Woodbridge, England, 2003.US$25.00*

The challenges of collecting antiquejewelry can be overwhelming. To makewise purchases, the collector must bearmed with a good working knowledgeof gems, metals, manufacturers, manu-facturing styles, jewelry periods, andmuch more. With this guide, Mr.Benjamin succeeds in giving the readerspecific information on all of these top-ics, as well as a broad overview ofantique jewelry categories that will behelpful to the novice collector.

The book is comprised of 19 chap-ters, including “Gemmology andGems in Antique Jewellery,” “TheEvolution of Jewellery from EarlyTimes to the 18th Century,” “Cameosand Intaglios,” “Reviving History inthe 19th Century,” and “Fabergé, Tif-fany, and Cartier.” References for fur-ther reading are provided at the end ofeach chapter so readers may expand ontheir studies of selected topics. Thebook is generously illustrated withphotographs of antique jewelry thatrange from the very simple to the mostsumptuous. Illustrations from old jew-elry catalogs are also included.

Especially appealing to this review-er were the chapters on “Enamels” and“Jewels of Sentiment and Love,” asthey are areas of personal interest, butthe entire book is filled with valuableinformation, such as why ironworkjewelry became popular, what featuresto look for in a cameo, and how plat-inum was integral to the developmentof the “garland” style of jewelry. Otheruseful features are short biographies ofnotable jewelers and a compendiumthat briefly discusses valuations, fakes,restorations, and cleaning.

Many enthusiasts find that theyare drawn to specific areas of antiquejewelry and focus their collectionsaccordingly. Starting to Collect An-tique Jewellery can help beginningcollectors decide which areas interestthem most, and it should be required

reading for anyone thinking of startingan antique jewelry collection.

JANA E. MIYAHIRA-SMITHGemological Institute of America


Starting to Collect Antique Silver By Ian Pickford, 192 pp., illus., publ.by Antique Collectors’ Club, Wood-bridge, England, 2003. US$25.00

This visually appealing book, writtenby antique silver expert Ian Pickford,covers a wide range of areas related tocollecting and caring for antique silver.

The author first discusses earlysilver sources, and how silver itemsand coins from other regions began toaccumulate in Britain as internationaltrade grew. Since silver was oftenmelted down during times of eco-nomic hardship, problems associatedwith variations in metal contentarose. The first solution was imple-mented in 1238, when sterling silver(.925) was established as the standardand wardens were appointed to test,assess, or assay silver. Pieces not up tothe standard were destroyed. Later,the hallmarking system enacted byKing Edward I in 1300 set a standardfor coinage and goldsmithing.

Mr. Pickford details the evolutionof this marking system, which essen-tially was the first form of consumerprotection and now allows us to dateand identify pieces of British silverwith accuracy. The reader is neverbored; scattered throughout are manyinteresting bits of related trivia, such asthe origin of the expression “born witha silver spoon,” silver’s anti-bacterialproperties, and why the price of silverdropped significantly with technologi-cal changes in the photographic indus-try. Entire chapters cover manufactur-ing techniques, highlight prominentmaker marks, and take in-depth looksat various utility items and vessels.Items covered include spoons, candle-sticks, tea and coffee items, drinkingvessels, boxes, and small collectibles.The numerous high-quality photos andillustrations clearly demonstrate whata collector should look for on a particu-

lar piece. The author strives to clear upany misunderstandings regarding plat-ing, and covers areas of concern such asfakes, forgeries, and alterations.

Although this book is gearedtoward collecting British silver andonly offers a few American examples, itis an essential and informative book forany novice or serious silver collector.

MARY MATHEWSGemological Institute of America


Within the StoneBy Bill Atkinson, 180 pp., illus., publ.by Browntrout Publishers, SanFrancisco, CA, 2004. US$39.95*

The first impression of Within theStone is one of beauty—it hits youimmediately with the cover photos. Aclose-up photograph of a fine dendriticagate from Madagascar adorns thefront and draws your interest. Theback cover features an intriguingphoto of pietersite, a brecciated rockcomposed primarily of fragmentedhawk’s-eye and tiger’s-eye quartz.These images draw you in, and thatinitial impression of beauty stays withyou throughout the entire book.

Although this is not a scientificwork or gemological text as such, thepleasure it conveys to the reader satis-fies the curiosity we all have fornature and makes us want to explorethis natural world beyond the limitsof the 72 magnificent illustrationsprovided. The original poems andessays by noted artists and scientiststhat accompany each of the full-pagephotos add to the unique experienceof this book as well.

Twenty-six different rocks andminerals are illustrated. The mostwidely pictured subjects are jasper,agate, petrified wood, and pietersite, sothe breadth of material covered is notextensive. These are photomacro-graphs taken with a camera and bel-lows mounted on a copy stand, asopposed to photomicrographs takenthrough a microscope. The magnifica-tion range is 1¥ to 9¥, and the polishedrock and mineral subjects vary from 1

272 BOOK REVIEWS GEMS & GEMOLOGY FALL 2004

to 10 inches (2.5–25 cm) in width. Forthis reviewer, this approach is part ofwhat is so special about the book. As aphotomicrographer, I can also see ineach image several different areasthat would have been captured won-derfully with the higher magnifyingpower of a microscope. From eitherperspective, the selection of subjectsis excellent.

Gemologists and earth scientistswill find additional enjoyment in the“Rock Descriptions” chapter, writtenby well-known mineral authorities Siand Ann Frazier and Robert Hutchin-son, which describes the mineralogicaland gemological aspects of each of theillustrated subjects. The book closeswith brief explanations on how thephotographs were taken, and how thebook itself was made. Nature is neverperfect, but near perfection wasachieved in these photos using carefulgray-scale balancing with a digital graycard for each subject. The computer-scanned transparencies and the directdigital captures were also “cleaned upand refined in Adobe Photoshop.”

Besides being a pleasure to read andstudy, this book also serves as an artcatalogue, since all of the photos areavailable as large matted and signedfine-art prints. With its large (12 ¥ 11inch) format, Within the Stone is visu-ally impressive. Mr. Atkinson refers tothe book as “a beautiful work of art,”and I agree completely. It will beenjoyed by anyone with a fondness forour natural world.

JOHN I. KOIVULAGemological Institute of America


Gems and Jewels: A Connoisseur’s GuideBy Benjamin Zucker, 248 pp., illus.,publ. by Overlook Press, New York,NY, 2003. US$60.00

Since this book’s original publicationin 1984, the field of gemology hasexploded with ever-increasing scien-tific sophistication, not to mention asuccession of dizzying discoveries

from Canada’s subarctic to the tropi-cal depths of the South Pacific.Readers seeking relief from photolu-minescence spectra and the geopoliti-cal dramas facing today’s gem mar-kets will find welcome respite inBenjamin Zucker’s reissued Guide.

Each of the chapters is dedicated toa single gem, beginning with the “BigFour” followed by pearl, amber, lapislazuli, jade, turquoise, opal, and garnet.To address the subtitle’s promise,Zucker introduces readers to eachgem’s distinguishing features. Thoughemphasis is placed on top-grade colorand geographic provenance, the authoralso discusses chemical variances andmicroscopic characteristics.

More than 200 color photographsillustrate the qualitative aspects of finegemstones as well as their recoveryfrom earth and sea. Though inconsis-tent in composition and photographicquality, the illustrations include manyhistorical and rarely seen pieces as wellas notable gemstone objects and jewel-ry dating back to the Bronze Age. Theexceptionally well-chosen color photosaptly serve this labor-of-love’s empha-sis on historical references and geo-graphic origins.

Gem lovers with a pulse willquickly find their interest piqued bythe book’s description of extraordi-nary pieces residing in far-flung muse-ums. As is the case throughout mostof the book, collections from Europe,the Near East, and elsewhere in Asiaare featured more prominently thanNorth American treasures.

Changes to this new editioninclude a bounty of new photos andsubstantial text revisions, all supportedby an updated introduction and index.While some readers may note occa-sional outdated terminology and omis-sions (such as a very thin discussion ofChinese freshwater cultured pearls inan otherwise thorough chapter on thatorganic material), the book is ideal forthose seeking a descriptive and lessquantitative approach to gem apprecia-tion from a well-traveled and respectedconnoisseur.

MATILDE PARENTEIndian Wells, California

OTHER BOOKS RECEIVEDThe Tourmaline. By A. C. Hamlin,107 pp., illus., originally publ. byJames R. Osgood & Co., Boston,MA, 1873; republished by RubellitePress, New Orleans, LA, 2004,US$75.00. The History of MountMica. By A. C. Hamlin, 123 pp.,illus., originally publ. by AugustusChoate Hamlin, M.D., 1895; repub-lished by Rubellite Press, NewOrleans, LA, 2004, US$85.00.

Until now, these two classic works ontourmaline were available onlythrough rare-book dealers. Fortunatelyfor tourmaline aficionados, they haverecently been republished as faithfulreproductions—even including theoriginal punctuation, spelling (witherrors), formatting, and “feel.” Thehardbound covers also mimic the orig-inals, as custom dies were created forthe stamping and embossing process.Color plates in both books weremeticulously matched to those foundin original copies and are printed onhigh-quality paper.

The Tourmaline, which containsfour color plates, chronicles the earlyhistory, world localities known at thetime, and physical properties (particu-larly color) of this mineral. Approxi-mately one-third of the book is devot-ed to the characteristics of tourmalinefrom Mt. Mica (Maine), including thecrystal forms, colors, and specific local-ities at this historic deposit. In closing,the author offers some creative ideason the origin of gems in general.

The History of Mount Mica con-tains eight drawings, five black-and-white photos (including a panoramicpull-out image), and color platesshowing drawings of 43 tourmalinecrystals from this locality. Theseplates, which occupy the latter half ofthe book, document the impressivecolor variations in tourmaline fromMt. Mica. The text provides a detailedaccount of the mining and productionat this classic locality, from its dis-covery in 1820 until the mid-1890s.

BRENDAN M. LAURSGemological Institute of America


GEMOLOGICAL ABSTRACTS GEMS & GEMOLOGY FALL 2004 273

COLORED STONES AND ORGANIC MATERIALSThe changing opal market. P. B. Downing, Lapidary Journal,

Vol. 57, No. 7, 2003, pp. 62–66.The opal market has seen many changes in the last 10 years.There has been a major decline in the production of naturalopal from Australia, and the new non-Australian (e.g., Ethiopia,British Columbia) sources have not provided adequate replace-ment. At the same time, worldwide demand, especially fromJapan, has weakened while prices have increased. The responseto the reduced supply and high prices has been a reshaping ofthe market for opal. As a result, the market is now seeing moreopal doublets, inlay, and intarsia in jewelry, as well as moreopal synthetics and simulants.

Most common today are inexpensive “boulder opal” dou-blets (thin pieces of opal—usually darkened to simulate blackopal—cemented to a brown ironstone base); cutters can easilyproduce four doublets from the same opal that would normal-ly yield only one solid cabochon. However, boulder doubletsare fragile and must be bezel set to reduce the risk of chipping.Another type of doublet, assembled primarily in LightningRidge, Australia, consists of thicker slices of opal with a com-mon opal (potch) backing. These are cut in a dome shape withthicker edges that are less prone to chipping. Inlay jewelrybecame more popular in the 1990s. Inlay, like intarsia, con-serves opal, as the rough used is only about 1–2 mm thick.Inlay and intarsia jewelry are currently being mass-producedin Asia.

Synthetic opal is gaining acceptance. The new GilsonCreated Opal is difficult to distinguish from natural opal and isnow used in quality jewelry. Opal simulants in a wide varietyof colors are also becoming increasingly available. MT

Gemological ABSTRACTS

2004EDITOR

A. A. LevinsonUniversity of Calgary

Calgary, Alberta, Canada

REVIEW BOARDJo Ellen Cole

Vista, California

Michelle Walden FinkGIA Gem Laboratory, Carlsbad

R. A. HowieRoyal Holloway, University of London

Alethea InnsGIA Gem Laboratory, Carlsbad

David M. KondoGIA Gem Laboratory, New York

Taijin LuGIA Research, Carlsbad

Wendi M. MayersonGIA Gem Laboratory, New York

Keith A. MychalukCalgary, Alberta, Canada

Joshua ShebyGIA Gem Laboratory, New York

James E. ShigleyGIA Research, Carlsbad

Boris M. Shmakin Russian Academy of Sciences, Irkutsk, Russia

Russell ShorGIA, Carlsbad

Maha TannousGIA Gem Laboratory, Carlsbad

Rolf Tatje Duisburg University, Germany

Christina TaylorBoulder, Colorado

Sharon WakefieldNorthwest Gem Lab, Boise, Idaho

This section is designed to provide as complete a record as prac-tical of the recent literature on gems and gemology. Articles areselected for abstracting solely at the discretion of the section edi-tor and his reviewers, and space limitations may require that weinclude only those articles that we feel will be of greatest interest to our readership.

Requests for reprints of articles abstracted must be addressed tothe author or publisher of the original material.

The reviewer of each article is identified by his or her initials at theend of each abstract. Guest reviewers are identified by their fullnames. Opinions expressed in an abstract belong to the abstrac-ter and in no way reflect the position of Gems & Gemology or GIA.

© 2004 Gemological Institute of America

274 GEMOLOGICAL ABSTRACTS GEMS & GEMOLOGY FALL 2004

Crystallographic position of Mn3+ in violet tourmaline. R.V. Shabalin and B. B. Shkurski, GemologicalBulletin, No. 9, 2003, p. 26–30 [in Russian withEnglish abstract].

A tourmaline from Madagascar with an unusual violetcolor (type vB 8/2 in the GIA color notation) and character-ized by a predominant absorption peak at 560 nm wasinvestigated using X-ray diffraction analysis, electronmicroprobe analysis, and energy calculations with aTanabe-Sugano diagram. This color was unique in a collec-tion of more than 40 colored tourmalines available to theauthors.

This tourmaline has a surplus of Al (7.73 formulaunits), which fully occupies the Z crystallographic posi-tions. Thus, Mn3+ ions, which prefer the Z positions andcause an absorption at 515 nm, were required to occupy Ycrystallographic positions, which resulted in the 560 nmabsorption peak. Fe3+, Li, and Al were also found in the Ypositions, whereas Na and Ca occurred in the X positions.Small absorption peaks at 450 and 415 nm were caused byelectron transitions due to Fe3+ and Mn2+, respectively.Ti, Mg, and K were not detected by electron microprobeanalysis. BMS

An interesting Australian abalone pearl. S. M. B. Kellyand G. Brown, Australian Gemmologist, Vol. 21,No. 12, 2003, pp. 498–501.

A detailed, illustrated description is given of a large cres-cent-shaped abalone pearl (95.79 ct) recovered from a blacklip abalone (Haliotis rubra) from waters off Clay Head,northern New South Wales. The fleshy body of theabalone was so distorted by the size and shape of the pearlthat the shell had to be broken for the pearl to be recov-ered. The multicolored iridescent nacre on the pearl wasunevenly distributed, so it could only be considered to beof low-medium quality. RAH

Relationship between the groove density of the gratingstructure and the strength of iridescence in molluscshells. Y. Liu, K. N. Hurwit, and L. Tian, AustralianGemmologist, Vol. 21, No. 10, 2003, pp. 405–407.

The iridescence of mollusk shells and pearls is caused bydiffraction. A previous study showed that the strength ofiridescence of a Pinctada margaritifera shell is qualitative-ly related to the groove density of the diffraction-gratingstructure formed by its aragonite tiles. In this study, thegroove density of a shell of the pearl-producing Pinctadamaxima was examined and found to be intermediatebetween the groove densities of the outer and inner sur-faces of P. margaritifera; the strength of its iridescence wasalso intermediate. In general, a groove density in the rangeof 80–300 grooves per millimeter can produce iridescencein the visible-wavelength range; a density of 300 groovesper millimeter yields the strongest iridescence, whereas adensity of 80 grooves per millimeter rarely produces irides-cence. The authors conclude that the strength of the iri-

descence of a mollusk shell is directly related to the groovedensity of the diffraction-grating structure, and not to thespecies of mollusk. RAH

Rock-forming moissanite (natural aa-silicon carbide). S. DiPierro, E. Gnos, B. H. Grobety, T. Armbruster, S. M.Bernasconi, and P. Ulmer, American Mineralogist,Vol. 88, Nos. 11–12, 2003, pp. 1817–1821.

The possibility of silicon carbide occurring as a natural ter-restrial mineral has been the subject of scientific contro-versy for nearly a century. This article reports the discov-ery of a unique volcanic rock that contains moissanite (6Hpolytype) as a significant mineral component (8.4% by vol-ume). The rock was found accidentally as a beach pebblein an unpopulated region along the coast of Turkey about150 km northwest of Izmir. The source outcrop of thispebble has not yet been located. The grayish blue rockexhibits a homogeneous, porphyritic texture consisting offine-grained brucite, calcite, and magnesite along withlarger macrocrysts of quartz and moissanite. The bulkwhole-rock chemistry is somewhat similar to that of akimberlite.

The moissanite crystals are hexagonal, 0.2–1.5 mm insize, and blue or black in color, with a metallic luster.Some crystals are transparent and display a subadaman-tine luster. They have a platy, tabular shape dominated bypinacoid (001) crystal faces. About one-third of the crys-tals examined contain one (or more on rare occasions)rounded black metallic inclusions of several phases,including Si and, in lesser amounts, various Fe-silicides.When viewed with transmitted light in thin section, themoissanite grains range from colorless to light or darkblue to almost black. Some crystals are pleochroic, andsome are likely to be twinned. Optically they are uniaxialpositive. A few grains exhibited yellow, blue, and redcathodoluminescence. The crystals are very homoge-neous chemically, with no elements (besides Si) detectedabove background levels by electron microprobe analysis.

The authors present various kinds of evidence to sug-gest this rock is a naturally occurring specimen, and notthe product of any industrial or scientific process. Theyconclude that the rock most likely formed at ultra-high-pressure conditions in the upper mantle or transitionzone, and was brought to the Earth’s surface during a vol-canic eruption. JES

DIAMONDSDiamonds: Time capsules from the Siberian mantle. L. A.

Taylor and M. Anand, Chemie der Erde: Geo-chemistry, Vol. 64, No. 1, 2004, pp. 1–74.

In this invited review, the authors report their systematicstudies on diamondiferous eclogite xenoliths from Siberia.The steps in investigating these are: (1) high-resolutioncomputed X-ray tomography of the xenoliths to give 3-D

images that relate the minerals of the xenolith to their dia-monds; (2) detailed dissection of the entire xenolith toreveal the diamonds inside, followed by characterization ofthe setting of the diamonds within their enclosing materi-als; and (3) extraction of diamonds from the xenolith tofacilitate further investigation of the diamonds and theirinclusions. In this last step, it is important to record care-fully the nature and relative positions of the inclusions inthe diamonds to maximize the number of inclusions thatcan be exposed simultaneously on one polished surface.Once the diamonds have been extracted, cathodolumines-cence imaging is performed on polished surfaces to revealthe internal growth zones of the diamonds and the spatialrelationship of the mineral inclusions to these zones. Inaddition, infrared analyses of nitrogen aggregation and car-bon and nitrogen isotopic analyses are performed on thediamonds. Such multiple lines of evidence indicate theultimate crustal origin for the majority of mantle eclogites.Similar pieces of evidence, particularly from d13C in P-typediamonds and d18O in peridotitic garnets, suggest that atleast some of the mantle peridotites, including diamondif-erous ones, as well as inclusions in P-type diamonds, mayhave a crustal protolith as well. RAH

DTC sightholders. P. Insch, Rough Diamond Review, No.3, December 2003, pp. 13–15.

As a result of the economic hardships brought on by theGreat Depression, consumer demand for diamondsdeclined drastically in the 1930s. In 1934, De Beers createdthe Central Selling Organisation (CSO, now known as theDiamond Trading Co. [DTC]) to stabilize the market byeffectively directing the flow of rough diamonds to themarketplace through a single-channel system. Central tothis system were “sightholders” (select dealers and manu-facturers in international cutting centers), who were allo-cated rough diamonds every 10 weeks at a “sight.” Thissystem worked well for several decades because De Beerssupplied, via the DTC and its predecessor, 80% of all gem-quality rough diamonds sold worldwide.

Recently, De Beers, which currently supplies 50–60%of global production, revamped their single-channel mar-keting system into an updated Supplier of Choice initia-tive designed to increase consumer demand for diamondjewelry and facilitate effective competition of diamondswith other luxury goods. Currently, there are about 80sightholders in this new rough diamond marketing for-mat as opposed to 125–300 in previous decades. Severalcriteria are considered in the selection process for newsightholders: financial standing; market position; distribu-tion, marketing, technical, and manufacturing abilities;and adherence to the Best Practice Principles of the DTC.Sightholders keep contracts with the DTC for two years,after which they must reapply to maintain their sight-holder status.

Three factors determine the type and volume of goodsoffered the sightholder: (1) the sightholder’s effectiveness

in marketing diamonds and diamond jewelry; (2) the sizeand quality of diamonds requested by the sightholder forthe next six-month period; and (3) the predicted availabili-ty of diamonds in the next period. Once all of these fac-tors are assessed, De Beers gives each client an “intentionto offer” list of the goods that client can expect. There areminimum prices for each box, and the sightholder (or arepresentative) may reject all or part of them. JS

Seeking the origin of carbon in diamond. P. Cartigny,Rough Diamond Review, No. 3, December 2003,pp. 39–42.

The study of carbon isotope variations in natural dia-monds is important to our understanding of how dia-monds formed. The isotopic ratio of 13C to 12C, expressedas d13C and reported as per mil (‰), is a deviation from aninternational standard for which d13C=0. The d13C valuesfor diamonds range from –38.5 to +2.7‰; however, twodistinct d13C distributions occur. One spans the broadrange of –38.5 to +2.7‰, whereas the second is restrictedto a narrow band of –8 to –2‰. These distinct d13C distri-butions, incorporated in systematic studies of diamond’sphysical properties and inclusions, are associated with twotypes of diamonds, eclogitic and peridotitic, respectively.Each type originates from different sources in the earth.

Eclogitic diamonds form from sedimentary carbonrecycled from the surface of the earth and subducted intothe mantle. Peridotitic diamonds form in the mantle fromcarbon originally in that location. Notwithstanding theabove explanations with respect to the two sources of car-bon in diamond, some uncertainties and ambiguities exist(e.g., data from nitrogen isotopes) that have encouragedsome researchers to favor a single source of carbon forboth types of diamond. DMK

Spectroscopic and morphological characteristics of dia-monds from the Grib kimberlite pipe. R. M.Mineeva, A. V. Speranskii, S. V. Titkov, O. M.Zhilicheva, L. V. Bershov, O. A. Bogatikov, and G.P. Kudryavtseva, Doklady of the Russian Academyof Sciences, Earth Science Section, Vol. 394, No. 1,2004, pp. 96–99.

Diamonds from the new (discovered in 1980) diamondifer-ous province in the Arkhangel’sk district of northwesternRussia contain an extremely high content of roundedrhombododecahedral and tetrahexahedral, as well as cubic,crystals compared to diamonds from Yakutia (Siberia).Generally, Arkhangel’sk crystals are nearly colorless withpale yellow and brown tints. Details of the crystal mor-phology of 15 crystals are tabulated together with data ontheir paramagnetic centers. Compared with diamondsfrom other pipes in the Arkhangel’sk district, those fromthe Grib kimberlite have a lower content of Group 1 dia-monds with a predominance of P1 centers. The majorityof the Grib diamonds are classified as Group 2 with a pre-dominance of P2 centers. The Grib kimberlite also con-



tains more N2-bearing Groups 3 and 4 diamonds com-pared to other kimberlites in the province. Since N2 cen-ters in diamond crystals form during plastic deformation,the authors conclude that this epigenetic process (i.e., plas-tic deformation) had a greater effect on diamonds from theGrib pipe than it did on those from other pipes in theArkhangel’sk district. RAH

GEM LOCALITIESDer Pegmatit von Anjahamiary bei Fort Dauphin,

Madagaskar [The Anjahamiary pegmatite near FortDauphin, Madagascar]. F. Pezzotta and M. Jobin,Lapis, Vol. 29, No. 2, 2004, pp. 24–28 [in German].

This zoned granitic pegmatite is situated about 80 kmnorthwest of Fort Dauphin (Taolagnaro) in southeasternMadagascar, far from the well-known and highly produc-tive gem pegmatite province in the central part of thecountry. It was first worked in the 1930s and has beenexploited intermittently since then by different owners.After the discovery of a large tourmaline pocket in 1991,Somema, a Madagascan company, began exploration inthe area as well as gem production from the tourmaline-bearing zones of the pegmatite. The tourmalines are main-ly deep to pale pink and grayish blue to intense blue. Apreliminary chemical analysis indicated that the blue partsare liddicoatite and the pink parts are rossmanite.Although the colors of some of the blue crystals resemblethose of Paraíba tourmalines, the analysis did not detectany copper. RT

Genesis of amethyst geodes in basaltic rocks of the SerraGeral Formation (Ametista do Sul, Rio Grande doSul, Brazil): A fluid inclusion, REE, oxygen, carbon,and Sr isotope study of basalt, quartz, and calcite.H. A. Gilg, G. Morteani, Y. Kostitsyn, C. Preinfalk,I. Gatter, and A. J. Strieder, Mineralium Deposita,Vol. 38, No. 8, 2003, pp. 1009–1025.

Amethyst geodes occur abundantly in a basalt lava flow(40–50 m thick) of Lower Cretaceous age in the Ametistado Sul region of the state of Rio Grande do Sul in southernBrazil. This area is famous for the production of amethystboth as geodes and cutting material. The geodes, typicallyspherical cap-shaped and sometimes vertically elongate upto 6 m in largest dimension, display an outer rim ofceladonite followed inwards by agate, colorless quartz, andfinally amethyst crystals. Occurrences of calcite and gyp-sum are also common within the geodes.

In this study, chemical and isotopic composition aswell as fluid inclusion data were analyzed to better under-stand the conditions of geode formation. The authors sug-gest that their genesis most likely occurred as the resultof a two-stage process. During the initial magmatic stage,numerous cavities are thought to have formed within thecooling lava from an immiscible fluid phase of lower den-

sity and viscosity than the surrounding basaltic magma.In a second, post-magmatic stage, the cavities were thenfilled with amethyst and other minerals (thus becominggeodes) at temperatures of less than 100°C, as determinedby fluid inclusion and stable isotope data. These mineralscrystallized from a circulating gas-poor aqueous fluid ofmeteoric origin that leached the necessary constituentsfrom highly reactive interstitial glass in the host basalt.This infilling process is thought to have continued for anextended period of time (perhaps 40–60 million years)after the eruption of the basaltic lava. JES

Jurassic to Miocene magmatism and metamorphism inthe Mogok metamorphic belt and the India-Eurasiacollision in Myanmar. M. E. Barley, A. L. Pickard,K. Zaw, P. Rak, and M. G. Doyle, Tectonics, Vol.22, No. 3, 2003, pp. 4-1–4-11.

For centuries, the Mogok Stone Tract in north-centralMyanmar has been famous as a source of ruby, sapphire,peridot, spinel, and a wide variety of other gems. Thisabundance of gem minerals originates mainly from theMogok metamorphic belt (MMB), a 50-km-wide zone ofmarbles, schists, and gneisses that are intruded by granitesand pegmatites. Northward, this belt can be traced to theeastern edge of the Himalayas, whereas to the south itextends into Thailand. Because of its location, the MMBoccupied a key position in the tectonic evolution of theregion that witnessed the convergence and collision offragments of the Paleozoic Gondwana and Eurasia conti-nents, culminating in the collision of India and the upliftof Tibet and the Himalayas.

This article presents results from high-resolution ion-microprobe analyses of grains of zircon that had been col-lected in granitic rocks along the MMB. The resulting U-Pbages suggest a complex magmatic and metamorphic histo-ry for the MMB, ranging from the Jurassic to the Miocene,that both pre-dates and post-dates the collision of Indiawith the southern margin of Eurasia. Earliest zircon ages ofabout 170 million years (My) were obtained from stronglydeformed granitic orthogneisses. Overgrowths on zirconsgive an age of about 43 My that is thought to representmetamorphic recrystallization during a period of thicken-ing of the Earth’s crust and associated metamorphism fol-lowing the initial collision of India and Eurasia (65 to 55My). The authors conclude that the MMB played a key rolein the network of deformation zones that accommodatedstrain during the northward movement of India and theresulting extrusion and rotation of Indochina. JES

Lambina opalfield: An update. J. Townsend, AustralianGemmologist, Vol. 21, No. 12, 2003, pp. 490–494.

The Lambina opal field, 10 km south of the LambinaHomestead, in the remote northern part of South Australia,has been a significant producer of precious opal only in thelast decade (e.g., the value of the production has rangedbetween $A7.3 and 10.2 million annually since 1999). At

Lambina, opalized sandstone is accompanied by opal intro-duced into cracks, in nodules, and replacing marine snails,belemnites, and bivalves. Good-quality opal from thislocality includes white (light) opal and crystal opal that dis-plays good play-of-color; about 50% of the opal does notfluoresce under long-wave ultraviolet light. Emphasis isplaced on the possible influence of paleochannels as con-duits for water movement (and hence silica movement) andthe deposition of opal adjacent to these ancient channels.

RAH

Les gisements de corindon: Classification et genèse[Corundum deposits: classification and origin] andLes placers à corindon gemme [Gem corundum plac-er deposits]. V. Garnier, G. Giuliani, D. Ohnen-stetter, and D. Schwarz, Le Règne Minéral, No. 55,2004, pp. 7–35 and 36–47 [in French with Englishabstract].

Several classifications of corundum have been proposedbased on crystal habits, geologic settings and origin of thedeposits, lithology of the host rocks, and chemical compo-sition (only rubies). These authors propose an improvedclassification based on (a) the lithology of the host rocksand (b) deposit types. In general, the two articles deal withprimary and secondary deposits, respectively.

Primary deposits are divided into two main groups:magmatic and metamorphic. These two groups are furtherdivided into subtypes, each of which is discussed in detailusing typical deposits as examples. Magmatic depositscomprise mainly basalts, such as those in Australia andThailand/Cambodia, but also syenite xenoliths in basalts(Loch Roag, Scotland), syenites (Garba Tula, Kenya), andmafic intrusions (Yogo Gulch, Montana). The metamor-phic deposit subtypes are: (1) pegmatitic intrusions inmafic, ultramafic, and carbonate rocks and their associatedmetasomatic rocks (Umba, Tanzania; Andranondambo,Madagascar); (2) marbles (Mogok, Myanmar; Hunza,Pakistan); (3) gneisses, granulites, and charnockites(Mysore, India; Ratnapura/Elahera, Sri Lanka); (4) amphi-bolites (Longido, Tanzania; North Carolina); and (5) ana-texites (Morogoro, Tanzania).

The discussion of secondary deposits is basically acomprehensive description of the customary classifica-tion of such deposits into eluvial/colluvial, alluvial, andlittoral deposits and their exploitation. Deposits in NewSouth Wales (Australia), Vietnam, Ilakaka (Madagascar),and Tunduru (Tanzania) are discussed in detail. The arti-cle contains several maps and diagrams and is richly illus-trated. RT

Rare gem mineral deposits from Brazil. Part 2: Lazulite andscorzalite. M. L. de Sá C. Chaves, J. Karfunkel, A. H.Horn, and D. B. Hoover, Australian Gemmologist,Vol. 21, No. 10, 2003, pp. 390–399.

Lazulite and scorzalite constitute a complete solid-solu-tion series between the magnesium and iron end members

of the lazulite group. Although they are valued for theirfine blue color, their relatively low hardness (~51¼2–6) andrarity in gem quality make them of interest primarily tocollectors of rare gems. This paper identifies the Braziliandeposits where these lazulite-group minerals occur, dis-cusses their genesis, and gives details of their gemologicalcharacteristics. Representative chemical analyses are givenfor lazulite-scorzalite series minerals from the MinasGerais and Rio Grande do Norte areas. Most specimensare translucent to opaque, best suited for cutting en cabo-chon; transparent crystals are small (most faceted stonesweigh < 1 ct) and typically have cracks and “feathers” thatcause them to fracture during cutting. RAH

Spectroscopic and related evidence on the coloring and con-stitution of New Zealand jade. C. J. Wilkins, W. C.Tennant, B. E. Williamson, and C. A. McCammon,American Mineralogist, Vol. 88, No. 8/9, 2003, pp.1336–1344.

Nephrite jade has long been known to occur at severallocations along the west-central coast of the South Islandof New Zealand, where it is recovered from rivers as cob-bles and boulders. The authors documented material fromseveral deposits by chemical analysis and several spectro-scopic techniques. Nephrite is a near mono-mineralic rockconsisting of randomly oriented, felted masses of micro-scopic needles of actinolite/tremolite. The analyzed sam-ples were generally similar in chemical compositionexcept for variations in iron content, as well as the pres-ence of trace elements such as chromium and nickel. Thecolor of these samples could not be directly correlated todifferences in composition.

Infrared spectra confirmed that the samples were com-posed principally of nephrite; chromian margarite (calci-um mica) was also identified in several samples. Opticalabsorption spectra were consistent with those in the pub-lished literature. The Fe3+/Fe2+ ratio increased for samplesthat were more highly weathered (which typically have abrown surface coloration of hydrated iron oxide that, inturn, is sometimes altered to a whitish rim if the materialwas exposed to more acidic weathering conditions). Theblack appearance of some samples is due to dispersedminute grains of a black opaque mineral (such as mag-netite or chromite). JES

INSTRUMENTS AND TECHNIQUESAn assessment of nuclear microprobe analyses of B in sili-

cate minerals. H. Skogby, P. Kristiansson, and U.Hålenius, American Mineralogist, Vol. 88, No. 10,2003, pp. 1601–1604.

Boron is a widespread but uncommon element in theEarth’s crust. It is an important chemical component incertain gem minerals (e.g., tourmaline) and a minor ele-ment in many others (e.g., some olivine, sillimanite,



idocrase, muscovite). Because it is a light atomic-weightelement, boron cannot be detected by standard electronmicroprobe analysis; as a result, the extent of its incorpo-ration in many silicate minerals is uncertain. This articledescribes an alternative quantitative method for themicroanalysis of the boron content in minerals over awide concentration range. The method involves measure-ment of alpha particles released by this element when itundergoes a nuclear reaction involving a proton. Themethod requires access to a beam of 600–800 keV protonsproduced by a proton accelerator. The authors presentresults of their analyses of several minerals with B concen-trations in the range of 1.9–8.8 wt.%, and discuss theadvantages and possible applications of this nuclear micro-probe method for analyzing boron concentrations in sili-cate minerals. JES

Large OPLTM diffraction grating spectroscope. T. Linton, A.Cumming, and K. Hunter, Australian Gemmologist,Vol. 21, No. 10, 2003, pp. 410–412.

This article describes a new, larger version of the well-known OPL diffraction grating spectroscope. Both are easyto use, have a fixed slit width and a fixed focus, and arecapable of producing high resolution and a wide linearspectrum. However, the new instrument enables increasedresolution of finer spectral features, as well as a 30%increase in the height of the spectra. The quality of theimages of complex absorption spectra obtained from thelarge OPL instrument matched the quality of those pro-duced by most other prism spectroscopes and was superiorin some cases (e.g., it was easier to observe the 450/460nm absorption that typifies blue Australian basaltic sap-phire, and the wider red spectrum allowed more accurateand easier recognition of absorptions in that spectralregion).

Gemologists who require vision correction mayencounter difficulties using this instrument, because thespectral image viewed through the eyepiece lens of afixed-focus spectroscope is set for users with “normalvision acuity.” However, the authors provide an inge-nious solution to this problem through the use of plasticlenses obtained from inexpensive “reading glasses.”Numerous other practical suggestions are made, such aswith regard to illumination, a stand to hold the spectro-scope, and the use of a polarizing filter to increase resolu-tion. The large OPL spectroscope is recommended forteaching purposes and for the practicing gemologist, as itwill show most spectra that are required for gemologicalinvestigations.

MWF

The petrographic microscope: Evolution of a mineralogi-cal research instrument. D. E. Kile, MineralogicalRecord, Special Publication No. 1, 2003, 96 pp.

The petrographic microscope is an instrument that isdesigned to observe and measure the optical properties of

minerals as a means of identifying them, in both unpolar-ized and polarized light. For over a century, it has been anessential tool in the development of the related sciences ofmineralogy and petrography. This beautifully illustratedspecial supplement to the Mineralogical Record, writtenby a research chemist at the U.S. Geological Survey inBoulder, Colorado, presents a history of the petrographicmicroscope from the beginnings of optical mineralogy inthe 1600s. It opens with descriptions of the main compo-nents of a petrographic microscope and of the optical prop-erties of minerals that can be observed with it. Then theparallel historical evolution of the instrument and opticalmineralogy are traced by citing the contributions of manyindividuals. Numerous color photographs illustrate thisevolution with changes in the design of the microscope.The manufacturing of these instruments is also described,as are many specifically developed accessories (e.g., thequartz wedge, waveplate, and universal stage). The finalsection discusses the evaluation and restoration of antiquepetrographic microscopes. This extensive and thorougharticle is recommended reading for anyone interested inthe development of optical mineralogy, which formsmuch of the basis for the tools and techniques of gemidentification. JES

Possibilities of laser ablation–inductively coupled plas-ma–mass spectrometry for diamond fingerprinting.M. Resano, F. Vanhaecke, D. Hutsebaut, K. DeCorte, and L. Moens, Journal of Analytical AtomicSpectrometry, Vol. 18, No. 10, 2003, pp. 1238–1242.

A homogenized 193 nm excimer laser with a flat-topbeam profile is capable of controlled ablation of dia-monds, and the sensitivity of ICP-MS suffices for thedetermination of more than 10 elements. For this study,31 diamonds from four mines (Premier, Orapa, Udach-naya, and Panda) were ablated in eight different spots for30 seconds, and the median of eight integrated signalswas taken as the representative value for statistical analy-sis. Under these conditions, the total mass of materialremoved from a single diamond was ~16 mg (leaving acrater diameter of 120 mm).

Nine elements were selected for fingerprinting purpos-es (Al, Hg, Na, Ni, Pb, Sb, Sn, Ti, and Zn), and various pat-tern recognition techniques (ternary plots, cluster analy-sis, and partial least-squares analysis) were used to classi-fy the data. For example, distinct differences in the dia-monds from the four mines were observed when data forNi, Ti, and Pb were plotted on ternary diagrams. Theseresults are considered as promising, especially for the par-tial least-squares approach, provided that appropriate datastandardization is carried out. The authors emphasize thepractical difficulties associated with determining theprovenance of diamonds by this multi-element tech-nique—particularly for alluvial diamonds that may betransported vast distances from their primary sources.

RAH

JEWELRY RETAILINGHot rocks. T. Treadgold, BRW [Business Review Weekly,

Australia], Vol. 26, No. 20, 2004, pp 98–101.As smaller diamonds and Chinese freshwater culturedpearls find favor in the mass markets, the rich are optingfor bigger and better diamonds and pearls. As a result, pro-ducers of Australia’s high-quality South Sea culturedpearls are maintaining their standards as supplies of lesserqualities of cultured pearls increase steadily, causing pricesto erode. Australian producers, who are limited by govern-ment quotas, must command a premium price for theirgoods or face financial ruin. Last year, Australian culturedpearls accounted for 1% of total world production by vol-ume but 30% by value.

Production of larger, high-quality diamonds is declin-ing, creating a shortage and higher prices. Wealthy con-sumers are competing for such goods, while prices formass-market diamonds have been languishing becauseless-affluent consumers are more price conscious. Brands,including BHP Billiton’s CanadaMark diamonds whichoriginate from the Ekati mine in Canada, are succeeding increating a premium-brand consciousness in that country.

RS

SYNTHETICS AND SIMULANTSFe and Ni impurities in synthetic diamond. Y. Meng, M.

Newville, S. Sutton, J. Rakovan, and H.-K. Mao,American Mineralogist, Vol. 88, No. 10, 2003, pp.1555–1559.

The distribution and incorporation mechanisms of Fe andNi impurities in General Electric Co. HPHT-grown syn-thetic diamond crystals were studied using synchrotron X-ray fluorescence (XRF) microanalysis and X-ray absorptionnear-edge structure (XANES) spectroscopy. The crystalswere small (~700 mm) and were selected from standardindustrial grit product in which Fe and Ni were used ascatalysts. Nickel is dispersed as atoms either in the dia-mond lattice or in interstitial sites. It is concentrated inthe {111} growth sectors relative to the {100} sectors. Incontrast, iron exists as micro-aggregate inclusions with noobservable growth-sector correlations. Further, the Fe isoxidized to the ferrous (Fe2+) valence state and is very like-ly bonded with oxygen as FeO. JES

Features of beryllium aluminate crystal growth by themethod of horizontally oriented crystallization. V.V. Gurov, E. G. Tsvetkov, and A. G. Kirdyashkin,Journal of Crystal Growth, Vol. 256, No. 3–4, 2003,pp. 361–367.

Synthetic chrysoberyl, including its variety alexandrite,has been grown mainly by the Czochralski pullingmethod. A new technique developed in Russia for growingsingle crystals of synthetic chrysoberyl—the horizontally

oriented crystallization (HOC) method—has two mainadvantages over the Czochralski technique. One is theability to use large molybdenum crucibles (in the form of“boats” up to 100 ¥ 35 mm made from sheet molybdenum0.2–0.3 mm thick) instead of expensive iridium crucibles.Another is that the process is relatively simple, with theoriented seed plate being moved horizontally through themelting zone at the rate of 1.5–3 mm/hr. In addition to thecrystallization container, the other main component of theHOC apparatus is a resistance heating system consistingof tungsten coils that allows the growth of crystals with amelting point above 2000°C.

Synthetic chrysoberyl crystal ingots grown by theHOC method are elongated along the [100] direction; theirsize and geometry are determined by the shape of the boat-like crystallization container. The crystals have growthstriations, localized zones of gas-melt inclusions, andnumerous metallic inclusions composed of minute molyb-denum crystals. Synthetic chromium-doped alexandritecrystals grown by the HOC method have several advan-tages over those produced by the Czochralski technique;for example, their chromium content is nearly constantalong the length of the entire synthetic alexandrite crystal.

TL

Growth of 15-inch diameter sapphire boules. C. P. Khattak,P. J. Guggenheim, and F. Schmid, in R.W. Tustison,Ed., Window and Dome Technologies VIII,Proceedings of SPIE, Vol. 5078, 2003, pp. 47–53.

This article describes the world’s largest transparent syn-thetic sapphire boules, which weigh approximately 84 kgand measure 380 mm (15 inches) in diameter. They arebeing produced on a routine basis by the Heat ExchangerMethod (HEM) at Crystal Systems Inc. in Salem, Massa-chusetts. These crystals are grown for use as special opti-cal windows in military and other high-technology appli-cations where high optical quality, compositional purity,and uniformity of physical properties are of crucial impor-tance. The use of very pure starting materials for the crys-tallization process has resulted in a synthetic sapphireproduct with impurity levels that are very near the detec-tion limit of the Glow Discharge Mass Spectroscopy(GDMS) technique (~5 ppm or less for the impurity ele-ments). Efforts are underway to grow 500-mm-diametersynthetic sapphire boules by this same technique. JES

35 years on: A new look at synthetic opal. A. Smallwood,Australian Gemmologist, Vol. 21, No. 11, 2003, pp.438–447.

Within a few years after the structure of synthetic opal wasestablished in 1964, Pierre Gilson Laboratories in Franceproduced the first truly synthetic opal. The history of theearly days of synthetic opal manufacture is reviewed in thispaper. The more recent production of synthetic and imita-tion opals from Japan, Russia, and China is described andtheir gemological properties are listed. Natural opal will



phosphoresce, but synthetic opals will not; the observationof a “lizard skin” pattern and no UV photoluminescence(i.e., the absence of phosphorescence) will confirm that anopal is synthetic. Dark varieties of synthetic opal will readi-ly absorb a drop of water on the surface. Mention is madeof industrial applications whereby the same-sized silica lep-ispheres of synthetic opal are coated with carbon; subse-quent etching out of the silica leaves a series of ordered“shells” of carbon for use in so-called “photonic” devicesthat trap certain wavelengths of light. RAH

TREATMENTSChange of cathodoluminescence spectra of natural dia-

mond with HPHT treatment. H. Kanda and K.Watanabe, Diamond and Related Materials, Vol.13, No. 4–8, 2004, pp. 904–908.

The recent introduction of high pressure, high tempera-ture (HPHT) treatments to change the colors of naturaldiamonds from brown to colorless or to yellowish greenhas generated a persistent demand for new gemologicalidentification techniques. To that end, several natural dia-monds (mostly type IIa) were examined before and afterHPHT treatment (6 GPa, 2000°C) using cathodolumines-cence spectroscopy over the range 220–320 nm to evaluatechanges induced by HPHT processes.

Three luminescence bands—2BD(G), 2BD(F), and5RL—were found to decrease dramatically in intensityduring HPHT treatment. For comparison, the band inten-sities were normalized to free-exciton (FE) lines, which arecommon in type IIa diamonds and not affected by HPHTtreatment. The authors attribute the 2BD bands to plasticdeformation on the basis of their association with mosaicstrain patterns. The 5RL band is reportedly due to naturalirradiation (i.e., a-particles) and quickly disappears withheating. Whereas the presence of the 2BD and 5RL bandsin association with FE lines suggests that a diamond hasnot been HPHT treated, the absence of the bands is notdiagnostic evidence for treatment because the bands donot occur in all type IIa diamonds.

Christopher M. Breeding

MISCELLANEOUSDepositional placer accumulations in coarse-grained allu-

vial braided river systems. J. P. Burton and P.Fralick, Economic Geology, Vol. 98, No. 5, 2003,pp. 985–1001.

Economically significant deposits of certain minerals,including gems such as diamonds and corundum, occur as

alluvial placers along braided river systems. The studyreported here was based on field work at several uraniumor gold placers in Canada, and on results obtained fromlaboratory experiments of moving sediment-water mix-tures designed to simulate flowing rivers. The object wasto identify the fundamental conditions present in gravel-dominated river systems that act to maximize the concen-tration of heavy minerals in alluvial deposits.

Alluvial placers form along bars, that is, ridge-likeaccumulations of sand, gravel, or other material at certainlocations in a river where a decrease in the velocity of themoving water allows for deposition of the transportedsediment load. Although similar in their geologic setting,two processes control the concentration of ore in placerdeposits. Erosional placers are produced by preferentialremoval by the flowing water of light minerals from themineral assemblage in a deposit. In contrast, depositionalplacers are formed by the preferential deposition of heavyminerals from the mixed assemblage of sediments mov-ing past the site. Along a particular river system, both pro-cesses will simultaneously occur in different parts of theriver channel, and will shift position over time. Erosionand deposition work together to produce a placer, and thedominance of one process over the other defines the plac-er type.

Data from this study indicate that a number of condi-tions are necessary, or at least desirable, for heavy miner-als to accumulate in coarse-grained sediments. These con-ditions include:

1. A low proportion of granule to small pebble-size rockfragments in the alluvium

2. A very heavy mineral population with a hydraulicbehavior in the flowing water that more closely resem-bles that of the pebble population compared to that ofthe quartz sand

3. Velocities of the flowing water capable of removing thecoarse-grained quartz sand by suspension

4. A change in the slope of the river that creates a zonewhere the heavy mineral population can no longer betransported by the flowing water and is deposited

5. Infrequent major flooding events that would destroyplacer deposits

6. Possible preconcentration of the heavy mineral popula-tion upstream in the river

These conditions, plus the presence of economicallyvaluable minerals in the sediment load, control the for-mation of exploitable placers deposits in longitudinal barsin braided river systems.

JES

Documents

Diamond Cut: A Foundation for Grading · of blue color in a diamond • Nephrite that mimics serpentine • Pink opal • Rubies clarity enhanced with a lead glass filler • Unusual