TABLE CONTENTS

FEATURE ARTICLES

2 Rubies from Mong Hsu Adolf Pelsetti, I(ar7 Schmetzer, Heinz-Jiirgen Bernhardt, and Fred Mouawad

" 28 The Yogo Sapphire Deposit Keith A. ~~chaluk

NOTES AND NEW TECHNIQUES

42 Meerschaum from Eskisehir Province, Turkey I

REGULAR FEATURES 52 Gem Trade Lab Notes Gem News Most Valuable Article Award Gems ed Gemology Challenge Book Reviews Gemological Abstracts Guidelines for Authors

ABOUT THE COVER: One of the most important ruby localities of the 1990s cov- ers a broad orea near the town of Mong Hsu, in northeastern Myann~ar(B~lrrna). The distinctive gemological features of these rubies are detailed in this issue's lead article. The suite of fine jewelry illustraled here contains 36 Mong Hsu rubies with a total weigh1 of 65.90 ct; the two rubies in the ring total 5.23 ct. jewelry courtesy of Mouawad jewellers. Photo by Opass Sultsumboon-Opass Suksuniboon Studio, Bangltolz, Thailand. Typesetting for Gerrls eS Gemology is by Graphix Express, Santa Monica, CA. Color separations are by Effective Graphics, Compton, CA. Printing is by Cadmus lournal Services, Easton, MD. 0 1995 Gemological Institute of America All rights reserved ISSN 0016-626X - Editor-in-Chief Editor Editors, Gem Trade Lab Notes Richard T. Lidtlicoat Alicc S. I

PRODUCHO~ Art Director Production Assistant Word Processor STAFF Lisa Jolto-Cleeson Gail Young Ruth Patchiclc

EDITORIN. G. Robert Crowningshield C. S. Hurlbl~t,Jr. Henry 0. A. Meyer REI'IEW BOARD New York, IVY Com briclge, A4A West Lofayefte, IN Alan T. Colli~ls Alan jobbins Kurt Nassau Loi~rlon,United l

SUBScRlITlONS Subscriptions in thc U.S.A. alc pricctl as follows: S59.95 for onc year (4 ~ssues),$149.95 for ~lirceycars (I2 issucs). Subscriptions sent elscwhere are $70.00 for one year, S180.00 for three ycars. Spccial annual suhscript~onratcs arc availnble tor all studcnts actively invol\letl in a GIA program: $49.95, U.S.A.; $60.00, elsewhere. Your stdent number 111trstbc listed at thc tinle your si~bscriptionis entcred. Single issues may be purchased for S15.00 in tlic U.S.A., S18.00 elsewhcre. D~scountsarc given for hullc orders of 10 or lnorc of any one issuc. A liniited numbcr of back issues of GdG arc also available for purchase. Plcase address all inquiries regarding subscripttons and the purclinse of single copies or hack issues to the Stlbscril,tions Departmunt. To obtain a Japancse translation of Gems d C;emolcgj~, contact the Association oi Japan Gem Tr~ist,Okachin~acl~i Cy Bldg., 5-15-14 Ueno, Tn~to-ku,Tokyo 110, lapan. Our Canadian gootls :~ntlscrvice registration nilmhcr is R1261.12892.

Gems ed (;~11101ogj~\~~cIc(lnics thc si~~~niissio~iof ~rticlcs on 311 aspccts of the ficld. Pleasc sce tlic Suggestions for Authors in this issuc of tlic iournal, or contact the editor for a copy. Lclters on articles p~~hlisiietlIn Gems cd Gelnology and other rclcvant rnattcrs are also wclcomc. Abstracting is pcrmittcd with credit to the source. Librarics are permitted to photocopy hcyond the li~n~tsoi U.S. copyright law for privatc use of patrons. Instructors arc permi[tetl to photocopy isolated articlcs for noncommercial classroom use withoi~tfee. Copying of thc photographs by any mcnns other than trad~tionnlplintocopy~ng tcch- nirloes (Xerox, etc.) is prohibited without tlic exprcss pcr~nissionof the photographer (where listed) or author of the ;ir~iclein which the photo appears (wherc no photographer is listcd). For other copying, rcprint, or repubIic:it~on pcr- mission pIease contact thc editor. Gems d Gelllology is published quarterly by thc Gemological Inst~tutcof America, a nonprofit educational organi- zation for the iewclry industry, 1660 Stewart Street, Santa Monica, CA 90404. Postmastcr: Rcti~rnilndelivcrahlc copics of Gems eJ Ge117olo~yto 1660 Stewart Strcet, Santa Monica, CA 90404. Any opinions expressed in s~gneda~ticlcs are understooil to be the opinions of the authors ant1 not of the pi~blislicrs. LETTERS

KUDOS FOR "COLOR GRADING SAPPHIRE-BEARING ALKALI OF COLORED DIAMONDS" BASALTS IN NIGERIA Congratulations to the authors of the report With grcat interest I read the article "Gem "Colored Grading of Colored Diamonds in the Corundunl in Allcali Basalt: Origin and GIA Genl Trade Laboratory," by Icing, Moscs, Occurrence," by Drs. Levinson and Cook, in the Shigley, and Liu, which appeared in the Winter Winter 1994 issue (pp. 253-262). However, I '94 issue of Gems & Geniology (pp. 220-2421, It would like to updatc the inforination in onc is a superb achievement. I am so impressed that I remarlc, on page 256: "Significantly, all reported feel coinment is appropriate. economic, and potentially economic, secoildary I was very concerned when I first learned occurrences of basaltic corundurn are spatially that the GIA was preparing an article on their associated with the alkali typc of basalt. (Coen- color-grading system for colored diamonds. The raads et al., 1990, illention two possiblc excep- measureinent of color for paints and other tions, in Nigeria and Southcrn China, but these opaque surfaces is a quite routine matter, but are not well documented.)" transparent substances present complicatio~~s. In the Iournal of Gemmology, 1990, Vol. 22, There are additional problems that arise with the No. 4 (pp. 195-2021, J. Icanis and R. R. Harding ineasurenlent of color in facetcd gemstones, published an article titled "Gemstone Prospects including the variations of color with 01-ienta- in Central Nigeria," in which we described thc tion, across the stonc, face up versus face down, occurrence of sapphircs and zircons of the Jemaa and so on. Much nonsense has been written on district, in central Nigeria. We coinpared the this subject. Jenlaa alkali basalts with thc occurrcnccs in I need not have worried; publication of this Australia, Thailand, and Kampuchea. article has put all illy concerns to rcst. The Froin our description, it is evident that the authors (as well as the ~nallyothers within GIA vast sapphire deposits in central Nigeria are not who helped) are to be congratulated on an excel- "a possible exception," but definitely belong to lent solution to a difficult taslc. They appear to thc alkali type of coruildum occurrencc. have covered all the problems that could arise in JAN KANIS, Ph.D. normal practice. I bclieve the terlninology grid, Veitsrodt, Germany which they based on the time-tested Munsell and ISCC-NCS approaches, is also lilcely to sur- vive the test of time. Their technique is extreme- Reply and Erratum ly well presented in the article; no doubt the edi- We regret that we imissed the paper by Icorundum crystal is 1.3 KURT NASSAU, Ph.D. cin (about 0.5 inches), not 13 cm (about 5 inches). Nassou Cons~ll~ants A. A. Levinson, Ph.D. and Fred Cook, Ph.D. Lebanon, New lersey Calgary, Alberta, Canadn

GEIMS W GEMOLOGY By Adof Pesetti, Karl Schmetzes, Heinz-Jiirgen Bernhadt, and Fred Mouawad

Large quantities of rubies-both rough and ince 1992, Mong Hsu has been a primary source of ruby faceted-fron~ a con~merciollyimportant available in Thailand (figure 1).Mong Hsu is a small town new source in Myanmar (Buirn~n)hmve been Ssituated in northeastern Myanmar (formerly Burma) in avaiIcrble on the Bangltolt inarltet since 1992. Shall State, which borders Thailand, Laos, and China. The nzby crystrrls from the Mong Hsu marble Untreated samples from this new source typically consist of deposit have clipymmidnl to barrel-shaped bicolored corundum, with dark violet to almost black sap- hobils and reverrl dark violet to almost black phire cores and ruby rims (figure 2). With heat treatment, "cores" 017d red "rims." With heat treatment, which removes their blue color componer~t, the violet cores can be converted to red. Large quantities of the cores become intense red. The rubies grew untreated corundum crystals are brought into Thailand at umder varying conditions in complex growth Mae Sai and, to a lesser extent, at Mae Hong Son (see figure sequences. The color distribution between 3). In 1993, about 200 buyers from Chantaburi (Thailand) cores and rims is related to a different incor- were spending several million U.S. dollars a month on poration of chrorniun~andlor titanium during Mong Hsu rough in Mae Sai ("Special report: Mong Hsu . . .," crystnl growth. Gemological, microscopic, 1993).Thus, the Mong Hsu ruby has become an important chemical, and spectroscopic properties pre- source of supply to the world inarlzet. sented here permit the separal.ion of faceted In September 1992, one of the authors [AP) joined a Mong Hs~inibies from their syntlietic and group of gemologists who traveled to Myanmar and other natural counterpclrls. Problems arising Vietnam at the invitation of the Asian Institute of hon7 artificinl fracture fillings nre also addressed. Geininological Sciences (AIGS),Banglzolz, to learn about the occurrences of rubies and sapphires in these countries (see Jobbins, 1992; I

2 Mong Hsu Iiubics GEMS & GEMOLOGY Spring 1995 Figure I. Since 1991, the Mong Hsu area in north- eastern Myanmar has p~o- dzlced large quantities of szlperb rubies. This suite contains 58.22 ct of Mong Hsu rubies (the largest is 2.62 ct). Courtesy of Moziawad lewellers;photo by Wicky Tjerk.

one of the authors (KS).Seventy of the 74 cut stones hand information and photographs were obtained in this lot were almost identical with respect to froill miners in the summer of 1993. Reports on the their gemological characteristics, and unlike mate- local geology and mining operations have been pub- rial from any other lznown locality; of the remain- lished by Hlaing (1991, 1993, 1994) and are summa- ing four stones, two had typical Mogolz features and rized below. two were typical of Thai mby. When it was deter- Mong Hsu is located about 250 km (150 miles) mined that the features of this new material were southeast of Mogok (figure3), at an elevation of 700 very similar to those of the Mong Hsu stones in above sea level. It can be reached from Taung acquired in Yangon (notwithstanding any potential Gyi, the capital of Shan State, by traveling over a change of color and other properties that might be rough road for about 14 hours (Hlaing, 1994). caused by heat treatment), the decision was made Secondary (alluvial)corunduin deposits are found in to carry out a complete study of these rubies. the terraces of Nan1 Hsu River, southeast of Mong Preliminary reports about some of the proper- Hsu township. These river terraces, where the first ties of Mong Hsu rubies have appeared in trade rubies were discovered and where mining and journals (Clark, 1993; Laughter, 1993a and b; prospecting started, are 4 lzm long and 0.8 lzm wide, Peretti, 1993; Milisenda and Henn, 1994)) and an trending froill northwest to southeast, and 80-160 initial research study was published by Smith and 111 deep. In 1992, about 2,000 miners were working Surdez ( 1994). these alluvial deposits. Additional secondary deposits have been found LOCATION 16 km (10 miles) farther southeast. These extend Although, because of security concerns, the authors over an area more than 100 lun2 between the two were not able to visit the Mong Hsu deposit, first- mo~mtainsHsan Tao and Loi PLaning.The "government

Mong HSLIRubies GEMS 6: GEMOLOGY Spring 1995 1994), literally chewing up the surrounding hills (figure 4). Alluvial ruby is also found at Loi Khan and Mong Sang, which lie southwest and south (respectively)of Mong Hsu (again, see figure 3). GEOLOGY AND MINERAL ASSEMBLAGES According to the geologic map of Myanmar (Earth Sciences Research Division, 1977), the Mong Hsu ruby deposits are situated at the contact of upper Paleozoic marbles and other Paleozoic rocks, including various types of n~etainorphosedsedi- ments. Hlaing (1991, 1993, 1994) reports that the major roclc types in the region of the primary ruby occurrences are mica schist, phyllite, and calcsili- cate rocks; the rubies occur in a marble belonging Figure 2. A distinctive featme of Mong Hszr ruby to this Paleozoic series. crystals 1s their deep violel (sopphire) core, which is surrounded by a red (ruby) rim. With heat treat- Additional information on the different rock ment, the violet core becomes a deep red. This types in the vicinity of the Mong Hsu mines was Mong HSLZruby "thin section" is 5.3 mm wide; obtained from the study of minerals that appear photo courtesy of John Emmett. mixed with the ruby rough from the market at Mae Sai and on the surface of the corundum crystals prospecting area" is restricted to a central 0.8 km2 themselves. These were identified by ineails of X- portion (flgure3). By 1994, about 500 joint ventures ray powder diffraction analysis, quantitative elec- between private individuals and the Myanmar gov- tron microprobe analysis, and a scanning electron ernment were operating in the region (Hlaing, microscope with energy-dispersive X-ray detector.

F'ig~rc3. This map of the Mong Hsu region in norlheastern Myrinmar (see insel) shows the areas of currenl mining activity. The allzzvial deposits (stip- pled areas), where he first rubies were fozmd, are in river terraces sozztheast of the town of Mong Hszz. Primary occurrences of nzby with adjacent secondary deposits were discovered around Loi Hsan Too. The pink region represents the area of gem potential suggested bj, prospecling in 1993. Adapted fron~Hlaing, 1993; courtesy of the Australian Gemmologist.

Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 They include green and brown chromium-bearing dravite (tourmaline),andalusite, almandine, quartz, and green, chroini~~m-and vanadium-bearing tremolite. One of the almandine crystals was over- grown with white mica. Overgrowths on the ruby crystals were identificd as fuchsite (green mica), white mica, and light green Mg-chlorite. Garnet, green tourmaline, tremolite, white mica, and quartz were also described by Hlaing (1993), with staurolite and pyrite mentioned as accessory minerals. The overgrowth of green tourmaline on Mong Hsu rubies noted by Hlaing (1993) was not found in this studyj rather, all light green to green overgrowths on rubies avail- able to us were identified as either fucbsite or Mg-chlorite. Figz~re4. Holes nnd shnfts dug to mine the gem- From these mineral assemblages, it is evi- bearing grave1.s penellate the hillsides s:~Mong dent that, although both the Mong Hsu and Hsu in this 1993 photo by V. Yothan~t. Mogolz occurrences are metnmorphic, they differ in the specific environment in which the rubies CUT STONES AVAILABLE IN THE MARKET formed (Earth Sciences Research Division, 1977; The vast majority of faceted Mong Hsu rubies Hlaing, 198 1; Keller, 1983; Hunstiger, 1990; found on the world marlzet are heat treated, and Kane and Kammerling, 1992; Ka~nmerlinget al., there is a wide ringe of qualities. Large quantities of 1994). At Mogolz, ruby occurs in situ in amphi- transparent stones with fine red color were avail- bolite-to-granulite facies n~eta~norphosedmar- able in 1993 and 1994 in sizes up to 0.7 ct (see, e.g., bles and calcsilicate marbles. In contrast, the figure 6).Stones of good to very good color and clari- nlineral assenlblages of Mong Hsu indicate mar- ty were found in the 1-2 ct range. Stones between 3 bles and metapelitic roclzs metamorphosed to and 5 ct were found to be nlostly of inedium to (lower temperature) amphibolite facies. Thus, good quality. These observations are consistent Mogolz represents a higher degree of inctarnor- with those reported recently in the trade press phisin than does the Mong Hsu mining area. ("Rains wash out Mong Hsu supply," 1994). Clarity inay be reduced in samples of all sizes by translu- MINING cent zones of dense white clouds or by the presence Milling of the secondary deposits is by the classic nlethods used in Mogolz (as described by, e.g., Kane Figure 5. ALthe Mong Hsu ruby occurrence, un and Kamn~erling,1992) and elsewhere in Southeast elaborate slzlicing system brii1g.y the gravels down Asia. In fact, a Mogolz miner is believed to have this hillside for processing. Photo by V. Yothavu t. been the first to discover rubies in Na~uNga Stream at Mong Hsu (Hlaing 19941, and inany inin- ers have traveled from Mogolz to work at the new locality. In and along the waterways, the gravels are reilloved and washed in siinple baskets. Ruby-bear- ing gravel layers (known as byon] in the surround- ing area are reached by: (1) digging holes from the surface into which the miner is lowered (and the gravels removed) by a simple rope and pulley sys- tem; (2)excavating horizontal tunnels into the hill itself, at the level of the gravel layer; and (3)open- pit mining, with the gravels sorted and recovered by soinetinles elaborate sluicing systems (figure 5). At some mining sites, inechanized sieves are used to work the gravels (Hlaing, 1994).

~MongHsu Rubies GEMS & GEMOLOGY Spring 1995 5 Fillings in large fissures and craclzs improve the apparent clarity of the stones considerably (Peretti, 1993; Milisenda and Henn, 1994).They result from heat treatment in borax or a similar substance, but heat treatment in chemicals is not necessary to alter the violet cores of color-zoned stones to red, as shown by heat-treatment experiments performed in Germany (see figures 7 and 8). According to infor- mation obtained in Banglzolz, some commercial Thai laboratories use a two-step procedure to heat treat Mong Hsu nibies. First, the samples are heat- ed, without the use of borax, to reinove the violet color. Then, the nlbies are heated in a borax con- Figure 6. Mong Hsu rubies are typicolly heat treat- tainer to fill fissures and thus enhance apparent ed, to prodrlce a well-saturated red hue as shown clarity. Consequently, some Mong Hsu rubies that here. Lilie most of the face~edstones prodzzced to have fractures filled with a foreign substance have date, these Mong Hsu rubies, from the study sam- been treated for both clarity and color enhance- ple, are small, 0.348 to 0.683 ct. Photo by Shane F. ment. In these cases, glass or other foreign fracture McClure. fillings are not simply an "inadvertent" by-product of heat treatment conducted to reinove the blue color component in the samples, as some in the of fingerprint-like inclusions and craclzs (see trade have claimed. "Microscopic Features" below). Because rubies with artificially filled fractures In inany of the faceted nlbies examined for this must be sold as treated, large lots of Mong Hsu study, foreign filli~~gs-suchas glass-lilze sub- rubies containing many stones with such fillings stances-were found in fissures and surface cavities were rejected by European dealers and returned to (again, see "Microscopic Features" below). Gem lab- Banglzok in 1993. In addition, some fracture-filled oratories in Asia have reported seeing glass fillings Mong Hsu rubies have been misidentified as flux- in more than 50% (and, one Thai laboratory, in as grown synthetic rubies. Some Thai companies have much as 90%) of the Mong Hsu rubies examined to tried to remove the fillings through acid treatment date ("Glass filled rubies increasing," 1994). Stones or by recutting the stones. These problems-and above 5 ct with excellent color and clarity, and no others, such as the high voluine of production- evidence of foreign fillings, appear to be extremely caused a steep drop in the price of Mong Hsu rough rare.

Figure 8. This 8-min-long Mong Hsu crystal has Figure 7. Heat treatment (in Germony)produced been cut to show the effect of heat treatment on the dran~oticcolor change (right) in this Mong the core area. The right half is the original Hsrr crystal, tvllich was originally almost black (untreated) control sample; with heat treatment, (left). Total lengt11, aboilt 6 mm. the left half is now a solid red

6 Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 at Mae Sai and Bangkok in 1993, to nearly half of its peak (see, e.g., Icoivula et al., 1993b; Icammerling et al., 1994; Hlaing, 1994).

CHARACTERISTICS OF MONG HSU RUBIES Materials and Methods. In addition to the six rough samples provided by MGE in September 1992 and the 74 faceted stones obtained in October 1992 from a Munich dealer, we subsequently saw and examined several parcels of untreated and heat- treated rough and faceted Mong Hsu rubies in both the Banglzok and European markets. Using as our guide the unusual genlological properties (e.g., growth features) of Mong Hsu rubies that we had determined by late 1992, we were able to select parcels of stones that we were confident were free of rubies from other localities for further examina- tion. As of summer 1993, we had selected approxi- mately 50 rough samples and 100 faceted stones- most less than 1 ct (again, see figure 6), with a few up to 2.5 ct-for this research project. In addition, during separate visits to northern Thailand in mid- 1993, two of the authors (AP and FM) saw large quantities .of untreated Mong Hsu rubies that had entered Thailand at Mae Sai. We purchased several parcels of selected san~plesfor this study. After pre- liminary examination of about 1,000 carats (in sizes up to 5 ct) of the inaterial obtained at Mae Sai, Figure 9. We selec~ed~hese untlaated pieces of which was top-quality rough, we selected a parcel Mong Hs~rr~rby ro~lgh from our study sample to of 23 rough, untreated pieces for further study (see, illustrate some of the many differentforms in e.g., figure 9). In addition, we examined a great which this materir~loccurs. The first two rows are number of the samples selected at Mae Sai after typical barrel-shaped, well-terminated they had been heat treated in Chantaburi, Thailand; crystals; the third row shows flat sanlples that are sometilnes seen, which possibly fornled in nrirrow from these, we selected nine for further study in veins; the waterworn crystals in the fourth row Europe. One of the authors (KS)also heat treated represent a minor proportion of the ruby lots. For about 15 Mong Hsu samples in Germany. an idea of size, nole that the crystal on the far left In summary, we had access to more than 200 in the upper row is approximately 8 mnl loilg. cut and about 100 rough gem-quality mbies, which we were confident were of Mong Hsu origin, for detailed gemological and mineralogc studies. Schiieider lnicroscope with Zeiss optics as well as We performed standard gemologcal testing on an Eiclzhorst vertical lnicroscope with Nikon optics about 50 of these samples (faceted stones and rough (thelatter with fiber-optic illumination). with polished windows). To characterize the inter- Solid inclusions were identified by Raman nal and external growth planes, we studied approxi- nlicroprobe spectroscopy, using an X-Y Dilor mately 200 samples total (about 60% of which were instrument, as well as by a Philips scanning elec- faceted) using a Schneider horizontal (immersion) tron microscope with a Tracor energy-dispersive X- microscope with a specially designed sample holder ray spectrometer (SEM-EDS). and with specially designed (to measure angles) eye- Bulk chemical analyses of five untreated rough pieces (Schmetzel; 1986a; Kiefert and Schmetzer, and five heat-treated cut stones of variable color sat- 1991; see box A); an additional 12 rough crystals uration were performed by energy-dispersive X-ray were examined with a standard goniometer. We fluorescence (EDXRF)using a Tracor Northern TN studied and photographed the inclusions using the 5000 system. The analyses yield the mean chemical

~MongHsu Iiubies GEMS & GEMOLOGY Spring 1995 7 BOX A: Determination of Growth Structures

The determination of a gem's structural properties, the stone in this orientation, which is easily accom- such as straight growth planes that parallel the exter- plished by interference figures seen with crossed polar- nal faces of the original crystal, or twin planes, is izers, the gemologist can determine the growth planes becoming increasingly important as an additional, eas- by tilting the gem in the direction the optic axis is ily performed method to characterize natural and syn- inclined vis a vis the microscope axis. The angle thetic gemstones. It requires a microscope, an immer- between the optic axis and a family of straight parallel sion cell, and immersion liquids (see, e.g., Kiefert and growth planes can be read on the sample-holder dial. Schmetzer, 1991, for a detailed discussion of the appa- For Mong Hsu rubies, traces of the hexagonal ratus used for this report). Also useful for the determi- &pyramid w are sharply outlined with the optic axis native procedures in a horizontal lnicroscope are (1) a inclined about 5" from the lnicroscope axis (as illus- bi

conlposition of a surface exposed to the X-ray beam faceted and two of the rough samples. A inore (approximately 2 mm in diameter). Sixteen ele- detailed examination, with four scans of 600 to 800 inents (Al, Mg, Si, P, I(, Ca, Ti, V, Cr, Mn, Fe, Ni, point ailalyses each, was performed on the third CLI,Ga, Zr, Nb) were analyzed with software pro- rough sainple (figure B- 1), which was extraordinary. vided by the manufacturer, which normalizes the We had the samples oriented so that the visible resulting oxide percentages to 100 wt.% totals. color zoning between the core and the riin could be Five faceted and three rough sainples were ana- traversed. lyzed by electron illicroprobe (CAMECA Camebax Spectral data of 25 representative samples, SX 501, usiilg an acceleration voltage of 20 1V; stan- including untreated rough and heat-treated cut dard materials of corund~ml(A1203), rutile (Ti02), stones, were obtained with a Leitz-Unicam SP 800 eslzolaite (Crz03),hematite (Fez03),and Mn and V UV-VIS spectropl~otometer.Infrared spectroscopy metals; and counting times of 60 seconds for each was carried out on 15 heat-treated samples using a element (necessary to detect traces of, e.g., iron). To Pye-Unicam 9600 FTIR spectronleter and a diffuse- evaluate the inhoinogeneous chenlical composi- reflection unit. tions of the sanlples, between two and seven tra- verses of 30 to 50 point analyses each were mea- Crystallography. Facet-cluality Mong HSLIrubies are sured across the cores and outer areas of the five typically well-terminated, barrel-shaped crystals

8 Mong Hsu Rubies GEMS & GEMOLOGY Spring 1905 Figue A-1. Mong Hstl nlby slice (nbout 4.5 nun wide), cut perpej~dicular to the c-tms:incli~~ed nbout 5" 1-0the c-urds (right),o planes tue shtup; inclined abotlt 30" (left), rand n planes (ueshmp. Inunersioj I.

'I heat-treatedFigure A-2. III and this {acetecl Mong Hstr rtlby, the intense red "core" is conlined lo growl11 ftices parallel 10 [hebast11 pinncoid c; I Iigh~erred rlrens ore conjined to two hexog- o11a1dipyrt~micls n and o (lefi)ajid LO t11e posi- tive rho1n11oher1ro11r (right). Viewperpenclic- ulnr to the c-axis;dur- ing (I rol rllioji of 1 kc Istones ~11rotig1130' obot~tthe c-tlxis, two dif{erent s11orp oullines o/growtli slnicl [Ires ore visible. Ijnmcrsion, n it~gnificd50x.

(figure 10).Two major habits were observed (figure ent red outer zones ("rims")and violct to alillost 11):The first is dominated by the hexagonal dipyra- blaclz centers (commonly called "cores")that appear mid o (14 14 28 3) and by the basal pinacoid c opaque to translucent (again, see figures 9 and 10). (OOOl),with a subordinate positive rhoinbohedroil r In thin scction, a corc that originally loolzed allnost (1071).The other habit shows additional hexagonal blaclz aid opaque will undoubtedly appear violet dipyramids n (22a).Only ininor ainounts of water- and transparent (see, e.g., figure 2). Depending on worn crystals were found in the lots examined. what part of the original crystal a fragment rcpre- Another forin of Mong Hsu niby is extreinely sents, great variability in color and color zoning is flat, possibly due to growth in narrow veins (again observed in lots of the rough. A inore detailed see figure 9). Crystals with this morphology, in gen- description of this extraordinary color zoning is pre- eral, are not useful for jewelry purposes and were sented in "Microscopic Features" below. therefore not included in this report. Heat treatment renloves the violet component of the cores to produce stones that are uniforinly Visual Appearance. Less than 5% of the rough red, as confirined by cxperiincnts carried out in Mong Hsu samples we examined were a uniform Gcrinany (see figures 7 and 8). Therc is, however, red or violet to allnost blaclz. Nlost stones from this some variation in the shades of red seen in the heat- locality show a distinct color zoning, with transpar- treated samples. I11 addition, in some of the rough

~MongHsu Rubies GEMS ek GEMOLOGY Spring 1995 9 Fipre 1 I. The crystal habit of Mong Hsu rubies consists of two hexagonal dipyramids, n (2233) and w (14 14 28 3), [hebasal pinacoid c (OOOl), and the positive lhombohedron r (1Oil). The crys- Figure 10. A well-formed Mong Hsu crystal con- tal drawn on ~11eleft is dominaled by w (2nd c sists of two hexagonal dipyramids, w and n, the faces, wilh a szlbordinale r plane; the crystal positive rhombohedron r, and the basal pinacoid shown on the right is dominated by w and c faces, c, as illzrstrated by this 5-inm sample. Note the with subordinate r ai~dn planes. violet core. heat treated commercially, we observed translu- Optical Properties. We found the variations in cent, highly reflecting, whitish zones that inade the refractive index ainong the different sainples to be stones unsuitable for faceting. As a result of these unusually high for both the ordinary and extraordi- whitish zones, a percentage of the material is reject- nary rays, with a slightly variable birefringence ed after heat treatment. Although distinct color between 0.008 and 0.010 (see table 1). A similar zoning is usually not visible with the unaided eye, large variation in R.I.'s was reported by Smith and microscopic examination reveals a highly charac- Surdez (1994) for rubies froin Mong Hsu, and was teristic type of color zoning related to specific recently describcd for rubies froin Malawi that growth structures in most of the heat-treated sam- revealed highly variable ainounts of trace elements, ples-with intense red related to core zones and especially chroiniuin (Bank et al., 1988). lighter red related to rims. We also found differences in R.I. between cores and rims of both heat-treated and untreated Mong Gemological Properties. The distinctive color distri- Hsu samples. For example, one rough sample that bution in Mong Hsu rubies is also reflected in some was sawn and polished revealed R.I.'s of no = 1.774 of their gelnological properties (table 1). for the core and 1.770 for the rim, ne = 1.765 for the core and 1.762 for the riin (birefringence = 0.009 and UV Fluorescence. The differences between cores 0.008, respectively). and rim areas are well illustrated by their reactions Therefore, depending on the orientation of the to long- and short-wave ultraviolet radiation. In table facet with respect to the optical axis and the untreated samples, the cores are inert or fluoresce portion of the original crystal the table encompass- light orange to light red, whereas the riins fluoresce es, different refractive indices are possible. If the intense orange-red to red. This difference is less table facet is largely confined to one single growth obvious after heat treatment: The cores of heat- zone (see "Microscopic Features" below and boxes A treated samples fluoresce orange-red to red, whereas and B), sharp shadow edges are observed for the the rims remain intense orange-red to intense red. ordinary and the extraordinary rays on the refrac-

Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 TABLE 1. Gemological characteristics of Mong Hsu rubies.8

Untreated samples Heat-treated samples Property Core Rim Core Rim

Color Violet to black Red Intense red Red Pleochroic colors Parallel to c Light bluish violet Orange-red Orange-red Orange-red to reddish violet Perpendicular to c Intense bluish violet Purplish red Purplish red Purplish red to reddish violet Fluorescence Long-wave UV Inert or light orange Intense orange red Orange red Intense orange-red Short-wave UV Inert or light red Intense red Red Intense red

Specific gravity (range) 3.990 - 4.010 Refractive indices (range)

"0 1.768 - 1.780 "e 1.760 - 1.770 Birefringence 0.008 - 0.010

a For the ordinary type of samples with violet to black cores and red rims.

tometer (table 1).This is often the case for samples Well-preserved Mong Hsu corundum crystals with table facets cut parallel to the c-axis or for are defined by c, r, n, and w faces. These growth fea- samples with small cores. If the table facet encom- tures can also be observed in faceted stones or in passes aL1inixed"area-that is, part is from the core fragments of rough stones as thin traces of growth and part is from the rim, the shadow edges are often planes parallel to the former external faces of the less distinct and in extreme cases no readings are mby. visible on the refractometer. Typically, this is the Because of the complex growth sequence estab- case with a stone on which the table facet is orient- lished for Mong HSLIrubies (see box B), crystals and ed perpendicular to the c-axis, with an intense red crystal fragments also reveal distinctive color zon- central core and a lighter red rim. All intermediate ing. Samples that are homogeneous in color (sug- situations are observed for faceted samples with gesting crystal growth in a single phase) occur only randoin orientation. rarely. These are, for example, red with c, r, n, and This unusual variation in refractive indices w faces (again, see figure 9) or dark violet to allnost within one crystal is usually caused by the chemi- blaclz with c, r, and w faces (again, see figure 7). The cal zoning between cores and rims, but it also may most common pattern in Mong Hsu rubies, howev- occur between different zones within cores and er, consists of a darlz, nontransparent core and a red outer areas (see "Chemical Properties" and box B). outer zone (see again Figure 2). Other gemological properties of the Mong Hsu An even more conlplex pattern is formed in rubies tested, such as pleochroism and specific grav- samples that contain two dark violet "cores" that ity (table 1))were consistent with those of natural are separated by a red layer along the c-axis (figure rubies from other localities. Note that the color of 12).In a view parallel to the c-axis, a darlz core in the untreated cores appeared intense violet perpen- the center is followed by a light red zone, a second dicular to the c-axis and somewhat lighter violet lighter violet zone (representing the second "core"), parallel to the c-axis. and an outer red rim. In a very few instances, we observed Mong Hsu rubies with a thin, intense red Microscopic Features. Growth Structures and Color rim confined to n or w faces (figure 13).The thin red Zoning in Untreated Mong Hsrl Rubies. By deter- rims represent the latest stage of growth for these mining the internal growth structures in a cut samples in which an enrichment of chromium toolz stone, the gemologist can reconstruct the habit of place. the original mby crystal. (Again, see box A for the A modification of the typical habit and color procedures used in this study.) Changes in the habit zoning in Mong Hsu rubies is shown in figure 14. during growth and color zoning characteristics also The violet-to-blaclz color in this extraordinary ruby can be studied (see, e.g., box B.) fornls only in growth zones parallel to basal planes,

1Mo11g Hsu Rubies GEMS& GEMOLOGY Spring 1995 11 BOX B: Color Zoning and Habit Changes- in Untreated on^ Hsu Rubies The most distinctive. characteristic feature of untreat- &ate (I) zones, which subsequently convert to violet ed Mong Hsu rubies is their color zoning. Although (V']zones. The I zones consist of alternating thln red two zones are usually evident-violet "cores" and red layers parallel to c, r, n, and w, and small violet layers "rimsu-the zoning in Mong Hsu rubies is actually parallel to c, r, and n (again, see table B-1; also see fig- somewhat more complicated than this. ure B-3), but not w. Most Mong Hsu ruby crystals consist of one (or, In svIalong the c-axis of the crystal described, rarely, two) violet to allnost black cores with intense the following cyclic sequence of growth zones is observed: red rims. An extraordinarv saln~lewith two dark vio- violet violet let "cores"separated by an intense red zone (figure B-l) layers layers best reveals this general scheme for color zoning relat- v 'T' n,r*v 'T' r,, ed to growth history. The duplicate sequence identi- v retl habit red habit red fied in this stone (table B-1)confirms the "single" char- layers change layers change layers acteristic sequence observed in most Mong Hsu rubies I (an exception is the stone dustrated in figure 14). Each of the two violet "cores"in this crystal con- Red growth zones (R)consist of red (ruby) layers sists of violet layers parallel to c, r, and n faces, which parallel to the basal pinacoid c, to the positive rhom- form in subsequent I and V growth zones. Over the bohedron r, and to two hexagonal dipyramids, n and two violet cores are red layers parallel to c, r, n, and w, w. Because all four faces (c, r, n, and w)have growth which grow in subsequent R and I zones. Those parts rates greater than zero, red layers form parallel to all four. of the "cores" that formed during V are usually dark Violet (V)growth zones consist of violet (sapphire) violet and translucent to almost opaque black; those layers parallel to c, r, and n. In V growth zones, only that formed during I are lighter violet and translucent these three faces reveal growth rates greater than zero; to transparent. w growth rates are not observed. Tlus indicates that w We have found the growth sequence R I V to dominates the external crystal form, although no vio- apply to most of the Mong Hsu rubies examined. Note let layers grow parallel to w . that in some crystals, stage R is very small. Others Along tbe c-axis of the ruby shown in figure B-1, which is schematically drawn in figure B-2, the V zones in the center of the crystal abruptly convert to R zones. The R zones then gradually convert to interme-

Figure B-1. This 10.5-mm- long untreated crystal, with two violet cores SLU- rouncled by red meas, ilh~strnt~sthe relotionship between growth condi- tions nnd hnbit change typical of Mow Hsu rubies.

Figure B-2.This schem- atic representation of habit change aid color 1 zoning in the center of the Mong Hszl ruby in figure B-1 shows the R (red), V (violet) and I (inter- mediate) growth phases seen in the following growth sequence (from bottom): I V R I' V' R'. The arrow indicates the pi- tion of a microprobe sun. were grown during a period that did not cover the full I R I V growth sequence; such a crystal may consist only of a dark violet, nontransparent core grown in stage V, and a red rim grown in stage R. Also notable in Mong Hsu nibies with the R I V growth sequence is the decreasing size of the hexago- nal dipyramid n. During growth phase R, relatively large n faces form (see, e.g., figure 11B). In the subsequent zones I and V, n is progressively smaller; in the end stage of phase V, n is absent (figure 11Aj. Consequently, with an abnipt change of growth conditions from stage V to stage R, there is also a distinct habit change. In summary, at the beginning of growth stage R, the crystal habit consists of c, r, n, and o faces (figures 10 and 11B).In the growth sequence R I V, the growth rate of n strongly increases until, at the end of stage V, it disappears completely. Concurrently, the growth rate of face o reaches zero, so o dominates the external crystal form (figures 11A and 7). To date, we have seen no systematic change in the r and c faces over the sequence R I V in the samples available. C The conlplex growth structure of Mong Hsu rubies is reflected by a complex chemical zoning. For the sample pictured in figure B-1, four microprobe tra- piswe B-3, e~nrsdpom~onoftheM~~~ H~~ verses, with 600 to 800 point analyses each, were per- figues B-I md 8-2 showspart of the I' growthzone, formed. The position of one of these scans across the w.& dternatingviolet and r~layc?rspar~elton, md sample (including 700 equidistant analysis points) is ,dlayerspnralleItoo. Inmersion, mwfid50x shown in figure B-2. It twice crossed three red o growthiones R', I', and R, as well as violet n and c growth layers belonging to the V area. The profile about 0.40 wt.% TiOz. This represents a growth zone revealed that chromium values are symmetrical in that was present during the formation of the second both areas right and left of the center of the crystal; in violet core V'. In the outer w growth zone of R', we general, chromium contents are higher in the violet recorded a continuous increase (from 0.63 to 0.97 core of growth zone V than in the R', 1', and R growth wt.%) in Cr203, which reaches values in the outer- zones. Agm, in the violet core, Mferences between areas most layers that are shnilar to the concentration in the related to the basal pinacoid c (1.20-1.35 wt.'Xo Cr203) core area that is related to n faces (see also figure 13). and areas related to dipyramidal n growth faces Average vanadium and titanium contents in the (0.90.95wt.% Cr203)were also measured. Chromium violet core are about twice those measured in the red contents in the red R and I' zones of the rim (0.63-0.75 rim. No distinct zoning of iron was observed in this wt.%) are generally lower than in the violet core. traverse. The other three sc,uls show even more com- In the small area between two adjacent w growth plex results, but these details are beyond the scope of zones in the red rim (i.e., between I' and R'), the scan this article. In summary, the colnplex color zoning reveals a distinct decrease from about 0.63 to 0.38 seen with the microscope is reflected by a complex wt.% Cr203, correlated to an increase from 0.05 to chemical zoning in the sample.

TABLE B-1. General outline of habit change in different growth phases (R, I, V) of Mong Hsu rubies.a

Variables R (red) I (intermediate) V (violet) crnw crn w c r n w Growth rate Mod. Mod. Slow Slow Mod. Mod. Mod. Alternating Mod. Mod. Very fast None slow/none Layers formed Yes Yes Yes Yes Yes Yes Yes Alternating Yes Yes Yes No yes/no Relative sizes Mod. Mod. Large Very Mod. Mod. Mod. Very Mod. Mod. Very small Very of faces large large or absent large Variety Ruby Alternating ruby/violet sapphire Violet sapphire Habit c, r, n, o (figure 11) c, r, n, rn c, r, o (figure 11) n is large n becomes smaller n is very small or absent

aHabit changes from R to I and from I lo V are continuous; the habit change from V to R is abrupt.

Mong MSLI Rubics GEMS & GEMOLOGY Spring 1995 13 Figure 12. This untreated piece of Mong Hsu rough Figzrrc 14. Uzllilze most of the other Mong Hsz~ revealed a complexgrowth history: A small blaclz rzlbies examined for [his strrdy, the violet core in core in the center is szlrrozlnded by a first red this untreated rough sample is confined only to layer, over rvhich is a second violet zone that, in the basal c face, while n and o growth planes are turn, is surrounded by a red layer. Viewparallel ro red. The size of the basnl plnne increases toward the c-axis, immersion, magnified 40x. the outer zone of the crystnl, in a wedge-shaped pattern. Viewperpendicular to the c-axis, ilnmer- sion, magnified 30x.

Figure 13. In this z~ntrentedMong Hsz~fragment whereas growth planes parallel to n and o are red. with n and o growth plunes, an intense red stripe We also observed that the c-plane increased in size co~lfinedto a na~zlrcrln face represents Lhe latest during growth, in a direction toward the outer faces growth stage. Viewperpendicular to the c-axis, of the crystal. Thus, the violet portion in the stone immersion, mamified 30~. appeared as a wedge-shaped pattern in the rough crystal, with the base of the wedge confined to the latest growth area. In another sample, which also had a wedge-shaped growth pattern confined to the basal plane, dark violet areas were developed as small stripes parallel to o, and lighter violet stripes were observed parallel to n, but areas confined to c and r growth zones were red.

Growth Structures in Heat-Treated Rubies. Some of the growth characteristics related to natural color zoning are not observed in cut Mong Hsu rubies, because part of the growth history of a crystal is lost during cutting and heat treatment turns the violet cores red. However, neither cutting nor heat treat- ment alters the characteristic internal growth planes, which we were able to identify in most of the faceted Mong Hsu rubies examined. Due to the fact that these same patterns and combinations of patteins have never, to the best of our lznowledge, been observed in natural rubies before, we feel they are useful to distinguish Mong Hsu rubies from rubies froin other localities. In other words: The individual crystal faces observed in Mong Hsu rubies have been identified in rubies from other sources, but the overall pattern of color zoning and habit-that is, the combination of faces

Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 Fipre 15.Lfhis faceted, heat-treated Mong Hsu nrby is ~ypicalof most of the cut Mong Hsu rubies we exa1nil7ed.An il7tense red core is confined to fr~ccsparollel to thc basol c plane, t17e slzc of the basal plane varies irreg~llarlyalong the c-axis, and Figure 16. 111 this fa~~~~cl,heat-treated Mong Hsu r~~by,wilh don7incint growth parallel to c in the the lighter red areas are confined to 11 and o core and w planes in the rin7 (0s is ~ypicalfor growth zones. View perpendiculnr LO thc c-axis, imnlersion, crossed polarizcrs, mognilled 40x. rubies fro177 this locality), only subordinate ond small 11 faces are observecl in the core. Viewper- pendicullar to the c-axis, immersion, magnified 50x. and the color zoning related to certain growth zones-has not. In addition, because of their dis- tinctive growth patterns, Mong Hsu rubies can be ning, predonlinantly in one direction parallel to one easily separated from their synthetic counterparts. rhombohedra1 r face. Wc observed particlcs con- Typical exaillples of growth structures that can finecl to intersection lines of twin plancs in only be found in heat-treated Mong Hsu rubies are two 01 our samples, which confirms that rubies shown in figures A-2, 15, and 16. In lnost of our with two directions of rl~ombol~edraltwinning par- samples, dark red color zones were confined to allel to two r faccs are extrcnlely rare from this basal c growth planes of variable size (figure 15). locality. These dark red zones are surrounded by lighter red areas with growth planes parallel to r, n, and (0 (fig- Solid Inclusions. Only rarely did we observe solid ure A-2). Because faceted stones represent only one inclusions other than whitish particles (see below) area within the original crystal, in the complex in the Mong Hsu rubies examined. These include growth sequence of Mong Hsu rubies (see box B] rutile and fluorite (figure 17; both identified by the growth zones confined to n faces may be Rarnan spectroscopy and SEM-EDS), as well as extremely s~nallor absent (figure 16). spinel, which was identified by Rainan spec- troscopy. Dolomite was identified by SEM-EDS Twinning. Twinning is encountered only infre- analysis in one crystal; it occurred as a scrics of quently in Mong Hsu rubies. The nlost coinlnon rounded, transparent inclilsions throughout the red type appears to be a repeated rhombohedra1 twin- portion. Dolomite was also identified by Sinith and

Mong Hsu Rubies Spring 1995 15 Flgurc 17. Fluorite (lelt)and spinel (rzght) are Figure 18. White Mg-chlorite wrrs identified in the anlong the /ew solid inclusions observed in Mong outelmost parts of some ~lntrerltedMong Hsu Hsu rubies. Fiber-optic illrnninatzon, frnnsmitfecl rubies. Fiber-optic illumination, reflected rllld and reflected light; diameter of the solid inclu- trozlslnitted light, n~agnified90x. sioz~sis approxin~ately0.2 mm. rough and heat-treated cut Mong Hsu rubies. These Surdez (1994)in Mong Hsu rubies; they found particles represent one of thc most characteristic apatite as well. inclusion features conlpared to rubies lroim other As noted earlier, white mica, fuchsite, and Mg- natural sources, and so they are useful in separating chlorite were found as overgrowths on some rough Mong Hsu rubies from those froin other sources or Mong Hsu specimens. Mg-chlorite and white mica froill synthetic rubies (see also Laughter, 1993a and also were identified by Raman spectroscopy (based b; Smith and Surdez, 1994). These inclusions are on the reference work of Prieto et al., 1991)as inclu- best resolved using fiber-optic illumination. Two sions in the outerinost parts of some rough crystals types are: (figure 18). It appears that they are, in most cases, Whitish streamers that are oriented perpendicu- removed during preforming of the rough before heat lar to growth planes (figures 19-21). They usually treatment. extend from the outermost edge of the violet Whitish Particles. Various types of small particles core or lie in close proximity to that area. They ("whitish dust"]also are found in both the untreated Figure 20. A whitish streamer (left)eznerges from the end o/ CI pseuclosecondory leather of fluid Figure 19. Tl~isw11i f ish streamer in a11 rrntrented inclusiol~sa long the blrrck core o/ this unlrealed Mong Hsu rnby appears to originate from a solid Mong Hsu. View olnlost prrrallel to the c-oxis, inclusion located at the bozrndary betweell the fiber-optic illuminafion, tra~~sinif~edand rellected violet core rind red rim. Fiber-optic illzm~ination, light, magnified box; photomicrograph by E. reflected light, m~lgnified100x Giibelin.

GEMS & GEMOLOGY Spring 1995 t;ipre 21. Clouds of snowflalze-lilze parlicles Figure 22. Often tlle whitish dust-like particles appear in the upper right of this untreated Mong seen in untreated Mong Hsu rubies occurred in Hsu ruby, together wit11 a streamer that runs per- zones confined to growth planes. This type of pendicurlar to growth structures (top left) and par- iilclusion was no1 removed by heal treatment. ticles confined to s g~owthareas (bottom left). View almost parallel to the c-axis, fiber-optic illu- View almost parallel to the c-axis, fiber-optic illu- mina tion, reflected ligllt, magnified 6Ox. mination, reflected light, magnified 80x.

appear to be initiated by the trapping of solid (fig- sionally in growth zones confined to the basal face. lire 19) or fluid (figure 20) incl~~sions,which tend Note that dense zones of whitish particles were to be concentrated in the boundary zone also observed in greas confined to the violet or between, core and rim. These initial crystal almost black core. In extreme cases, the whitish defectscdo not coinpletely heal during subse- particles are so dense that the stone appears seini- quent growth, and new defects-which also act translucent, with large whitish reflecting areas, as traps for fluids-are coiltinuously formed. which inalzes it unsuitable for the jewelry marlzet. Ultimately, they appear as a series of lines con- In one heat-treated sample, we observed dense, ori- sisting of small reflecting particles, which are ented, needle-like particles. We also observed the oriented perpendicular to the growth planes of formation of such needles in another sample heat that particular zone. treated in Germany; they appeared in an area, con- Whitish dust, resembling clouds of snow- fined to the rim of the stone, that had originally flalzes (figure 21), frequently can be seen irreg- been transparent red, as well as in violet zones. We ularly dispersed in large zones of a crystal. do not yet know the exact nature of these needles. This type of inclusion is often confined to cer- Fluid Inclusions. In contrast to solid inclusions, tain growth zones (figure 22), usually related fluid inclusions were frequently seen in both to the w plane (figure 21). untreated and heat-treated Mong Hsu rubies, in Whitish particles of a con~pletelydifferent type are cores as well as rims. For the most part, they repre- formed by heat treatment. These particles form sent various primary fluids trapped in single cavi- dense areas, in some cases in w zones that grew ties as well as in pseudosecondary (figure 20) and after the originally violet core (figure 23)) and occa- secondary "feathers" or "fingerprints." They are

Figure 23. Violet to almost blaclz zones in the center of this Mong Hsu crystal in its natural state (right) were completely gone after heat I treatment (left), but whitish particles had formed in w growth zones outside the center. View almost parallel to tile c-axis, fiber-opiic illumination, reflected and transmitted light, magnified 60x. I

Mong Hsu Rubies GEMS 8: GEMOLOGY Spring 1995 17 the transparency of the stone. As demonstrated by heat-treatment experiments in Germany, no chem- icals are needed to remove the blue color coinpo- nent coinpletely from the Mong Hsu stones. As noted earlier, though, comlnercial treaters often use various chemicals, such as borax, during a second heating process to fill cracks and fissures exposed at the surface and thus enhance apparent clarity (Hughes, 1988; Peretti, 1993; Henn and Bank, 1993). Borax or similar substances call act as a flux to dis- solve alumina and can cause, at least partly, a recrystallization or healing of open fracture planes (Hinni, 1992; Koivula et al., 1993a; Milisenda and Henn, 1994).Mica or chlorite present in these open Figure 24. nts~llingfealhers ("fingerprints1l)-actu- cavities or fissures will dissolve in the presence of ally interconnec~ingtubes and isola~eddots of fluid inclusions-were commonly seen in both borax and form borosilicates that are then trapped heat-treated (as here) and untreated Mong Hsu as artificial glassy fillings. The formation of a crys- rubies. Immersion, magnified 60x. talline phase in fractures of treated Mong Hsu rubies has also been observed, and the compound was identified by means of X-ray powder diffraction related to the cracking of the ruby and subsequent healing by fluids. as aluminum borate (H. A. Hanni, pers. coinin., If this type of inclusion forms early in crystal 1994). We observed fissures and cavities containing growth close to the core, the outline of the core is these foreign fillers in many Mong Hsu samples sometimes seen parallel to the border of the "feather" obtained from the trade. In some treated rubies, we (figure20). Another "feather"in a heat-treated Mong saw flow structures in the glassy fillers of heavily Hsu ruby (figure 24) consists of isolated droplets included samples, which easily identified them as and interconnecting tubes. It formed later than the foreign material (figure 25). In contrast, only careful example illustrated in figure 20. microscopic examination revealed the presence of foreign substances in other, partly recrystallized Fracture Fillings Produced by Heat Treatment. fractures (figure 26). The exact composition of vari- Heat treatment may create additional fractures in ous filling materials, however, cannot be identified the ruby because of the decrepitation of solid mat- by microscopic examination. In areas in which the ter or fluids trapped in small cavities. This reduces filling material reaches the surface of the ruby, the filled fissure can be recognized by its reduced luster Figwe 25. Flow structures can be seen in this fis- sure, which has been partially filled with a solid Figure 26. In some partly recrystallized fractures, foreign material. During heat treatment for claril y it is difficult to locate the foreign substance (possi- enhancement, the host Mong Hsu ruby was bly borax or an aluminum borate) to which the placed in contact wilh borax or a similar sub- ruby was exposed during heat treatment. slance. Immersion, magnified 60%. Immersion, magnified 60%.

Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 TABLE 2. X-ray fluorescence analyses and refractive indices of Mong Hsu rubies.a

Untrealed rough Heal-Lreated faceted

Oxide A1203 99.300 Cr203 0.460 FeO 0.006 T102 0.107 "2O3 0.038 Ga203 0.010

Refractive indices A "0 -b "e

a The columns do not folal 100 ~wt.%due to lracos of IvlnO, K20, flAg0. CaO, and 80,.No lraces of CuO, NiO, Zr02, P205, or Nb205 were detecled. b No sharp shadow edges obsefved on the refraclometer.

compared to that of the polished ruby when viewed For the eight samples analyzed for which refrac- with reflected light (see, e.g., Kane, 1984; Scarratt tive indices could be measured on the refractolneter and Hardmg, 1984; Scarratt et al., 1986; I-hni, 1986). (again, see table 2), a reasonable correlation was found between R.I.'s and chromium concentrations, Chemical Properties. The methods applied to deter- and a good coincidence was obtained in a plot with mine the rubies' chemical properties provide ana- the sum of traci-element concentrations (calcu- lytical dara that represent different-size areas with- lated as Cr203 + Ti203 + V203 + Fe203 + Ga203] in the samples. X-ray fluorescence analysis reveals versus refractive indices (figure 27). Consecluently, an average composition of that part of the ruby (in an area that can be measured in millimeters) that was exposed to the X-ray beam, sucll as part of the table of a faceted stone. The electron microprobe analyzes areas with diameters in the micrometer Figlire 27. Refrmctive indices were correlated wiih chemical conlposition in ihe eight Mong Hsu range. Thus, traverses with several point analyses rubies analyzed by X-ray fluorescence for which across the polished surface of a ruby indicate the refractive indices could be measured (see table 2). chemical variability of a sample. For certain trace The correlations are expressed here as the sum o/ elements, such as gallium, the more sensitive X-ray trace-element concenlrtitions; red stluares = ordi- fluorescence analysis is required to obtain cluantita- nary ray (n,), bl~~clzsquares = extraordinary ray tive data. (n,). A distinct increase in refractive indices is The reliability of the data obtained for this caused by incre~isingtrace-element contents. study by X-ray fluorescence and electron micro- probe analyses is supported by the similarity in 1.775 8 I average trace-element concentrations reported by 1.774 - ¤ both methods. 1.773 - L 1.772 - X-r~lyFluorescence An~llysis.The results of X-ray 1.771 - fi 1.770 - 8 fluorescence analyses of various rubies (five U untreated and five heat treated) are shown in table 1.769- 5 1.768- + 2. The untreated rubies were con~posedof rough 2 1.767- fragments of crystals with plane polished faces. As 1.766 - . . a can be seen from table 2, the rubies contain signifi- 1.765 - cant trace-element concentrations of Cr203, FeO, 1.764 - Ti02, V2O3, and Ga203. Large differences in the 1.763 - trace-element concentrations ainong the various 1.762 - . 1.761 Il/IIIIIIII samples were measured, but we saw no significant 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50 differences in trace-element amounts between SUM OF TRACE ELEMENT CONCEIJTRATIONS (WTS) treated and untreated stones.

Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 19 variations in refractive index are related to varia- less extreme microscopic red color zoning, tions in trace-element contents. although the sample was faceted with its table perpendicular to the c-axis. Accordingly, the Electron Microprobe Analysis. As the results shown chemical variations were less pronounced than in table 3 indicate, systematic variations within the in other samples with a strong visual color zon- samples were observed for Ti02 and Cr203 and, in ing . some samples, for V203; only the statistical varia- tions of the analytical instrunlent were found for Type B: A distinct variation in Cr203 was found FeO and MnO. to correlate with a distinct Ti02 zoning in heat- On the basis of the relationships between treated samples 2 and 5 (figure 28B). High con- Cr203 and Ti02 (in combination with FeO), we centrations of Cr203 and Ti02 were restricted to have identified three basic types of chemical zoning the relatively intense red core. in Mong Hsu rubies to date (see also box B): Type C: A distinct zoning of Ti02 perpendic- Type A: A distinct variation in Cr203 that is not ular to the c-axis, inversely correlated to Cr203 correlated to titaniunl was found for samples 1, in the outer zones of the rim, was found in 3, and 4 (figure 28A). These rubies had high untreated sample 6 (figure 28C), which had a Cr203 concentratioils in the intense red zone homogeneous violet core and a light red rim. (that represents the core of the original crystal) Ti02 concentrations were higher in the core and lesser quantities of Cr203 in the lighter red than in the rim. Cr203 was higher in the outer- "rim" (samples 1 and 3).In sample 4, we found a most part of the rim. No systematic variations

TABLE 3. Electron microgrobe analyses of Mona Hsu rubies

Faceted sample number" Rough sample number"

Variable 1 2b 3" 4 5 6d 7e

Orientation Table Table Table Table Table Cut parallel Cut perp. parallel c-axis perp. c-axis perp, c-axis perp, c-axis parallel c-axis c-axis c-axis Visual appearance Dark red core, Dark red core, Dark red core, Red core, Dark red core, Dark violet core, Complex zoning lighter red rim lighter red rim lighter red rim lighter red rim lighter red rim lighter red rim dark violet core, light red rim Number of scans 3 4 4 4 2 5 3 7 Number of analyses 109 138 138 138 78 208 108 288 Direction of scans Perp, c-axis Perp. c-axis Perp. c-axis Pcrp, c-axis Perp. c-axis Perp, c-axis Parallel Perp, c-axis c-axis Analyses in wt.% (range) A1203 98.42-99.44 98.17-99.72 96.53-98.97 98.15-99.60 98.15-9966 98.33-99.78 98.40-99.90 98.45-99.91 Cr203 0.26-0.89 0.42- 1.05 0.90- 2.86 060- 0.89 0.39-0.98 0.40- 1.03 0.42- 0.69 0.40- 1.34 "24 0.00-0.13 0.000.11 0.00-0.12 0.00-0.07 0.00-0.10 0.00-0.08 0.00-0.07 0.00-0.08 Ti02 0.09- 0.28 0.05- 0.23 0.00 0.14 0.06- 0.27 0.04- 0.31 0.00- 0.38 0.04- 0.34 0.03- 0.51 FeO 0.00-0.03 0.00-0.02 O.O(r0.04 0.00-0.03 0.00-0.02 O.O(t0.04 0.00-0.03 0.00-0.04 MnO 0.00-0.03 0.00-0.03 0.00-0.02 0.00-0.02 0.00-0.04 0.00-0.03 0.00-0.02 0.00-0.03 Chemical zoning Cr Crand Ti Cr Cr Cr and Ti Ti1 Cr No distinct Complex zoning Area of high Core Core Core Core Core Corelouter rim zoning of Cr and Ti concentration Correlation No Yes No No Yes No No No between Ti and Cr

a All laceled saniples had been heal Ireald; bolh rough samples were unlrealed. Perp. =perpendicular b~eeligure 288. See ligure 286 See ligure 28A. see ligure A- 1.

20 Mong Hsu Rubies GEMS 8 GEMOLOGY Spring 1995 were found in this sample in a direction parallel In some samples, wc found a correlation to the c-axis. between V2O3 and Cr203; that is, high chromium and relatively high vanadium in the intense red In sample 7 (table 3), the complex growth and color cores, with lower chromiurn and vanadium in the zoning involving c, n, r, and w faces (figure A-1) lighter red rims. No correlation was seen between revealed an extremely complex zoning of both iron and titanium or between iron and chromium. Cr203 and TiOz that does not fit any of the simple Thus, it can be concluded that the color zoning types described above. correlates with a systematic variation in Cr203 and/or Ti02 (see also box B]: The violet (untreated) and intense red (heat-treated)cores of Mong Hsu rubies have significantly higher Cr203 concentra- 3 0- tions than the lighter red "rim" layers around them. A We also identified zones with high Ti02 concentra- 2 5- tions (relative to the rim portion) in the violet s (untreated] and red (heat-treated]cores of some -z 20- ..q,..,. ..q,..,. Mong Hsu rubies (type C). In some san~ples,we 0 k 1.5- found both high Cr203 and high Ti02 concentra- k tions in the core. 0e g 10- [\t-.)m.==A=*,. In summary, it appears that the main chemical zoning of Mong Iisu rubies, between crystal corc a o 5 and riin zones, is due to greater amounts of chromi- um and/or titaniunl in the areas confined to the I 5 10 15 20 25 30 35 40 intense red or violet cores. Between distinct zones SCAN POINTS within the rims, chronliun~and/or titanium values also may vary (see box B, as well as figures 13 and B 28C). 1 0- The trace-element concentratioils of Mong HSLI ae- rubies can be compared to those of natural rubies -Z 08- from other deposits of commercial importance, I +0 including Luc Yen (Vietnam), Morogoro (Tanzania], 06- \ I ,VLt- Mogolt (Myanmar),Kenya, Sri Lanlta, Malawi, and 8 .'= Thailand (Harder, 1969; Schmetzcr, 1986b; Bank et 04- .;;".I ~h...m=.h. al., 1988; Tang et al., 1988, 1989; Hanni and Schnletzer, 1991; Kane et al., 1991; Dele-Dubois et

0 1 1 I T- O 5 10 15 20 25 30 35 40 Figure 28. These three electron microprobe tra- SCAN POINTS verses reveal the differences in Cr203contents (blacli squares) and Ti02 contents (red squares) C between core and rim areas of three Mong Hsu 1 0- rubies. Salnples A (no. 3, toble 3) and B (no. 2, trrble 3) are both f~celecl~n~l heal-treote~l; sanq~lc :-0I 08-;, C (no. 6, table 3) n an untreated polished platelet.

I- Sample A shows hlgl~erCr203 in the core than tl~e 2 06- -\ I- rim, and no Ti02zoning. Sample 6 shows Cr203 Z LU concentrations correlated wth a distillct Ti02 0 04- "-v'.-g?rvAf I zonlng, wth highel Cr2On and TiO2 col~tentsin the core and lowel VOILICS in the rlnl S~mpleC 0 2- reveals increased T102 in the core, increasing 1 Cr203 toward the outer zones of the riin, and an 0- I 1 I I I I inverse correlation between C%03 zoning ond 0 5 10 15 20 25 30 35 40 SCAN POINTS variatio~ain TiO? All scans were perforlnetl in a direction perpen~llc~rlnrto the c-0x1s

Mong Hsu Rubies GEMS rk GEMOLOGY Spring I995 21 Although there may be some overlaps with other WAVELENGTH (nm) natural rubies originating from marble deposits, 400 500 600 700 trace-element contents of individual samples, in coinbination with other characteristics, can also be helpful for locality determination. Spectroscopic Features. Visible and Ultraviolet Speclroscopy. Absorption spectra of heat-treated Mong Hsu rubies and of the outer rims of untreated specimens were typical of those seen in low-iron rubies, with no iron-related absorption at 450 nm. Absorption characteristics in the ultraviolet were siinilar to those of iron-poor rubies from marble- 25 20 15 WAVENUMBER (1000 cm 'I) type deposits (Bosshart, 1982; Smith and Surdez, 1994). The spectra of untreated samples with violet Figure 29. Tl~eseabsorption spectra were recorded cores revealed additional absorption features in the in three Mong Hszr rubies. The top is fron~a heat- treated faceted stone with its table oriented paral- red and yellow area, between about 800 nm and the lel to the c-axis, 2 3 mm thick (sanlple I, table 3). broad chromiunl absorption band in the green The middle is from the violet core of an zln~reated range, which are superimposed on the ruby absoip- slice (about 2.5 mm thick) cut parallel to the c- tions (figure 29). These absorption features are axis (sample 6, table 3). The bottom is froin an described as: untreated slice (abozlt 1 mnl thick) cut perpendic- a broad absorption in the spectninl parallel and ular to the c-axis with violet core and red rim (see perpendicular to the c-axis, rangng from about sample 7, table 3, figure A-I). In the red LO yellow 800 nin to about 550 nnl without a distinct spectral range of untreated samples, (1 broad absorption without distinct nlaximzlm and an absorption maximum, and absorption band at 6 75 nm are szlperimposed on a polarized absorption band in the spectmm per- the chromium (ruby)spectrum. pendicular to the c-axis, with a maxiinuin at 675 nm, that is, in the range of the well-lcnown chromium lines at 693,669, and 659 nm. al., 1993).The Cr203 concentrations in Mong Hsu rubies can be extremely high. The FeO concentra- In different violet samples measured, the relative tions are relatively low compared to those of rubies intensities of the bands in the red to yellow arca from other marble-type deposits such as Morogoro vary. In some, the 675-nm band was wealcer than or Luc Yen. The upper end of the Ti02 range is the 693-nm Cr3+ absorption line; in others, the 675- much higher than the values found in rubies froin nin absorption, which is strongly polarized, exceed- these same occurrences. The V203 concentrations ed this well-ltnown chroinium absorption in inten- are similar to those reported for rubies froin Mogolc, sity (figure 29). but can be much higher than those published for These additional absorption features in the red, nibies from some other marble-type deposits, such particularly the broad absorption, are responsible as Morogoro and Luc Yen. The range of FeO, Ti02 for the color in the cores, that is, light violet parallel and V2O3 concentrations found in Mong Hsu rubies to the c-axis and intense violet perpendicular to c. is different from those of rubies from basaltic roclts Thus, the violet color in Mong Hsu rubies is caused (e.g., Thailand). by the superilnposition of a red (ruby) component The combination of relatively high Cr203, caused by cl~romiuinand a blue component that is Ti02, and V2O3 along with relatively low concen- removed by heat treatment. The blue component trations of FeO has so far not been reported for any consists of two different absorption features, which of the various types of synthetic rubies (Tang et al., are more intense in the spectrum perpendicular to 1989; Muhlmeister and Ilevouard, 1991; Peretti and the c-axis. Smith, 1993; HBnni et al., 1994). Violet sapphires and purplish red rubies are In summary, the trace-element patterns of known from various localities, such as Ratnapura, Mong Hsu rubies are useful to distinguish faceted Sri Lanlca, and Umba, Tanzania. In all types of vio- samples from their synthetic counterparts. let samples froill various occurrences, the blue

Mong Hsu Rubies GElMS & GElMOLOGY Spring 1995 component of the violet color is due to a broad chromium and relatively low manganese concen- absorption band in the red area, which can be at trations (see tables 2 and 31, it is conceivable that least partially removed by heat treatment, cven at only the small 675-nin absorption band close to the relatively low temperatures such as 1000" or small chromium fluorescence line at 693 11111 will 1200°C, which have been used historically for the be observed. This assignment, however, needs fur- heat-treatment of corundum (Bauer and Schloss- ther experimental research for confirmation, espe- macher, 1932). cially for the explanation of a possible stabilization An extremely broad absorption band in thc red- mechanism by charge compensation and its reac- to-green spcctral region (betwecn 800 and 500 nm) tion to heat-treatment. has been seen in blue sapphires from various locali- In summary, the spectra of the violet cores in ties and is generally assibmed to an FeZ+/Ti4+charge- ~MoilgHsu rubies consist of the well-lznown absorp- transfer absorption. The blue color in this type of tion lines and absorption bands of C13+ in corun- low-iron sapphire (i.e., without a specific Fe2+/~e3+ dum, on which are superimposed a broad Fe2+/Ti4+ absorption band in the near infrared) can also be charge-transfer absorption and an additional line at partially removed by heat treatment at low temper- 675 nm, the nature of which is not yet lznown. The atures; some of these sapphires turn colorless when influence of one or both additional absorption fea- heated (Schmetzer and Bank, 1980).If the iron-tita- tures in the red on the violet color of different cores nium charge-transfer al7sorption is superimposed on varies, and additional research is necessary (for a ruby spectrum, the color of the sample is altered example, by a combination of microscope absorp- from red to purplish red, purple, or violet, according tion spectroscopy and microprobe analyses) to to the relative intensity of the two color-causing understand the cause of color in these highly zoned components in the spectrunl (Schmetzer and Bank, samples. 1981). Given the spectral characteristics of Mong Hsu Infrared Spectroscopy. IR spectroscopy of transpar- rubies and their titanium zoning, the additional ent, heat-treated samples revealed spectra with broad absorption in the rcd and the resulting violet sharp absorption lines in the 3000 to 3500 cm-I color of the cores of these rubies is consistent with range-either with two maxima, at 3233 and 3310 the presence of such a blue sapphire component. cm-1 (flgure301, or with one n~axhnunlat 3310 cm-1- In the absorption spectruin of untreated Mong or spectra with a complete absence of infrared Hsu samples with violet cores, however, we also absorptions in the range mentioned. Smith and observed an additional polarized absorption band at 675 nm (figure 29), which is not fully understood at present. In soine of the samples, this absorption Figzire 30. This infrared spectrum of (1 lle(11-treated band exceeded the 693-nm chromium line in inten- Mong Hsu ruby reveals sharp absorption lines at sity. The only correlation found in the literature for 3233 und 3310 cin 1, which are assigned to OH- stretching vibrations. These lines are chal-acteris- this band is an absorption in the spectruin of Mn4+ tic for OH groups in rubies, which are related to in corundum (Geschwind et al., 1962; Crozier, str~lcturaldefects. Such lines are not found in 1965; Potnau and Adde, 1976).Mn4+ appears in syn- flux-grown syn~heticmbies. thetic flux-grown and Verneuil-grown corundum crystals, which are doped by manganese, with MgO added for charge compensation. Mn4+ is isoelec- tronic with Cr3+ and has three d electrons. It reveals, 0.771 - in addition to the sharp 675-nm band, a broad 0.60 - P absorption at about 470 nm and an absorption edge 0W - 22 - in the ultraviolet range, the low-energy tail of a 0.40 - 0 - which extends to the visible area, causing an allnost cn - continuously increasing absorption from about 500 ma - nm toward smaller wavelengths. If a broad absoip- 0.20 - tion band of Mn4+ at 470 nm is present in the spectnun of chromium-bearing conindum, this absorption is, -0.015~ll111111,1,,11, no st probably, hidden between the two doininant 4000 3500 3000 2500 color-causing chromium absorption bands. Thus, in WAVENUMBER (cm') a sample of NIong Hsu niby with high amounts of

~MongHsu Rubics GEMS & GEMOLOGY Spring 1995 2 3 Surdez (1994)reported seven sharp bands of varying A distinct color zoning confined to specific intensity at 3189 (wealz),3233 (medium),3299 (very growth stnictures, with one or two violet "cores" weal<),3310 (strong), 3368 (very wealz), 3380 (very surrounded by a red "rim." weal<),and 3393 cm-1 (very weal<)in the absorption Spectroscopic features in the red to yellow por- spectra of Mong Hsu rubies. Similar features have tion of the visible spectrum, with absorption been found in the infrared spectra of Verneuil- bands that are reinoved during heat treatment to grown synthetic rubies and sapphires that were change the cores froni violet to red. doped with vario~istrace elements in different con- The presence of whitish particles in certain centrations and, at least partly, annealed in a hyclro- growth zones, formed as a result of heat treat- gen atinosphere at high ten~peratures(Borer et al., ment. 1970; Eigenniann and Giinthard, 1971; Eigenniann et al., 1972; Bluin et al., 1973; Volynets et al., 1972, We found that the color zoning in our samples is 1974; Beran, 1991; Moon and Phillips, 1994). closely related to a con~plexchemical zoning con- Consequently, tlie sharp absorption bands in the fined to growth layers formed parallel to the basal infrared spectra of Mong Hsu rubies are assigned to pinacoid, to the positive rhombohedron, and to two OH-stretching vibrations and indicate that hydrox- hexagonal dipyramids. We observed a distinct yl groups have been incorporated in the crystal growth sequence whereby red and violet areas structure of soine heat-treated samples. fornled in various growth cycles, with a spccific Smith (1995) recorded an absorption spectrum habit change between different growth zones. The in soine untreated samples that consisted of several variation in physical properties, such as refractive broad absorption bands. He assigned this spectrunl indices, is closely related to the chemical composi- to microscopic or subi~iicroscopicinclusions of tion of the samples. Although those properties of diaspore, AlO(0H). the crystals that are related to different growth con- OH-stretching vibrations were also ineasured ditions during the forillation of the rubies-that is, previously in the infrared spectra of a few untreated growth zoning, color zoning, and chemical zoning ruby and sapphire samples froni Sri Lanlza related to temperature and/or pressure and/or (Schinetzel; unpublished), in a ruby froin Sri Lanlta chemical composition of tlie environment-are and a blue sapphire from Montana (Beran, 1991), well understood. Only preliminary n1odels are and in untreated blue Australian sapphires (Moon presently available to provide a detailed explanatioil and Phillips, 1994). They were not foulid in tlie of the cause of successive growth cycles (Peretti and spectra of flux-grown synthetic rubies (Belt, 1967; Mouawad, 1994). Volynets et al., 1972; Peretti and Smith, 1994). Nor is thcre a comprehensive explanation for For practical gen~ology,the presence of OH- those properties of Mong Hsu rubies that change related absorptioil lines in the infrared spectrum of with heat treatment. Lilcewise, no model is avail- an unlznown ntby indicates that the sample is not a able that can explain all features related to thc flux-grown synthetic ruby, although it may be a change in W-visible and 1K spectroscopic proper- Verneuil-grown or hydrothermally grown synthetic ties, which are closely related to the color change sample (either of which is illore readily identifiable and possibly also to the forination of the whitish froin natural rubies than the flux-grown material) particles. or a natural stone. For discussion of the difference The distinctive properties of Mong I-isu ntbies between Vemeuil- and hydrothermally grown syn- are useful in separating faceted saiiiples froiii their thetic ruby, see Belt (1967),Beran (1991),and Peretti synthetic counterparts and also in establishing the and Smith (1993, 1994). locality of origin. The most prominent diagnostic properties of faceted, heat-treated Moiig HSU rubies SUMMARY AND CONCLUSION require careful inicroscopic examination, using Large cluantities of rubies froin the new deposit at immersion techniques in conjunction with fiber- Mong Hsu have been widely available since 1992. optic illumination, Key features include growth Most are heat-treated before they enter the jewelry structures confined to a distinct color zoning trade. Mong Hsu rubies are easily recognized by between cores and rims; different types of whitish their distinctive n~icroscopicproperties. They have particles and whitish streamers are also of diagnos- a number of features that thus far have not been tic value. Specialized laboratory techniclues, such as reported for nibies froni other occurrences. These XRF analysis and IR spectroscopy, provide addition- include: al diagnostic information. Probleins for the trade

24 Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 arise, however, from the large nuillbers of stones I(. V. Gems Co., Bangltok; Dr. H.A. Hiinni, SSEF (Basel) with fractures that appear to have been filled with a and University of Basel, Switzerland; R. E. Icane, Helena, foreign material, especially partially healed frac- Monlona; H.M. Graf, Grafgen~,Wintertl~ur, Switzerland; tures with glassy and/or crystalline fillers. and R. Bicler, Ernst Fiirber Company, Munich, Gcrntany. For technical assis~ance,we are gratefirl 1-0 C. P. Smith, Giihclin Cemn~ologicalLaboratory, Lucerne, Switzer- land (infrared spectroscopy); Prof. Dr. W. Stern, Geo- Aclznowledginents: The aiithors ore grntef~ilto the fol- cheinical Lahoralory, U~~iversityof Basel, Switzerl~nd lowing persons or companies who kindly submitted (XRF-~malysis);Dr. I. Dubessy, CREGIJ, Vandoeuvre- Mong Hsu nrbies for the present investigation: Mo~lawad Les-Nancy, France (Roman speclroscopy); Pro/. Dr. R. Bangltolz Co., Bangltok, ?hail(~nd;13. Ho, Asian Institute Guggenheim and M. Diiggelin, Institute of Geology, of Gen~mologicalSciences (AIGS), Bangkok; Myanma University of Basel, Switzerland (electron microscopy); Gems Enterprise, Yangon, Myanmar; R. I

Glass filled rubies increasing (1994). lewellery News Asia, REFERENCES No. 119, July, pp. 66, 68, 70. Hanni H.A. (1986) Behandelte Korunde nit glasartigcn Banlt H., Hcnn U., Lind Th. (1988) Rubine aus Malawi. Fullungcn. Zeitschrlft der De~rtschenGen~mologischen Zeitschrift der Deirtschen Gen~mologischenCesellschnft, Gesellschnft, Vol. 35, No. 314, pp. 87-96. Vol. 37, No. 314, pp. 1 13-1 19. Hanni H.A. (1992) ldcntification of fissure-treated gem- Baucr M., Schlossinacher K. (1932)Edelsteinlz~rr~de. 3. Aufl., stones. lournnl of Gemlnology, Vol. 23, No. 4, pp. Bernharcl Tauchnitz, Leipzig, p. 209, 498. 201-205. Bclt R.F. (1967) Hydrother~nalruby: Infrared spectra and X- Hanni H.A., Schmetzcr I<. (1991) New rubies from the ray topography, lorrrntrl of Applied Physics, Vol. 38, No. Morogoro area, Tanzania. Gems d Gemology, Vol. 27, 6, pp. 2688-2689. No. 3, pp. 156-167. Beran A. 0991) Trace hydrogen in Verneuil-grown corundum Hinni H.A., Schmetzer I<., Bernhardt H.-J. (1994) Synthctic and its color varieties-An IR spcctroscopic study. rubies by Douros; A ncw challenge for gemologists. Genls E~rropeonlourno1 of Milleralogy, Vol. 3, pp. 971-975. d Gemology, Vol. 30, No. 2, pp. 72-86. Blu~nH., Frey R., Giinthard Hs.IH., Ha Tae-Kyu (1973) Ab Harder H. (1969) Farbgebendc Spure~lcleme~ltein natfirlichen initio scf study of OHO" systein and its relation to the I

Moi-rg Hsu Rubies GEMS & GEMOLOGY Spring 1995 25 Kane R.E., McClure S.F., Iamethyst. Iournal of Gemmology, Koivula J.I., Icorundums. Neucs lahrbuch fiir Minerologie leweldiam, Vol. 4, No. 5, pp. 38-41. Abhandlungc~~,Vol. 139, No. 2, pp. 216-225. ~MilisendaC.C., Henn U. (1994) Neues Rubinvorltommen in Sclimetzer I<., Bank H. (1981)The color of natural corundum. Myanmar (Burma]. Goldschmiede und Uhrmacher Neues lahrbrrch fir Mineralogie Monatsheftc, Vol. 1981, Zeit~~ng,Vol. 92, No. 4, pp. 147-148. No. 2, pp. 59-68. Moon A.R., Phillips M.R. (1994) Defect clustering and color Smith C.P. 119951 A contribution to understanding the in Fe,Ti:alpha-Al,03. lourno1 of the American Ceromic infrared 'spec;ra of Mong Hsu rubies, lournil of Society, Vol. 77, No. 2, pp. 356-367. Gernrnology, Vol. 24, No. 5, pp. 321-335. Muhlmeister S., Devouard B. (1991)Determining the natural Smith C.P., Surdez N. (1994)The Mong Hsu ruby: a new type or synthetic origin of rubies using energy-dispersive X-ray of Burmese ruby. /ewelSiam, Vol. 4, No. 6, pp. 82-98. fluorescence (EDXRF). In A.S. Keller, Ed., Proceedings of Special report: Mong Hsu ruby fact sheet (1993).leweldion~, the International Gemological Symposiun~1991, Vol. 4, No. 5, p. 33. Gemological Institute of America, Santa Monica, CA, pp. Tang S.M., Tang S.H., Tang, T.S., Retty A.T. (1988) Analysis 139-140. of Burmese and Thai rubies by PIXE. Applied Peretti A. (1993) Foreign substances in Mong Hsu rubies. Spectroscopy, Vol. 42, No. 1, pp. 4448. /ewelSiom, Vol. 4, No. 5, p. 42. Tang S.lM., Tang S.H., Mok I<.F., Retty A.T., Tay T.S. (1989) Peretti A., Smith C.P. (1993)A new type of synthetic ruby on A study of natural and synthetic rubies by PIXE. Applied the market: offered as hydrothermal rubies from Spectroscopy, Vol. 43, No. 2, pp. 219-223. Novosibirsk. Australian Gemmologist, Vol. 18, No. 5, pp. Volynets F.K., Sidorova E.A., Stsepuro N.A. (1974) OH groups 149-157. in corundum crystals which were grown with the Peretti A., Smith C.P. (1994) Letter to the editor. Iournal of Verneille technique. Iournol of Applied Spectroscopy, Vol. Gemn7ology, Vol. 24, No. 1, pp. 61-63. 17, pp. 1626-1628. Peretti A., Mouawad F. (1994) Fluorite inclusions in Mong Volynets F.K., Vorob'ev V.G., Sidorova E.A. 11972) Infrared Hsu ruby. IewelSiom, Vol. 5, No. 4, pp. 136137. absorption bands in corundum crystals. Iournol of Applied Potnau J., Adde R. (1976)Crystalline field parameters of Crz+ Spectroscopy, Vol. 10, pp. 665-667. and Cr44 in corundum. lournal de Physique, Vol. 37, pp. 603-6 10.

26 Mong Hsu Rubies GEMS & GEMOLOGY Spring 1995 BACK ISSUES OF Spring 1987 Fall 1991 "Modern" Jewelry: Retro to Abstract Rubies and Fancy Sapphires from Vietnam Infrared Spectroscopy in Gem Identification New Rubies from Morogoro, Tanzania A Study ol the General Electric Synthetic Jadeite Bohemian Garnet-Today Iridescent Orthoamphibole from Greenland Winter 1991 Summer 1987 Marine Mining of Diamonds off Southern Africa Gemstone Durability: Design to Display Sunstone Labradorite from the Ponderosa Mine Wessels Mine Sugilite Nontraditional Gemstone Cutting Three Notable Fancy-Color Diamonds Nontransparent "CZ from Russia The Separation of Natural lrom Synlhetic Limited quantities of these issues are still available Spring 1992 Emeralds by lnfrarcd Spectroscopy Gem-Quality Green Zoisile The Rutilaled Topaz Misnomer Kilbourne Hole Peridot Fall 1987 Fluid Inclusion Study of Querktaro Opal An Update on Color in Gems. Part I Natural-Color Noncondudive Gray-to-Blue Diamonds The Lennix Synthetic Emerald Peridot as an Interplanetary Gemstone Kyocera Corp. Products that Show Play-of-Color Summer 1992 Man-Made Jewelry Malachile Gem Wealth of Tanzania lnamori Synthetic Cat's-Eye Alexandrite Gamma-Ray Speclroscopy and Radioactivity Winter 1987 Dyed Natural Corundum as a Ruby Imitation The De Beers Gem-Quality Synthetic Diamonds An Update on Sumitomo Synthetic Diamonds Queen Conch "Pearls' The Seven Types ol Yellow Sapphire and Their Fa11 1992 Stabilily to Light Ruby and Sapphire Mining in Mogok Bleached and Polymer-Impregnated Jadeite Summer 1988 Radiation-Induced Yellow-Green in Garnet The Diamond Deposits of Kalimantan. Borneo An Update on Color in Gems. Part 3 Winter 1992 Pastel Pyropes Determining the Gold Content of Jewelry Metals Three-Phase lnclusions in Sapphires from Sri Lanka Diamond Sources and Production Sapphires from Changle, China Fall 1988 An Economic Review 01 Diamonds Spring 1993 The Sapphires of Penglai, Hainan Island, China Queensland Boulder Opal Iridescent Orthoamphibole lrom Wyoming Update on Diffusion-Treated Corundum: Detection of Treatment in Two Green Diamonds Red and Other Colors Winter 1988 A New Gem Beryl Locality: Luumaki, Finland Gemstone Irradiation and Radioactivity De Beers Near Colorless-to-Blue Experimental Aniethyst from Brazil Gem-Quality Synthelic Diamonds Opal from Opal Butte, Oregon Summer 1993 Kyocera's Synthetic Slar Ruby Flux-Grown Synthetic Red and Blue Spinels Spring 1989 lrom Russia The Sinkankas Library Emeralds and Green Beryls of Upper Egypt The Gujar Killi Emerald Deposit Reactor-Irradiated Green Topaz Beryl Gem Nodules from the Bananal Mine Fa11 1993 "Opalite:" Plastic Imitation Opal Jewels of the Edwardians Summer 1989 A Guide Map to the Gem Deposits of Sri Lanka Filled Diamonds Two Treated-Color Synthetic Red Diamonds Synthetic Diamond Thin Films Two Near-Colorless General Electric Type Ila Fall 159 Fall 1M Fdl199J Grading the Hope Diamond Synthetic Diamond Crystals Diamonds with Color-Zoned Pavilions Winter 1993 Fa11 1989 Russian Gem-Quality Synthetic Yellow Diamonds Polynesian Black Pearls Heat Treating Rock Creek (Montana) Sapphires The Capoeirana Emerald Deposit Garnets from Allay, China Brazil-Twinned Synlhetic Quartz Spring 1994 Thermal Alteration of Inclusions in Rutilated Topaz The Anahi Ametrine Mine, Bolivia Chicken-Blood Stone from China lndaia Sapphire Deposits of Minas Gerais, Brazil Winter 1989 Flux-Induced Fingerprints in Synthetic Ruby Emerald and Gold&ures of the Atocha Summer 1994 Zircon frordi Range, Australia Synthetic Rubies by Douros ue 4 Emeralds from the Mananjary Region, ~ipectroscopyin Gemology Madagascar: Internal Features Mildly I Rhinestones Synthetic Forsterite and Synthetic Peridot Spring 1990 Complete your back issues of Update on Mining Rubies and Fancy Sapphires in Gem Localities of the 1980s Gems & Gemology NOW! Northern Vietnam Gemstone Enhancement and Its Detection Fall 1994 Synthetic Gem Materials of the 1980s Single Issues' $ 9.95 ea. U.S. $ 14.00 ea. elsewhere Filled Diamonds: ldenlification and Durabilily New Technologies of the 80s Complele Volumes:* Inclusions of Native Copper and Tenorite in Winter 1990 1986,1987,1989, Cuprian-Elbaite Tourmaline, Paraiba, Brazil The Dresden Green Diamond 1991, 1992, 1993, 1994 $ 36.00 ea. vol. U.S. $ 45.00 ea. vol. elsewhere Winter 1994 Identification of Kashmir Sapphires Three-year set $ 99.00 U.S. $125.00 elsewhere Color Grading of Colored Diamonds in the GIA A Suite of Black Diamond Jewelry Five-year set $160.00 U.S. $200.00 elsewhere Gem Trade Laboratory Emeraldolite Ruby and Sapphire from the Ural Mounlains, Russia 1 .I08 discounl lor GIAAnnul Fund Qnors al Ihe Doasl8t.s Circle level and absve. Spring 1991 I Gem Corundum in Alkali Basalt Age. Origin, and Emplacement of Diamonds Emeralds of Panjshir Valley, Afghanislan D~WIteUtra(111)42l-12#,ela.PIPr Some issues from the 1984-1986 Summer 1991 p1q Iw-m,rrt. 110 volume years are also available. Please call the Fracture Filling of Emeralds: Opticon and "Oils" ( (310) 453-4478 OR WRITE: G&G Ssubscmtions illA Subscriptions Oflice lor details. Emeralds from the Ural Mountains, USSR Trealed Andamooka Matrix Opal ORDER NOW! By Keith A. Ivlychaluk

Yogo Gzzlch, discovered more than 100 years he state of Montana, in the western United States, ago, is one offour mnjor sapplke-producing hosts several large and economically important sap- areas in Montana, United States. Yogo sap- phire deposits. The four major deposits-at Yogo Gulch, phires are lznown for their uniform, well-satu- Wssouri River, Rock Creek, and Dry Cottonwood Creelz- rated blue color; relative absence of inclusions were all discovered in the late 1800s by gold prospectors. and zonationj and high luster and brilliance in Three of the deposits are large secondary (placer)occur- both artificial and nntwal light; they do not rences for which the original source roclzs have not been reqzrire heat treatment. Rozzgh crystals, usual- ly flat with low cutting retention, generally conclusively identified. At Yogo Gulch (henceforth referred weigh less than one carat (but have been to as Yogo), however, the sapphires are found in sitrz in a reported up to 19 ct). Unlike the other lamprophyre dilze system, with only minor related placer Montana deposits, which are secondruy, Yogo deposits. Yogo is one of the few major sapphire mines sapphires are mined directly from a lampro- worldwide where sapphires are extracted directly from their phyre host roclc. There are at least six known host roclz. dikes, five sapphire-bearing, at Yogo. From Not only are Yogo sapphires unique among Montana 1895 to 1994, the Yogo deposit produced nn sapphires in terms of deposit type, but they also exhibit dis- mtimated 18.2 million carats of rough that are tinctive gemological features. The vast majority of Yogo believed to have yielded more than 500,000 sapphires are naturally the same uniformly saturated blue carats of cut stones. Considerable reserves (figure I), whereas sapphires from the placer deposits of remain. western Montana are predominantly pale green, blue, or yellow before heat treatment. The relative absence of inclu- sions, fractures, and color zoning also distinguishes Yogo sapphires from other Montana sapphires. Unlilze sapphire-producing areas elsewhere in the state, which have received little geologic investigation since the significant contribution made by Clabaugh (1952), Yogo Gulch has been studied since the late 1970s by various geol- ogists (see, e.g., Meyer and Mitchell, 1988; Brownlow and Komorowslzi, 1988; Dahy, 1988, 1991; and Balzer, 1994). The author was the first geologist to study the Vortex mine, an extension of the Yogo sapphire deposit developed in 1987 (Mychalulz, 1992). Between 1990 and 1992, the author collected and studied 200 rock and 100 rough sap- phire samples from the Vortex mine and tailings from the nearby English and American mines. Representative sam- ples of various roclz types, and the minerals separated from them, were examined by thin-section microscopy and X-ray diffraction analysis. The author also analyzed core-drilling logs and samples from drill holes made by Vortex Mining Company in 1993. Further, the author visited the Vortex mine many times between 1990 and 1995, and was the first

Yogo Sapphire Deposit GEMS & GEMOLOGY Spring 1995 Figure I. The sapphires from the historic deposit at Yogo Gulch, Montana, are noted for their deep, uniforn color and their brilliance when cut. Unlilie most sapphires on the murket today, they are not heat treated. Typically, thozrgh, they are small he rough aver- anes- less than 1 ct). The loose Yogo sapphires illus- trated here weigh 0.93- 1.34 ct; they and the pin set with Yogo sapphires Lu'e courtesy of Mac Madex and American Gem Corp. The Yogo sapphire in the ring weighs 1.04 ct; it is courtesy of Robert E. IZone. Photo 0 Horold ed Erica Van Pelt.

geologist to view the lowest worlzings of the 61-111- of the Little Belt Mountains; it is about 33 lzln (20 deep mine in 1994. miles) from the historic lead-zinc-silver mining This article reviews the history of the sapphire camps of Neihart and Hughesville. The mine is deposit at Yogo Gulch and the distinctive gemologi- accessible all year by U.S. Forest Service gravel cal properties of the sapphires. The geology and roads leading south froill Utica. Utica itself is locat- occurrence of the sapphires is discussed, iilcludiilg ed on Montana Route 239, approxilnately 28 1zn1 exalnination of various theories as to how they (17 miles) southeast of Stanford, the county seat, were formed and einplaced. Attention is also given and roughly 125 kin (78 miles] southeast of Great to historical production and ore-grade variability of Falls. the entire Yogo deposit. The eastern portion of the deposit is located in grass-covered rolling hills, whereas the western por- LOCATION AND ACCESS tion is situated in rugged and heavily forested ter- The Yogo Gulch sapphire deposit is located about rain. Yogo Creek, in Yogo Gulch, flows across the 25 lzm (15 miles) southwest of Utica, Montana, in deposit in the west and has carved a canyon Judith Basin County (figure 2). The deposit lies through ancient limestone fonnations. Yogo Creek within the legal land description of T.13.N1 R.ll.E, is a tributary of the Judith River, which in turn sections 20 through 24, on the northeastern flanlzs flows into the Missouri River.

Yogo Sapphire Deposit GEMS & GEMOLOGY Spring 1995 29 Figure 2. This map of Montana shows the loco- tions of the state's lour major sapphi~e-producing localities: (I) Yogo Gulch, (2) Missouri River (Eldor- ado, Spoltane, French, Gruell's, Cheyenne, Dana's, and American bars), (3)Rock Creek, (4) Dry Cotton wood Creek. Modified from Zeihen, 1987.

WYOMING 1 I

HISTORY OF MINING AT YOGO GULCH carat for first-quality stones, $1.25 pcr carat for sec- Sapphires were first discovered in Montana-in the ond quality, and $0.25 per carat for everything else Missouri River near Helena--on May 5, 1865, by Ed (Voyniclz, I987a). Tiffany & Co. eventually bccalne Collins. Gold prospectors later found sapphires in an important buyer of Yogo sapphires and manufac- Dry Cottonwood Creelz north of Butte in 1889, and tured some important pieces of reportedly Yogo in Rock Creelz near Philipsburg in 1892 (Clabaugh, sapphire jewelry, including the fainous Iris Rrooch 1952; again, see figure 2). In general, the sapphires (figure 3). fro111 these three deposits proved to be of low-satu- In 1896, Jim Ettien, a local sheep herdcr, discov- rated colors; natural, deep blue stones were rare. ered several hundred carats of sapphires around bad- However, the last significant sapphire discovery in ger and gopher burrows aligned along a lincar Montana, at Yogo Gulch in 1895, yielded superb depression that marlzed the surface exposure of an blue stones. The history and general geology of igneous dike (now called the "A" dilze; see figure 4) Yogo are discussed in detail by Clabaugh (1952)and on the flatlands above Yogo Creelz (Weed, 1899).He Voyniclz (1987a). Except as noted, the author has stalzed the first claims to the dike that year. Soon drawn the following account from these two others, notably John Burlze and Pat Swceney, stalzed authors. more claims to the dilze. By 1897, Hoover and his Placer gold was discovered by prospectors in partners had bought out Ettien's claims and formed the upper end of Yogo Creelz in 1866, but the area the New Mine Sapphire Syndicate. was not given serious attention until 1878. Several I11 1898, London gein merchants Johnson, nlore years passed before, in 1895 gold prospector Wallzer and Tolhurst Ltd., acquired the majority Jake Hoover began to collect the translucent blue interest in the New Minc Sapphire Syndicate and stones that were being trapped in his sluices in began an intensive mining and marketing effort. lower Yogo Creelz. He sent samples to an assay They initially concentrated on removing sapphire- office, which then forwarded them to Dr. George F. bearing ore from surface outcrops of the middle sec- Kunz, at Tiffany & Co, in New Yorlz City tion of the A dike by means of hydraulic mining; (Clabaugh, 1952). The stones were identified as sap- this area is now referred to as the English Cut (fig- phires of unusual quality, and Kunz sent Hoover ure 5). By 1902, the syndicate had two underground and his partners a check in the amount of $3,750 operations, the English (or British) mine and the (about twice what his gold operation had paid). This Middle mine, working the dilte. figure was based on the current London prices paid The Anlerican Sapphirc Company of New Yorlz for rough sapphires from Southeast Asia: $6 per began operations on the western third of the A dilze

Yogo Sapphire Deposit Spring 1995 by purchasing Burlze and Sweeney's clainls in 1904; it opeiled the American mine-also underground- in 1905 (Clabaugh, 1952). In 1909, its capital exhausted, the firm reorganized as the Yogo Ainericail Sapphire Company. The New Mine Sapphire Syndicate acquired Yogo American in 1914. Although the syndicate never reopened the Aillerican mine, they recovered the $80,000 pur- chase price under the supervision of the company's renowned mine manager, Englishman Charles Gadsden, by reworltiilg the mine's tailings. The New Mine Sapphire Syndicate continued underground mining at the English and Middle iilines until 1923, when a severc storm destroyed equipment and infrastructure alike. The operation continued by processing stoclzpiled ore until 1929, when econon~icfactors (including the loss of the marlzet for industrial-quality corundunl to synthet- ics, and taxation by both the U.S. and British gov- ernments) finally forced it to close. After more than 25 years of inactivity, the New Mine Sapphire Figure 3. One of the most fomous pieces o/ Yogo Syndicate sold the property to An1erican interests sappl~irejewelry is,this "lrls Brooch," a corsage in 1955. Total production froin 1895 to 1929 onlament coutaining 120 Yogo sapphires (as amounted to 16 illillion carats of sapphires, of well as dlnnlonds, demantoid gnrnets, and which about 2.25 million carats were gem quality- topaz) that was mclnufactnred by Tiffony d Co. with an estimated value between $20 inillion and at the turn of the century. Pl~otocourtesy of ~ke $30 inillion (in 1952 dollars; Clabaugh, 1952). Walters Art Gallery, Baltimore, Maryland. Various groups attempted to reopen Yogo between 1955 and 1968, but met with little success; the gcology of the deposit was not well undcrstood, to rcestablish Yogo sapphires in the U.S. marltet. As mining costs were high, and marketing was liinit- a result of an aggressive pron~otionalcampaign, ed. I11 1968, Sapphire Village 111~.purchased the which guaranteed that their sapphires were not property. The firin raised capital by subdividing heat treated, Intergem sold 4,000 carats of cut sap- agricultural land near the mine and selling the resi- phires and $3 million of finished jewelry in 1984 dential lots with the right to dig limited quantities (Voyniclt, 1987a). However, higher taxes, greater of sapphire-bearing ore froin the eastern portion of regulation by the state government, and persistent the A dike (Voyniclz, 1987a). In 1972, Chikara financial problems forced Intergem into insolvency I

Yogo Sapphire Dcposit GEMS & GEMOLOGY Spring 1995 31 N

rn a*'

I figure 4. The Yogo sapphire deposlt consists of at least six subparallel lan~prophyredilzes, labeled by the azlthor as "A", "B", "C", etc. All but the "B" (commonly called the "barren" dllze) are known to be sapphire bearing. The vast majority of sapphlre production at Yogo has been derlved from the nlaln A dllze, il is this dike thal 1s com- monly ieferred lo when the Yogo sapphire deposit is discussed. The insct Irovides detail on the area that is currently active. The dike system has iiltr~~dedinto limestone o/ the Mission Canyon for- mation (Mm)and shales of the younger Kibbey (Mk)and Otter (Mo)formations, probably along (1 p1.e-exist- ing fazllt or fracture. Dahy (1988) mapped several faz11~sparalleling the dilze syslenl to both the north and south (where upthrown side = U and downthrown side = D). Map n~odifiedfrom Dally (1991).

.. sample was partially processed during the summer of 1994 at Roncor's on-site plant (P. Eclzer, pers. comin., 1994), but the results are confidential. =-, P =-, Cyprus-AMAX was unable to negotiate a new lease arrangement with Roncor; their lease expired on January 3 1, 1995. (P. Eclzer, pers. comm., 1995).

Figure 5. The sapphire-hearing A dilze, which I has produced the b~~lkof Yogo sapphires, aver- ages 2.4 m wide and has a lzl~owl~length of 5 km (not continuous). The upper portions have weathered to n soft clay-lilze material. The A- dike section shown here was hydrarllicolly mined by the British New Mine Sapphire Syndicate at the tun1 of the century; it is now referred to as the English Cut. Photo by the author.

32 Yogo Sapphire Deposit GEMS & GEMOLOGY Spring 1995 Figure 6. ego1)lzie sapphires ale Itnown for Figure 7. A sn~nllportlon o( Yogo snpphl~es(ire the~rwell-saturnled color rind brilliance when violet to purple, like the 1.27-ct stone shown faceted. Thesc three stones, w111chrange from 17ere wilh lwo b11ze Yogo coul~/eipar~s(2 22 0.50 to 1.69 cl, were cut fro111 Vortex mnle and 2.77 ct). Note thrit all three stones lire innterir~l.Courtesy of Vortex Mining Co.; rounrlecl, witllout distinct crystal frrces. This IS photo O GIA find Tino Hnmmid. typicnl of the Yogo roz~gh.CourLesy oj Vortex !Mining Co.; photo O GIA and Tino Hrrmmid.

The Vortex Mining Company was formed in shades of violet or purple (D. Brown, pers. comm., 1984 by a group of local Utica, Montana, prospcc- 1995; figure 7). George F. Kunz used the term corn- tors. Vortex's exploration effort rcsulted in signifi- flower blue to describc the color of Yogo sapphires, cant discoveries on the west cnd of the deposit, referring to the coi~~nloi~garden flower of that including sevcral new sapphire-bearing dikes and name (Centaurea cyonzzs). Stoncs with distinct red, associated brccciation zones (Voyniclz, 19S7a and 13; pinlz, or green hucs are cxtreincly rarc and are usu- Dally, 1988, 1991; Mychalulz, 1992). The Vortex ally too slnall for faceting (M. Ridgeway, pers. illine began operations in 1987, and it is currently comm., 1993). Hughes (1990)stated that Intergem the only active underground imiile at Yogo. Vortex recovcred only two "true" rubies and 10 green sap- Mining is also involved in jewelry manufacturing as phires from 300,000 carats of Yogo rough, none of well as mining. Both Roncor and Vortex Mining which were suitable for faceting. New Minc continuc to market Yogo sapphires as the world's Sapphire Syndicate illiile manager Charles Gadsden only sapphire that is guaranteed not to be hcat found only three or four rubies between 1895 and treated. 1929 (Clabaugh, 1952). Because Yogo sapphires are typically a well-sat- GEMOLOGICAL CHARACTERISTICS urated, unilor~nblue-rather than pale or zoned-it Color. Yogo sapphires are fanlous for their uniform has not been necessary to heat treat the stones for blue color, general abseiice of inclusions and zona- the commercial nlarltet (Voyniclz, 1987a; L. Perry, tion, as well as for their vivid luster and brilliance pers. comm., 1994). Hughes (1990, p. 305) statcd, in both artificial and natural light (see, e.,g., figures "More amazing than the color itself is thc great con- 1, 3, and 6). Approxilnately 97% of all Yogo sap- sistcncy of color from one stone to the next. phires are "cornflower blue" and 3% are various Virtually all are of the same even-blue hue." It is

Yogo Sapphire Deposit GEMS & GEMOLOGY Spring 199.5 33 Komorowslzi, 1988; Hughes, 1990). Giibelin and Koivula (1986) did note small inclusions of pyrite, dark mica, calcite, analcime (figure 9), and rutile (figure 10). Dunn (1976)also identified inclusions of spinel. Giibelin and Koivula (1986) reported that many Yogo inclusions appear similar to those found in Thai rubies, a conclusion later supported by Hughes (1990).

Figzzre 8. Some Yogo sapphires show a distincl Crystal Shape and Size. Doininant crystal forms are alexflndrite efJect, lilte the flpproximfltely 1-cl short rhombohedra1 prisms terminated by the basal stone in the cenler of this photo. 11 changes pinacoid (Clabaugh, 1952; Hughes, 1990; DelRe, from purple (similar to the stone on theright) in 1994),although Yogo sapphires typically show little incandescent lighl to blue (similar to that of evidence of their original ciystal shapes (again, see the stone on the left) in day or fluoresceni figures 7 and 8). Most of the stones recovered are light. Photo by Iohn I. Koivula. rounded (figure 1l), chipped, abraded, pitted, or bro- lzen into shards and wafers. The apparent inodifica- this blue color that has been likened to that of sap- tion of the original crystal shapes has been phires from Kashmir (Sinlzanlzas, 1976), although explained by partial dissolution (resorption)by the other gemologists find that the color appearance host magma (Clabaugh, 1952; Dahy, 1988), mechanical abrasion during dike emplacement and of Yogo sapphires is very distinctive (R. E. Icane, brecciation (Mychalulz, 1992), and to some extent pers. comm., 1993). Dichroism in Yogo sapphires can be quite pro- by mining and recovery illethods (e.g., blasting and nounced (Clabaugh, 1952; Allen, 1991): light green crushing of the ore). It should be noted, though, that pei-pendicular to the c-axis, blue parallel to the c- most Yogo sapphires also have "flat" shapes, as indi- axis. Soille chromium-bearing Yogo sapphires cated by the crystal faces that can still be seen. exhibit an alexandrite effect, appearing blue in day Intergem consulting geologist Deliner Brown (pers. or fluorescent light and red (Balzer, 1994) or purple comm., 1995) has hypothesized that, because the (figure 8; J. I. I

34 Yogo Sapphire Deposit GEMS & GEMOLOGY Spring 1995 bearing igneous dike (Clabaugh, 1952; Meyer and Mitchell, 1988; Dahy, 1988; Brownlow and Komorowslzi, 1985). Howevel; the deposit actually consists of a complex set of subparallel lainpro- phyre dilzes (a lamprophyre is a group of darlz, por- phyritic igneous rocks that usually contain phe- nocrysts of dark mica, pyroxene, or olivine in a fine- grained, crystalline groundmass and associated breccia zones. It appears that the emplacement of these dilzes has been influenced by faulting and the development of lzarst in the host liinestone (karst is a type of topography forlned on or within limestone by dissolution; it is characterized by sinlzl~oles, caves, and underground drainage). Six of the dilzes, labeled "A" through "F" for the purpose of this arti- cle, are shown in figure 4 and discussed below. Figure 1 I. Rozrgh Yogo sapphire ci ystals often clppear rozlnded, brolzen, or flattened. Tlie Main (A)Dike. The A dike has produced the bulk of Vortex !nine specin~erlshown here is cr clossic Yogo sapphires to date. Hence, before the author's extllnple of (1 Yogo sopphire ~liatwas clie~ni- research, it is the only dilze that was studied in any cally rounded (actually, partially resorbed) detail. The A dilze is about 5 km long and <1 to 6 m during LmnsportcJtroll in the host n~rlgnzn The (average 2.4 m] wide (Clabaugh, 1952).The dike has face~ing-qualitysapphire meoszrres 5 1n1n long intruded linlestone (Mission Canyon formation) x 2 nzm deep. Photo by Maha DeMaggio. and shales (fibbey and Otter formations), probably along a preexisting fault or fracture. Only very lim- Most rough Yogo sapphires weigh less than one ited contact-metamorpl~iceffects are seen in the carat, although stones up to 19 ct have been found. host roclzs, indicating relatively quiclz emplace- Clabaugh's list of some larger rough Yogo sapphires ment and cooling. According to Dally (19911, the included illdividual stones weighing 10 and 12 ct. dilze may be separated into three en echelon (over- Nevertheless, Hughes (1990)stated that only 10% lapping or step-lilze] segnlents in a zone oriented of the rough stones recovered from Yogo exceed one S75"W; each segment has been significantly mined, carat. both on the surface and underground. The three original underground mines at Yogo-the English, The flat shapes of most of the Yogo crystals Middle, and American-are all located on the A recovered thus far is a major drawback for gem cut- dike. They have provided the vast majority of geo- ters. Cutting retention for a standard brilliant cut logic and mineralogic data available to date on the averages 20%. The largest rough Yogo sapphire, 19 Yogo deposit. ct (found in 1910), was reportedly cut into four stones, one of which weighed 8.5 carats (Clabaugh, "Barren" (B) Dike. The B dike lies approximately 1952). According to Voyniclz (1987a), the largest 183 in to the north of the A dilze and n~nsparallel known cut Yogo sapphire weighs 1.0.2 carats and is to it for about 1.6 lzm. The apparent absence of sap- presently in the collection of the Snlithsonian phires in this dike has been explained by the fact Institution. However, some gemologists feel- that it is a inafic lainpropl~yre(specifically, a because of its inclusio~~s-that the source of this ininette], whereas thus far sapphires have been stone has been wrongly identified. It has exsolution found only in ultramafic lalnprophyre (Dahy, 1988). needles of rutile in a hexagonal zoned pattern, and Given the limited sampling that has talzen place to large liquid-and-gas C02fluid inclusions, both of date, though, further research is needed to verify which are not found in Yogo sapphires (J. I. Koivula, this theory. pers. coinm., 1995). C Dike. Approxin~ately46 m south of the A dilze, GEOLOGY near the American mine, is the sapphire-bearing C Overview. The Yogo Gulch deposit has traditionally dike (Dahy, 1988). Discovered at the same time as bcen described, for siinplicity, as a single sapphire- the A dilze, the C dilze was not developed because it

Yogo Sapphire Dcposit GEMS h GEMOLOGY Spring 1995 35 found it to be approxiinately 0.6-1.2 in wide and proved a length of 152 111 (P. Eclzer, pers. con~m., 1994).Lilze the C dike, all D-dike inaterial has been altered to yellowish clay minerals; no original, unweathered dilze rock has been found. As of 1957, the largest ro~ighstone recovered from the D dilze was 6 ct, whereas the largest cut stone weighed 2.4 ct (Zeihen, 1957). Currently, there is no mining activity at this &lie.

E Dilte. Across Yogo Creek froin the American mine is the new Vortex mine (figure 12).This thin, highly weathered dilze was initially discovered froin a set of prospect pits nlade in 1924. Vortex Mining began develop~nentin 1987 (L. Perry, pers. comm., 1994). Subsequent worlz deterinined that the dilze within the Vortex mine is not an extension of the nlaiil A dilze, but rathcr is a separate dilze, laheled "En here (again, see figure 4). In the upper lcvcls of the Vortex mine, the E dilzc is only a few centimc- ters wide; but along the 30-111 section of dilze cur- rciltly exposecl by mining at the 61-in level, the dilzc is about 0.5 n~ wide, strilzes S4S0W, and is highly weathercd to a reddish brown color. Accordi~lgto core drilling, unaltered dilze material exists at a depth below 100 in (total proven depth is Fig~rre12. The Vortex i~~ii~ewtrs first 11iscov- 110 in). Currently, all Vortex ininc production is erzd trrounr1 1924, wlleil rr ::!-oL~~Iof niinms, derived from the E dilze and associated breccia 1.e1tr~edto the old Yogo Alnericnll Snl~l~hire zo11es. Coinptrny, stri~ksevc~rrrl prospect pits in to 1111s cli/f. Vortex Miilii~gCo. Oegnll dcvclopi~lelllo/ [helong-ol~tu~doncrl site in 1987. Tile laoi11 F Dike. In 1993, core drilling by Vortex Mining s/lt!f/,pictured here, hos ilow i~c(rchec1tr depth reveillcd this new dilze. Although underground tuu- of 61 111. Snpp11ire.s are bciilg recoverccl fron~trt neling has not yet reached the F dilze, drilling indi- lcr~stoilc dike. Photo by !he (r~r~l~or. cates that it is approximately 1 m wide ant1 strilzes S60°W, 011 an intersect angle with the E dilze to the soutl~west.Lilze the E clilze, F has been proved to is rnucl~thinner (15 cin) than the A dike (a11 average exist as deep as 110 in. of 2.4 111 wide). However, the American Sapphire Coinpany recovered some sapphires from a small, 6-in-long, adit that was driven into it in 1902 (P. Secondary Deposits. Erosion of the dilze systenl has Eclzer, pers. co111111., 1994).Thcre is 110 current pro- created ininor secondary sapphire deposits. duction froin this dilze. Nor has any mineralogic or Colluvial deposits on hillsides helow the English geologic description of the dilze ever bccn pub- mine 2nd English Cut were quiclzly mined out at lished, although the author has observed that the the turn of thc ceiltury by the New Mine Sapphire dilze material has been significantly altered by C02- Syndicate (Voyniclz, 19872).In addition, Yogo Crcelz rich grouild water to yellow clay minerals. cuts through the A clilze at the Amcrican mine, and has scattered sapphires as far as 4 lzm tlownstream. D Uile. The D dilze is located on the north riin of However, the sapphire-bearing gravels within Yogo Kelly Coulee, at the west end of the Yogo deposit. Creclz have never heen coininercially ininecl (L. Its orientation is slightly different-due east-wcst- Periy, pers. coinin., 1994). Because the secondary froin that of the other dilzes describetl herc (which occLlrrences are so linlited, Intergem believed that are slightly inclined to the south; Voyniclz, 198717). the primary deposit had only recently been cxposed Vortex inined this dilze in the inid-1980s. They by erosion (Voyniclz, 19873).

Yogo Sappllirc Deposit GEMS W GEMOLOGY Spring 1995 ings. Brecciated roclz within these lzarst features Yogo magma, as it rose toward the surface, cap- may or may not contain sapphires, depending on a tured fragments of the corundum-bearing gneiss complex set of conditions, including whether the and transported it upward as xenoliths. The corun- breccia was formed before or after dilze emplace- dum crystals were eventually released into the ment. magma as foreign fragments (i.e., xenocrysts). Dahy There are at least six main brecciated zones further hypothesized that magmatic heat naturally within the Yogo sapphire deposit; four are located "heat treated" the corundum into uniform blue, along the A dilze, includiilg the Kelly breccia (the gem-quality sapphire. Much of the con~ndumwas letter "K" in figure 4). Although sapphires have later resorbed into the inagma as it rose to the sur- been mined from the Ibreccias of the American mine some sapphires and rounding, pitting, and etching produced little or no sapphire (Voyniclz, 1987a). others. Evidence to support this theory includes the Breccias are also quite common within the Vortex discovery of a xenolith-containing corundum, mine, mainly between the E and F dilzes. These feldspar, augite, and spinel-which Dahy (1988) breccias are sapphire-bearing, with grades as high as interpreted as a n~etamo~-phosedclay or bauxite. 5 ct/t. The sixth breccia zone occurs on the west Another possibility is that the sapphires crystal- end of the D dilze, on the ridge overloolzing Kelly lized in an earlier magma and were subsequently Coulee. These breccias have never been exploited, borne upward; they would still be called xeilocrysts so it is not known whether they contain sapphires. because they did not crystallize in the transporting Originally, the Kelly and Vortex mine breccias were magma. Meyer and Mitchell (1988) also presulned interpreted to be volcanic breccias related to dia- that Yogo sapphires were xenocrystic. treine activity (Dahy, 1988, 1991; Mychaluk, 1992), New evidence supporting the xenocrystic ori- although Balzer (1994)suggests that they may sim- gin of the Yogo corundum crystals comes from ply be lzarst features such as collapsed sinkholes. therrnodynalnic modeling of the crystallization sequence of the Yogo magma, which is now repre- ORIGIN OF THE YOGO SAPPHIRES sented by the rock ouachitite. Such models can be The origin of sapphires in the Yogo dike system has calculated today with computer programs that use generated much discussion and debate over the past thermodynamic data for multico~nponent(Si02, century. Many of the first geologists to study the A1203, FeO, and other chemical constituents) sili- deposit concluded that the sapphires inust have cate liquids (magmas).Co~nputer sin~ulations of the been forlned directly froin the Yogo magma as phe- crystallization of the Yogo magma by J. Nicholls, nocrysts (Pirsson, 1900). These early worlzers envi- University of Calgary, using the program described sioned the silica-deficient Yogo magma incorporat- by Ghiorso and Saclz (1995)suggest that corundum ing large amounts of Al-rich shales as it rose toward could not have crystallized directly from the Yogo the surface. The magma consun~edthe aluminum magma. Therc arc, however, some caveats to the and subsequently crystallized corundum directly conclusions drawn to date; additional research is from it. Later researchers, notably Clabaugh (1952), needed to resolve this question. suggested that the Yogo inagrna incorporated frag- ments of a lzyanite-bearing gneiss, instead of shales. MINING AND PRODUCTION The lzyanite, a source of aluminum, was then con- Mining and Processing. Gem-quality Yogo sap- sumed by the magma and later crystallized as phires are extracted directly from their host roclz. corundum. Baker (1994)agrees with this latter phe- C02-rich surface and ground water has weathered nocryst model and argues that inclusions he portions of the lamprophyric dilze roclz to a soft, fri- observed within Yogo sapphires, such as C02 gas able nlaterial that disintegrates in water (owing to and analcime, could only have been forlned directly its content of montmorillonite and other clay min- from the Yogo magma. erals]. Unweathered dilze roclz, on the other hand, is However, it is also possible that the sapphires in a hard, competent state, so mining requires care- were incorporated as inclusions in the Yogo magma ful blasting. To extract the gems from the roclz, (i.e., are xenocrysts). A detailed model describing early sapphire miners exposed both types of inateri- this theory is presented by Dahy (1988).He envi- a1 to rain, snow, and frost 011 large wooden "weath- sioned that metamol-phic roclzs (e.g., gneiss) below ering floors" (Clabaugh, 1952). Within six to 12 the Yogo area already contained conlndum, as they months, the dilze rock would decompose, at which do in other parts of Montana (Clabaugh, 1952).The point the sapphires could be recovered by simple

38 Yogo Sapphire Deposit GEMS & GEIMOLOGY Spring 1995 Figure 14. At the Vortex mine processir~gplant, ore is loaded into the yellow hopper, which feeds into the large rotat- ing tron~mel.Heavy chains in the troznrnel Ilelp break down sap- phire-bearing ore, which is tl~enfed into a jig behind the bulldozer. Final clean-up of the sluice tray in lhe jig is done by hand. Photo bv I the author.

gravity methods such as sluicing. Today, sinall pro- Between 1902 and 1929, 200,000 tons of ore were cessing plants with rotating trominels break down extracted from the English mine, with a grade of the sapphire-bearing ore and trap the sapphires in 20-50 carats of rough per ton (Clabaugh, 1952). At modified jigs and sluices that are usually used in 8-15 in below the surface, the Middle mine con- placer gold recovery (figure 14). Both steam and nects with the English nline (figure 7 in Clabaugh, dilute HC1 acid have been used experimentally to 1952); the English mine was worked to a maximum decrease ore-processing time, altl~oughneither depth of 76 m. Neither mine has been commercial- method has been fully implemented (P. Eclzer, pers. ly operated since 1929, although the old Middle comn~.,1994). Large amounts of goethitelhematite mine worlzings were the site of a 1993 Cyprus- (limonite)cubes are also recovered in the processing AMAX bulk sample (the results of which have not jigs. The cubes may be heated to change the been made public). limonite to a inagiletic phase for easier removal Between 1901 and 1914, first the American from the jig concentrates, a technique used by the Sapphire Co. and then the Yogo American Sapphire New ~MineSapphire Syndicate at the turn of the Co. extracted a total of 3 million carats of sapphire century (Voyl~iclz,1987a). Final sorting of the con- rough from the American mine, which was worlzed centrate for sapphires is done by hand. to a depth of 91 m. The mine was not operated again until the early 1970s: Between 1970 and 1973, Production. Most of the Yogo sapphires recovered to Sapphire International Corp, produced 300,000 date have come from six locations in the deposit: carats of rough, of which 15%--or 45,000 carats- English Cut, English mine, NIiddle mine, American was gem cl~~ality(Dal~y, 1988). In 1974, a further mine, lntergem Cut, and Vortex mine. Initially, sap- 250,000 carats of rough was produced (Sinlzanlzas, phires were extracted by the New Mine Sapphire 1976). Therefore, a total of at least 3.55 million Syndicate froin surface exposures of the A dilze- carats of rough have been recovered from the and the colluvial gravels weathered from them-by American mine. means of hydraulic mining. This initial mining The three underground mines and hydraulic (roughly 1895-1901) occurred mainly at the English operations on the English Cut produced an estimat- Cut, where early miners had to construct a 15-lzm- ed 16 million carats of rough sapphires between long flume to bring in the water needed for process- 1895 and 1929, according to Clabaugll (1952), who ing. Clabaugh (1952) reported production figures compiled detailed production statistics for this peri- ranging from 296,862 carats in 1598 to 777,550 od. Fourteen percent, or about 2.25 million carats, carats in 1901. of these stones were gem quality. The next phase of mining at Yogo, which was Intergem undertook a large-scale strip-mining also by the New Mine Sapphire Syndicate, occurred operation along the east end of the A dilze (now underground at the English and Middle mines. referred to as the Intergein Cut). Between 1980 and

fogo Sapphire Deposit GEMS & GElMOLOGY Spring 1995 1986, Intergel11 produced an estimated one million SULMMARYAND CONCLUSION carats of rough sapphires froin this cut (Dahy, Over 100 years of intermittent operation, the Yogo 1988). Because, as discusscd above, the sapphire sapphire deposit has produced an estinlated 18.2 content can vary greatly fro111 one part of a dike to illillioil carats of rough sapphires, of which about another, ore gradcs ranging fro111 5 to 50 carats of 2.55 million were gein quality. More than a half rough per ton liavc been reported (Voyniclz, 1987a). nill lion carats of cut sapphires have entered the Cyprus-AMAX also toolz a bullz sample from thc n~arlzetplacefrom this locality. Approxiinatcly 97% Intergcin Cut in 1993, but little othcr developlnent of these gci~lsare "cornflower" blue, whereas 3% work has occurred there since 1986. are various shades of violet; there is no coininercial Currently, the Vortex n~ineis the only fully heat treatment of this material. Yogo sapphires also operating underground inine at Yogo, albeit relative- contain fewer inclusions ancl fractures than sap- ly snlall conlpared to past operations. As of January phires froin most other localities, although only 1995, the mine was 61111 deep wit11 one 50-m-long about 10% oi rough Yogo sapphires exceed one tunnel at the 61-m level, three 91-nl-long tuilnels at carat-their main drawback as a gemstone. Cutting the 18-m lcvel, and one 12-111-long adit at ground retention of these typically flat-shaped crystals level (L. Perry, pcrs. comm., 1995). Although first average 20%, producing nlany stones in the 0.10- to opencd in 1987, the Vortcx inine only recently 0.50-ct range. Marlzcting high-quality sapphires of hcgan full-time ore extraction. Allen (1991) stated this size has been difficult for past and present pro- that the Vortex mine is capable of producing 400 duccrs. Profitahility is further hainpercd by the carats of rough sapphires a day, of which half are extra costs associated with undergroi~ndnlining suitable lor cutting. The mine produced 5,000 and and subsequent ore processing, which are not 12,000 carats, respectively, of gem-quality sapphires incurred with alluvial sapphire deposits. in 1992 and 1994. Thcre was no 1993 production Gcologic investigations have revealed that the because of a core-drilling and minc-development deposit consists of at least six subparallel ultramafic program (L. Per~y,pers. comm., 1995).As notcd ear- lamprophyre dilzes-not just one as is popularly lier, the orc grade at the Vortex minc also varies recorded. Sapphires have been discovered in all the greatly, froin 0 to 70 carats of rough per ton. dilzcs except the B, or "barren," dilzc. The deposit Roncor, the currcnt owner oi the English, has been influenced hy thc developinent of karst Middle, and American mines and the English ancl and hy pre-existing faults and fractures in the host lntcrgem Cuts, has not undcrtalzen any under- limestone, as well as by multiple magma intru- ground or hyclraulic mining. Rather, most of sions. There is grcat variability in g-adc within the Roncor's production is derived froin reprocessing dilze: Some areas produce little or nothing, and oth- old mine tailings and left-over ore from the ers produce as much as 70 carats of rougli sapphire lntergein era. With this tcchniclue, Roncor has pro- per toil of ore. duced a reported 30,000 to 50,000 carats of rough sapphires annually (Verhin, 1993). There is consiclcrable tlisagreemcnt in the liter- Total production for the deposit's 100-year ature as to whether the sapphires originated ia the operating period has hcen cstiinated by the author, magma (as phenocrysts) or werc part of "foreign" by updating Clahaugh's figures, to be approxilnately roclzs transported to the surface by the inagitla 18.2 million carats of rough sapphires, of which (xenocrysts).Recent thermodynamic coinputer 14%-or about 2.55 nlillion carats-were of gcill modeling lcnds soine support to the theory that the quality. Assuming an average cutting re ten tion of Yogo sapphire crystals were xenoclysts. Howcvcr, 20%, the Yogo deposit has probably produced the original source of the Yogo sapphires, if truly 51 0,000 carats of cut sapphires. xcnocrysts, renlains a mystery. The geology of the deposit indicates that the Sapphire reserves at Yogo appear to be consider- depth of thc A dilzc is considerably greater than the able. 011 the basis of reports in the literature and his 9 1 111 reached thus far. Givcil that the specific gavi- experience, the author estiinates that thcy proha- ty of the sapphires is higher than that of the host hly-if not practicably-cxceed twice what has rock, there inay bc even more sapphires at lower already been produced. Because of greater activity Icvels. Onc can conclude, therefore, that thc by groups such as Vortex, Roncor, and Cyprus- reserves arc significant, pcrhaps twice what has AMAX, the author suspects that production will already been recovered, although the cost of recov- gradually incrcase during thc next decade, yielding ery nlay be prohibitive. inany more of the now-legendaiy Yogo sapphires.

40 Yogo Sapphire lkposit GEMS k GElMOLOGY Spring 1995 REFERENCES Allcn K.M. (1991)Thc Yogo Sapphire Dcposit-New discoverics perature and pressures. Contrihcrtions to Minerology and crcate Illore interest in Anlerica's fincst gemstone. I1etrology, Vol. 1 19, No. 2, pp. 197-212. C:en~ologictrl Digest, Vol. 3, No. 2, pp. 9-16. Giibelin E.J., IAugust, pp. 25-29. Dunn P. (1976)Gem notes. Gcn~seJ Gemology, Vol. 15, No. 7, Weed W.H. (1899) United Strrtes Geological Survey Atlas, Little p. 200; Relt Morrntoins Folio No. 56, USGS, Washington, DC. Ghiorso M.S., Sack R.O. (1995)Chcmical mass transfcr in mag- Zcihen L.G. (1987) Thc sapphire dcposits of Montana. tn D.C. matic process IV. A revised and internally consistcnt ther- Lawson, Ed., Directory of Molltono Minii~gB~terl~rises lor modynamic nlodel for the interpolation and extrapolation of 1986, Nlontana Bureau of Mincs ant1 Gcology Bulletin 126, liquid-solid equilibria in magmatic systems at elevatccl tenl- Great Falls, MT, pp.2040.

CAPITALIZEON THE ROWING DEMAND r FOR COLOREDSTONES +++ 1 *.*WITHTHE GIA LEARN-AT-HOME COLOREDSTONES PROGRAM I + Learn to gracle colored stones using the GIA Colored Stone Glacllng System + Gain the sltills and procluct knowleclge you neecl to sell more cololed stones + Learn how to test and ~dent~fygems + Practice with stones GIA lends you + Ea~nthe GIA Colored Stones, Gem Idelltificatlon ancl Colored Stone Grading Certificates

FOR \IORt li\lEORhlh~lO\ OR TO ChltOLL TODAY Call Toll-Free (800) 421-7250, Ext. 292 Outsicle the U.S (310) 829-2991, Ext 292 GGCSS MEERSCHAUMFROM ESKISEHIR

By Kadir Sariiz and Iskender Isik

fiis study p~pesentsa model for the o~qqinofthe fihisehi~~)T&~rp/zej\ sepiolite deposits. IC~~olvnas "wceerschaum" in thegem tmde ~vhercit occurs as compact weasses (especially nod~tles),sepiolite is mined by local fhrmers using basic iweplew~e~trand traditional t~~anclingmethods. 7%;best aodub are carved into objects s~chas pi~ebo~vls, b?.ncelcts, and rcecklaces. rfie sepiolite nodules occulFirc th Pliocene-age Imisehi?.conLq~ome?.ates) assaciated ~vitj~dolomite, nzagnesitc) opal-CT, and lizardite. Fib.~.o&~sin nzo?pholqqj5 they probably farmed in relatively shallo~vwater under alkaline conditiorc~;as a result of tile ~feplace~~entofnzzzngnesitc~~.nvels that were subjected to pore maters with hiqh SiO 2 concentmtions.

Sepiolite, M&Si6015(OH)2.61-120, is a clay mineral has been a popular omainental gein material. It is that has inany industrial uses. One of the earliest best lcnown as a carving inaterial for pipe bowls (fig- was soap, because the coinpact material has a ure l), but it has also been used for objects such as grcasy texture and lathers when first recovered. caineos and jewelry. White, compact, inassivc sepiolite is called meer- Although inecrschauin may have been used as schaum (from the Gerinau 111eer [sea] and schurtm early as the days of the Roinan Empire (Ball, 1950; [foam]);it has a very low specific gravity (dry, it can Ece and Coban, 1994)) its use as a carving inaterial float 011 water) and low hardness. Because it is easi- has been documented since the late 1600s, when it ly carved and polished, ineerscl~auinhistor~cally was inined for the lnanufacture of pipe bowls, cigarette holders, and building inaterial from the sepiolite deposit in Vallecas, Spain (Galan and Castillo, 1984). At present, the Spanish sepiolite ABOUT THE AUTHORS deposits, the largest in the world, are mined for Dr. Sariiz is associate professor of geology, Depatment of Mining industrial uses only; there are no recent reports of Engineering, Osmangazi Universty, 26030 Eskisehii, Turkey. Dr. gem-quality (nodular) inaterial froin Spain. lsik is associate professor of geology, DepaJtment of Ceramics Engineering, Dumlupinar Universty, 43 100 Kutahya, Turkey. Relatively new occurrences of sepiolite, some of Acknowledgments: The authors thank Dr. Rifat Bozkvrt for provid- which contain inaterial classified as meerschaum, ing samples; Dr. Fdi Coban for performing the X-ray dimaction have been reported in the United States (California, ana4ses; Mr. Sedat Koncak and Mr. Muharrem Wmaz for allow- Nevada), Spain, Kenya, Japan, Russia, and China ing US to photograp17 their meerscl?aumarticles: and Mr. Necdet Altinay for information on current distribution practices. Dr. (Singer and Galan, 1984; Jones and Galan, 1988)) Herbert Rupp and Ms. Hannelore Schmrdl, of the Austrian but the published reports on these areas do not Tobacco Museum, and Dr. G. Nerdermayr, of the Natural History nlentio~linaterial suitable for carving. Musum of Venna, kind4 provided photographs of histork meer- scl7aum carvings. All other photos are by the authors. Within Turlcey, virtually all ineerscl~aumpro- duced is from Eslcisehir Province. Althougl~another Gems & Gemology, Vol. 31, No. 1, pp. 42-5 1. O 1995 Gemological Institute of Amefica deposit of meerschauin nodules has been identified in Yunalc Township, Koilya 13rovince (150 lcm [93

Notes and New Tcchniclucs GEMS & GEMOLOGY Spring 1995 F1gul.e I Made for the 1873 Viennese M'orld's Fair, this intricately carved 111eerscl1a~lnlpzpe boc\7l1s decoraled with amber, turq~roise,and enonleled sllva Courtesy of the Austrian Tobacco Museuzn, Vienna;pholo by Tlloznns Reinagl.

miles] southeast of the city of Eslzisehir; Yeniyol to Budapest (Hungary) and throughout Gerinany, and Oztunali, 19851, to date there is no lznown priinarily for carving into objets d'art (figure 3) and coinmercial production fro111 that area. In the vicin- articles for smolzers (figure 4). Historically, meer- ity of Eslzisehir (the provincial capital), there are schauin was lznown as "Vienna stone" in the three major and two ininor sepiolite districts (figure European marlzet [Uzkesici, 1988); even today, 2).TurlzmentolNorway, The Eskisehir sepiolite deposits are lznown to Denmark, and other European countries. have been exploited since the 18th century. For Sepiolite usually occurs in sedimentary roclis, much of this time, the district's meerschaum was in forins ranging froin massive to earthy, as thin exported to Vienna, Austria, as preforlns, as well as layers, disseminated, and as concretions. Although

Notes and Nclv Techniques GEMS & GEMOLOGY Spring 1995 Sei lroi So ulcuk ,haurn Bits sepelci .Margi Fiprc 2. This !nap o/ he S - /+ Esltischir region of Turkey shows the loca- tion of the five niain sepi- olitc disfricts identified to date in this historic n~eerschouzn -producing area. Thc yellow orco sur- ro~imdingthe villagcs of Turltmentolzal and Goltceoglu-show17 liere in dark yellow-WAS studicrl cxtcnsi~~elyby ~raduyusl the authors (see Sariiz, (Atiziye) 1 990). Lines i?~rrrl

the Eskisehir sediineiltary sepiolite deposits have which it has hosted several international festivals) been exploited for more than 200 years, their geo- and for traditional follz dancing. The main indus- logic origin has not been studied in detail. The tries of Eslzisehir Province are farming, mining, objective of this study was to develop a genetic carving, clothing, briclz, and tile manufacture. model for the geologic origin of the deposits at Turkmentolzat-Gokceoglu. In the course of this research, we also studied current inining practices, Figure 3. During tke 19th cent~lrj~,mecrschaunl was crlso used for cameos, as illustraied by this the gemological properties of meerschaum, and the 37.5 x33.5 cm carving of Emperor Franz joseph I of fashioning and distribution of this ornamental gem Austria. Co~utesyof t11e NALZUA~History MLISELLNIof Vie~mrl. material.

LOCATION AND ACCESS The meerschaum is mined from shafts (or holes) in an area between the villages of Turlzmentolzat and Golzceoglu, approximately 20 Izm (12.5 miles) east of the city of Eskisehir (again, see figire 2),. This region is about 1,000 m (3,280 feet) above sea level; vegetation is subtropical. The mean temperature is 11.3"C (about 52°F).The inean rainfall in the region is 250 to 300 mm (10 inches to 1 foot), with cool summers and cold winters (rainy and snowy). The city of Eslzisehir is located about 250 lzm west of Anlzara, the capital of Turkey, and 360 lzm south- east of Istanbul by road. Access to the inining area is possible by car via Highway E90 from Eslzisehir to the center of the villages, and then by horse or by foot (and, in some cases, by car) via dirt trails for about another 1-2 kin. Eslzisehir was founded in the first millennium before Christ by the Phrygions, as they settled on the banks of Porsulz River. It has an active artistic and cultural life, with world-class perforillances of the ballet, theater, opera, and folk dancing. The city is especially well known both for ineerschaum (for

Notcs and Ncw Techniques GEMS & GEMOLOGY Spring 1995 Figure 4. Corved ill 18th- 1 can t~~syViennrr, these 1necl.sc11nzr111cignr hold- . crs (icp~ctn~ythologrcul, hnnting, alld historic scenes. Note the cl~ange 111 coloi conscd by 11ico tine sta~nil~g.Comtesy o) g the Azrstrian Tobaccc Muselrm, Vlcnncl;pholo 'J&, 17~1Tllolnns Xelnagl

MATERIALS AND METHODS logical testing (i.e., dcter~ninationof optical and Detailed geologic napping of the study area was physical properties). co~npletedduring the sumimer of 1990 (Sariiz, 1990).Descriptioils and study ol about 100 samples GEOLOGY AND OCCURRENCE from the conglomerates in which the nleerschau~n Geology of the Esltisehir District. The oldcst roclzs deposits occur provided the primary data for this in the district belong to the Upper Paleozoic study. Karatepe group, which is composed of dcep marine Polarizing microscopy, X-ray diffraction (XRD) (oceanic trench) roclzs that have undergone meta- analysis, and scanning clectro~l~nicroscopy (SEM) morphism (high pressure and low temperature) as a were used to identify, study, and characterize meer- result of subduction. The group is divided into three schaum and other minerals within the deposit. units: the Sarilzavalzcali marble, the I(ara1zaya meta- XRD analysis was done at the Department of morphic~(calc schist and quartz schist), and thc Geological Engineering of Istanb~~lTechnical Golzceoglu forn~ation(cluartz-bearing muscovite University using a Philips Model 1140 X-ray pow- schist, glaucouite schist, marble, "schisty" diabase, der diffractometer. SEM analysis was done at the and phyllite). This group is overlain by the Department of Metallurgy of the Uiliversity of Turlzn~entolzatophiolites (inafic and ultrainafic Texas at El Paso with a dual-stage scanning elec- igneous roclzs associated with a geosyncline), which tron microscope (Model ISI-DS 130). are probably Triassic in age and are conlposed of Carved ~neerscl~aurnarticles provided by the dunite, serpentinite, and gabbro. The T~irlzinentolzat nleerschau~ncompany Isandstones) and

Notes ant1 New Tcchniques GEMS &. GEMOLOGY Figure 5. Tl~esecross-sections from the Tzrrkmentoknt-Go1rceogI~1district sl~owthe lo~atioriof the stloll- grrrph~csequence of roclcs (see lext for details) and he sepiolite (meerschaun~)deposits. Localions ofcross- sections are lobeled "A" arld "B" in figr~re2.

the Yorulzkaracaoren liinestones (mostly massive that the ~nagnesiteforincd first and later was and partly recrystallized). These, in turn, are over- replaced by sepiolite. In addition, we observed that lain unconfor~nablyby the Pliocene-age Iinisehir lnagnesite gravels, like thc sepiolite nodules that conglon~eratesand 1Coyunyatagi limestones; the form froin them, are randomly distributed in differ- former contain the econo~nicallyiinportant sepio- ent levels of the Iinisehir conglon~erates,which lite nodules as well as peridotite, diabase, magne- suggests that the gravels were transported froin the site, and serpentinitc gravels and sand-size grains of n~agnesitedeposits in the region. quartz, feldspar, and biotite, all of which are Although bedded sepiolite is found in solme cemented by dolomite. The conglomerates are parts of the Eskisehir district, this material is typi- 3540 in (about 115-130 feet) thick; the gravels (as cally an undesirable gray color, is contaminated by distinct froin the larger nodules) range up to 4 cin (about 1.5 inches) in diameter. The Turlzn~entolzat ophiolites and the Jurassic roclzs are cut by dilzes F~gure6. This view of the extrncted Imisehir con- and stoclzs of Paleocene-age Icarlik quartz diorite ylomeratcs shows the 1~11itemeerschani~.~ nod- porphyry (figure 5). ~rlesnnd frngmcnls, the dmlz grrry selpenlinile, The present gcologic structures (e.g., grabens) of light gray silica, o11d ~hcbrow11 dolornltic the study area formed duriilg a inajor deforinational cenlents tho1 bind the niincrals. (tectonic)stage in the co~nplexgeologic history of the Aegean region during lower Miocene time. This was a result of the Arabian platforin being pushed toward the Anatolian (Asia Minor) plate (Sengor, 1980).

The Sepiolite Nodules. At Eslzisehir, sepiolite nod- ules occur in the Inlisehir congloinerates (figire 6). Typically, the nodules are 8-10 cm in diameter, but soine are as large as 25 cm. In natural for~n,they are gray to white, compact, and soft as some soaps. Soine sepiolite nodules, when brolzen, display a inagnesite core (figure 71, indicating a genetic rela- tionship between the two minerals-in this case,

Notcs and New Tcchniqucs GEMS h GEMOLOGY Spring 1995 Figure 7. Mllgnesiie (I&) cores oncl rims ore visible in tl~csetwo sepiolite (Se) nodrrles: left = 10 cm wicle, right = 8 cm rvicle.

clay and other materials, has a relatively high spe- are intimately intergrown, and terminations are cific gravity, and is subject to nlultiple craclziilg as rare. The ends of the fibers are bent slightly. it dries. Some has bcen used for carving, but the product is of very low quality. Gem-quality meer- Occurrellce of Sepiolite. Singer and Galan (1984) schaum occurs only in the sepiolite nodules. and Jones and Galail (1988)made the most recent On the basis of microscopic studies, we identi- and authoritative .compilations on all aspects (his- fied quark and dolomite grains up to 30 pin in thc tory, crystal structure, chen~istry,synthesis, phys- sepiolite nodules. XRD patterns recorded the pres- ical properties, geologic and geographic occur- ence of dolomite, magnesite, and ininor amounts of rences, industrial applications) of sepiolite, opal-CT (opal-cristobalite)and lizardite (a illember Mg4Si60,5(0H)2-6H20. They show that it is of the lzaolinite-serpentine group) in the nodules. closely related to palygorslzite (varietal naines for The ceinent of the congloillcrates contains which include "mountain leather" and "mountain dolomite and minor amounts of opal-CT and ser- wood"), (Mg,A1)2Si4010(OH)-4H20, with which it pentinite. frequently occurs. Sepiolite shows a fibrous morphology in the Both minerals have been found in a number of scanning electron micrograph (figure 81, with fibers geologic environments. These include: (a)inarine as wide as 0.5 ,urn and up to 35 pm long. The fibers enviroilinents (sediments)ranging from deep seas to shallow lagoons; (b)continental environments rang- ing froill soils, calcretes (caliche),and alluviuim (all characteristically allzaline and in arid or semi-arid Figwe 8. This scanning electrol~photomicrograph regions); (c)continental saline and allzaline lake sed- of n sepiolite nodzlle shows the fibrous texizne iments; and (d)associated with igneous roclzs as typicol of this mnteriol. Scale bm= 2.00 pm. veins and alteration products, some of which may be related to hydrotherinal alteration of Mg-rich (ultramafic) roclzs (Singer and Galan, 1984; Jones and Galan, 1988).In addition, Ece and Coban (1994) discuss the diagenetic replacement of magnesite cobbles and pebbles by sepiolite with specific refer- ence to these (Eslzisehir)deposits. Because sepiolite may form in many environments, it is not surpris- ing that this mineral is found in several physical states, ranging froin compact (meerschaum) to earthy (as in soils). The random distribution of sepiolite nodules within the Imisehir congloi~leratescan be related to ground-water fluctuations over time as well as to

Notes and New Techniques GEMS & GEMOLOGY Spring 1995 47 bearing solutions from which the Eslzisehir sepio- I lite deposits fomled: 1. Under allzaline conditions (pH = 8-8.5), when the Si02 concentration of the pore waters in the inagnesite gravels reached saturation, sepiolite fom~edin situ as a replacenlent of the inagnesite gravels. (Sepiolite does not forin when the Si02 concentrations in pore waters are low.) I 1 2. When pore waters were supersaturated with respect to SO2, both sepiolite and opal-CT forined together, as evidenced by the presence of opal-CT areas associated with sepiolite in the I nodules (figure 9). 3. As no talc was detected in our samules, it call be 1 . r assumed that dolomite-unlilze magnesite-in I the ilodules did not react with the Si02-saturated Figure 9. Areas of opal-CT (OP)were foiin~lto be pore waters. Lf such a reaction had talzen place, associated with sepiolite (Se) in he nodules talc would have been produced (Mueller and (here, 13 crn wide), Saxena, 1977). the proximity of the conglomerates to fault systems MINING that enhance the flow of fluids. Also, clay layers The Turkinentokat-Golzceoglu district contains inay have influenced sepiolite distribution, because hundreds of shafts that are helieved to have been they affect the ability of pore water (subsurface worked internzittently for sepiolite. Traditioilal water in the voids of the roclz) to penetrate the for- illethods for mining n~eerschaum,using siinple mation. On the basis of the field investigations and hand tools, still appear to be the most economic. on cheinical, inineralogic, and petrographic analy- Specifically, shafts about I m in diameter and auy- ses performed on sepiolite nodules (taking into con- where from 10 to 75 m decp are dug into the ground sideration the stability diagram of Wollast et al. to reach the sepiolite-containing beds. When min- [1968]),we detennined three main points about the ers reach a layer that may have economic anlounts physico-chemicalbehavior of the M~+~+s~o~(~~)+H~o-of sepiolite, they drive a productioil gallery (hori- zontal "drift" or opening), approxinlately 120 cin Figu1.e 10. A "spinningwheel" (rope and pulley) (about 1.5 yards) high and 150 cm wide, in the system is zlsed lo remove bnclzets of sepiolite direction that the sepiolite nodules appear to occur. nodules and waste materials froin one of many Because the conglomerates bearing the sepiolite mining "holes"operating in the Eskisehir district. nodules are strongly cemented by the dolomitic material, miners can operate in the gallery without adding any structural supports. To extract the sepiolite, ininers use the tradi- tional room-and-pillars style of digging. Both waste materials and sepiolite ilodules are raised to the sur- face in buckets by a siimple "spinning wheel" (rope and pulley system) at the top of the shaft (figure 10). Colnmoil extraction tools include picks, shovels, and wedges to pry the sepiolite from the conglomer- ate. Because wet sepiolite is very soft, ininers must be extremely careful not to damage the nodules during extraction. Explosives are not used. Zn the Eslzisehir district, meerschaum is inined by small coinpallies and groups (2-5 people, figure 11). Because most of the miners are farmers, the deposits are worked most intensively after the

Notes and New Techniques GEMS 6: GEJMOLOGY provided by the State Organization of Planning (Prime Ministry, Republic ol Turlzey), we estimate that illeerschaum reserves in the district exceed six million lzilograins.

THE GEMOLOGICAL PROPERTIES OF MEERSCHAUM The Turlzmentokat-Golzceoglu ineerschauin nod- ules have distinctive inineralogic and physical prop- erties (table l), as well as size, shape, and texture- that malze them particularly suitable for carving decorative objects such as pipe bowls (figure 12)) objets dlart, and jewelry (figure 13). Properties such as sorption, chemical composi- tion, the white color, high porosity, and low specif- Figure 11. A group of n~inersreturn 10 I(art7tepe ic gravity, malze ineerschaunl ideal for use in pipe village (vlsihle In the baclzgromici) at the szd of bowls. Although soine manufacturers produce sepi- a ~vorlzday.Note the mining tools: a wedge, a olite articles in tourist venues elsewhere in Turlzey, shovel, a11d (1 carl>irlelan~p. Eslzisehir Province is the center of sepiolite manu- facture. farming season, in late fall through early spring CARVING MEERSCHAUM (November through April). In addition, the relative- The first step in imaking a sepiolite carving is to ly shallow depth of the water table in some areas pick the right nodule. Only by careful selection can means that flooding of the tunnels can be a inajor defects such as craclzs, color irregularities, and problem, especially in the spring. inclusions of other ininerals be avoided in the final product. Sepiolite can be easily calved as long as it PRODUCTION is somewhat moist. Because the nodules typically Meerschaum is believed to have been exploited in lose some of their natural inoisture during the often the region since the tinle of the Roinan Empire. lengthy carving process, they usually have to be From the end of thc 18th century until the begiu- soalzed in water periodically to maintain sufficient ning of the 20th) meerschaum was inined exten- softness. Eslzisehir carvers use as illany as 40 types sively in Eskisehir, and the rough was exported to Europe and the United States. According to the State Institute of Statistics, Prime Ministry, Republic of Turlzey (1990 and 1992))Eslzisehir pro- TABLE 1. Physical appearance and gemological properties duction at that time averaged an estimated 12,000 of gem-quality sepiolite (meerschaum) from Eskisehir, Turkey. cases a year (one case = 12 kg). Between 1913 and 1952, sepiolite production fell to a total of 6,600 kg, Color While (rarely, cream) while a total of 176,856 lzg was produced from 1954 Refractive indicesa n, = 1.525-1.529, n, = 1.506-1.520 with a b~relrinyenceof 0.009 to 0.019 to 1973. To support the local calving industry, the Specific gravily 2.0 (on compacl material)b Turlzish governillent has banned the exportation of Fluorescence Inert to both long- and short-wave rough ineerschauin since 1972. Production reported UV radialion for the period since then pealzed at about 525 cases lnlernal characteristics Quartz, dolomite, and magnesite in 1975 and dropped to a low of 40 cases in 1992. inclusions (all identified by X-ray diffraction) (Note that these figures represent production Hardness (Mohs scale) 2-2.5C recorded by government sources. Because much of Other Floats on water and adheres to tongue the meerschaunl is extracted outside of government in its raw staie supervision, actual production is undoubtedly alvlkroscoaic observation !./ilh immersion as reoorled bv Sag~rogluand much greater.) About half of this reported produc- Ccgulii (1 972, p. 877). tion is believed to be mined froill the Turlz- blvieerschaum floals on water !ithen dry Porous rnalerial has a specific gravrly 01 0.508-0.894. mentokat-Golzceoglu region (Necdet Altiuay, pers. C~VVl~enhealed meersclia~/mloses ils rnoislure and hardens lialher. comm., March 1995). On the basis of inforillation

Notes and New Techniques GEMS h GEMOLOGY Spring 1995 49 Figure 12. The beau- tiful meerschaum bowl 011 lhis 40-cm long pipe was carved in Esl

of lznives and other tools, which they usually man- ished meerschaum article a creamy white color and ufacture themselves. shiny appearance. Meerschaum turns a progressive- Once a nodule has been selected, the carver ly darlzer yellow when used as a pipe bowl bccause n~ustchoose a design that best fits the natural it absorbs nicotine from the tobacco. shape of that nodule and minimizes waste. From this point on, the single inost important point in DISTRIBUTION deciding on a design is the carver's vision and imag- In Eslzisehir and nearby villages, all meerschaum ination. With talent, experience, and patience, the products are manufactured by independent artisans artisan can produce a variety of intricate designs. at their worlzshops or homes. Currently, there are After the nodule has been given the desired or about 150 nleerschauin carvers in Eslzisehir requested form, it is partially dried slowly in the Province. There are no large-scale, mechanized sun. Subsequently, the piece is carefully hand carving operations in Turlzey. To educate meer- carved. Once the meerschaum calving is complet- schaum carvers and miners, and to raise future ed, it is dried in an oven at 110°C for two hours. carvers, the local government established a ineer- Next, it is polished with a very fine abrasive and schaum school in Eslzisehir in 1989. However, it is then i~nmersedin heated, liquid beeswax for a few not currently in operation. minutesj last, it is polished with a very soft cloth. To purchase meerschaum objects, one can con- This waxing and polishing treatment gives the fin- tact either the Chamber of Trade of Eslzisehir, as they publish requests in their journal, or any one of Figure 13. In Eskisehir, rneerschaun~is also the several companies operating out of Eslzisehir carved illto lewelry, like his suile featuring a (e.g., Oztas, Altinay, Altintas, or Iconcak). neclzlace, bracelet, and earrings. Tl~eineer- Meerschaum articles from Eslzisehir artists reported- scha~zmin the blncelet is 14.5 x 0.50 mln. ly have sold at high prices. For example, one by noted pipe-malzer Isinail Ozel was auctioned for $18,000 five years ago (Necdet Altinay, pers. comm., 1995). According to Mr. Altinay, the owner of Altinay Co., collectors of lneerschaurn carvings con- sider three maill factors: the date of manufacture (the older, the better), the carver (the more fanlous the better), and the artistiy of the piece itself.

CONCLUSIONS On the basis of a geologic map of the Turlzmentoltat-Golzceoglu district completed by the senior author, study of the local stratigraphy, and field observations, as well as X-ray diffraction, scanning electron ~nicroscopy,and petrographic analyses, the authors have concluded that the sepi- olite (meerschaum) nodules probably formed at

Notcs and New Techniques GElMS 23 GEMOLOGY Spring 1995 shallow depths and in alkaline conditions, in the ber of expert carvers in Turliey, especially in vicinity of paleo-shorelines of a lacustrine basin Esltisehir Province. Some of their worlzs of art have (large, inland lalie). They are the result of replace- been sold at auction for significant sums. ment of transported magnesite gravels when these The worldwide decline in the popularity of gravels were subjected to pore waters with high sinoliing has undoubtedly affected the demand for Si02 concentrations. Sepiolite's properties make it meerschaum pipes. However, sepiolite has many especially suitable for carving and polishing when it other industrial uses, which suggests that mining is occurs in a compact, massive forin, as meerschaum. lilzely to continue. One potential use for meer- Although significant quantities of meerschaum schaum could be as an alternative to ivory, which is were mined and exported from the 18th through banned in many countries. Both have a similar the early 20th centuries, production appears to have color, talie a high polish, and are easy to carve and declined greatly in recent decades. Since 1972, gov- cut. It must be noted, however, that ivory is harder ernment regulations have limited export to only than meerschaum and-unlike meerschaum- fashioned meerschaum. However, there are a num- bends without brealzing.

REFERENCES Ball S.H. (1950) A Rornnn Book on Precio~lsStones. Turlunentoltat-Icaratepe (Eskischir)(in Turltish). Bulletil~of Gemological Institute of knerica, Los Angeles, CA. the Minernl Research ond Explorntion Institute [Anltara], Ece O.I., Coban F. (1994)Geology, occurrence, and genesis of Vol. 110, pp. 77-96. Esltisehir sepiolites, Turltey. Cloys nnd Clay Minernls, Vol. Sengor A.1M.C. (1980)The Fundnmentnls of [he Ncotectonics o/ 42, No. 1, pp. 81-92. T~rrlwy(in Turltish). A bulletin of the Geological Society of Galan E., Castillo A. (1984).Sepiolite-palygorsltite in Spanish Turltey, 40 pp. 'Tertiary basins: Genetical patterns in continental environ- Singcr A,, Galan Er, Eds. (1984) Palygorskite-Sepiolitc: ments.,In A. Singer and E. Galan, Eds., Pnlygorskite- Occ~lrrences,Genesis, ~lndUses, Developments in Sepiolite. Occ~lrrences,Genesis, nnd Uses, Developments in Sedimentology 37, Elscvier, A~nstcrdam. Sedi~nentology37, Elsevier, Anstertlarn, pp. 87-1 24. Uzltesici T. (1988)The Stnte of the Mecrschnum lnd~lstry(in Joncs B.F., Galan E. (1988)Sepiolite and palygorsltite. In S.W. Turltish). First International Meerscha~~mFestival of Bailey, Ed., I-lydrons Phyllosilicntes (Exclusive o/ Micns), Eskischir, pp. 21-22. Reviews in Mineralogy, Mineralogical Society of America, Wollast R., Macltenzie F.T., Briclter O.P. (1968)Experimental Washington, DC , Vol. 19, Chap. 16, pp. 631-674. prccipitation and gcncsis of scpiolitc at carth-surfacc condi- Mueller R.F., Saxena S.K. (1977) Chemical l'ctrolo,~y. Springer- tions. American Mineralogist, Vol. 53, Nos. 9-10, pp. Verlag, New York. 164551662, Sagiroglu G., Cogul11 E. (1972).The Determination o/ Minerals Yeniyol M., Oztunali 0. (1985)The mineralogy and the genesis with the Polarizing Microscope (in Turltish). Technical of the Yunalt sepiolite (inTurltish). In l'rocecdings o/ the 2nd University, Istanbul, Gumussuyu, lstanhul. Notional Clay Symposi~lrn,Hnccttepc Univcrsity, Anltara, Sariiz K. (1990)The occurrence of magnesite tleposits at pp. 171-186.

ceaury La the present day. Whether you are imres~din karahg EDITORS Robert C. Kommerling and C. W. Fryer GEM TRi-.&DErr-.& .- - Giil Gem Trade laboratory, West Coast CONTRIBUTING EDITORS GIA Gem Trade laboratory, East Coast G. Robert Crowningshield Thomas Moses Ilene Reinitz Karin Hurwit GiA Gem Trade laborato~/,West Coost Mary 1. Johnson Shone F. McClure

ALEXANDRITE, with Pleochroic alexandrite (from the Elahera gem lab detect signs of treatment, they Twi~~ncdGrowth Zones field in Sri Lanlca) has been described bring the suspect stone to the A 1.21-ct faceted stone, with a green- before in Gems 0) Gemology Identification and Research lab for to-purple color ciiange, was sent to (Gunawardene and Rupasinghe, further exaininatioil and the issuance the West Coast laboratory for identifi- Sumnler 1986, pp. 80-95), the of the report. Although some stones cation. We foillid the optical proper- twinned areas in this stone were exhibit clear evidence of filling, the ties to be typical for alexandrite- unusual in that they were brightly breaks in other diamonds have biaxial optic character, with refractive colored and irregularly shaped, with a ambiguous features and thus require inclices of 1.743 to 1.750. The pres- vermiform, or worm-like, appearance additional investigation. Graders also ence of fluid (including two-phase) when observed througl~a inicroscope routinely show us stones that are not iiic~usionswas consistent with natu- with an attached polarizer. We con- filled, but inigbt be confused with ral origin. Also characteristic of many cluded that this feature was due to ones that are, so that we can docu- natural aiexandrites were the pro- twinning and not to color zoning, nlent thein for research and ecluca- nounced gowth zoni~lgand twinning. because the regions traded colors as tion pill-poses. Chrysoberyl, especially alexan- the polarizer was rotateti [figure 1, Last Fall, the West Coast lab drite, is stro~~glyplcochroic, typically left and right). ML! received for grading a 0.57-ct mar- displaying reit, orange, and green quise-cut dian~o~idthat had a IIIOS~ when viewed in different orienta- i~nusualfracture. The break exhibited tions. We have seen tivinnetl syn- DIAMOND both thin-film iridescence and orangy thetic crystals where one twin Untreated or Fracture-Filled? brown natural staining (figure 2), two appeared red, while the other looked Because the CIA Cell1 Trade features that might be confused wit11 green because of thc different orien- Laborato~yhas a policy of not issuing flash effects in filled diamonds. (For tations of the individual crystals. grading reports on fracture-fillet1 dia- more 011 such features, see the sec- Most of this natural stonc was of monds, any diamond siibinitted for tion on "Techniques to Identify one twin orientation, wit11 o~llya grading that ive determine to be so Fracture Filling,:,"pp. 169-1 73, in the sinall region near the culet helonging treated is instead issued an identifica- article "Update on Filled Diamonds: to the other twin. Although many- tion report. Therefore, once staff ldentification and Durability," Gems layer (polysynthetic) tivinning in genlologists in the diamond-grading d Gemology, Fall 1994.) This is the first diamond we have exanlined that contai~ledtwo such potentially con- Figme I. Wit11 a polarizer, tl~etruit~ platies in tllis 1.21-ct nlexandrite fusing features in a single unfilled look like brjghtly colored "~~orms"that cha~~gecolors lrs tl~epolal-izer break. However, this separation also rotates from olie orlcotation (left)to another (rirrl~t). sliows a "featlie~y"appearance, a fea- ture typical of an unfilled break and one that we have not seen to date in a filled break. H CI<

Edilor's /?ole. The ii~ilialsat Ihe end 01 each ilem iden- lily the conlfibuling editor !;tho provided lhal ilem Gems & Gemology, Yo/ 331. bio I. pp. 52-58 @ 1995 Gemological Iiatitule 01 Anierica

Gcm Trade Lab Notes GEiMS & GEMOLOGY explain the depth of the stone's body plane. Then there was a triangular color. inclusion that was thought to have Spectroscopic examination pro- originated as lonsdaleite that revert- vided strong hints that the color had ed to graphite (Fall 1994 Lab Notes, been artificially enhanced. We saw a pp. 185-186). Now, the West Coast strong 415-nm line, but not the rest staff has observed yet another type of of the Cape spectrum-that is, no triangular inclusion, this one in a lines at 435, 453, 466, and 478 nm. 1.74-ct marquise diamond. The pair of lines at 4961503 nm were A dianlond grader who saw the visible (with the 496-nm line being stone remarlzed that the reflective the stronger of the two), but there triangular features were reminiscent Figure 2. Both the thin-film iri- was no 595-nnl line. These features of platinum platelets seen in some descence and the naturiil color are consistent with treated color in a synthetic gems. However, platinum staining in this unfilled fracture diamond, but they still do not consti- platelets are opaque; these inclusons in a diamond might be confused tute proof of treatment. were not. Further, we are not aware indivi&zrrlly with the flash Infrared spectroscopy, however, of ally synthetic diamonds that were eljects typically associated with was conclusive. Weak, but definite, grown in platinunl crucibles, so there a filled fracture. Magnified 25x Hlb and Hlc pealzs in the infraretl is no reason why they would contain spectnlm of this type-Ia stone led us platinuin platelets. Synthetic dia- to conclude that the color was not monds may contain a niclzel-iron Treated Old-Mine Cut natural. It probably resulted from just flux, but we have yet to see such Just as a particular saturation and a light "lcleavage tative of natural trigon growth (or dis- ultraviolet radiation. There was inod- solution] feat~~reson the octahedral erate-to-strong yellow fluorescence face of the phantom crystal. The high (with one slightly blue zone) to short- Figure 3. The triangular growth relief of these trigons is attributed to wave UV, and no phosphorescence to features in this diamond form a gas that was trapped in triangular either wavelength. Although these plane associated with the octahe- voids that formed at the interface of responses were consistent with natu- dral face of an enclosed pl~anton? the enclosed octahedral face and the ral color in a diamond, nlicroscopic diamond crystal. Ma~nified30s. host diamond. Cheryl Y.Wen~ell examination failed to reveal features that we expect to see associated with such a color. For instance, there were SYNTHETIC DIAMOND, no irrarliation stains, although we did Treated-Color Red find four unusually glassy "melted- The Fall 1993 Gelns d Getnology lool

Gcrn Trade Lab Notes GEMS & GEMOLOGY Spring 1995 660 nm-including lines at ahout that enlerald (beryl) is uniaxial, with 595 and 635 nm, and an emission two distinct refractive indices for line at about 580 nm. Such a spec- light polarized in different directions. trum is typical of treated pink to red When a noncrystalline material [such color in both natural and synthetic as an oil, resin, or glass) and a crys- diamonds. talline solid have crossing dispersion On the basis of these test results, curves-refractive indices that match we identified the round brilliant as a at only one wavelength of light (i.e., synthetic diamond, treated color. at one color)-we see flashes of the RCK a11d SFM color at which the illclices match in brightfielcl illun~inationand flashes of its spectral-complementary color EMERALD, with Unusual in darlzfield illumination. [For a dis- Flash-Effect Colors cussion of flash-effect optids, see BOX Figure 4. The dark red colol. of this The Spring 1990 Lab Notes section B, pp. 156-157, in Icammerling et al., 0.14-ct (about 3.23-3.27 x 2.07 reported on an emerald that exhibited "An Update on Filled Diamonds: mn~)synthetic diamond is the orange and blue "flash effects" from Identification and Durability," Gems result of treatment. large filled fractures (pp. 95-96). Since d Gemology, Fall 1994.) Because then, both East and West Coast labs enlerald has two distinct refractive have seen this pair of flash-effect col- indices, however, blue (brightfield) In late October 1994, another ors in a number of clarity-enhanced and conlplen~elltaryorange (dark- dark red, 0.14-ct round brilliant (fig- emeralds. field) form one pair of flash colors, ure 4) was submitted to the West Last year, the West Coast lab seen in one optic orientation, where- Coast lab for determination of color observed sonle "new" flash effects in as orange-to-pinkish purple (dark- origin. Our suspicions were in~medl- an einernltl submitted for identifica- field) is seen in thc other orientation. ately aroused because red is excep- tion. A series of fractures extended The expected brightfield color, green tionally rare in natural diamonds, the length of this emerald-cut stone. to blue-green, is probably maslzecl by even as a treated color. We quickly When we examined them across the the emerald's body color. confirmed these suspicions when a stone's width, they displayed an In this stone, we also noted hand magnet picked up the speci- orange-to-pinkish purple flash effect sonle irregular, highly reflective bub- men-a reaction typical for synthetic (figure 5).However, when we viewed bles-as well as some white, cloudy diamonds that contain large metallic these same fractures down the length areas--in the filler: both of these fea- flux inclusions (again, see the above- of the stone, the flash effect was blue tures have been seen in other filled referenced Fall 1993 report). and orange. emel-alds. Magnification revealed a large One possible explanation for this RCK, IJotricia Moddison, and ML] metallic-appearing inclusion under unusual behavior relates to the fact the crown facets and clouds of pin- point i~lclusionsco~lfined to wedge- shaped areas. These latter zones were Figure 5. Filled fractures viewed across the width of this clarity-enhanced further defined by their green, hazy elnerald produce a clemly visible orange-to-pinkish purple flash effect in appearance; there was also a fairly darkfield illuniinotion. Different flosli colors were seen when the some distinct greenish zone in the center breolzs were es~~lnineddown tl7e length of the stone. Mogz~ified15x. of the round brilliant. When this syn- thetic dianlo~ldwas examined per- pendicular to the table (face up) with r both long- and short-wave UV radia- tion, we noted a cross-shaped area with moderate green fluorescence; the remainder was inert to long-wave UV, but fluoresced a faint orange to short-wave UV. These characteristics are all typical of synthetic tlianlonds (with the fluorescence colors consis- tent with treated-color red synthetic diamonds seen to date). Exalnination with a desk-model prisnl spectroscope revealed several absorption lines between 500 and

Gem Trade Lab Notes GElMS & GEMOLOGY mer-treated jadeites we have encoun- Spring and Summer 1984 Gem Trade tered thus far. Lab Notes, pp. 48 and 108, respec- MLJ and Dine DeGhionno tively). In December 1994, the West Treated and Untreated Coast lab was aslted to identify five Beads in One Necklace oval, translucent, mottled-green A conlprehensive report on bleached cabochons, which ranged from 6.38 and impregnated jadeite in the Fall to 7.12 ct. All had an appearance 1992 Gems ed Gemology (Fritsch et similar to that of good-quality al., pp. 176-187) described features jadeite. However, standard gemologi- Figure 6. In reflected light, the bot- diagnostic of this treatment. The Lab cal testing-including a 1.61 spot R.I. and an S.G. of 2.95 or 2.96-identi- tom of tl~ispolymer-impregnated Notes section has also reported on fied the five stones as nephrite. jadeltc cabochon shows strong evi- jade objects treated in this fashion, Magnification revealed dye concen- dence of selective grain etching. includiilg entries on necklaces con- trations (figure 7) like those seen in Magnified 25x. structed of treated beads (see, e.g., dyed jadeite and other siinilarly treat- Fall 1994, p. 187). Until recently, ed gem materials. The absorption however, our testing had shown that spectrum observed with a deslt- all of the beads in these necltlaces model prism spectroscope revealed a JADEITE were treated. broad (lye band in the red region, Bleached and Impregnated, with In November 1994, the West much like that seen in dyed-green Distinctive Surface Texture Coast lab was aslzed to identify a jadeite and quartzite. In these stones, The West Coast laboratory received double-strand ~~eclzlacewith 69 beads (7.88-5.08 mm) on one strand and 74 the diagnostic absorption band was for identification a 6.78-ct translu- centered at about 660 nm. All five cent, mottled-green cabochon that beads (8.18-5.49 mm) on the other. Standard geinological testing proved specimens were therefore identified revealed properties tmical for iadeite as dyed nephrite. R CIC jade, with an R.1, of 1.66 (by the spot that the translucent to semi-translu- method) and an S.G. of 3.34. cent, mottled green-and-white beads Chromium lines, seen with a spectro- were jadeite jade. A desk-model Blister PEARL scope, proved that the green color prism spectroscope showed chrome Attached to Shell lines, proving that the green color was natural. The stone was inert to Pearls submitted for identification was natural. One unusual feature short-wave UV radiation, but some frequently pose testing challenges. was noted during testing: So~neof areas fluoresced a faint green to long- Usually it is a mounting that inter- the bends on both strands fluoresced wave UV. feres with a complete examination, a weak-to-moderate, mottled yellow The gemstone had an unusual, but occasionally it is something dif- to long-wave UV radiation, while etched appearance in reflected light ferent. One such challenge was pre- others were inert. (figure 6): The aggregate stiucture of sented to the West Coast lab in the Most polymer-impregnated the cabochon appeared to be made up forin of a shell with an attached gray jadeite fluoresces. However, some of many interloclzing grains in vari- baroque pearl (figure 8). Our client untreated material may also fluorescc ous orientations, with certain grains stated that the item came from the preferentially eroded. Because of this a weak yellow to long-wave UV. In surface texture, we suspected that our experience, we usually do not get the stone had been bleached and such varied reactions from beads on Figme 7. At 15x magnification, polymer impregnated. Infrared spec- one necklace. Further testing, using infrared spectroscopy, proved that concentrations of green dye are troscopy coilfirmed this treatment by easily seen in this nephrite jade revealing several pealzs hetween 2800 there were treated (impregnated) and cabochon. and 3200 cin-1; the relative positions untreated beads in both strands. This and intensities of these peaks was our first encounter with such matched those of "polymer 5" in ICA mixed strands. RCIC (International Colored Gemstone Association) Laboratory Alert No. 75, 1993, issued by S. McClure and E. Dyed Green NEPHRITE Fritsch. Therefore, we identified the Of the two jades (jadeite and neph- stone as impregnated jadeite jade, rite), jadeite is much more common- natural green color, sometimes ly color enhanced [usually dyed Itnown in the trade as "B jade." green). Nevertheless, we occasionally Interestingly, the S.G. of this stone find evidence of green dye in nephrite was higher than that of most poly- submitted for identification [see, e.g.,

Gc~nTratle Lab Notes GEMS W GEMOLOGY Spring 1995 55 thickness of the shell. The X-radio- graph showed no bead nucleus, but rather a series of col~centricrings near the center of the pearl, proving natural origin (figure 9). Origin of color was determined using long-wave UV radiation. The mottled, weak, chalky reddish brown fluorescence of the pearl closely matched that of the grayish lip of the attached shell [again, see figure 8). Together with the physical appear- ance and the absence of any evidence of dye when examined with magnifi- cation, this verified the natural color of the pearl. Our final report stated that the pearl was a natural solid blis- ter pearl, with natural color. Of further interest, especially since we hatl not previously seen ally used in this context, were a series of CAT scan (Computerized Axial To~nogaphy)images of the pearl ant1 shell provided by the client (see, c.g., Figure 8. The grny brrroque pearl attached to this sl~ellmeasures figure 10). Altl~oughwe know of no 12.1 x 11.5 122111. data regarding CAT scan analyses of pearls, we do Itnow that axial tomog- raphy uses X-rays to lnalze a different Gulf of California, near La l'az, question first. Thc pearl's nacre type of ilnagc than that obtained by Mexico, on the east coast of Baja formed a continuous bridge to the conventional X-radiography. Con- California. The client wishcd In inothcr-of-pearl laycr 011 the shell, sequently, an attempt to interpret a Iznow: (I) if the pcarl was natural or proving natural attachment. CAT scan by conventional X-radio- cultured, (2)if its color was natural, Natural or cultured origin of the graphic analysis could well lead to an and (3)if it was assenlbled or natural- pearl was dctcrmined by X-radiogra- erroneous conclusion. For instance, ly attacl~cdto thc shell. phy; a slightly longer than usual the CAT scan in figure 10 displays Examination with low-powcr film-cxposure time was required in what appears to be a roughly spl~eri- m:~gnilicationanswered the last order to compensate for the extra cal "core" surrounded by an outer layer, giving the illusion that this natural pearl is bead nuclca~ed.Just as dental X-radiographs are inade- Figure 9. Conca~tricrillgs scer~ Figure 10. Altlzoz~ghthls CAT quate for resolving tho fine detail near the center of the penrl in scon imnge clearlj7 shorvs ll~e necessary for pearl identification, tl2i.s X-radiograph show that it is porn1 of attacllnle~~tof the pearl CAT scans cannot be used inter- natural. to the shell, the circular wl~lte changeably with specialized pearl- area could easily DL' ~n~smter- testing X-radiography. preterl ns the betrd nucleus of n Cheryl Y. Mrentzell cultured pearl

SAPPHIRE, with Diffusion- Induced Color and Star Recent rcports in Bangkok jewelry publications indicate that corundulll with diffusion-induced stars is now appearing in the trade, with a large quailtity soon to be released on the worlcl market. For example, the labo- ratory of the Asian Institute of Gemological Sciences (AIGS)in

Gein Tradc Lab Notes GElMS & GEMOLOGY Spring 1995 F,,,,e 12. When the stone in figure 1 I is immersed in methylene iodide, there appears to be a "red" color confined to the surfice.

Banglasterism is located at, or just the iron present in the cabochon such stones typically reveal heavy beneath, the surface (see "AIGS Finds reacted with the titanium in the dif- absorption bands in the 450-470 nnl More Stars," Iewellery News Asia, fusion process, resulting in a blue range when examined with a spectro- June 1994, p. 74). Unlike inost natu- component that ultinlately produced scope. ral asterism, the individual needles the purple hue. GRC Green star sapphires are even in the induced-star sapphires us~ially less common. One such natural stone cannot be detected at lower inagnifi- was reported in the Spring 1989 Gem cations. SYNTHETIC STAR SAPPHIRE Trade Lab Notes section (p. 39). It The East Coast Gein Tracle with an U~iusualColor was thus with interest that, in late Laboratory has also exanlined a In our experience, green is an uncom- Fall 1994. the West Coast lab exam- slightly different diffusion-treated mon color for gem sapphires. Those ined a ring set with the translucent, star corundum. The 23.26-ct purple we do encounter are almost invari- asteriated oval cabochon shown in stone in figure 11 illustrates the ably iron-rich stones, such as the figure 14. The stone had a spot R.I. of cloudy appearance of the induced ones from Australia, that have been 1.76, and a uniaxial optic figure was star. Routine examination revealed easily resolved in the polariscope. that the starting inaterial was natu- Examination with a desk-model spec- ral. On close examination with the Figure 13. These red spots and troscope failed to reveal iron-related stone inlinersed in methylene iodide, "cloud" seen on the dome of lhe features, although we did detect an we noted a thin layer of red that stone in figure 11 at 45x ~nagnifi- absorption line at 670 nm, which we appeared to be confined to the sur- cation provide proof that the red attributed to cobalt (see below). face of the stone [figure 12). On the surface color, lil

Gem Tradc Lab Notes GEMS h GEMOLOGY Spring 1995 For further documentation, GIA Russia," Gems d Gemology, Sum- Research performed chelnical analy- mer 1993, pp. 81-98), sis using energy-dispersive X-ray flu- A 1.37-ct orangy rcd oval mixed- orescelice (EDXRF) spectroscopy. cut spinel, which arrived in the West This revealed cobalt (Co)as the prob- Coast laboratory in Decemher 1994, able coloring agent. Using detailed presented a slightly cliffcrcnt identifi- UV-visible absorption spectroscopy, cation challengc. It contained several we had previously docu~nented inclusions: octahedral crystals (possi- cobalt (as Co3+) as the sole coloring bly negative crystals] scattered agent in a rarely encountered non- throughout the stone, as well as large phenomenal green synthetic sapphire feathers, some with brown iron produced in Switzerland. The non- staining. Aillong the latter was a phenomenal green synthetic sap- prominent dendritic iron stain (figure phires produced in thc U.S. that GIA lhis synthetic star sapphire (about 15) that could be mistalien by an Research and the GIA Gein Trade 14 x. 12 x5.90 lnm) is probably unwary gemologist for the yellowish Laboratory have studied are colored dzle to Co3+. white to yellowish brown flux seen hy a combination of cobalt (Co3+)and in some synthetic spinels. The stain vanadium (V3+).To date, we have not probahly resulted froill iron-rich flu- doculnentecl cobalt as a coloring Figure 15. Tliis dendrite in (1 1.37- ids penetrating a secondaly fracture agent in natural sapphires. ct natural spinel could be con- that for~llctlafter the c~ystalgrew. In On the basis of these tests, we firsed with a flzlx inclusion in n addition, according to the Mulil- identified this cabochon as a synthet- svnthelic suinel. Ma~nified20x. meister et al. article cited above, ic star sapphire. sonle flux-grown synthetic spincls HC1< and Emmanuel Fritsch contain "pyramid-shaped phantoms in near-perfect aligllnlcnt with the external faces and cdges of the octa- hedra." These could also be confused SPINEL, Natural with a at first glance with the octahedral- Dendritic Inclusiorl crystal inclusions in this stone. Frequently, identification of spinel MLI requires advanced testing techniques. Many spinels do not contain charac- II teristic inclusions that could be used to distinguish natural stones from their synthetic counterparts, espe- cially from flux synthetics. In these ~.~lvlcClureprovided ligures 4 and 7: Figures &and 14 cases, EDXRF spectroscopy is needed Yere laken by Maha DePAaggio. The X-radiograph ill to ~naltethe separation (see, e.g., I ligure 9 is by Cheryl Y Vlenlzelland Palricia Maddison Mul~lmeisteret al., "Flux-Grown (figure 10 :was provided by the clienl). Nicholas DelRe Synthetic Red and Blue Spinels from 2 supplied /he pholographs in ligures 1 1-73.

Gem Trade Lab Notes GEMS & GEMOLOGY I11 what has become an eagerly awaited annual pil- more reminiscent of swap meets than sophisticated grimage, Gem News editors traveled to Tucson, jewelry industry events. And, throughout the almost Arizona, again this February to attend the inany over- two-weel< experience, there are classes and lectures lapping trade shows in this Southwestern desert com- for those who wish to colnbine formal education with munity. In doing so, they joined thousands of others- buying. including gemologists, retail jewelers, gein and miner- Helping the editors and regular contributors with al dealers and collectors-in what is often referred to this Gem News section were Dr. Sang-Ho Lee, Shane as the "Tucson experience." F. McClure, Thomas M. Moses, and Cheryl Wentzell, This experience has something for anyone with of the GIA Gem Trade Laboratory; and Dr. James E. an interest in minerals and the gein materials that are Shigley, Dr. Ilene Reinitz, Sam Muhlmeister, and Yan fashioned from them. The large gem shows, such as Liu, of GIA Research. Because of the scope the 1995 the A1ne;ican Gem Trade Association (AGTA) show Tucson shows, this report will continue in the Sum- at the Tucson Convention Center, draw many retail- mer issue. ers froin all over the world. When that show closes, the Tucson Gein and Mineral Society takes over the same venue for what was historically the main show, DIAMONDS the nucleus around which the others have grown. New diamond cuts. Two new trademarl

Gem News GEMS & GEMOLOGY Spring 1995 riglrre 3. Note the disancr spoKes ln this 59.04-ct elongated trapiche emerald crystal Fig~~re2. This 2.46-ct. cat's-eye emerald from Colombia. Courtesy of Guillernlo Orliz; (7.38-7.89 x 5.61 mm) is from Santa Terezinha, photo 0 GIA and Tino Hammid. Brazil. Courtesy of David I

GEMS & GElMOLOGY Spring 1995 Among the attractive trapiche emeralds seen diffuser1 chatoyant band. We also saw carvings in non- were a matched pair of heart-shnpetl cabochons with a pl~enomenalblack obsidian and carvings exhibiting a total welght of 24.71 ct, offered by Eminent Genls, of golden sheen. One vendor had nlodels of tetragonal New Yorlc. We saw an exccptional 59.04-ct colunlnar crystals fashioned from obsidian, approximately 12.5 trapiche emerald crystal (figure 3). It was interesting cm (5 inches) long. to note the great variation, along its length, in the Other types of nat~lralglasses were seen in both concentrations of carbonaceous inclusions protlucing rough and fashionetl form. These included moldavite, the "spolandradite garnet reportedly from the Rcpublic of Mali in Western Africa. A number of deal- ers offered similar material at Tucson, also described Figure 4. lri~fescent"rain bow" hematite is now7 as fro111 Mali. Thomas M. Schneider had n dozen or so being zlsed in jewelry. These pieces were fash- faceted stones that ranged from light greenish yellow ioned by Bill Hel~er;courtesy of Zoltan David. to mediunl brown, ant1 a 200-ct parcel of s~nall Photo 0 GIA and Tino Hammid. "cobl>ed" rough that would produce finished stones of I ct or less. Mr. Schneider said that approximately 300 kg of these garnet crystals had just been received in Australia. Sonle 3 kg (15,000 ct) of very gemmy facet- grade material was quicl

Fashioned natural glasses. Iridescent "rainbow" obsid- ian, which we first described in the Sum~ner1993 Gein News section (p. 133),was available from several dealers and in a variety of fashioned forms. Carvings included both geo~netricand representational shapes, an exanlple of the latter being fish figurines sporting - iridescent scales. Slightly dollled tablets, free forms, and oval cabochons were alnong the pieces fashioned for jewelry. Some of the oval cabochons, offered by Genl Marlteting International, of Roclzvillc, Maryland, were fashlonetl to show the iridescence as a somewhat

GEMS & GEMOLOGY Spring 1995 the birefringence was somewl~athigher than the 0.008-0.009 that is typical for transparent iolite. Another vendor, Saint-Roy Gems and Minerals, of Antananarivo, Madagascar, and Paris, France, was offering petrified wood from the Morondavo region. We also saw two feldspar gems-yellow orthoclase and phenon~enallabradorite-as wcll as seinitranslu- cent-to-opaque, well-formed hexagonal tabular crys- tals of ruby ant1 transparent purple scapolite crystals. Soine blue sapphires of good color were available from Madagascar. A report on their genlological prop- erties will appear in the Summer 1995 Gein News section.

Iolite and other Orissn gems. Last year's Tucson report mentioned at least eight gein n~aterialsfrom the Indian state of Orissa. Many of these were seen at the show again this year, including significant ainounts of iolite, much of it calibre cut. We found Figure 5. This 1.33-ct iolite (7.63-7.77 x5.11 that in the great majority of cases where the source of mm) ~vasfashioned from moterial lz~inedat (7 iolite was identified, that sourcc was Orissa. new loc~llitj~,Amhovolnhe, ill the {ar sonth of The Orissa area is actually producing ',I 11 even Madagascar. Photo by Maha DeMaggio. greater variety of gems, according to Asholt Iagate for additional strength. In amounts of green and yellow beryl, eineraltl (including this baclted form, according to 1Mr. Heher, the materi- some crystals of gootl-but not exceptional-color), al appeared to be quite durable and hat1 presentetl no amethyst with good color, green tourn~alineand indi- probleins during setting. colite, cat's-eye sillimanite, darlt brown sphene, pinlt Rainbow henlatite has drawn the attention of zircon, darlt green apatite, and pinlt fluorite of very some respected jelvelry designers. Zoltan David, of good color. Rhodolite and hessonite garnets were both Austin, Texas, inclicated that he was tlesigning a line available in large quantities. The l~essonitewas gener- of jewelry incorporating it. ally highly transparent, laclting the "roiled" appear- ance of Sri Lanltan stones. Also seen was some irradi- Iolite aid other gems from Madagascar. One gein we ated blue topaz, the starting material tor which report- saw for the first time this year was iolite frorn edly came froin Orissa. Madagascar. The inaterial was only tliscovered in Cat's-eye quartzes from Orissa were also avail- 1994, at a localitv callecl Ainbovoinbe in the far south able. According to Michael Randall, Gem Reflections of the island nation, according to a representative of of California, San Ansellno, his material canle froin Le Mineral Brut, Saint-Jean-le-Vie~~x,France. an area that also produces the cat's-eye chrysobcryl. In Although many crystals have been found, he reports, general, cat's-eye chrysoberyl seen this year was of far they are typically very highly included, greatly retluc- better quality than that seen in past years. We also ing the yield of facet-grade material. Nevertheless, the saw a few cat's-eye alexandrites, some with a fairly firin was offering faceted stones up to 2-3 ct as well as clarlt body color and prolninent color change. some interesting small (about 1 cin) cubes, fashioned Templc Tratling, of Encinitas, California, had to display the material's strong pleochroisin. cat's-eye diopsides from Orissa with a much purer and One round brilliant (figure 5) was purchased for somewhat lighter color than what is typically seen. gemological characterization. Properties tleternlinecl When illuminated from above, these stones exhibited were: R.I. = 1.536-1.548; birefringence = 0.012; S.G. a bright, 1necliun1-dark-green body color and a lighter, (determined hydrostatically) = 2.59; strong trichroism distinctly green chatoyant band. This inaterial in colorless, light bluish violet, and mediunl violet; remintled the editors inore of cat's-eye tourmaline inert to both long- and short-wave UV ratliation; and a with an atypically sharp eye than of the usual Indian wealt absorption spectruin typical of iolite (see, e.g., cat's-eye diopside. R. T. Liddicoat's Handbook of Gem Identificatioll, 12th ed., 1990, p. 146). All these properties were con- More ineteorite products. In our 1993 Tucson report sistent with those reported in the literature, although [pp. 55-56), we noted the availability of extraterrestri-

GEMS & GEMOLOGY a1 materials. These included jewelry set with "Gibeon class" iron-niclzel ineteorite slices from Namibia and free-form "gems" fashioned froin the large pallasitic meteor fount1 in Esquel, Argentina. Both at the 1994 show and again this year, the editors saw these and several other meteorite inaterials being offered by the firin Robcrt Haag Meteorites, of Tucson. Aillong the specimens were pieces from the pallasitic meteorites that struck near Mount Vernon, Kentucky (in 1868)) and near Brahin, Russia (discovered in 1810). Also offered were pieces of the niclzel-iron ineteorite that fell near Odessa, Texas, about 50,000 years ago; this material exhibits the same "Widmanstatten" lines that are seen in the Namibian inaterial (see Sumrner 1992 Gein News, figure 6, p. 133). New "products" fashionecl from nickel-iron meteorites include spheres, flnger rings, g~ritarpicks, and buttons. This year, we also saw a inaterial that Figure 6. R~lssiois the SoLlrce of this polished ly reselnbled pieces of a peridot-containing nicltel-iron 58.22-ct "strawberry" cILlflrtz cr~stol.Courtesy yallasitic llleteorite. TIlese speci1ne-l~of peritlot in of Ilidifh Whilehead; ph0loOG1A alld Tino basalt matrix froin Sail Carlos, Arizona, had been coated with shellac. Although the purpose of the treatment was to keep the small, fragile peridot crys- tals from separating from the matrix, the coating gave the basalt a metallic appearance. Tucson by the fi'rin Gebr. Henn, of Idar-Oberstein, Germany. This exceptioilal cushion-cut Pakistani gem weighed 309.90 ct. Gebr. Henn also had another Miscellaneous notes on peridot. As is often the case, cushion cut of 108.37 ct and a 125.25-ct round bril- inuch of the larger high-quality peridot seen in liant cut. Tucson was reportedly from Mynninar. Gem Tech, of Many crystal specilllens of Pakistani peridot were New Yorlz City, showed off three attractive stones- available at the mineral show. We noted that some weighing 45.87, 49.47, and 61.16 ct. Andrew Sarosi, of had very large, acicular inclusions (see, e.g., Winter Los Angcles, had a nicely cut rectangular cushion 1994 Ge~nsed Gemology, p. 275). weighing 57.82 ct. Direct Gems, of New York, had a 47.25-ct cushion cut and three other stones in the 20+-ct range. "Strawberry" quartz and other Russian gems. Natural Peridot from a new source, Pakistan (Fall and and synthetic gem nlaterials are coming from Russia Winter 1994 Gem News, pp. 196 and 275, respective- and other republics of the former Soviet Union at ly), was seen for the first time at the Tucson shows what appears to be an ever-increasing rate (see also this year. Dealers included John Bachman, of Boulder, entries elsewhere this section). This trend was again Colorado, who had two well-cut 27-ct stonesj Jonte evident at Tucson. The drusy uvarovite garnet Berlon Gems, which had about a dozen stones in the described in our Tucson '94 report (Spring 1994 Gems 3-8 ct range; and Shades of the Earth, of Phoenix, eS Gemologj~,pp. 53-54) was available in both rough Arizona, which provided information for the previous- and tablet forin (including matched pairs for use in ly cited Gem News entries and had a largc selection of earrings) from several dealers. Interestingly, some this material. Pala International, of Fallbrook, dealers were marketing this material by the square California, had a nice selection of about 50 stones. centimeter. The largest weighed 74.5 ct. "Peridot from Pakistan," One Russian gem seen at Tucson for the first a report in the January 1995 Gein Spectrum, the firm's time was "strawberry" quartz. Short crystals with pol- newsletter, said that the material is similar to peridot ished faces (figure 6) were offered by Judith from Myaninar and Egypt (Zabargad Island) because it whitehead, of San Francisco. The color-causing inclu- forms in pockets in high-temperature veins. This sions in this material seemed to be a lighter and inore inode of forlnation explains why the new Pakistani saturated color than that seen in similar material material often occurs in large, transparent, euhedral from Mexico. With magnification, however, the inclu- crystals-in contrast to peridot from localities such as sions appeared to be goethite and lepidocrocite, the San Carlos. Arizona, where the inaterial forms as same minerals found in the Mexican material. xenocrysts in basalts. The largest faceted peridot we Also encountered was Russian amazonite, as have seen to date from this locality was offered at large tumbled pieces and steep-angled pyramids (fig

Gem News GEMS & GEMOLOGY Spring 1995 63 132; emerald-in-matrix from Brazil, fashioned as carv- ings and spheres, was also available]. Faceted Russian gem materials included sinall quantities of demantoid garnet, alexantlrite, and emerald.

Miscellaneous notes on sapphires. Malhotra Corp., of New Yorlz City, had an interesting 1.38-ct einerald- cut hicolored sapphire that was about two-thirds blue ant1 one-third light yellowish brown. The border between the two colors was distinct. The House of Williams, Loveland, Colorado, had a small quantity of blue sapphire rough containing yellow cores. This rough inaterial reportedly came froin Burundi about three years ago. Apparently, the unsettled political cli- lnate (especially in neighboring Rwanda] is limiting developnlent of the deposit. Both ruby and sapphire from gem gravels in Sierra Leone, West Africa, were being inarlzeted by Figure 7. This arn, lite pyr :d is frorn on Tideswell Dale Roclz Shop, of Derbyshire, United unspecified locality in Russia; the slmoss agate, jasper, Siberian nephrite, some range; he described them as a "wine" or "burgundy1' calcite-rich lapis lazuli, snlall quantities of jadeite, color. yellowish green serpentine, and charoite. We were A sinall percentage of stones from this region also shown a few cabochons cut from Uralian emer- exhibit a color change, ranging fro111 weak to distinct. alds in their feldspar matrix. These appeared to be In general, the change was from grayish bluish green very similar to en~erald-in-matrixcabochons fro111 to reddish brown. This is more reminiscent of the North Carolina (Summer 1993 Gems d Gemology, p. color change displayed by vanadium-doped synthetic

GEMS W GEMOLOGY Spring 1995 Figure 8. Tl~esefive sap- phires, ranging fro117 1.32 to 2.14 ct, show different col- ors ill il~cclndescellr(left) and fluorescent (right) light. They are all reported- ly from o new deposit in sourherl~Tcznzcznia. Cozlrtesy of DW Enterprises; photo 0GIA oncl Tino Haminid.

color-change sapphires and some alexandrites, rather Faceted sphalerite and other collector stones from than the violet-to-purple effect shown by some sap- Canada. Tucson is the dace where collectors of rare phires froin other sources-Sri Lanka and Montana, gems congregatein search of the new and unusual. for exalnple. We saw a number of new gem nlaterials that we had Bill Marcue, of DW Enterprises, Boulder, not documented before, as well as some exceptional Coloratlo, subsequently loaned the editors 11 sap- examples of already known species. phires (1.06-2.16 ct] fro111 southern Tanzania. The One of the editors (EF) was shown an interesting stones all showed sonle color change between incan- collection of gems fro111 Quebec and elsewhere in descent and daylight-equivalent fluorescent light (see, Canada by Montreal collectors Guy Langelier and e.g., figure 8). With incandescent light, colors ranged Gilles Haineault. Many of these gems were fro111 Mt. from pink-brown through brownish pink and purple; St. Hilaire, outside Montreal; several of these are untler fluorescent light, colors iilcluded yellowish noted below because of their quality or rarity. brown, grayish violet, and grayish greenish blue. Faceted spl~olerite,ZnS, fro111 Mt. St. Hilaire is Docurnentation of both the faceup colors and color not new, but the 3.98-ct hexagonal cut in figure 9 is change was conlplicated by strong color zoning. For the first we have see11 in this saturatetl, medium instance, anlong the distinct colors noted with magni- bluish green, devoid of any yellow overtones. The fication (in incandescent light) were yellow, green- nicest green color previously seen was a medium-dark blue, blue, purple-pink, and brownish pink. Other fea- yellowish green. (Other colors include yellow, brown, tures seen with magilification iilclude dark red-brown red, orange, and black; it also nlay be colorless.) The crystals (possibly rutile), exsolution needles (also nlost R.1, of this particular stone was over the linlits of the liltely rutile; some showed twinning), pinkish orange gemological refractometer. S.G. was measured at 4.10, crystals that loolted like garnet, and light brown crys- typical for low-iron sphalerite. We observed a weak tals similar in appearance to clinozoisite crystals orange fluorescence to long-wave UV radiation only. found in Molltana sapphires fro111 Dry Cottonwood In thc hand-heltl spectroscope, we saw weak, moder- Creek. ately broad bands at about 480, 500, and 560 nm-and Gemological properties for these stones were: R.1. a stronger one at 590 nin. There was total absorption of o = 1.770 to 1.771, E = 1.761 to 1.762, ant1 birefrin- above 620 nm. EDXRF ailalysis revealed that the gem gence = 0.008 to 0.009; S.G., determined by hydrostat- was almost a pure zinc sulfide (which confirmed the ic weighing, of 4.00-4.03. All the stones appeared red identity of the stone). There was a sinall peak for cal- through the Chelsea color filter, were inert to both cium, and traces of manganese and iron. long-snd short-wave UV radiation, and showed an Origin of color was established by UV-visible absorption spectrum conlbining absorption features spectroscopy, which showed sinall absorption bands typical of ruby and pink sapphire with iron-related at 475, 491, 564, and 591 nm, with total absorption lines at about 450, 460, and 470 nm. Energy-dispersive between 660 and 740 nin. These features come from X-ray fluorescence analysis confirmed the presence of Co2+. None of the iron-related absorptions ltnown to both chroiniuln and iron. cause a yellow-to-brown color component were

Gem Ncws GEMS & GEMOLOGY Spring 199s 65 Shortite is a rare sodium-calcium carbonate, Na2Ca2(C03)3,which occurs very sparsely in the Mt. St. Hilaire deposit. Most crystals are less than a mil- limeter in longest dinlension and yellow. However, some exceptionally large and transparent crystals have been found and faceted (the largest-lznown gem is a 3.52-ct yellow square cut). We examined a 0.23-ct, slightly greenish yellow, cut-corner rectangular mixed cut (again, see figure 9). Such stones are strictly col- lectors' items, as the crystals are water soluble. We noted a moderate pleochroism from very light yellow to greenish yellow and colorless. R.1.l~were a = 1.530, and both P and y very-close to 1.568, for a birefrin- gence of approximately 0.038. S.G. was measured hydrostatically (quiclzly and without damage to the Figure 9. Rare stones from Mt. St. Hilaire stone] at 2.58. The stone was inert to UV radiation incl~~de(cloclzwise lrom upper left) remondite-(Ce) and did not show anv s~ectralfeatures in the hand- (1.70 ct), sugilite (4.14 ct), sphalerite (3.98 ct), , - held spectroscope. With magnification, we noted sev- manganotychite (0.28 ct), shortite (0.23 ct), and eral Ilealed fractures with liquid or two-phase inclu- sernndite (0.79 ct). sions. We confirined the identity of this gem by exploring its cheinistry with EDXRF analysis. A peak for sodium established that this element was a major constituent; calcium was also present in large observed, as one would expect since only a trace of amounts. We did not detect any other element. iron was found. It is not surprising that EDXRF did not detect cobalt, although we believe cobalt-a very Mangnnotychite, Na6Mn2(S04)(C03)4,is a very strong absorber of light in even minute amounts-to rare mineral in itself. Rarer still are large transparent be the cause of color. A few parts per million Co24, crystals that can be faceted. This species also is water which is below the detection limit of our instrument, soluble, making faceting a real challenge. We studied is sufficient to cause this intensitv of coloration. a 0.28-ct emerald cut [again, see figure 9). The stone Finally, the trace of Mn2+ present may explain the appeared brown, but on close examination that color orange luminescence. was found to be due to a superficial coating, presum- Serandite, Na(Mi1~+,Ca)~Si~0~(0H),is one of the ably resulting from alteration of the surface. We found best-lznown minerals found at Mt. St. Hilaire. It usu- no pleochroism, and the R.I. was 1.552. The specific ally crystallizes as blade-lilze, opaque, slightly pinltish gravity, measured hydrostatically (also quiclzly), was orange ("salmon") crystals. Very rarely, parts of these 2.83. The stone was inert to UV radiation: it showed a crystals are gem quality. Most faceted serandites are sharp band at about 414 nin in the hand-held spectro- very small, or not totally transparent. The 0.79-ct scope (presumably due to Mn2+). Magnification stone we studied recently is unusual for its size, re- revealed healed fractures and negative crystals. markable transparency, and very bright orange color EDXRF analysis confirmed the presence of sodiunl, (again, see figure 9). Its indices of refraction are manganese, and sulfur as major components. However, it also revealed fairly large amounts of iron, approximately a = 1.679, P = 1.680 to 1.68 1, and y = 1.71 1. S.G., measured hydrostatically, was 3.47. The although we would estimate the iron concentration to stone was inert to UV radiation. In the hand-held be significantly lower than the manganese concentra- spectroscope, it showed total absorption up to about tion. This confirmed that the gemstone was nlangano- 420 nm, and a strong, broad band centered at about tychite, but not a pure end member. 520 nm. UV-visible absorption spectroscopy con- Sugilite, ISouth Africa. We were surprised erately broad band centered at 520 nm. These features to discover that large single crystals of light purple are typical of Mn2+ absorption. EDXRF analysis sugilite had been discovered in 1994 at Mt. St. Hilaire. demonstrated that silicon and inannanese were major We examined several dozen such crystals, which coinponents (sodium cannot be dvetected with o'ur ranged from a few millimeters to over 2 cm. They instrument at the expected concentration], with exhibited a hexagonal pyramid habit, truncated at the ininor calciunl and zinc. The predominance of man- top by a basal plane. Some crystals contained portions ganese over calcium in the compositioil is typical of that were transparent and suitable for faceting. We crystals from Mt. St. Hilaire. borrowed one such crystal for further study.

Genl News GEMS & GEMOLOGY Spring 1995 Only a very weal< pleochroisin in slightly differ- ent intensities of purple was detected. In the hand- held spectroscope, the crystal showed a strong absorp- tion at about 420 nm. It was inert to UV radiation. EDXKF analysis showed silicon, iron, and sodium to be major compollellts; potassium as a minor con- stituent; and traces of zinc, rubidium, calcium, titani- um, and zirconium. We obtained polarized UV-visible absorption spectra of the crystal using a Hitachi 4001 spectropho- tometer and calcite polarizers. The two spectra were very similar, with two intense doublets at 350-363 nm and 413-419 nm, and two weak, broad bands cen- tered at approximately 555 and 770 nm. Such an absorption spectrum is typical of Fe3+ in octahedral coordination. Although the color was very similar to that of manganoan sugilite from South Africa, the cause of color was different. The band at 555 nm induced the purple color, while the doublet at 413-419 nin was the feature seen in the hand-held spectroscope. The 365-nm peak was stronger than the Figure 10. Under its Swarogem brand name, 350 peak in the Elc direction compared to the D. Swarovslzi and Co. is manufactz~ringnnd E llc direction. We could not see any significant dif- marketing calibrated natul.al peridot, amethyst, ference in the intensity of the 555-nm band between citrine, and rhodolite in novel bubble-wrap the two orientations, which explained the very weak paclznges. Photo O'Harolcl d Erica Vnn Pelt. pleochroism. Leifite, Na2(Si,A1,Be)7(0,0H,F]14,a rare alltali pegmatite mineral, is generally found as sinall acicu- lar crystals unsuitable for faceting. We studied a light purplish pink, 2,30-,-t lnodified fan-shape cut. The duced at Tucsoil by an Austrian firm well known for pleochrois~nwasfaint,fromaslightlylightertoa its~~ntlleticgelnsandcut-cr~stal~roducts. slightly darker pink. The optical character was uniaxi- Applying its expertise in state-of-the art automat- al positive. R.I.'s were o = 1,517 and E = 1,522; s,~. ed cutting, D. Swarovslti and Co., of Wattens, ~~rol, was about 2.62. The stone was inert to both long. and is initially offering faceted peridot, rhodolite garnet (in short-wave UV radiation and did not show any notice- "orchid," "pink rose," and "raspberry"], anlethyst able absorption bands in the hand-held spectroscope. ("lilac" and "violet"), and citrine ("saffron" and "golden"). Numerous needles parallel to the optic axis (and per- Each type will be available in 50 sizes and shapes, pendicular to the table] were seen with magnification, including round and "princess" cuts (2-4 nlm), square some arranged in a step-like growth pattern. step cuts (2-2.5 mm), ovals and pear shapes (5 x 3 inm EDXRF confirmed the presence of the lnajor coin- and 6 x 4 nlm), and lnarcluise and baguette cuts (5 x ponents: sodium, aluminum, and silicon. Also detect- 2.5 mln and 6 x 3 mm]. According to a recent press ed were potassium, manganese, iroll, zinc, gallium, release, all Stones Are CLlt to a maximunl of O.lO-lnm rubidium, and cesium. ~~-~i~ibl~and near-infrared tolerance. The Conlpany experimented with larger absorption spectroscopy showed a sharp band at about sizes but foulld the yield too low given the clualit~ 375 nm (typical of Fe3+),with a complex broad band- standards, according to Swarogelll Product Manager the cause of the color-centered at about 485 nm, and St~hellKahler. Swarogem has the capacity to cut another broad band centered at about 1160 nm. We do about 300,000 synthetic and natural gemstolles a day, not know the reason for the 485-nm broad band. he said. Again seen this year from Mt. St. Hilaire was the To meet promised consistency in color and quali- so-called burbankite, which has been shown to be ty, m~lchas possible Swarogeln buys each gem actually remondite-(Ce), Na3(Ce,La,Ca,Na,Sr)3(C03)5material (in extremely large quantities) from a specific (also in figure 9). For more on this material, see region: For example, the peridot is from Arizona, and the Winter 1992 Gem News, pp. 270-271. the citrine is from South America. Initially they ~~sed rhodolite garnet from India, but thought the overtones too bluish gray. They switclled to rough froin an Swarovslci debuts machine-cut gems. A new line of unnained locality in Africa, according to Mr. I

Gem News GEMS & GEMOLOGY Spring 1995 67 topazes with images-such as Abraham Lincoln and the Great Seal of the United States-carved intaglio- style into their table facets. The firm also had some sizable faceted aquamarines from this source, includ- ing a 108-ct stone. Tais-Panin also fashions copies of famous gems, according to a brochure from the firm. Also exhibiting was gemologist Vladislav Iavorskii, froin Kiev, Ukraine. Arnong materials that he offered were fashioned beryls and topazes from the Wolodarsli pegmatite, including a 105-ct faceted aqua- marine egg, some bicolored topaz, and a 35.40-ct orangy brown egg faceted from topaz.

Miscellaneous notes on tourmaline. The bright tour- malines from Paraiba, Brazil, were scarce this year. What limited quantities we saw werc primarily in the greenish blue to blue-green range ant1 in melee sizes. In their absence, people loolietl for similarly coloretl tourmali~lefro111 other sourccs. Ron Ohm Exotic Stones, of Carmel, California, was offering material fro111 Sahia, Brazil, with a color similar to, but lcss saturated than, that of Paraiba tourmaline. We also saw some fine-quality tournlaline with a color closc to that of the greenish blue Paraiba stones. Reportedly Figure 11. This 141.79-ct bicolored topaz was carved it came from the Ara~uaiarea of Minas Gerais, Brazil. from lnaterlal mined from the Wolodarslc pegmatite, Good blue-to-greenish blue stones are also coining in the Ukrmlne. Cozlrtesy of Tzlr~nalied Herschecle, from the area of Itambacuri, a town about 35 km Sonibcl, Florzda, photo by S11ane F. McClz~re. so~~thwestof Tebfilo Otoni, according to Jerry Manning, of MClM Gems, Middletown, Ohio. He said that about 300 ct of exceptional material was pro- duccd in late 1994. He loanetl us a 30.17-ct oval modificd IZahler said. Clarity grading is lnore narrow. The com- pany offers only one clarity grade: eye-clean in the faceup position. Stones come in distinctive, sealed bubble-wrap Figure 12. This 30.17-ct tourmaline (18.82 x pacliages, reminiscent of tamper-proof pacliaging for 17.79 x 13.42 lnm) was mined near the town of over-the-counter pills (figure 10). In the fall, Itambacuri in Minas Gerais, Brazil. Courtesy of Swarogern hopes to offer calibrated machine-cut ruby MCM Gems; photo by Shane F. McClure. and sapphire in small rounds and squares.

Topaz and beryl from the Ukraine. In the Spring 1993 Gem News section, we reported on large greenish yel- low to yellowish green beryls from a major pegmatite at Wolodarsli, Uliraine (pp. 54-55). A later entry (Summer 1994, p. 128) described topaz, sonle of it bicolored, from western Russia and the Ul

Gem News GEMS & GEMOLOGY Spring 1995 brilliant that was recovered from this area in Nov- ember (figure 12). EDXRF analysis revealed that the stone contained 110 copper, which is the primary col- oring agent responsible for the vivid blue-to-green col- ors of the Paraiba material. As for other colors of tourmaline, Fernando Otavio da Silveira, of Braz-G-Can International Trading, Rio de Janeiro, had some interesting bicol- ored Brazilian tourmaline-matched pairs cut perpen- dicular to the crystal's c-axis, lilze waternlelon tour- maline is typically cut. However, unlilze watermelon tourmaline, this material had a medium green core surrounded by a light "mint" green outer layer.

INSTRUMENTATION Inclusion-viewing system. Gemstone inclusions, which frequently provide diagnostic information for Figure 13. The GEM MicroVision system experienced gen~ologists,call also be fascinating to i~~clndesa CCD camera that is attached to a observers of any level of sophistication. Photo- microscope eyepiece, a high-resolution video micrographs have even been used to marlzet amber monitor, and a color printer. Courtesy GIA with insect "incl~~sions"(see, c.g., Summer 1994 Gem GEM Instrrrments. News, p. 124). Recently, GIA GEM Instruinents introduced a system for displaying the internal world of gemstones. At 'l'ucson, Scottish gemologist Alan Hodgkiilson The GEM MicroVision system collsists of a CCD showed the editors a magnetic device developed to camera, which attaches to the eyepiece of a geinologi- detect the magnetic properties of certain synthetic cal microscope. The camera is conilected to a high- diamonds. Called the "Magnetic Wand," it consists of resolutioil video lnonitor and a color printer (figure a small, powerful neodyn~iuiniron boron magnet 13).The GEM MicroVision system can display just about 5 inm in diameter that is mounted on a 60-mm- about anything that the inicroscope "sees" under vari- long wooden rod. According to literature provided to ous lighting inethods (darlzficld, brightfield, fiber- the editors with a sample magnet, neodymium iron optic, polarized). Calibrated and proportion-indicating boron alloy is the most compact magnetic material eyepieces can also be used. The system adjusts to available. Experiments by Mr. Hodglzinson proved iniinic different color temperatures, allowing the that the instruineilt is very effective in detecting mag- operator to show true colors or to enhance contrast. netism in a range of synthetic diamonds. Preliminary Such a remote-viewing technique has obvious testing by the editors on De Beers, Sumitorno, and advantages over a traditional microscope. For exam- Russian gem-quality synthetic diamonds supported plc, more than one persoil can watch, and individual Mr. Hodglzinson's findings. features can be preserved for future reference as pho- In our test, seven synthetic diamonds were indi- tographs or on videotape. vidually suspended from a thread. The inagnet was then brought close to each stone to see if there would New magnet for gem testing. Until recently, mag- be an attraction. Five of the stones were clearly drawn netism was a property with little practical application to the magnet. An alternative method, suggested in in gel11 testing. By and large, magnetic minerals are the product literature, is to make a small "raft1' of few and far between. Even fewer are used in jewelry. plastic foam and float the gem in a glass of water. The Now, however, that has changed. diamond is then checlzed-actually, drawn across the Testing for magnetism is being used to help iden- water-with the inagnet. tify gem-quality synthetic diamonds, as those exam- Dr. Williain Hanneman, of Hanneman ined to date have been grown in iron-niclzel fluxes. Gemological Instruments, Poulsbo, Washington, is Iilclusions of such flux-sometimes so small that they marketing the product and offered some precautions are not resolvable with a standard gemological micro- about it. Do not carry the magnet in or near your wal- scope-can cause the stones to be attracted to a strong let. This could result in erasing the magnetic strip on inagnet (see, e.g., "The Gemological Properties of credit cards and magnetic door lzeys. For the same Russian Gem-Quality Synthetic Yellow Diamonds," son, lzeep the inagnet away froin computers and com- Gems d Gemology, Winter 1993, pp. 226-248). puter diskettes.

GCIIINews GEMS W GEMOLOGY Spring 1995 chromiun~,but less than is typical for ruby-as would be expected given the lighter color. Titnniuln was also detected, however, in a concentration proportionally much lower (relative to chrolnium) than that typical of Icashan synthetic rubies.

More Russian synthetics and simulants. A number of synthetic and imitation gem materials from Russia were available. A firm new to the show this year, the Morion Company, of Cambridge, Massachusetts, offered both rough and cabochon-cut synthetic opals produced at the VNIISIlMS facility in Alexandrov, Russia (for more on this facility, see the Winter 1994 Gem News, pp. 279-2801. This inaterial was available in both black and white body colors; it showed great variation in both color and distribution of play-of- color, ranging from multicolored pinpoint patterns to broad flashes of a single hue. Morion also offered rough Czochralslzi-pulled Figcirc 14. Tllis 0.84-ct Kashan synthetic pink synthetic alexandrite, hydrotherlnal synthetic emer- sappilire is an example of lighter-toned mate- ald, and flux synthetic spinel [both red and blue]. The latter two materials were produced at the Institute of rial offered at Tucson. Photo by Shane F. 1Monocrystals in Novosibirsk. Synthetic quartz and McClrrrc. split boules of flame-fusion synthetic corunduln were available in a broad range of colors as well. Among the inlitation gem inaterials (that is, SYNTHETICS AND SIMULANTS those with no natural gem counterpart) being offered - by Morion was cubic zirconia (CZ)in a wide range of Faceted Ki~shansyuthetic rubies and sapphires. Ruyle colors, including a color-change type. Both the blue Laboratories, of Dallas, Texas, was offering faceted and various shades of green CZ were produced at the Kashan flux-grown synthetic rubies and pink sap- Physical Institute, Russian Academy of Sciences, phires in four quality grades, based on their clarity, according to Dr. Leonid Pride, a geologist and presi- and in calibrated sizes in a variety of cutting styles. dent of the firm. He said that the acronym in Russian They were being sold as "Kashan Created Ruhy," fur- for this institute is PHAIN, from which the Russian ther described in the firm's marlzeting flier as "per- trade name "Pkainite" was derived for CZ. Other n~issivelygrown stones." According to the company's manufacturetl nlaterials offered by Morion included presitlent, Steve Ruyle, they are using this tag line to VNIISIMS-produced pink yttrium alumi~luillgarnet distinguish thls solution-growth product from what (YAG]and gadoliniunl galliuin garnet (GGG)in sever- he calls "forced-growth" nlelt syntl~etics. al colors. Mr. Ruyle further informed us that although all their Tucson material was fro111 old stoclz, new pro- duction has begun ant1 the first colnn~ercialmaterial Figure 15. These syn~hericzincites (1.35-3.26 is expected to be available later this year. Mr. Ruyle ct) are accideiltal by-products of at2 indrls~rinl added that the new material will be grown in a fur- kiln in Silesia, Poland. P110to by Maha DeMaggio. nace he designed, using essentially the saine n~ethoci and flux as that used by Icashan in the past. We had not seen many of the higher-quality, lighter-toned Icashan products, so we purchased for examination a 0.84-ct round, mixed-cut synthetic sap- phire [figure 14). We determined that the properties of this transparent purplish pink stone were consistent with those reported earlier for Kashan synthetic rubies [see, e.g., "Some Aspects of Identification of Icashan Synthetic Rubies," by U. Henn and H.-W. Schrader, louri~alof Gcmniology, Vol. 19, No. 6, 1985, pp. 469-4781. Trace-element chemistry was determined using EDXRF analysis. This revealed the prcsence of

GEMS & GEMOLOGY Spring 1995 "Recrystallized" synthetics. Just before the show, the material. Apparently, it was an accidental by-product editors learned that TrueGem, of Las Vegas, Nevada, of an industrial kiln in Silesia, Poland, that was ~~sed was beginning to market what it called "recrystal- to produce zinc-based paint. The synthetic zincite lized" Czocliralslzi-pulled synthetic ruby and synthet- forined spontaneously and raiidomly by vapor deposi- ic piiilz sapphire. These are being sold under the trade- tion in the air vents of the Iziln's chimney due to inarlzed names of "TrueRuby" and "Truesapphire." some undetermined error in the commercial produc- The rationale behind this non~enclature,according to tion process. Larry ILapidary lournal, 1983, Vol. 36, No. 12, pp. Yet another firm, A. G. Japan Ltd., was selling "Agee 1974-19791, author Marie Kennedy reported on some Emeralds," which a company-provided flier described yellow and red-orange "amberu-colored crystals found as "refined and recrystallized." The flier said that they at the Blackwell Zinc Smelter in Blaclzwell. crush Coloinbian rough emerald into a fine powder, Olzlahoiila, when an old furnace used to produce zinc "purify" the powder with a laser process, and then re- ore concentrates was torn down. crystallize it hydrothermally. The editors caiiiiot ~,ommentat this time on the New source of synthetic emeralds. The editors came feasibility of any of these processes. However, with across another firm that was inarlzeting hyclrotherinal respect to nomenclature, we would gelnologically synthetic emerald as both rough crystals and fash- classify all these illaterials as synthetic,^. In all three ioned gems. Sold as "Maystone, Siberian-Created instances the end products have been crystallized- Emeralds," the product was being offered by Russia- ii~anufactured-artificially in the laboratory. based Asia Ltd.. of Novosibirsk. This material is Dro- duced at a facility other than that used by the airu us Svntlietic zincite from Poland. The Winter 1985 Gein joint venture, according to the firm's executive direc- Trade Lab notes contained an entry on faceted yellow tor, Oleg Yu.Yachny. synthetic zincite (p. 237). The client who had submit- ted the stone for identification said that it was a by- product of an industrial process used in Poland. Also ENHANCEMENTS mentioned was that synthetic zincite 11as been pro- duced experimentally by at least two methods: Diffusion-treated sapphires. Although they seemed to hydrother~naland vapor growth. receive little attention, small quantities of blue diffu- This year the eclitors encountered a considerable sion-treated sapphires were available froin a few amount of this material. While most were single crys- firms, including Budsol Merchandising, of Tacoma, tals up to 15 cin (6 inches) in length, and complex Washington, and Super Shine Gems, of Nugegoda, Sri crystal aggregates, faceted stones were also being sold. Lanlza. The latter also had a 64-ct parcel of small dif- The material ranged from a medium-toned yellow fusion-treated corundums in the pink-to-red-to-purple through orange and dark orangy red, with a few yel- range. All these colors (plus orange) of diffusion-treat- lowish green crystals. ed corundums were included in the firm's price list ill Danuta and Jacelz Wachowialz, of Minerals and both calibrated and noncalibrated sizes, in two color Gemstones, Kralzow, Poland, shed more light on this and two quality grades each.

Ccnl Ncws GEMS & GEMOLOGY Spring 1995 71 Aice S. Keller, Editor

Inchoosing the Most Valuable Articles of 1994, our readers focused on the "how" side of gemology. By "how," I mean the nuts-and-bolts, hands-on type of practical information that a jeweler or appraiser, for example, can put to use in~n~ediatelyafter digesting one of these articles, information that teaches ganolo- gists "how" to avoid costly mistakes. Not surprisingly, first place went to an article in the Fall 1994 issue that deals with how to spot the most talked-about diamond treatment of this or any year: "An Update on Filled Diamonds: Identification and Durability," by Robert C. Kammerling, Shane F. McClure, Mary L. Johnson, John I. Koivula, Thomas M. Moses, Emmanuel Fritsch, and James E. Shigley. Second place goes to the first comprel~ensivearticle to explain bow GIA color grades colored diamonds: "Color Grading of Colored Diamonds in the GIA Gem Trade Laboratory," by John M. Kmg, Thomas M. Moses, James E. Shigley, and Yail Liu, which appeared in the Winter issue. Reiterating the importance of how to identify synthetics, read- ers selected for third place "Synthetic Rubies by DOLU-OS:A New Challenge for Gemologists," by Henry A. Hinni, Karl Schinetzer, and Heinz-JiirgenBernhardt. The article appeared in the Summer 1994 issue. The authors of these three articles will share cash prizes of $1,000, $500, and $300, respectively. Photographs and brief biographies of the winning authors appear below. Congratulations also to Janet S. Mayou, of Watsonde, California, whose ballot was randomly chosen from all those submitted to win the five-year subscription to Gems ed Gemology.

FIRST PLACE

ROBERT C. KAMMERLING SHANE F. McCLURE MARY L. JOHNSON JOHN I. KOIVULA THOMAS M. MOSES EMlMANUEL FRITSCH JAMES E. SHIGLEY

Vice president of research and development at the GIA Gein Trade Laboratory in Santa Monica, California, Robert C. Kainmerling is an associate editor of Gems et, Gemology and coeditor of the Gel11 Trade Lab Notes and Gem News sections, the latter along with Mr. Koivula and Dr. Fritsch. A regular con- tributor to gemological publications worldwide, Mr. Kan~nlerling coauthored-with Dr. Cornelius Hurlbut-the book Gemology. He has a B.A. from the University of Illinois. Shane F. McClure is supelvisor of identification ser- Front from left,Mary L. Johnson, Shane F. McClure; vices in the GM G~~ ~~~d~ rear from left, Robert C. IZammerling, Enlmanzlel ~~b~~~~~~,Santa Mr. Fritsch, and lohn I. I

Most VaIuable ArticIe GEMS & GElMOLOGY Spring 1995 19 years of trade and laboratory experience, is vice president of identification ser- vices at the GIA Gem Trade Laboratory in New York City. He attended Bowli.ng Green Uiliversity, Oho, and is a contributing editor of Gem Trade Lab Notes. He specializes in pearl identification and origin-of-colordetermination for colored diamonds. Manager of GIA Research Ernmanuel Fritsch specializes in the applica- tion of spectroscopy to gen~ology,the origin of color in gem materials, and treated and synthetic gems. A native of France, he has an advanced degree in geological engineering from the Geology School in Nancy, France, ;uld his P11.D. froin the Sorbonne in Paris. Janles E. Shigley, who received his doctorate in geology from Stanford University, is director of GIA Research. He has written many articles on natural, treated, and synthetic gems, and directs research on all aspects of identi- fying and characterizing gem materials. Ianles E. Shigley

SECOND PLACE

JOHN M. KING THOMAS M. MOSES JAMES E. SHIGLEY YAN LIU

John M. King is laboratory projects officer at the GIA Gem Trade Laboratory, New York City. Mr. lting holds an M.F.A. from H~tnterCollege, City University of New Yorlt. With 17 years of laboratory experience, he frequently lectures on col- ored diamonds and various aspects of laboratory grading pro- cedures. Yan Liu, a color researcher with GLA Research, has an M.S. in"co1or science from the Rochester Institute of Technology, New York, and an M.S. in color optics from Shandoilg College of Textile Engineering, Qingdao, China. The photographs and biographies for James E. Shigley and Thomas M. Moses appear under the first-place section. lohn 1M. King Yon Lizz

THIRD PLACE

HENRY A. HANNI KARL SCHMETZER HEINZ-JURGEN BERNHARDT

Henry A. Hlnni, who has a Ph.D. in mineralogy from Easel University, Switzerland, has worlted at the SSEF Swiss Gemmological Institute, now in Basel, since 1980 and became its director in 1990. Also an associate professor of gelnology at Henry A. Hiinni the Mineralogical Institute of Base1 University, his specialties are applied gemology and modern analytical methods. Karl Schmetzer is an independent gemology researcher and consul- tant based in Petershausen, Germany, near Munich. He spe- cializes in the identification of natural and synthetic gemstones, especially corundum and beryl. He has a Ph.D. in mineralogy and crystallography from the University of Heidelberg in Germany. Heinz-Jiirgen Bernhardt also has a P11.D. in mineral- ogy and crystallography from the University of Heidelberg. The German member of the coinmission on ore mineralogy of the International Mineralogy Association, he is head of the Ruhr University Electron Microprobe Laboratory, Bocl~um, Heinz-Pirgen Bernhardt Germany.

lMost Valuable Article GEMS e\ GEMOLOGY Spring 1994 Gems & Gemology C+H+A+L+L+E+N+G+E Gems & Gemology articles over the past year spanned many topics about colored gems, from revealing gen~ologrcaleannarlzs of a new synthetic ruby to dispelling myths about ametrine. Yet, it was diamond-arguably the industry staple-that seemed to be the center of attention. In particular, the presence of increasing numbers of fracture-filled lamonds in the trade prornpted GIA to challenge some of the world's top gemologists to update ways of identifying this treatment. Now, once again, as part of GIA's contin- uing education program, we challenge you. Based on information from the four 1994 issues of Gems etl Gemology, the following 25 questions call 011 you to demonstrate your laowledge about developments in gemology. Refer to the feature articles and Notes and New Techniques in these issues to find the single best answer for each question, then mark your choice with the corresponding letter on the response card provided in this issue (sorry, no photocopies or facsimiles will be accepted]. Mail the card so that we receive it no later than Friday, August 18,1995. Be sure to include your name and address. All entries will be aclcnowledged with a letter and an answer key. Score 75% or better, and you will receive a GIA Continuing Education Certificate. Earn a perfect score of loo%, and your naine will also be fea- tured in the Fall 1995 issue of Gems d Gemology.

Note: Qzrestions are token froni only 3. A~netrinecrystallizes as part of: the fozzr 1994 issues. Choose the sin- A, hydrothermal brecciation. C. induced by a flux-melt pro- gle beg onswer for each question. B. hydrothermal synthesis. cess to reseinble those in C. volcanic ~ctivity. natural rubies. D. ~netamorphosisof silica D. a by-product of flux growth 1. The bicoloration of ametrine sediments. crystals from the Anahi mine that is usually removed in in Bolivia is the result of: cutting. 4. Noteworthy of the Indaia, Brazil, 6. Douros synthetic rubies can be A. natural color zoning. corundum deposits is: R, heat treatment of iron-bearing identified by: A. the high percentage of purple amethyst. A. veil-like flux inclusions. to pink sapphire. C. irradiation of amethyst to B. typical chromium absorption R. the high percentage of color- produce color centers. features and penetration change sapphire. D. heat treatment of colorless twinning. C. the sapphire's excellent C. chemical analysis and growth quartz. reponse to heat treatment. patterns. D, their occurrence in silu in D. no consistent combination of 2. The corunduin deposits of basalt. laboratory testing methods. Indaia, Brazil, have been found exclusively as: 5. "Fingerprint" patterns in 7. Douros synthetic rubies are A Precambrian intrusive bodies. Verneuil synthetic ruby are: grown by: R,primary megacrysts in basalt. A. found in all flame-fusion A, hydrothermal synthesis. C. eluvial deposits associated synthetic rubies. B. unseeded flux growth. with river gravels. B. always different from those in C, the Czochralslzi method. D, alluvial or colluvial deposits. natural rubies. D. flaine fusion.

74 Gems 01 Genlology Challenge GEMS ei GEMOLOGY Spring 1995 8. Synthetic forsterite call be 14. The durability of fracture-fillet1 20. Color description includes the separated from natural peridot: diamonds: three attributes: A. by Cr features in the former's A. is not of concern, because A. hue, density, and intensity. spectrum. fracture filling is permanent. R. tone, saturation, and hue. B, by the lower R.1. and B. is a major concern in the C. hue, tone, and spectrum. S.G. of peridot. daily wear of diamond jewelry. D, saturation, hue, and refraction. C, on the basis of color. C. is subject to damage by D. All of the above. prolonged or repeated exposure to typical jewelry 21. Gein cor~undumfrom the Ural cleaning methods. Mountains of Russia occurs in 9. The most common solid inclu- D. is not affected by ultraviolet association with marble, as sion found in emeralds from radiation. does corundum from: Mananjary, Madagascar, is: A. Myanmar. A. quartz. 15. Careful observation is reauired B. Montana, Unitetl States. B. mica. to distinguish flash effec;s in C. Thailand. C. feldspar. filled diamonds from: D. Australia. D. apatite. A, opalescence. B. adularescence. 22. Color grading of colored diamonds 10. Fluid inclusions in emeralds C. thin-film iridescence. requires a controlled from Mananjary, Madagascar: D. color banding. environment in which: A. resemble those of emeralds A. the ambient light contains no from ItabiraINova Era, Brazil, 16. Copper iilclusions in Paraiba blue wavelengths. and Zambia. tourmaline: B. ambient hunlitlity is minimized. B, always distinguish A, are proof of treatment. C. reflection of incident light is Mananjary emeralds from B. are situated according to the prevented. those of African localities. trigonal symmetry of the host D. lighting is from a standard C. occur only very rarely. crystal. source. D. Both A and B are correct. C. are not found in other gems. D, result from nucleation before 23. Most gem corundunl today 11. Commercial mining of growth of the host. coines from: corundum in Vietnam: A. Mogok, Myanmar. A. has been accelerated at 17. Gein corundum from the Ural B. primarily marble-type most major mines since Mountains of Russia : deposits. 1992. A. is plentiful on the gem C. alluvial deposits in Sri Lanlza. B. dropped off greatly in the market. D. secondary deposits in Luc Yen district in 1994. B, is not currently mined Southeast Asia ant1 Australia. C. is carried out exclusively cominercially. by the Vietnamese government. C, is ~niiledprimarily by hand. 24. The term that best describes D. is actively discouraged by D, is marketed through the Vietnamese government. Thai dealers. corundum's relationship to alkali basalt inagma is: 18. The GIA Gem Trade A. vug. 12. The most useful instrument Laboratory has added these B. hololith. for identifying fracture filling in terms to its color grading termi- C. xenocryst. diamonds is: nology for colored diamonds: D, metamorph. A. a microprobe. A. frincy vivid and fancy darl<. B, an infrared spectrometer. B. fancy deep and fancy intense. 25. A new treated flame-fusion C. a binocular gemological C.fancy vivid and fancy light. microscope. synthetic ruby is characterized D. ~AIICYdeep and foncy vivid. D. an immersion cell. by its: A. included silk resembling 19. When the GIA Gem Trade that in natural ruby. 13. The firm that fracture-filled a Laboratory grades a faceted B. three-dimensional cellular particular diamond can be fancy colored diamond, the network of flux-filled identified conclusively by: stone is viewed: fractures. A. DXRF analysis. A. face up. C. fingerprint-like inclusions B. microscopic features. B. through the pavilion. unlike those seen in natural C. X-radiography. C. edge up. ruhy. D. None of the above. D. Both A and B are correct. D. Both B ant1 C are correct.

Gen~sel) Gemology Challenge GEMS ei GElMOLOGY Spring 1995 75 GEMS, THEIR SOURCES, remains, in its new fifth edition, an h- DESCRIPTIONS AND valuable resource for every gemologist. RICHARD T. LIDDICOAT IDENTIFICATION Chairman of the Board 5TH EDITION Ge~nologicalInstitute of America By Robert Webster Sanla Monica, California Rev. by Pctcr Rend, 1,026 pp., illus., Rut terworth-Hcincmann, Oxford, OTHER BOOKS RECEIVED England, 1994. US$120.00* Hey's Mineral Index, 3rd Ed., by A. M. This monuinental gemology text, Clnrk, 852 pp., Chapman d Hall, flrst published in 1962, went through London, 1993, US$79.95". Hey's An three editions by the late Robert units of measurement have been Index of Mineral Specics &st appeared Webster, and then further revision in converted to metric. The major in 1950; since the second edition in 1983 by the renowned Basil changes are in the fields of new syn- 1974, the number of valid mineral Anderson. With Mr. Anderson's thetics, substitutes, and enhancements. species has increased to about 3,500 death, Peter Read was asked to pro- In general, the additions have (with some 600 added in the last decade duce an updated edition. For this been handled very well, although it alone). This volume includes the daunting task, he wisely called on is easy to find relatively minor names of vahd species, as well as vari- several others whom he considered annoyances in almost any effort of eties, synonyms, misspellings, some qualified in the various disciplines this magnitude. For example, the dia- meteorite and rock names (if they covered by this comprehensive vol- mond chapter pays too little atten- mght be confused with minerals), and ume. Enlisted were Grahaille Brown, tion to Russia, and especially more-about 15,000 entries total. of Austraha; C. R. Cavey, Roger Hadng Botswana, as sources. The map of The bulk of the book consists of Alan Jobbins, I. Mercer, Michael southern Africa, unchanged from the an alphabetical listing of 'all the names, O'Donoghue, R. Sanderson, and C. R. fourth edition, shows neither with the following given for valid Woodward, all of Great Britain; C. Jwaneng, arguably the most impor- species: chemical compositionj original Dominy, of Canada; Ulrich Henn, of tant dianlond mine in the world, nor reference and type locality (except for Germany; Richard Hughes, Robert the new Venetia mine. There are lninerals lrnown in antiquity); ciystal- C. Kammerling, John I. ICoivula, and more serious probleins as well. lographic data; inineral group affilia- j. R. Rousc, of the Unitecl States; plus Fracture filling in diainonds is given tion; synonyms; and a brief coininen- Kenneth Scarratt, now of Thailand. only five lines, plus a before-and-after tary, if warranted. In a separate part, In his original edition, Webster illustration, and the coverage of syn- minerals of similar chenucal composi- divided the publication into two sep- thetic diamonds omits some recent tion are grouped together-e.g., ele- arate volun~es:The first described all developments, such as thin films. An ments, sulfides, silicates. the gem materials and their substi- illustration also carried over from Many gelnological and varietal tutes, and the second dealt with gem earlier editions is the "correct" cut trade terms are included (e.g., copal, fabulite, maw-sit-sit, sunstone), identification. Apparently, users for a round brilliant, with a 39 O pavil- complained that they wanted all of ion angle and a 41" crown, angles lit- unlike other excellent compilations the information in one volume. As a tle used for diamonds. Red beryl is that are not intended to be as com- result, from the second edition on, almost dismissed under coverage of prehensive (e.g., Fleischer and his work was consolidated into a sin- pink beryl. Mandarino, Glossary of Mineral gle huge entity. (The downside to Part of the problem may be that Species; and Nickel and Nichols, this coilvenience is that it malzes for the newest revision has added only Mineral Reference Ahnual). an unwieldy whole, and places great 10 pages to the hasic text, plus 10 This hook wdl be a welcome addi- strain on the binding of a book that more to the appendices, over the tion to the library of any gemologist will undoubtedly be frequently used.) fourth edition. (The Anderson revi- interested in mineral nomenclature This fifth edition maintains the sion added 75 pages.) Most illustra- 'and chemistry. It contains a wealth of "two part" organization of the first. tions are black and white, and there information that can he found else- As expected, there are relatively few is a paucity of color plates. With such where only with great Miculty. changes in the chapters on propertics. an important work, it would have A. A. LEVINSON Exceptions are that the "Composite been nice to see more full color and University 01 Cnlgary and Artificially Colo~lredGemstones" an easing of length restrictions. Cnlgn1~7,Alberta, Cnnnda and "Colour in Gemstones" chapters Still, on the wl~ole,this major have been reorganized to "Composite update is a careful and competent revi- *This book is available for purchase at the GIA Bookslore, 1660 Stew& Str'eel, Santa Gemstones" and "Colour and Colour sion. Webstcr's Gems: Their Sources, Monica, CA 90404. Telephone (800) 421- Enhancement." Tliroughout, various Descriptions and Identification 7250, ext. 282.

Book Reviews GEMS & GEMOLOGY Spring 199s REVIEW BOARD Ernrnanuel Fritsch Mary L. Johnson Jana E. Miyahira GIA, Sanla Monica GIA, Santa Monica GIA, Santa Monica Charles E. Ashbaugh Ill Robert C. Kammerling Isotope Products Laboratories Juli Cook-Golden James E. Shigley GIA Gem Trade Lab, Santa Monica GIA, Santa Monica Burbank, California Royal Oak, Michigan Michael Gray A. A, Levinson Christopher P, Smith Andrew Chrislie Missoula, Montana University of Calgary Gubelin Gemmological Laboratory GIA, Santa Monica Galgar): Alberta, Canada Lucerne, Swilzerland Patricia A. S. Gray Lorella B. Loeb Carol M. Slocklon Jo Ellen Cole Missollla, Montana GIA, Santa Monica Visalia, California Los Angeles, California Professor R. A. Howie Elise B. Misiorowski Rolf Tatje Nanette D. Colomb Royal Holloway University ol London GlA, Sanla Monica Duisburg University Sydne): Australia United Kingdom Gary A. Roskin Duisburg, Germany Maha DeMaggio Karin N. Hurwit European Gemological Laboratory Robert Weldon GIA, Santa Monica GIA Gem Trade Lab, New York Los Angeles, California Radnot Pennsylvania

COLORED STONES AND M. A. Menzies; "Where's the Proton? Symmetry and ORGANIC MATERIALS Structure Variations in Topaz," P. H. Ribbe and S. C. Eriksson; "Topaz: Environments of Crystallization, EM-TGMS-MSA Symposium on topaz. Mineralogical Crystal Chemistry, and Infrared Spectra," E. E. Foord, Record, Vol. 26, No. 1, 1995, pp. 63-71. L. L. Jacltson, J. E. Taggart, J. G. Crock, and T. V. V. Held in conjunction with the 41st annual (February King; and "Items of North American Mineralogical 1995) Tucson Gem and Mineral Show, the Tucson and Gelnological Note During 1994," M. Gray. Fifty- Mineralogical Sy~nposiumon Topaz was sponsored by four crystal diagrams accompany the abstracts. LBL the Friends of Mineralogy, the Tucson Gem and Mineral Society, and the Mineralogical Society of The mineralogy, geology and occurrence of Topaz. M. America. Mineralogical Record published 13 abstracts A. Menzies, Mineralogical Record, Vol. 26, No. of talks given at the symposium, all authored by the 1, 1995, pp. 5-53. speakers themselves. Topics and authors include: "The This article addresses questions such as: How does Occurrence of Topaz in Northern New England topaz form? What are the conditions required for good Peginatites," C. A. Francis and L. C. Pitinan; "The crystals? How does the geologic environnlent influ- Occurrence of Topaz in the Southeastern United States," R. B. Cook; "Colorado Topaz," P. J. Modreski and T. C.. Michalslti; "Notes on the Occurrence of This section is designed to provide as complete a record as prac- tical of the recent literature on gems and gemology Articles are Topaz in Idaho," L. R. Ream; "Topaz from the selected for abstracting solely at the discretion of the section edi- Sawtooth Batholith, Idaho," M. A. Menzies; tor and his reviewers, and space limitations may require that we "Occurrences of Blue Topaz in the Peginatites of the include only those articles that we feel will be of greatest interest Peninsular Batholith, San Diego County, California," J. to our readership. Fisher; "Pink Topaz from the Thomas Range, Juab Inquiries for reprints of articles abstracted must be addressed to County, Utah," E. E. Foord, W. Chirnside, F. E. Lichte, the author or publisher of the original material. and P. H. Briggs; "Topaz Rhyolites in Arizona and the The reviewer of each article is identified by his or her initials at the Southwest," D. M. Burt; "Topaz and Beryl-bearing end of each abstract. Guest reviewers are identified by their full Gem Pegmatites of the Alabashlta-Mursinlca-Adui dis- names. Opinions expressed in an abstract belong to the abstrac- ter and in no way reflect the position of Gems & Gemology or GIA. trict in the Ural Mountains, Russia," P. Lyckberg; "Geology and Occurrence of Well-Crystallized Topaz," 0 1995 Gemological Institcrte of America

Gemological Abstracts GEMS & GEMOLOGY Spring 1995 77 ence topaz-crystal features and associations? Dr. Outicallv observable fibers in concentrically banded Menzies taclcles these questions by first discussing thc agates of geode origin are composed of much finer tectonic regimes or settings that influence the parent fibers, in which quartz crystallites 8-100 nm long are magma from which topaz crystallizes. The text and staclced together parallel to <1120> or <1100>. Brazil- accompanying diagrams explain orogenic (crustal com- twin lamellae structures are frequently seen in grains pression] and anorogenic (crustal extension] processes, longer than 30 nm. Uniformly spaced systematic stria- easing the lay reader into the complicated subject of tions consist of a cyclic alternation in quartz grain deposit type and occurrence. sizes, the smallest being 6 nm. Coarse quartz or The deposit types are divided into three sections: amethyst represents the final stage of agate (lining volcanic (rhyolites), magmatic (granites, ongonites, type) formation. and topazites), and late to post-magmatic (pegmatites). The fine fibrous texture in opal from Beltane, Each section discusses specific topaz localities. Two California, is the result of cristobalite crystallites tables help the reader malcc quick and easy cross-refer- (8-20 nm] staclmalachite. 0. V. Il'in, World o/ pp. 348-35 1. Stones (produced in Moscow in English), No. 4, This article reports on the author's investigation into 1994, pp. 3-9. the comparative radiolucency (i.e.,X-ray transparency] Malachite's popularity as a gem and ornamental stone of diamond and its imitations when contained within results from its combination of rich green colors and the human body. The unusual investigatioil was variable textures and structures. This article illus- promptell by an actual case in the Australian courts trates, in 12 striking color plates, some of malachite's where a thief attempted to switch a diamond with an many different forms (e.g., radiating spherulite, hollow imitation in a jewelry store by swallowing the dia- pseudostalactite, reniform]. Seven of the specimens mond. The diamond was subsequently located in the shown came from two classic localities in the Ural thief's body by means of X-radiography. During the Mountains, the Gumeshevskii mine and the trial, questions about the differences in X-ray opacity Mednorudyanslcii mine. The author discusses the ori- of diamond and its imitations were often raised. gin of the various forms, and concludes that many nat- The author's experiments used diamond and a ural factors control the textural patterns in malachite. number of imitations, including colorless quartz, These factors include whether growth was restricted, topaz, synthetic spinel, synthetic sapphire, lead glass, the accumulation of impurities, and the roclc type in CZ, YAG, synthetic rutile, and strontium titanate. which the copper-rich solutions precipitated. AAL Procedures and equipment are described in detail. The stones were X-radiographed in the stomach, bowel, and Nanometre scale textures in agate and Beltane opal. T. bony pelvis of a "I'ixy whole-body phantom," essen- Lu, X. Zhang, I. Sunagawa, and G. W. Groves, tially a medical mannequin with "internal organs" Mineralogical Magnzine, Vol. 57, No. 1, 1775, pp. manufactured to radiologically simulate those of a 30- 103-107. year-old female.

78 Gemological Abstracts GEMS & GEMOLOGY Spring 1995 The data showed that the radiopacity of the study In general, age-dated diamonds have proved to be samples was inversely related to the atomic weight of older than the pipes that brought them to the surface. their principal elemental constituents: the lower the Zircons associated with the Mhuji Mayi Iziinberlite, atomic weight of the principal atoms, the higher the but not included within diamonds, have been age- radiolucency of the sample. Most of the materials fell dated at 70 million years old, consistent with other into two distinct groups: a relatively radiolucent group measurements of the age of the lzimherlite pipe. The (diamond, topaz, quartz, and synthetic spinel) and a authors believe that the 628 million years represents radiopaque group (synthetic rutile, CZ, YAG, and the age of the host diamond, which makes it the strontium titanate). It was especially difficult to locate youngest diamond age reported to date. Note that the the materials with low atomic weights (such as dia- use of the ion microprobe for U-Pb age-dating of indi- mond) in the radiographs taken within the simulated vidual diamonds is limited to diamonds that contain abdomen. Furthermore, when X-ray tube voltage was included zircon or other uranium-rich phases. MLI increased above 70 IzV, diamonds and simulants with low atoinic weight became transparent to X-rays; the GEM LOCALITIES resulting images had no diagnostic value. This article is well illustrated with X-radiographs Almandine garnets from Vietnam. R. C. Kainmerling and contains tables that present the data succinctly. It and J. I. Koivula, Australian Gemmologist, Vol. is a most interesting example of forensic gemology. 18, No. 11, 1994, pp. 356-358. R CI< The far north of Vietnam is a new locality for gem- quality almandine garnets. This article provides a Story on the "Cross of Asia" fancy yellow radiant cut detailed gemological study of garnets from this new diamond. A. Szymanslzi, Archiwiem Mineralogiczne, deposit, describing n~icroscopicfeatures as well as the Vol. 50, No. 1, 1994, pp. 8-10, results of standard physical tests. Since pyrope garnets had been tliscovered in southern Vietnam, the authors This brief article reports on the autl~or'sexamination wanted to determine exactly what type was collliilg of the famous Cross of Asla yellow diamond in from the north. Chelnical analysis proved that the October 1993. Found in the Jagersfontein minc, South samples were essentially almancline, with minor Africa, in 1902, the original piece of rough weighed 280 pyrope, grossular and spessartine components. /EC ct. It was cut first to 142 ct, and was later recut three more times. At the time the author examined it, the Brazil-a gemstone giant reawakens. R. Weldon, Radiant-cut diamond weighed 79.12 ct and measured lewelers' Circular-Keystone, Vol. 166, No. 2, 28.67 x 22.21 x 15.77 mm. The play of colors on the February 1995, pp. 105-120. table facet create the outline of a Maltese cross. Althougl~national inflation, world recession, ant1 local RAH governments have conspired to stifle gein mining and sales in Brazil, recent marlcet liberalization and politi- Zircon from the mantle: A new way to date old dia- cal and social changes have opened up new and excit- monds. P. D. Kinny and H. 0. A. Meyel; 1994, ing opportunities for the country's mineral riches. A lo~~rnalof Geology, Vol. 102, pp. 4755481. con~n~ercial-freezone is now being developetl in Age dating of diamonds has been performed in the past Tedfilo Otoni, in the gem-rich state of Minas Gerais, to by measuring the isotopic contents of disnlond inclu- help promote the gem, jewelry, and lapidary industries. sions, notably samarium (Sm) and neodymium (Nd) This informative article also describes the current isotopes in garnet and clinopyroxene inclusions. mining and supply status of various geins intligenous However, because diamond inclusions are small, and to Brazil, including emerald, tournlaline, topaz, aqua- rare-earth elements are uncommon in the minerals marine, Itunzite, chrysoberyl, and quartz. Copiously present, inclusions in nlany diamonds must be mea- illustrated with wonderful photographs by the author, sured at the same time to get an average age date for a the article provides current insights that will be diamontl region. In this article, the authors use a dif- invaluable to anyone interested in Brazil ant1 its cor- ferent technique, uranium-lead (U-Pb)age dating with nucopia of mineral riches. IEC an ion microprobe, to determine the age of a s~nglezircon incl~~sionin a lcimberlitic dirunond from Mbuji Mayi, Ure. A brief report on Indonesian opal. G. Lanlbert and G. The analyzed zircon measured 0.14 x 0.10 mnl and Brown, Australian Gemmologist, Vol. 18, No. showed a zoned 0.02-mm rim with cathodolumines- 1 I, 1994, pp. 359-361. cence, which suggests that it originated froin a melt. According to a fact sheet provided the authors by the Three areas in the zircon core were measured twice for Directorate of Mineral Resources in Bandung, Java, total uranium, total thorium, total lead, and the iso- opal is found in at least three regions of Indonesia: The topes 235U, 238U, 20GPb, and 207Pb. The zircon had a Labalc District of West Java (including black, white, high uranium content (350-650 ppm U). Six measure- colorless, and brownish "tea" opal), the Benglzulu ments gave an average age of 628 million years, with a Province on the northwestern coast of Sumatra (pre- 2-C error (95% confidence level) of + 12 nlillion years. dominantly white potch), and in southeast Sulawesi

Gemological Abstracts GEMS & GEMOLOGY Spring 1995 79 Province (translucent-to-opaquegreenish white potch). indicates the presence of variable amounts of C02 in The authors examined three cabochons of inateri- quartz (72.7 ppm), garnet (141.2ppm), and chrysoberyl a1 from the first deposit, the only one currently being (51.6 ppm). Gem localities in Kerala contain vast mined commercially. Two were of the "tea" type, untapped potential. At present, though, most mining is while the third is described as blaclz. Properties deter- done illegally. A map of gem occurrences is included, mined were: spot R.1.-1.385 to 1.395; S.G.-1.81 to as are photographs and brief details of chrysoberyl, sap- 2.09; Chelsea filter reaction-green; long-wave and phire, ruby, topaz, and zircon. RA H short-wave UV fluorescence reactions-various shades of green (except the blaclz opal, which was inert to Hsa-Taw grecn tourmaline. U Tin Hlaing, Australion short-wave UV); and very slight phosphorescence only Gemmologist, Vol. IS, No. 11, 1994, pp. in the blackspeci~nen.Allstones exhibited play-of-color, 352-353. described and illustrated for each. There is also a pho- Green tourmaline has been mined for about 25 years at tomicrograph of prominent flow structures seen in one two deposits in Myanmar's southeastern Kayah State. of the tea opals. One site is located about 10 nliles (16.5 lzm] north- R CI( and the other about 30 miles (50 kin) south-of the town of Hsa-Taw. The de~ositsoccur in a belt of A contribution to understanding the infrared spectra of Paleozoic metamorphic rocks. Crystals average about rubies from Mong Hsu, Myanmar. C. I? Smith, 5 ct; after cutting, the best-color stones average 1-2 ct. Journal of Gemmology, Vol. 24, No. 5, 1994, pp. Common crystal forms include triagonal prisms ter- 321335. minated by rhombohedra1 and pyramidal faces, with A study of 31 rubies from the new deposit at Mong and without a pinacoid. Short prismatic crystals are Hsu, Myanmar, revealed infrared features distinctive more comlnon than long ones, and vertical striations of heat-treated as comuared to non-heat-treated stones on prism faces were absent in about 25% of the crys- from this locality. Features were found in the mid- tals examined by the author. infrared range between 1900 and 3400 cm-1 in some Gemological properties are reported to be typical (but not all) of the untreated specimens examined. for elbaite tourmalines (although specific values are These were related to the presence of diaspore and not given]. Surface-reaching brealzs are typically filled boehmite inclusions (including submicroscopic ones), with reddish brown iron oxides. Internal features although only the former has been confirmed by X-ray noted include color zoning, trichites, mica, and solid diffraction analysis. In some heat-treated specimens, black inclusions of iindetermined identity. strongly pleochroic features between 3100 and 3600 The report includes color photographs of both cm-1 were found. These were related to structural OH rough and cut specimens, as well as a slzetch map of groups bonded within the corundum lattice. Other Myanmar that shows the general location of the heat-treated samules showed no structural OH. The deposits. XCK author concludes that the structural OH groups found in heat-treated Mong Hsu rubies were formed as the The 1995 ICA World Gemstone Mining Report. I. Z. result of alteration and dehydration of diaspore parti- Eliezri and C. Krernlzow, ICA Gazette, December cles in the host rubies during heat treatment. He sug- 1994, pp. 1, 12-19. gests that the features observed may be useful to dis- This relatively optinlistic report begins by stating that tinguish natural from heat-treated material. He warns, the greater liberalization of trade in many countries however, that infrared evidence of diaspore should not expanded the potential inarlzet for gemstone sales be used alone to identify locality. CMS around the world, as well as the potential for gem min- ing. It then gives a brief overview of new mines and the Gemstone mineralization in southern Kerala, India. R. current production status of the more important col- D. Menon, M. Santosh, and M. Toshida, Iournal ored geins in selected countries. Highlights are as follows. of the Geological Society of India, Vol. 44, No. 3, Ruby procluction-quality and size-is up in India 1994, pp. 241-252. and Myanmar; production is doing well in Cambodia Within a gem field that extends over an area 70 x 35 (despite the sporadic supply caused by turmoil in the lzm2 in south ICerala and the adjacent region of Tamil mining areas), and large quantities of lower-grade ruby Nadu, many gem minerals occul; including are coming from Sierra Leone. Fewcr Thai rubies are chrysoberyl (cat's-eye and alexandrite], ruby, sapphire, available due to heavy mining over the past few emerald, topaz, zircon, and amethyst. The primary decades. Production is down in Vietnam as well, mineralization is in zoned, complex peginatites of Pan- because too many buyers have been sold parcels salted African age, elnplaced in granulite-facies metapelites with synthetics. (lzhondalites), and variably weathered and laterized; Demand for Australian sapphire rough was secondary gem deposits are found in stream gravels stronger over the last two years. New sapphire deposits and placers. Thermal decrepitation of fluid inclusions have been found in Russia (Siberia), Myanmar, and

80 Ge~nologicalAbstracts GEMS & GEMOLOGY Spring 1995 Laos. A large new deposit of very heat-treatable blue ing on air temperature, and duration and intensity of sapphire from southern Madagascar is causing excite- light). Heating causes the color to facle rapidly. Some ment. There was a new find of pastel-colored sapphires crystals do not fade con~pletelyto colorless, but reveal in Tanzania and a new deposit of "white" sapphires in a colorless core ant1 a distinctly. pink- to pinkish red Sri Lanlca. outer rim. The initial yellow-orange "sherry" color Production is up for peridot, due to new finds in inaslcs the pinkish red zone, which is stable to light the Unitecl States [Arizona), Ethiopia, Palcistan, and and heat. Vietnam, as well as increased production in China. Bulk chemical analysis by ICP-AES-using gem- Emerald production is up, with new deposits in quality, colorless crystals and color-zoned crystals- South Africa, the Ural Mountains in Russia, China showed that the pinlc inaterial contains inore Mn (20 (first ever), Tanzania, Somalia, and Brazil. Zambia has ppm) and Fe (0.08%) than the colorless material (<8 reopened the famous Kainakanga mine (with reduced ppm and 0.05%, respectively); the colorless material mining elsewhere in that country).Regular production contains more Cr (17 ppm) than the pink material (1 1 continues in Colombia. Large amounts of greenish yel- ppm). Cr is not believed to be a chroinophore in this low beryl are coining from the Ulcraine, with small case. Chemical analvsis bv laser-ablation ICP-MS of amounts from Finland. individual spots on a color-zoned crystal showed Mn Fire-opal production is up in Canada, with large to be concentrated in the outer color zone; Cr and quantities being mined in Mexico. There was less opal other possible chroinophores were not detected. from the south Australian fields, but Lightning Ritlge Trivalent Mn and Fe are lilcely substituting for the A1 toolc an upturn. The United States produced various in the topaz crystal structure. This is believed to be the types of opal from volcanic deposits in Oregon and first documented occurrence of pink-red topaz that is Arizona. There is a new yellow dendritic opal froin colored by Mn + Fe. LBL Zambia. New finds of garnet occurred in Namibia, Mali, Reniarkable finds pf minerals of beryllium: From the Ethiopia, and China. A new variety was found in the Kola Peninsula to Primorie. I. V. Pelcov, World of Sierra Madre Mountains of Mexico. Massive green Stones, No. 4, 1994, pp. 10-26. grossularite is being mined in China. Beryllium is a relatively rare element in nature, yet 82 As for amethyst, production of high-quality mate- minerals (as of 1991) contain it as an essential compo- rial increased in Zambia. Bolivia held its own in gen- nent. Forty-five beryllium minerals are found in the eral production. CIS [Commonwealth of Independent States). There are Madagascar and Nigeria continue to supply con- classic localities for beryl and chrysoberyl (and their siderable amounts of aquamarine. Fine blue tourma- many gem varieties), but some of the more unusual line is coming from Namibia. There was a major new beryllium species sought by gem collectors, such as find of tourmaline in Minas Gerais, Brazil (at Morro phenacite and euclase, are also found there. Redondo), plus a new variety of color-change tourma- This article lists all 82 lcnown berylliuin minerals line from India. (with their chemical compositions]. It also has a table Tsavorite production continues in Kenya, with a containing the type locality (worldwide) and the first new fincl reported about 1,000 kin northwest of the place each of the berylliuin mii~eralsIcnown in the CIS previously lcnown localities, in Lolcirima. was found there. A map shows 72 important locations The report also noted new finds of lcornerupine, (such as type specimens for new species) ranging from sillimanite, and epidote in India; of scapolite in China; the Kola Peninsula [northwest Russia) to Primorie (in of alexandrite in Brazil; and of rhodochrosite in the the Far East); genetic types of geologic formations (e.g., United States. CEA granitic pegmatite, greisen) in which the ininerals occur are also indicated. Twelve museum-quality spec- inlens (and one collage of cut heryls) are illustrated in Pink topaz from the Thomas Range, Juab Connty, color. Of 83 references, 80 are in Russian. Those inter- Utah. E. E. Foord, W. Chirnside, F. E. Lichte, and ested in berylliuin minerals will find this a worthwhile P. H. Uriggs, Mineralogical Record, Vol. 26, No. compilation. AA L 1, 1995, pp. 57-60. Pink topaz was first rcported in the Tho~nasRange in 1934. These gems are found only as float crystals and Siberia and Far East: A brief ~nineralogicalguide. A. A. so have becn exposed to sunlight. Although the crys- Evseev, World of Stones, No. 4, 1994, pp. 42--54. tals may be as long as 4 cm, transparent, gem-quality This article rcports on various aspects of "notable" crystals with uiliform pinlc to pinkish red color are inineralogic occurrences in the vast reaches of Russia usually less than 2 cm. Crystals from newly opened lying east of the Ural Mountains. As "notable" occur- cavities are a yellow-orange "sherry" color and typi- rences, the authors include those famous for museum- cally fade to colorless in a few days to weeks (depend- quality specin~ens,typc localities of new species, and

Getnological Abstracts GLIMS a GEMOLOGY Spring 1995 81 first occurrences of mineral species in Russia, among use the terms jadeite jade or pure jade versus impure others. The article is divided into four major parts. jode for polyinineralic material. The most common The first part, "Mineral Localities," lists the 114 admixture is jadeite and ltosmochlor, with a continu- selected occurrences, which include individual mines ous series from essentially pure jadeite to 85% lzos- (e.g., the Mir lzimberlite pipe), massifs [e.g., Murun, the mochlor. The latter mineral is well lznown to gemolo- type locality for charoite], districts (e.g., Noril'slt, gists as a major constituent in Maw-sit-sit, also from famous for metallic ore minerals), and even a volcano Myanmar. The authors present an excellent overview (Klyuchevsltaya Soplta) in Kamchatlta, noted for fuma- of this con~plextopic, including chemical analyses and role minerals. A map of the localities is included (note X-ray diffraction data that emphasize the difficulties that the map ltey was not bound in the issue but that gemologists face in identifying jade and jade-lilze appears on a separate sheet of paper). Minerals found at green specimens. Minerals found as components in the each locality are listed (58 for the Noril'slz district specimens studied include jadeite, lzosinochlor (once alone]. lznown as ureyite), edenite, richterite, tremolite, The second part, "Minerals," is essentially a cross- chromite, magnesiochromite, and enstatite. This is an index that alphabetically lists 89 well-ltnown minerals important paper that should be read carefully by and gives their important localities (from the first part gen~ologistsinvolvecl in the trade and identification of of the article, as described above). For some minerals, jades. CMS the author includes interesting bits of history or obser- vations from the Russian perspective. INSTRUMENTS AND TECHNIQUES The third part, "Review of Mineralogical Finds," ROSIGEM Optics Model RFA 322 Refractometer. A. is a brief history of mineralogic discoveries in Russia Shields and B. Neville, Australian Gemmologist, starting about 11,000-12,000 B.C., when the prehis- Vol. 18, No. 11, 1994, pp. 354-355. toric people of Imoonstone: A new moonstone variety. H. The evaluators tested various gems with the Harder, Jozrrnal of Gemmology, Vol. 24, No. 3, instrument and found that it produced clear readings, 1994, pp. 179-182. with the high-quality scale in 0.005 intervals being Moonstones with a "smoky" body color have recently particularly helpful in reading R.l.'s of stones with low been coming from two localities in Sri Lanlza-near birefringences. The narrow width of the hemicylinder lmbulpe in the Central Mountains and in the was an advantage when testing prong-mounted stones Balangoda region. Chemical composition revealed a (presumably because the prongs would clear the hemi- higher iron content than is typical for colorless moon- cylinder on either side). stone. Much of the material is heavily included, most- Although there were some criticisms (for example, ly with clay minerals. The blue adularescence shows some light lealzed between the plate housing, prism, up particularly well when the body color is dark. The and outer case), the evaluators felt that the instrument author proposes the variety name "smoky moonstone" would appeal to many in the field. R CK for this material. CMS

Studies on kosmochlor, jadeite and associated miner- als in jade of Myanmar. Win Htein and Aye Mye JEWELRY RETAILING Naing, Joz~rnolof Gemmology, Vol. 24, No. 5, The China market: Poised to take off. G. Holmes, 1994, pp. 315-320. lewelers' Circnlar-Keystone, Vol. 166, No. 2, In proper gen~ologicalusage, the term jade applies only February 1995, pp. 124-131. to nephrite and jadeite. However, it is well known that Are you an entrepreneur? Do you want to get in on the individual specimens are rarely pure. in this discussion ground floor of an economy that is about to explode? of jadeite from Myanmar, the authors meticulously Then consider China! The Chinese are not only ready

82 Gemological Abstracts GEMS & GEMOLOGY Spring 1995 to talze back Hong Kong, they are already worlzi~lg World gold fabricators stage second half con~eback. toward being number one in the world jewelry marltet Retail [eweller, Vol. 32, No. 838, January 12, by the 21st century. 1995,. -p. 5. With nearly 1.2 billion people, almost half Worldwide gold jewelry production recovered dramati- between the ages of 15 and 39, China is more than just cally in the second half of 1994, rallying from an another marltet. One indication of how fast things are exceptionally subdued first half to end the year just progressing there is De Beers's advertising outlay: from 1.3% down from 1993. $100,000 in 1990 to over $3.4 ~nillioilin 1994. Last According to a report from Gold Fields Mineral year, China consumed 220 tons of gold. The World Services (GFMS),the United Kingdom and Switzerland Gold Council predicts that if the consumption of gold were anlong the strongest performers in Europe; they, reaches the level of their Taiwanese neighbors, China along with Italy, helped final figures for Europe show will soon be importing over 10,000 tons annually! only a 1% decrease. Italy's jewelry manufacturing was Mr. Holines predicts that China will be the leader fueled by an increase in exports in the second half in both jewelry ~nanufacturingand retailing, but the while, in Switzerland, hallrnarlting figures for watch two ends of the business will be intertwined. cases increased 16% over 1993. GFMS reported a 4.6% Hundreds of new jewelry plants have opened in the increase in U.K. jewelry consumption, based on hall- past three to four years. Sixty diamond cutting firms marking figures, imports faring slightly better than are scattered around the country. Private businesses doinestic fabrication. In comparison, however, France are doing well, with plenty of discretionary incoine and Germany perforined poorly. Jewelry production in being spent. Sales in Shanghai's leading department Germany suffered from weak retail sales and increased store were $230 rnillio~lin 1993, up 45% from the pre- competition from imports. Production in France ceding year. declined by more than 17% in the first 10 nlonths of But not all is wonderful. China is still struggling 1994, which CFMS attributed to overproduction the to provide decent housing, drinkable water, and work- preceding year. . ing sewage systems in Inany areas. Bicycles remain the GFMS estimated that jewelry consuinption in the standard mode of transportation, banlts are inefficient, United States was up at least 5% for the first three and import duties are still as high as 90%. quarters of the year. In Japan, increased confidence saw However, the Asian Wo11 Stl.eet journal Weelzly retailers restocltiilg newer designs after an earlier reported that China is anxious to cut the red tape to reduction in inventory levels. MD encourage major retailers to set up shop. And with one-fifth of the world's population in one spot, you can bet the retailers will be coming. Today, gold jewelry is SYNTHETICS AND SIMULANTS considered the third most important purchase in a Characteristics of natural and synthetic diamonds Chinese household, after a color TV and a refrigerator. observed in cathode luminesce~lcephotographs This maltes the potential for gold sales phenomenal. (in Japanese). All Japan Gen~ologyAssociation, De Beers hopes to convince Chinese consumers that Four Seasons of Jewelry, No. 116, August/ diamond, not gold, is forever. GAR Septeinber 1994, p. 37. This brief item points to the possibility that synthetic PRECIOUS METALS diamonds will sooil appear more frequently in the mar- Platinum apples of the asteroids. J. S. Lewis, Natzllz, Izet, and it analyzes the usefulness of cathodoluinines- Vol. 372, Deceinber 8, 1994, pp. 499-500. cence (CL] photographs to separate natural and syn- thetic diamonds. The CL photograph of a natural dia- In the jozlrntil of Geophysical Research (Vol. 99, pp. mond displays a pattern of numerous concentric 21 129-21 141, 1994), Jeffrey S. Kargel analyzed the eco- squares, indicating that it grew slowly. The CL photo- nomic potential of asteroids in Earth-crossing orbits. graph of a synthetic diamond is characterized by sim- The average near-Earth asteroid contains abundant ple geometric patterns. The author concludes that this iron, nicltel, and cobalt, carrying much higher concen- technique is one of the no st effective ways to separate trations of (dissolved]platinum metals than the richest synthetic from natural diamonds. Hirniko Nnlza known ore body on Earth. A kilometer-sized metal- bearing asteroid could be worth as inuch asnUS$5 tril- lion (5 x 1012 US dollars] at present market prices," Hydrothermal ruby (in Japanese). M. Shigeoka, Gemmology, although its impact on the strategic metals marlzet Vol. 25, No. 300, September 1994, pp. 8-10, lowers this value to $320 billion. Rather than trying to This is an inforinative report on the appearance, gemo- divert metal-rich asteroids from impacts with Earth, logical properties, and identifying characteristics of ILargel suggests, we should go mine them. This short five faceted hydrothermal synthetic rubies (0.808- article is a good analysis of the policy aspects of 2.553 ct] that were purchased in Bangkok and report- Kargelts paper. ML/ edly manufactured in Russia.

Gemological Abstracts GEMS & GEMOLOGY Spring 1995 83 As a whole, the rubies were dark red and hat1 that it could be heat treated. Thai buyers obtained rather low transparency due to their wavy growth large quantities at low prices. Today, there are more marlzs, which were very similar to those fountl 111 than 200 heat-treatment furnaces operating in Sri hydrothermal synthetic emeralds. The author believes Lanlza. The authors use a Sri Lanlzan-produced that the relatively high R.1. readings were due to the "Lalzmini" furnace, fueled by a mixture of oxygen and stones' high iron content, which also explains their butane. dark appearance. Their specific gravities ranged from The main focus of the report is a classification sys- 4.00 to 4.03, and most samples fluoresced a strong red tem for heat-treatable Sri Lanlzan corundums. Group 1 to long-wave ultraviolet radiation, with moderate red material is treated under reducing conditions to devel- fluorescence to short-wave UV. op blue coloration. Within this group are two principal The article is accompanied by three absorption varieties, "gi~eda"and oltil. The former is further spectra. Nine photographs (six photoi~~icrographsand divided into four subvarieties: sillzy (which has a char- two laser toinographs) illustrate the stones' appear- acteristic sheen), milky (slightly bluish in transinitted ance, their growth marlzs, characteristic feather inclu- light], waxy (translucent with a waxy luster), and dun sio~lsobserved in one sample, and copper inclusions (colorless or slightly bluish and allnost transparent). seen in another. Hinliko Nmka Ottil is transparent corundum with various patterns of blue-and-colorless zoning. It is divided into two basic A new type of synthetic ruby from Russia. U. Henn, types: Ethul ottir (which has patches of blue color Azrstralian Gemmologist, Vol. 16, No. 1 1, 1994, within the stone and is further divided by the specific pp. 362-364. zoning distribution into dot ott77, iri ottu, and normal ottu) and pito ottil (blue color restricted to the stone's Beginning with a brief review of gemstone synthesis in surface). Russia, the author proceeds to describe his examina- Group 11 corundum can be heat treated under oxi- tion of a single 138-ct "boule-lilze" synthetic ruby dizing conditions to either eliminate blue or enhance crystal, represented to him as having been grown yellow. This general category includes bluish ri~by hydrothermally in Novosibirsk, Siberia. The crystal (which appears purple clue to the presence of both blue had a small area with a nletallic coating and displayed and red chromophores), red "guedo" (pink to red darlz, eye-visible, dendritic inclusions. Other proper- corundum with characteristics such as a silky appear- ties determined were: R.1.- E = 1.762, o = 1.770; bire- ance), and ICowangu yellow snpphire (corundum that is fringence-0.008; S.G.-4.00; long-wave UV fluores- slightly yellow in reflected light). cence-strong red; short-wave UV-weak red. For each type of illaterial described, the authors Magnification revealed that the dendritic flux was provide basic heating conditions, inclurling tempera- polycrystalline. Also noted were clouds of bubble-like ture and length of heat treatment. This interesting flux residues in partially healed fractures. report provides insight into the current state of lznowl- Spectrophotometry revealed absorption features in edge in Sri Lanka about enhancing that country's the visible range due to Cr3+. Chemical analysis corundum. RCIZ revealed the expected aluminum and chromium oxides, while analysis of the metallic coating showed the presence of tungsten and molybdenum; these latter elements were also detected where darlz dentlritic MISCELLANEOUS inclusions in fractures reached the surface. Microscopic examination of mineral grains in forensic The author concludes that the synthetic ruby is a soil analysis: Part 2. N. Petraco, American new type of flux-grown product. RCK Laboratory, Vol. 26, No. 14, September 1994, pp. 33-36. TREATMENTS This article discusses further the identification, with a polarized-light microscope, of common mineral grains Heat treated cor~induinsof Sri Lanka: Their heat treat- found in soil samples (part 1 appeared in the April 1994 ment. T. G. Pemadasa and M. V. Danapala, issue of American Laboratory and was abstracted in Australian Gemmologist, Vol. 18, No. 11, 1994, the Winter 1994 issue of Gems d Gemology). The pp. 346-347. author uses the Michel-Levy interference chart to This report, by two employees of Sri Lanka's State address the topic of interference colors observed Gem Corporation, bcgins with a brief history of that between crossed polarizers. This short but inforillative country's "gueda" sapphire [spelled geuda by this jour- article contains a dozen photomicrographs and several nal and other U.S. publications, such as Kurt Nassau's excellent references. With this information, readers Gemstone Enhancement].Until the 1970s, Sri Lankan can learn to identify microscopic minerals found in miners often discarded this off-color material, unaware their owl1 backyard. CEA

84 Gemological Abstracts GEMS W GEMOLOGY Spring 1995 Authors must rise the following guide- English. We will consider articles already for review]. The various conlponents of lines when submitting a manuscript to published in languages other thnn English the manuscript should be prepared and Gclns cd G~cn~ology,or the paper may be only on a case-by-case basis and only if arranged as follows: returned unreviewed. Please contact the the author(s1 inform us at the time of sub- E,ditorls office if you have any questions mittal when and where the article was Title Page. Page 1 should include: (a] the about these guidelines or any other first published. It is our policy not to pub- article title; (b]the full name of each aspects of your topic. lish feature articles, review articles, or author; (c) each author's affiliation (the Notes on single stones or research based institution, city, and state or country on a single samplc (unless of extraordi- where he/she was working when the arti- nary historical importance, such as the cle was prepared); and (d) acknowledg- Gems d Gemology publishes original ments of persons who helped perform the articles concerning the study of gem- Hope diamond). Authors may not publish the same material elsewhere for at least research, prepare the report, or do the stones and research in gelnology and photography, etc. relatcd fields. Appropriate topics include three monihs after the nail date of the [but are not liniited to) colored stones, issue in question, wiihout express written Abstract. Page 2 should repeat the title of diamonds, gemological i~lstruincntsand pernlission from the Editor's office. the article, followed by an abstract. The identification tecluliques, gem localities, Gems d Cen~ology;ilso includes these abstract (no inore than 150 words for a gem enh.ulcemcnts, gem substitutes (sim- regular sections: Gem T'deLab Notes-- feature or review article, 75 words for a ul,mts and synthetics], gcmstoncs for the reports of interesting or unusual gem- Note] should state the purpose of the collector, jewelry manufacturing arts, stones, inclusions, or jewelry encountered article, what was done, and the main jewelry history, and contemporary trends in the GIA Gcn~Trade Laboratory; Book conclusions. in the trade. Ren~emberthat these are Reviews--as solicited by the Book 'Text. Papers should follow a clear outline general guiselincs only. Your best source Reviews editor (publishers should send with appropriate headings. For example, of appropriate topics is Gems d one copy of each book that they wish to for a research papel; the headings should Gelnology itself, liead the journal to have reviewed to thc Editor's office]; be: Introduction, Previous Studies, acquaint yourself with thc breadth of Ge~nologicalAhstrocts-summaries of Materials and Mcthods, Results, material coveretl. important articles recently published in Discussion, and Conclusion. Use other ~Manuscril~slnay be submitted as: the gemology li teraturc. heads and subheads as thc subject matter Featwe Articles-full-length articles warrants. Also, when writing your article, tlescribing previously unpublisl~cdstudies please avoid jargon, spell out the first and laboratory or field resc,arch. Such arti- ~nentionof all abbreviations, nnd present cles should be no lo~lgerthan 6,000 words All material (including tables, figure leg- your material as clearly and concisely as (24 double-spaced, typed pages) plus ends, and references] MUST bc typed, possible. For general style (grammar, etc.), tables, illustrations, and references. double spaced (no exceptions], one side see The Chicugo M~rrunlof Style only, 011 8.5 by 11 inch (21 by 28 cm] [University of Chicago Press, Chicago]. Review Articles~omprehensivercviews white paper (no colors) with 1.5-inch (3.8 Papers that describe original research of topics in the field. Length of text cni] margins. Each page sho~~ldbc consec- must include a ~Matei.ialsand Methods should not excced 8,000 words (32 dou- utively numbered-including thc first section that contains, at a minimuin, the ble-spaced, typcd pages]. and those containing figure legends, refer- numbers and descriptions of all samples ences, ctc. Use a fresh typewriter or print- Notcs & New Techniques-brief prelinii- examined, and the techniques and instnl- er ribbon; one that prints black, not gray. mentation used to obtain the data. nary con~~nunicationsof recent discover- Please avoid sending draft-quality dot- ies or developments in gemology and matrix printouts. It is strongly recom- lieferences. References should be used for related fields (c.g., new instruments or mended that the article also Ile submitted any information that is talten dircctly identification techniques, gem ~ninerals on a floppy disk in: Microsoft Word 4 froin another publication, to document for the collector, and lapidary techniques). through 6 (DOS),Wordstar 3 through 5.5, ideas and facts attributed to another writ- Articles for this section should be approx- Wortlperfect 5+, Rich Text Format (RTF], er, and to refer the reader to other sources imately 3,0003,000 words (4-12 double- Microsoft Word for Windows (2 or 61, for additional idonnation on a particular spaced pages!. Windows Wr~te,Microsoft Word 4+ for subject. References must be cited in the Gel11 News-very brlef (100-500 words, Mac~ntosh,or ASCII (DOS and ~Macin- body of the text (in parentheses], with the one-half page to two double-spaced pages) tosh). Please indicate which forinat is last name of the author(s)and the year of items on currcnt evcnts in the field or used. publication; add the appropriate page unusual gem materials. ldcntify the authors on the title page nutnber when citing a direct quote or a only, not in the body of the manuscript or specific illustration or set of lumbers or To be considered for publication, all con- figures, so that author anonymity can be data. An example would bc: (ICaminerling tributions to Gems d Gemology must be maintained with reviewers [the title page et al., 1990, p. 33). The list of references original and not previously published in is removed before the ni.muscript is sent at the end of the paper should be typed

Guidelines for Authors Gems &Gemology Spring 1995 85 double spaced in alphabetical order by ternis and figl~resuscd in the table arc Send all subn~issionsto: thc last name of the senlor author; ~f consistcnt with those used in thc tcxt. Al~ccS. Keller, Editor thcre 1s more than one paper by that Gems ed C:enlology author or same group of authors, list Line Illustrations. We prefer that all line art (graphs, charts, etc.) be scnt to us on 1660 Stewart Street those papers in chronological order, start- Santa Monica, CA 90404 ing with the oldest paper first. List only computer dislc, acconipanied by hard copy. (Please contact 11s about acceptablc those rcferences actl~allycited in the text Copyright, In vicw of U.S. copyright law, [or in the tables or figurcs). A paper that formats.) If this is not possible, linc art should be professionally drawn ant1 the we must ask that submittcd manuscrivts lacks a referencc scction or is missing be accolnpanicd by thc following state- information in the reference section- original sent to us. Submit black-and-white pho- mcnt, signed by all of the authors: publishers, dates, authors, etc.-may be "Upon publication of (titlc of article) returned unreviewed. tographs ant1 photomicrographs in the final desired size if possible. Wherc app1.o- in Gems d Gelnolog): I (wc)transfer to Include the following information, the Gc~nologicalhistitutc of hnerica all in the order given here, for each refcrence: priate, please usc a bar or othcr scale marker on thc photo, not outside it. elcctronic-use, scrialization, and ancillary (a) all author names (surnames followed rights, titlcs, and intcrest to the work, by initials); (b)the year of publication, in On thc back of cach illustration, affix a labcl with thc appropriate figure including copyright, togethcr with lull parcntheses; (c)for a journal, the full title right ant1 a~~thorityto claim worldwide of the article or, for a book, the full title of number and tlic article title. Do not trim, mount, clip, or staple illustrations. copyright for the work as p~iblishedIn the book cited; and (dl for a journal, the this journal. As author(s),I (we)rctain the full titlc of the journal plus volume num- Color Photographs. For originals, 35-117111 right to cxcerpt (up to 250 words) and ber, issue number, and inclusive page slides or 4- by 5-inch transparencics are reprint the material on request to the numbers of the articlc cited or, for a book, ideal. For review purposes (see bclow) we Gemological Institutc of Amcrica, to the publisher of the book and the city of need an additional three complcte sets oi nialte copics of the work for ~iscin class- publication. Sample referenccs are: color prints, slide dupcs, or high-quality room tcacliing or for internal distribution Kammerling R.C., 1Coivula J.I., ICane R.E. color Xcroxes submittcd with the copies within my (our)place of employment, to (1990) Gemstonc cilhanceinent and ol the paper. We rcscrve the right to reject use-alter publication-all or part of this its detection in the 1980s. Gems ed photographs that are not in keeping with material in a book I (we) havc authored, Gemology, Vol. 26, No. 1, pp. 32-49. the production standards of Gems d to prcscnt this material orally at any Armstrong J.T. (1988)Accurate quantita- Gemology. Pleasc contact the Editor's function, and to veto or approve perniis- tivc analysis of oxygen and nitrogen office if you need help finding appropriate sion grantcd by the Gcrnological Institute with a Si/W multilayer crystal. In photographcrs or specimens. of America to a third party to rcpublish all or a substantial part of thc article. I D.E. ~chbur~,~d.; Microbeail] "Call Outs." Figurcs and tables ~iiustbe (wc) also retain all proprietary rights Analysis-1988, San Francisco Press, callcd out at the appropriate placc in tlie other than copyright (such as patent San Francisco, CA, pp. 301304. tcxt. Figures and tablcs must be num- rights]. I (we] agrec that all copics of the Liddicoat R.T. (1989) Hondbook of Gel11 bered consecutivcly, starting with the article made within thcsc terms will Identificatioi~,12th cd., 2nd rev. first mention in thc text. printing. Gemological lnstitutc of include noticc of thc copyright of the America, Santa Monia, CA. Figure Captions. Type figure captions Gemological lnstitutc of America. This Personal communications (for infor- dottble spaced on a separate pagels). Each transfcr of rights is made in view of tlie mation obtained from someone with par- caption should clearly explain, in com- Gcrnological institute of America's ticular expertisc) shoultl be citcd in the plete sentences, the significancc of the cfforts in reviewing, editing, and p~~blisli- hody of thc text only, as follows-(G. figure and any syml~ols,arrows, numbers, ing this material. Rossnian, pers. comni., 1994). Permission or abbrcviations used therein; it should As author(s), 1 (we)also warrant that must be obtained fro111 the people cited to be consistent with the text. Whcrc a this article is my (our)original work. This use their names for this purpose. Such magnification is appropriatc, please article has been submittcd in English to resources should also he listed, with their include it in the caption. (For tlic purpos- this journal only and has not been puh- affiliations, in the Aclcnowlcdgments sec- es of this journal, the magnification for a lished elsewhere." tion. photomicrograph should refer to thc No payment is made for articles pub- magnification at which the image was lished in Gems d Genlology. However, Tables. Tables can present a large aniount photographed.) for cach article tlic author(s) will receivc of detail in a relatively small space. Con- 50 free copics of that issue. sider using onc whenever the bulk of information in a section thrcatcns to ovcr- Because your manuscript will be revicwed whelm the text. by at least tlircc separate revicwers, we iManuscripts are examincd by the editor, Type each table double spaced on a must have three complete copies of the editor-in-chief, technical editor, and at separate sheet. If tlie table must cxceed paper and tlvee coinplete, properly labclcd least tlvee reviewers. Authors will remain one typed page, plcase duplicate all head- sets of all illustrations. Color Xeroxcs of anonymous to the reviewers. Decisions of ings on the second sheet. Numhcr tables the color illustrations are acceptablc as the cditor are final. All material accepted in the order in which tlicy are cited in the long as they are of reasonable quality. We for publication is subjcct to copy editing; text. Every table should havc a title; evcry cannot consider any sub~nissionthat does authors will receivc gallcy proofs for column (including tlie left-hand column) not include these three copies of the revicw and are held fully responsible for should have a heading. Please make sure manuscript and figures. the content of thcir articles.

86 Guidelines for Authors Gems W Gemology Spring 1995