See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/271198687

The Geology and Genesis of Gem Corundum Deposits

Chapter · February 2014

CITATIONS

23READS

2,308

5 authors, including:

Some of the authors of this publication are also working on these related projects:

minerals from Latium View project

no title View project

Daniel Ohnenstetter

Centre de Recherches Pétrographiques et Géochimiques

141 PUBLICATIONS   2,788 CITATIONS   

SEE PROFILE

A.E. Fallick

University of Glasgow

713 PUBLICATIONS   16,465 CITATIONS   

SEE PROFILE

Andrew Fagan

University of British Columbia - Vancouver

10 PUBLICATIONS   36 CITATIONS   

SEE PROFILE

All content following this page was uploaded by A.E. Fallick on 11 July 2018.

The user has requested enhancement of the downloaded file.

GIULIANI ET AL. �– GEM CORUNDUM

29

CHAPTER 2: THE GEOLOGY AND GENESIS OF GEM CORUNDUM DEPOSITS Gaston Giuliani1, 2, Daniel Ohnenstetter2, Anthony E. Fallick3, Lee Groat4, Andrew J. Fagan5,4 1 Université Paul Sabatier, GET, UMR 5563 CNRS-IRD-CNES-UPS, 14 avenue Edouard Belin, 31400- Toulouse, France; e-mail: giuliani@crpg.cnrs-nancy.fr 2 Université de Lorraine, CRPG, UMR 7358 CNRS-UL, BP 20, 54501- Vand�œuvre-lès-Nancy cedex, France 3 Scottish Universities Environmental Research Centre, Rankine Avenue, East Kilbride, Glasgow G75 0QF, Scotland, UK 4 Department of Earth and Ocean Sciences, University of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada 5 True North Gems Inc., Vancouver, BC, Canada

Mineralogical Association of Canada Short Course 44, Tucson AZ, February 2014, p. 29-112

INTRODUCTION The aim of this review is to present the state of the art on the geology and genesis of gem corundum deposits, including the main characteristics and classifications, to an audience of non-specialists, academic students, professional geologists and gemologists. The careful detailed descriptions and studies on the geology, mineralogy and structures of the deposits carried out over the last century provides a foundation to unravel their genesis through new field and laboratory research including petrography, mineralogy, gemology, fluid inclusions, and stable and radiogenic isotopes. Descriptive geology coupled with recent advances in geology of gems permits decoding of how some of these historical and fabulous deposits formed as well as to estimate exploration potential in the future. New oxygen isotope data are presented here for the first time.

HISTORY, GEOGRAPHIC DISTRIBUTION, AND ECONOMIC SIGNIFICANCE The term corundum likely originates from the Sanskrit "kurunvinda" meaning "hard stone", that became kurund in Dravidian popular language (corundum is kurund in German). This term might also have originated from the Tamil word kurmidam (Anthony et al. 1997). The term "ruby" is derived from the Latin word ruber. In Sanskrit, it was called

ratnaraj, queen of precious stones and symbol of permanent internal fire. In Greco-Roman times, the reference to fire was common; in 400 CE Theophraste (Eichholz 1965) related ruby to anthrax (coal). Two thousand years ago Pliny (Littré 1850) called all the red stones carbunculus (diminutive of carbo) which became escarboucle in old French. In the 11th century, Marbode (Hernault 1890) separated "three times" three types of carbuncles that correspond to the three hues known for ruby (Myanmar, Thailand, and Sri Lanka), red spinel (balas, ruby�–spinel, and pleonaste), and red garnet (pyrope, almandine, and spessartine). The exact origin of the term sapphire is unknown; it probably comes from the Sanskrit sauriratna, became sappheiros to the Greeks, then sapphirus to the Romans. Up to the 13th century this term referred to all blue gemstones, especially lapis lazuli. Marbode used the term hyacinthe to refer to all colored sapphire. In the 18th century the contribution of the French scientists to the characterization of the corundum as a mineral species was major. In 1784, Romé de l'Isle showed that ruby and sapphire were similar minerals. In 1787, Brisson published the specific gravities, and in 1805 René-Just Haüy (Brard 1805) establishes the chemical formulae of both gemstones and formally united all varieties under corundum (Haüy 1817). In the 19th century,

GIULIANI ET AL. �– GEM CORUNDUM

30

Chenevix in 1802 first established the chemical composition of ruby and sapphire. The work of Rose allowed the characterization of corundum as practically pure alumina (Rose 1843). Between 1894 and 1921, the field studies in Madagascar of Lacroix and his team allowed the mineralogical inventory of the island. In 1922, Lacroix published his three-volume book on the Mineralogy of Madagascar, a fascinating treatise on Malagasian mineralogy and especially on corundum, its occurrence, typology, and mineralogy. It was the first book to present a geological overview of worldwide corundum deposits. Today, blue gem (translucent or transparent) corundum is called sapphire, and all other colors including yellow, orange, violet, green, purple, peach-apricot, and brown commonly called fancy sapphires are referred to by color (e.g., yellow sapphire, orange sapphire, etc.). The valuable other colored sapphire is the orange-pink or pinky orange variety called padparadscha after the lotus blossom. Since 1990, gem corundum has been the subject of several reference books on the gemology, mineralogy, and geology of this fascinating gemstone. The following were written by Hughes: Corundum in 1990 gives an invaluable comprehen-sive gemological and geological overview of gem corundum; Ruby and Sapphire in 1997 updated the first edition and completed the story and lore of ruby and sapphires, and Ruby & Sapphire: A Collector's Guide by Hughes & Manorotkul in 2014. Ward (1992) gave a general overview of gem corundum in his book Rubies and Sapphires while Themelis (1992) documented all the secrets of heat treatment applied to ruby and sapphire. Bowersox & Chamberlain (1995) provided a wonderful and very

informative review of the corundum and spinel deposits from Afghanistan in their Gemstones of Afghanistan. The English edition of the book Geology of Gems written by Kievlenko (2003) offered an interesting overview on the geology of gem corundum worldwide. The geology and genesis of gem corundum deposits was reviewed by Giuliani et al. (2007a) in the first edition of the Mineralogical Association of Canada, Short Course 37; on the Geology of Gem Deposits (Groat 2007). Themelis (2008) wrote the first state of the art on the Gems and Mines of Mogok, the ruby-sapphire-spinel Mogok Stone Tract famous for centuries. Finally, the special issue edited by Ore Geology Reviews (Graham et al. 2008a) on the Genesis of Gem Deposits offered six papers on the topic of corundum deposits in contrasted geological environments (Garnier et al. 2008; Graham et al. 2008b; Rakotondrazafy et al. 2008; Simonet et al. 2008; Sutherland et al. 2008; Yui et al. 2008). In the European literature, in 1998 Weise published in Extra-Lapis Ruby, Sapphire, Corundum in German, and Cesbron et al. (2002) detailed the mineralogy and crystallography in Corundum and Spinel from the French special issue of the Journal Minéraux & Fossiles. More recently, the French Journal le Règne Minéral published a review on the geology and classification of corundum deposits (Garnier et al. 2004a), a review on marble deposits from central and southeast Asia (Garnier et al. 2006a), a synthesis on the French corundum occurrences in the Saphirs et rubis de France (Giuliani et al. 2010), and finally the state of the art on corundum deposits in the Neoproterozoic metamorphic Mozambique belt (Giuliani 2013).

Plate 1 A. Typical mineral association of purplish sapphire from the Zazafotsy deposit (Ihosy region), Madagascar. The sapphire (1.6 x 1.4 cm) is associated with garnet (orange) and potash feldspar (white) in a biotitized feldspathic gneiss (black). Collection MNHN of Paris (N°108-1747). Photograph courtesy of L.-D. Bayle © le Règne Minéral. B. The ruby-bearing anyolite from Longido, Tanzania, is used for carving. The hexagonal prism of ruby (5 cm across) is contained in a gangue of green zoisite and black amphibole (tschermakite). Photograph courtesy of G. Giuliani. C. Polished slab of a colored zoned sapphire crystal from Garba Tula, Kenya. The blue and yellow sapphire (2.3 x 1.5 cm, 19.6 grams) was hosted by a monzonite dyke. The colors and chemistry are similar to those found for the green and yellow sapphires found in placers formed in basaltic environments. Collection J. Saul. Photograph courtesy of L.-D. Bayle © le Règne Minéral. D. Ruby crystal from Ampanihy, Madagascar. The ruby (6 x 4.8 x 1cm) was hosted by an amphibolite. Collection J. Béhier. Photograph courtesy of L.-D. Bayle © le Règne Minéral. E. Ruby in marble from the mine of Baw Padan, Mogok, Myanmar. The crystal is hosted in a gangue of calcite. Collection F. Barlocher. Photograph courtesy of L.-D. Bayle © le Règne Minéral. F. Transparent yellowish sapphire (2.3 x 0.7 x 0.5 cm) presenting blue zoning. Placer of Balangoda, southern Sri Lanka. Collection F. Lietard. Photograph courtesy of L.-D. Bayle © le Règne Minéral. G. Sapphires related to the Mont Coupet basaltic field in the French Massif Central, Saint-Eble, Haute-Loire. The highest crystal on the left: 12 x 8 x 8 mm. The cut stones: 5 x 4 mm. Collection and photograph courtesy of L.-D. Bayle © le Règne Minéral.

GIULIANI ET AL. �– GEM CORUNDUM

31

GIULIANI ET AL. �– GEM CORUNDUM

32

GIULIANI ET AL. �– GEM CORUNDUM

33

Corundum is an industrial raw material, prized for its optical, mechanical and chemical qualities, and in jewelry for its gem varieties, ruby and sapphire (Plates 1 and 2). Natural corundum was mined in antiquity as emery, a mechanical mixture of corundum and magnetite, used as a grinding and polishing compound. The ancient Greeks called emery smyrnis after the harbor city of Smyrna, which served as the main exportation for Naxos emery in the Cyclades Islands (King 1865). The Romans called emery naxium and Pliny in his Natural History (Littré 1850) reported the use of emery as an abrasive for lapidaries and gem engravers. Corundum was mined in North Carolina (USA) during two periods (Yurkovich 1985): the main mining period, between 1871 and 1900, took place in the Clay and Macon counties accounting for all of world production until 1893. An estimated 3,720 to 7,000 tons of ore was produced and the main site of extraction was the Corundum Hill Mine. The second mining period, 1914�–1919, reactivated the mining operations at Corundum Hill. South Africa was also a main producing country in the early twentieth century (Kupferburger 1935). The corundum was obtained from desilicated pegmatite in Northeast Transvaal (Robb & Robb 1986). Production fell from 3876 tons per year in 1918 to 74 tons per year in 1979, and the use of synthetic abrasives hastened the closure of the mines. Today, natural corundum abrasive is used in the fabrication of window glass for watches, and manufacturing ceramics. Alumina powder is used as a refractory mineral for traditional, composite ceramics and for neoceramics due to its high temperature of fusion (T > 2000°C). Alumina has a high resistance to chemical attack and abrasion, high electric and mechanical resistance, is inert in reductive and oxidative atmospheres, and has a low dielectric loss.

The main major and minor commercial, industrial, and historical world sources of corundum are reported in Figure 2-1. The high-value gem ruby and sapphire are important mineral resources which contribute significantly to the domestic product of many countries. In 2005, Africa was responsible for around 90% of world ruby production with 8,000 tons (Fig. 2-2) in addition to 42% of sapphire with 10,700 tons (Yager et al. 2008). Highest quality ruby crystals are from central and southeast Asia. Myanmar, with the Mogok Stone Tract, has produced 'pigeon blood' rubies since 600 CE (Hughes 1997), and the Mong Hsu deposit, first reported by Hlaing (1991), has ruby which was studied in detail by Peretti et al. (1995, 1996). Other Asian producers include Thailand (Chanthaburi-Trat mines) and Cambodia (Pailin). Ruby deposits in Vietnam (Luc Yen, Yen Bai, Quy Chau), Afghanistan (Jegdalek), Pakistan (Hunza valley), Azad-Kashmir (Batakundi, Nangimali), and Tajikistan (Kukurt, Turakoluma, Badakshan) have become significant producers as the traditional areas in Thailand have become depleted. Gem quality ruby is also extracted in Tanzania (Umba-Kalalani, Morogoro, Mahenge, Winza), Kenya (Mangare), Mozambique (Montepuez, Ruambeze, M'sawize), and Madagascar (Vatomandry, Andilamena, Didy). Blue sapphire from Indian Kashmir (Sumjam) is the source of the world's best quality, as well as those from Myanmar and Sri Lanka. Australia, Thailand, Cambodia, Nigeria, Laos, United States, Madagascar, and China are also traditional sources for blue sapphire. Colored sapphire is found worldwide, but Sri Lanka is thus far the only country to produce excellent quality crystals of all colors, including salmon colored padparadscha. Colored sapphire deposits were discovered in Tanzania (Tunduru and Songea) and in Madagascar (Ilakaka, Ambondromifehy, Vatomandry). In 2005

Plate 2 A. Pink sapphire crystals (foreground) with layering in marble. Revelstoke ruby occurrence in British Columbia, Canada. Photograph courtesy L. Groat. B. Cut stones from the Revelstoke ruby and sapphire occurrence. From the bottom to the top the weight of the sapphires are respectively of 0.38, 0.30, 0.20, and 0.14 carats. Photograph courtesy of B.S. Wilson. C. Non-gem ruby in phlogopitite matrix, minor kyanite (blue) and plagioclase (white) also observable, hosted in leucogabbro subjected to metamorphic metasomatism. This material was extracted at surface from the main ore zone at the Aappaluttoq ruby deposit, southwest Greenland, Denmark. The weight of the different cut stones are respectively from left to right: on the bottom 2.37, 0.71 and 4.03 ct, in the middle 0.73, 0.42, and 0.31 ct, and on the top 2.53, 0.77, 5.70, 1.26, and 1.34 ct. Photograph courtesy of True North Gems Inc. D. Faceted rubies and pink sapphires from the Aappaluttoq deposit, Greenland, Denmark. Photograph courtesy of True North Gems Inc. E. Amphibolitic gneiss hosting ruby in the M'sawize deposit at Niassa, Mozambique. The ruby is associated with plagioclase (white), phlogopite (brownish-yellow) and amphibole (greenish). Photograph courtesy of V. Pardieu © GIA. F. A 40-grams silky ruby crystal from Montepuez, Mozambique. Photograph courtesy of V. Pardieu © GIA.

GIULIANI ET AL. �– GEM CORUNDUM

34

FIG. 2-1. Main major and minor commercial, industrial, and historical world sources of corundum, ruby and sapphire.

Madagascar: 1- Ambondromifehy, Anivorano, Nosy Be, Ambato peninsula; 2- Andilamena, Didy, Vatomandry; 3- Ankaratra (Antsirabe-Antanifotsy region, Soamiakatra); 4- Ilakaka-Sakaraha; 5- Tranomaro (Andranondambo). Ethiopia: 6- Kibremengist, Dilla. Kenya: 7- Turkana, 8- Garba Tula; 9- Kitui; 10- Mangare (John Saul mine). Tanzania: 11- Umba valley; 12- Longido; 13- Winza; 14- Morogoro, Mahenge; 15- Songea; 16- Tunduru. Malawi: 17- Chimwadzulu. Mozambique: 18- Montepuez (Namahumbire/Namahaca), M'sawize, Ruambeze. Zimbabwe: 19- O'Briens (verdites). South Africa: 20- Barberton (verdites); 21- Leydsdorp (Cobra pit). Cameroon: 22- Adamawa district. Nigeria: 23- Kaduna plateau. Brazil: 24- Indaia. Chile: 25- El Salvador. Colombia: 26: Mercaderes. United States: 27- Plumas; 28- North Carolina; 29- Montana. Canada: 30- Revelstoke; 31- Baffin Island. Denmark: 32- Greenland (Fiskenæsset district). Scotland: 33- Loch Roag. Germany: 34- Eiffel. Switzerland: 35- Campo Lungo. France: 36- French Massif central. Italy: 37- Piemont. Czech Republic: 38- Jizerska Louka. Slovakia: 39- Cerová Highlands. Norway: 40- Froland. Finland: 41- Kittilä. Macedonia: 42- Prilep. Greece: 43- Xanthi; 44: Naxos. Turkey: 45- Aidin province. Tajikistan: 46- Turakuloma, Badakhshan. Afghanistan: 47- Jegdalek. Pakistan: 48- Hunza valley; 49- Batakundi, Nangimali. India: 50- Sumjam; 51- Orissa, Kalahandi; 52- Karnakata (Mysore); 53- Andhra Pradesh (Salem district). Sri Lanka: 54-Ratnapura, Elahera, Kataragama. Nepal: 55- Chumar, Ruyil. China: 56- Qinghai; 57- Yuan Jiang; 58- Hainan island (Penglai); 59- Fujian (Mingxi); 60- Jiangsu; 61- Shandong (Changle). Vietnam: 62- Luc Yen, Yen Bai; 63- Quy Chau; 64- Da Lat, Ban Me Thuot, etc. Myanmar: 65- Mogok; 66- Mong Hsu. Thailand: 67- Kanchanaburi, Bo Phloi; 68- Chantaburi-Trat. Cambodia: 69- Pailin; 70- Den Chai-Phrae; Laos: 71- Ban Huai Sai. Japan: 72- Ida. Australia: 73- Rosville-Lakeland Downs; 74- Lava Plains; 75- Anakie fields; 76- West-Brisbane fields; 77- New England fields (Inverell); 78- Cudgegong, Barrington; 79- Tumbarumba; 80- Western Melbourne fields; 81- Tasmania; 82- Harts range. New Zealand: 83- Westland (Hokitika). Russia: 84- Podgelbanochny, Shkotovo plateau; 85- Primorye, Nezametnoye; 86- Polar Urals (Hit Island); 87- Ural mountains (Kuchin-Chuksin-Kootchinskoye); 88- Karelia.

the Canadian company True North Gems Inc. collected ruby and pink sapphire from the Aappaluttoq deposit near Fiskenæsset (Qeqertarsuatsiaat), Greenland (Rohtert & Ritchie 2006, Fagan et al. 2011, Fagan 2012). Other countries known to produce gem corundum sporadically include Brazil (Malacacheta, Indaia), Canada (British Columbia, Baffin Island), Colombia (Mercaderes), Malawi (Chimwadzulu), Russia (Urals), and the United States (Idaho). Rubies from southeast Asia (Vietnam, Myanmar) are cut and sold in Bangkok. African and

Malagasy rubies are treated and cut in Bangkok and sold at Idar-Oberstein in Germany. Bangkok is also the main center of distribution for sapphires from Australia, Thailand, and Cambodia. Sapphires from Sri Lanka are sometimes negotiated in Colombo, but the greater part is exported to Bangkok. Today, Hong Kong also remains an important trade center for ruby and sapphire. The ruby and sapphire industry is a very lucrative business. Hughes (1997) devoted a chapter to the corundum gem market and pricing, �‘How to judge the quality of fine ruby and sapphire and how

GIULIANI ET AL. �– GEM CORUNDUM

35

FIG. 2-2. Global production of sapphire (A) and ruby (B) in 2005 with the percentages (%) for African countries (modified

from Yager et al. 2008). Global BGY sapphires production by country (C) and by type of deposit comparing BGY sapphires in basaltic fields to other geological types (D).

to define a price of sale?�’ The term �‘fine�’ may vary from one store to another, and among different countries, and so too the price. When buying natural quality ruby and sapphire, the criteria are based on the geographic origin, but also on the clarity, the color, the cut, and the weight in carats of the gem (Hughes 1997). Enhancements and treatments are also of importance in the final evaluation, and it is very rare to find gem corundum that has not been heat-treated (Themelis 1992). Thailand has great expertise in heat treatment and today crystals of corundum are routinely heat-treated in Thailand and Australia. Drucker (in Shigley et al. 2000), at the end of 1990, explained that the price of ruby fell after controversy over its treatments. The medium price of sale went from 4000 to 6000 USD/ct between 1990 and 1992, and the price remained stable up to 1997. The discovery of the Mong Hsu deposit in 1991 provided a new global source of excellent Burmese rubies, but the availability on the market of heated remnants of Mong Hsu rubies caused a fall in the price, to less than 4000 USD/ct, between 1997 and 2000. Recent discoveries and exploitation of ruby associated with biotitite in amphibolite such as at the Aappaluttoq (Greenland) and Montepuez (Mozambique) deposits will ensure the supply of gem ruby in the near future (Plate 2). They represent huge tonnage of ruby, approximately 400 million cts for the Aappaluttoq deposit, with around 5% classed as gem grade (Fagan 2012).

Sapphire has not been affected by such controversy and the prices have been stable around 2500 USD/ct since 1991. In 2001 the input on the gem market of a new kind of heat-treated orange sapphire has caused much controversy and the issue of erroneous certificates bearing the name of natural padparadscha (McClure et al. 2002, Fritsch et al. 2003). Natural sapphire from Songea, Ilakaka and even Burma and Thailand were heat-treated at 1800°C and in an oxidizing atmosphere in the presence of Be. Beryllium diffused from the outside of the stones inward, adding a yellow and orange component and more clarity to the color of the natural stone. Nowadays this sapphire is certified diffusion-treated sapphire not to be confused with natural untreated and heat-treated padparadscha. The prices for fine untreated padparadscha sapphires are quite demanding and only a few have the opportunity to purchase them. Even heated padparadscha sapphire is quite expensive if it is good quality. Ruby and sapphire are the most important colored gemstones in today�’s world gem trade; together they account for over 50% of global colored gem production (Hughes 1997). Ruby is perhaps the world�’s most expensive gemstone; the best Burmese rubies are more valued than an equivalent-sized awless colorless diamond (Walton 2004). The world record price at auction paid for a single ruby is USD 210,000 per ct in 2012

GIULIANI ET AL. �– GEM CORUNDUM

36

for a 32.08 ct Burmese ruby (total USD 6.74 million). The world record price paid at auction for a single blue sapphire is USD 3.84 million for a spectacular 26.41 ct cushion shaped Kashmir sapphire ($145 342 per ct). Recently discovered ruby and sapphire from the Didy mine in Madagascar (Pardieu 2012) also commands high prices for natural unheated ruby. A set of eight faceted rubies ranging from 7 to more than 14 ct exceeded an estimated market value of USD 10 million (Peretti & Hahn 2013). MINERALOGY, PHYSICAL PROPERTIES AND GEOCHEMISTRY Corundum belongs to the hematite group (X2O3) of rhombohedral oxides comprising hematite (Fe2O3), corundum (Al2O3), eskolaite (Cr2O3), karelianite (V2O3), and tistarite (Ti2O3). There are no solid solutions between any of the five species but they have the same type of structure. Hematite group mineral structures are based upon hexagonal closest packing of O atoms, with cations in octahedral coordination. Corundum crystallizes in the 3 2/m class of the rhombohedral system with, for the hexagonal multiple unit cell, a between 4.76 and 5.04 Å, and c between 12.99 and 14.01 Å (Cesbron et al. 2002). The smallest unit cell volume of the group is for corundum (rAl

3+ = 0.54 Å) and the largest for hematite (rFe

3+ = 0.65 Å). The network of O2�– ions (rO

2�– = 1.40 Å) is formed practically by a hexagonal closed packing formed by layers of ions arranged successively in the order ABAB...etc. (Fig. 2-3). All the layers are stacked two-two. Four

stacked atoms of O leave between them a tetrahedral site whereas six contiguous atoms of O leave between them a bigger octahedral site occupied by Al3+. As there are two atoms of Al for three atoms of O, two octahedral sites in three are occupied in each layer of Al. The chromophorous elements of corundum, such as Cr3+, enter the corundum lattice in place of Al. Euhedral crystals can present different faces (Fig. 2-4) that correspond to seven crystalline forms (Cesbron et al. 2002): the pinacoid {00.1}, the first order hexagonal prism {10.1} and second order {11.0}, the hexagonal prism {kk.0}, the hexagonal dipyramid {hh.l}, the ditrigonal scalenohedron {hk.l} and the rhombohedron {h0.l}. The first five crystalline general forms are also present in the classes that belong to the hexagonal system. Corundum can also crystallize in a particular texture called 'trapiche' (Fig. 2-5), formed by six skeletal arms which separate six sectors of growth (Sunagawa et al. 1999, Garnier et al. 2002a). Corundum has a hardness of 9 on the Mohs scale and the specific gravity is between 3.98 and 4.06. The lustre is adamantine to vitreous, transparent to translucent. The cleavage is denied in the literature but it is present in synthetic colorless sapphire (in Hughes 1997). Cleavage is found along the rhombohedron {h0.l} and the prism {kk.0}. Parting is identical to cleavage and fracture is conchoidal to sub-conchoidal when the break does not follow the cleavage. The melting point of corundum is 2044°C and its boiling temperature as 3500°C.

FIG. 2-3. A, Hexagonal close packing of O2�– spheres. In the second layer, the spheres are located in the B position, at the contact with three spheres of the lower layer, the third one overlapping in the first one. In a cubic-centered faces, the spheres of the third layer would occupy the C position. B, Octahedral site formed by the superposition of three spheres in the position B on three others in position A. Metallic ions are placed in the C position and they are in octahedral coordination. At the left is a vacant tetrahedral site in the structure of corundum (Cesbron et al. 2002). C, The structure of corundum looking down the c axis (Hughes 1997). D, Perspective view of the corundum structure. Only two of three octahedral sites are occupied in each Al3+ layer.

GIULIANI ET AL. �– GEM CORUNDUM

37

FIG. 2-4. Crystalline forms of the 3 2/m class of the

rhombohedral system (after Cesbron et al. 2002). A, positive rhombohedron {10.1}; B, negative rhombohedron {01.1}; C, hexagonal dipyramid {hh.l}; D, pinacoid {00.1}; E, hexagonal prism of first order {10.1}; F, hexagonal prism of second order {11.0}; G, dihexagonal prism {hk.0}; H, ditrigonal scalenohedron {hk.l}.

Corundum is uniaxial negative. Pure corundum has a refractive index n = 1.7687 and n = 1.7606. Colored corundum with Fe and/or Cr atoms has a variation respectively between 1.766 and 1.780 for of n and 1.758 and 1.772 for n . The birefringence is extremely consistent at 0.008�–0.009. Additional optical and physical properties of corundum as well as experimental work on its solubility, dissolution

and phase equilibria are summarized by Bowles (2011). Pure corundum is colorless; the array of gem corundum colors is due to impurities (chromo-phores) that substitute for Al (in the form Al3+) in the internal structure. Hughes (1997), Fritsch & Rossman (1987, 1988), Emmett et al. (2003), Peretti et al. (2008), Schwarz et al. (2008) and Rossman (2009) provided excellent reviews and articles on current understanding of corundum coloration. Chromophorous elements are metallic ions from the transition elements family of the periodic table. When chromium (in the form Cr3+) replaces Al3+, a red color results in the crystal but only if the Cr2O3 content is between 0.1 to 3.0 wt.% (Hughes 1997). However concentrations of 9.4 wt.% were detected in the ruby of Karelia (Bindeman & Serebryakov 2011)) and up to 13 wt.% in the ruby hosted in a rock called �‘goodletite�’ within serpentinite from Westland in New Zealand (Grapes & Palmer 1996). The 'goodletite' is composed of ruby, tourmaline and emerald green fuchsite and named for its discoverer William Goodletite. The optical absorption spectrum of ruby consists of two large absorption bands (Fig. 2-6). They are situated at highest energy with two transmission windows at 480 nm centered in the

FIG. 2-5. Trapiche textures. A, Trapiche ruby from Mong Hsu. B, Texture in trapiche emerald (left) and trapiche ruby (right) following a basal section (Sunagawa et al. 1999). Longitudinal section parallel to the c axis in a trapiche ruby (bottom). C, Detail of the core of a trapiche ruby from Mong Hsu (Garnier et al. 2002a).

GIULIANI ET AL. �– GEM CORUNDUM

38

FIG. 2-6. Typical polarized UV�–Vis absorption spectra of sapphire and ruby. A, polarized spectra (E//c) of four sapphires of

different deposits (Notari & Grobon 2002). The interpretation of color, indicated for iron (Fe3+, line at 388nm and Fe3+ pairs, lines at 450nm and 377nm), and Fe and Ti (Fe2+ Ti4+, band at 565nm). B, polarized absorption spectra (E//c) of Winza ruby (Schwarz et al. 2008) and orangey-pink sapphire from Madagascar (Peretti & Günther 2002). The spectrum of ruby shows the well-known Cr3+ absorption bands at 405�–410 and 560nm, and the Cr "doublet" at 694nm. The sapphire from Madagascar is reminiscent of padparadscha color (Fe + Cr), with for Cr (Cr3+, band centered at 405�–410 nm and 554 nm), for Fe (Fe3+, line at 388 nm, and Fe3+ pairs, lines at 450 nm, and bands at 540 nm), and for Ti (Fe2+ Ti4+ charge transfer band centered at 565 nm).

blue, and at 610 nm centered in the red. As the human eye is more sensitive in the red just above 610 nm than in blue, ruby appears red. In the corundum structure, Cr3+ occupies a site where the distance between the metallic ion and O is short (1.913 Å), and in consequence the Cr3+ ions have high electrostatic repulsion and the absorption feature is shifted to higher energies than for other Cr-bearing minerals such as green Cr-zoisite where the average metal�–O distance is 1.967 Å (Dollase 1968), causing the red color. Chromium-related red fluorescence under ultraviolet light and sometimes even in daylight, combined with the red color of ruby, is the cause of the particular fire effect found in Burmese and Vietnamese rubies. Substitution of Al3+ by iron (Fe2+or Fe3+) and titanium (Ti4+) generates the blue color of sapphire by a mechanism of heteronuclear intervalence charge transfer between metal ions, i.e., Fe2+�–O�–Ti4+ (Fig. 2-6) The heat treatment of colorless �‘geuda�’ sapphires produces blue stones because during the dissolution of inclusions of rutile and spinel at high temperatures, Fe and Ti are released into the crystal and form pairs of Fe2+�–O�–Ti4+, which then allow intervalence charge transfer as Fe2+ Ti4+ (Themelis 1992). Color in colored sapphires is due to the classical substitution of Al3+ by a combination

of other elements such as Fe, Ti and Cr (Fritsch & Rossman 1987, Hughes 1997), and charge transfer involving Fe2+, Fe3+, Ti4+ (Fig. 2-6), and color centers (Fristch & Rossman 1988). Less than 0.01% of Fe and Ti are necessary to obtain the blue color of sapphire. Color from fuchsia to reddish can be ascribed to sapphires (Caucia & Boiocchi 2005, Ralantoarison et al. 2006, Andriamamonjy et al. 2013) and they then resemble ruby; however, the concentration of Fe is always higher than Cr and the fluorescence is very low because Fe quenches the fluorescence effect. In marble deposits, a crystal of ruby can be made up of regions of ruby and regions of pink sapphire but in each zone, the content of Cr2O3 will always be higher than the content of Fe2O3. The definition of ruby (like emerald; Giuliani et al. 2004) has to take into account not only the color, but also the chemistry (Cr2O3 > Fe2O3), and the absorption spectra features. The orange-pink to pinkish-orange colored corundum variety is called �‘padparadscha�’ (Notari 1997), a term derived from the Sanskrit padmaraga, signifying the color of the lotus flower. The color is due to the presence of Cr3+ in octahedral coordination (pink hue) and of Fe3+ in intervalence charge transfer with O2�– (yellow hue; Schmetzer & Bank 1981).

GIULIANI ET AL. �– GEM CORUNDUM

39

THE GEOLOGY OF GEM CORUNDUM DEPOSITS Corundum forms in mafic and siliceous geological environments, but it is always associated with rocks depleted in silica and enriched in alumina. In the presence of silica, Al is preferentially incorporated into aluminosilicate minerals such as feldspars and micas. The mineral association of corundum commonly consists of plagioclase, sapphirine, biotite, phlogopite, amphibole, pyroxene, and carbonate minerals (Garnier et al. 2004a). Gem corundum is rare because its requires an environment enriched in alumina and impoverished in silica, but also the presence of Cr, Fe, and Ti to substitute for Al in the structure (Muhlmeister et al. 1998, Abduriyim & Kitawaki 2006), and thermobarometric conditions favorable for its crystallization and stability (Kievlenko 2003, Simonet et al. 2008). Corundum can be hosted in different rocks type but two major geological environments are favorable for the presence of gems (Simonet et al. 2008). These are (1) amphibolite- to medium pressure granulite-facies metamorphic belts. The gem corundum field corresponds to a domain above 3 kbar, and temperature between 500 and 800°C (Fig. 2-7). The lithologies are alumina-rich and/or silica-poor rocks such as marble,

aluminous gneiss, mafic and ultramafic rocks (M�–UM), or juxtaposed felsic and silica-poor rocks (limestone, marble, M�–UM) affected by the circulation of fluids at their contact and altered by metasomatism (desilicated pegmatite, skarn, etc.); and (2) alkaline basaltic volcanism in continental rifting environments. Marble is a silica- and alumina-depleted rock, and under these conditions the transport of alumina by a fluid phase seems necessary for the crystalliz-ation of corundum; however, Al is usually considered to be an immobile element. In meta-morphic rocks, corundum crystallizes at high temperatures and high to medium pressure such as in garnet�–clinopyroxene assemblages in clino-pyroxenite and metagabbro from the lower crust at a temperature around 1100°C and a pressure about 20 kbar (Rakotosamizanany 2009). Corundum rarely occurs in granite; for example, Wilson (1975) described blades of blue sapphire associated with pink andalusite in the Land's End granite in Cornwall. The author suggested that the corundum and andalusite formed in xenoliths of pelitic rocks assimilated by the granite. Sapphire is also described in Chilean skarns formed in the thermal anomaly accompanying the porphyry copper intrusions in El Salvador and Los Prelambres mines (Singer et al. 2008).

FIG. 2-7. P�–T conditions for the formation of

corundum in metamorphic deposits (modified from Simonet et al. 2008). P�–T fields of North Carolina (Tenthorey et al. 1996), Mangare (Mercier et al. 1999a), Morogoro (Altherr et al. 1982), southern Kenya (Key & Ochieng 1991, Simonet 2000), Hunza (Okrusch et al. 1976), Sri Lanka (De Maesschalk & Oen 1989), Greenland (Garde & Marker 1988), Kashmir (Peretti et al. 1990) with three P�–T boxes corresponding to the evolution of the fluids in the sapphire crystals from the center (c), to intermediate (i) and outer (o) zones, Urals (Kissin 1994), and Mong Hsu (Peretti et al. 1996).

GIULIANI ET AL. �– GEM CORUNDUM

40

During the past two last decades, our knowledge of the formation of corundum deposits has improved significantly (see papers in Groat 2007, Graham et al. 2008a). However, there exists little genetic data on primary gem corundum deposits, i.e., gem corundum that is hosted by a mother rock, certainly insufficient to understand the formation of all deposits and to elaborate precise genetic models. Magmatic deposits have been studied the most. The geology and mechanisms of formation of basalt deposits in Asia and Oceania have been the focus of several studies (Coenraads et al. 1990, 1995, Levinson & Cook 1994, O'Reilly & Zhang 1995, Guo et al. 1996a, 1996b; Sutherland & Coenraads 1996, Sutherland et al. 1998a, 1998b, Limkatrun et al. 2001, Sutherland & Schwarz 2001, Sutthirat et al. 2001, Garnier et al. 2005, Graham et al. 2008b, Sutherland et al. 2009a, 2009b, Sutherland & Abduriyim 2009, and Abduriyim et al. 2012), as well as their ages of formation (Coenraads et al. 1990, Guo et al. 1992, Coenraads et al. 1995, Khin Zaw et al. 2006, Sutherland et al. 2008, Graham et al. 2008b). Sapphire associated with lamprophyre at Yogo Gulch in Montana (USA) was studied by Voynic (1985), Meyer & Mitchell (1988), Brownlow & Komorowski (1988), Gauthier et al. (1995) and Harlan (1996). Sapphire hosted in monchiquite at Loch Roag in Scotland was described by Jackson (1984) and Upton et al. (1983, 1999, 2009). In Kenya, Simonet et al. (2004) studied in detail the Garba Tula sapphire deposit hosted by dykes of syenite. Studies of albitite dikes in serpentinized lherzolite in the Pyrenees, France, has shed new light on the origin of corundum and associated xenoliths in alkali basalts (Monchoux et al. 2006, Pin et al. 2006). Finally, sapphire and ruby inclusions in diamond from Brazil were reported by Watt et al. (1994) and studied by Hutchinson et al. (2001, 2004). Models of formation of the other types of deposits, i.e., metamorphic, metasomatic�–hydro-thermal, have been insufficiently developed. Indeed, the most famous deposits are not necessarily accessible due to government policy or military conflicts (e.g., Azad-Kashmir, Afghanistan, Myanmar, Tajikistan, Colombia), and others (such as Madagascar) have not been systematically investigated (Rakotondrazafy et al. 1996, Rakotosamizanany et al. 2009a, 2009b, Peretti & Hahn 2013). The Mogok Stone Track in Myanmar was mapped by Iyer (1953), but modern geological investigations were lacking until Harlow et al. (2006), Kyaw Thu (2007), Themelis (2008), Yui et

al. (2008), and Nissimboim & Harlow (2011). The Jegdalek ruby deposit in Afghanistan, described by Al-Biruni in the 11th century, was only recently studied by Bowersox & Chamberlain (1995). Garnier (2003) and Garnier et al. (2008) reported detailed geological and geochemical studies of ruby deposits from the Hunza valley (Pakistan), Nangimali (Azad-Kashmir), and Luc Yen-Yen Bai (northern Vietnam), and proposed a new genetic model for marble-hosted ruby. Likewise, the economic interest in blue and padparadscha sapphires found in placer deposits from Sri Lanka has stimulated local geologists to characterize the parental rocks to guide exploration in the Highlands Precambrian series of the island (Rupasinghe & Dissanayake 1985). Further, the majority of geological studies are performed with economic objectives and the reports remain confidential (Bassett 1984, Malik 1994). Because of this most published research is focused on the gemology and mineralogy of corundum, and details of the geological context, mechanisms of formation, and the origin of the mineralizing fluids are scarce (Okrush et al. 1976, Peretti et al. 1990, Mercier et al. 1999a, Giuliani et al. 2003, Garnier et al. 2008, Schwarz et al. 2008). Oxygen isotopic data for corundum are now available (Gauthier et al. 1995, Pomian-Srzednicki 1997, Upton et al. 1999, Yui et al. 2003, 2006, 2008, Giuliani et al. 2005, 2007a, 2007b, 2009, 2012a, Khin Zaw et al. 2006, Sutherland et al. 2009a, 2009b, Uher et al. 2012). The global distribution of corundum deposits is closely linked to collision, rift and subduction geodynamics. Three main periods of corundum formation are recognized worldwide as described below (Giuliani et al. 2007a, Graham et al. 2008b, Stern et al. 2013). The Pan-African orogeny (750 �– 450 Ma) Primary ruby and sapphire deposits found in the gemstone belt of eastern Africa, Madagascar, India and Sri Lanka are linked to the collisional processes between East and West Gondwana (Fig. 2-8), during Pan-African tectonometamorphic events (Kröner 1984). The metamorphic ruby and sapphire deposits from southern Madagascar have numerous geological similarities with those from eastern Africa (Mercier et al. 1999a, 1999b, Giuliani et al. 2007b Rakotondrazafy et al. 2008, Feneyrol et al. 2013), Sri Lanka (Rupasinghe & Dissanayake 1985, Silva & Siriwardena 1988, Dharmaratne et al. 2012), and South India (Santosh & Collins 2003). Based on correlations between gem- and graphite-

GIULIANI ET AL. �– GEM CORUNDUM

41

FIG. 2-8. Juxtaposition of Antarctica, Sri Lanka, India, Madagascar, and Eastern Africa in a tight-fit reconstruction of

Gondwana with the location of the main gem corundum deposits (modified after Windley et al. 1994, Dissayanake & Chandrajith 1999, Mercier et al. 1999b, Collins & Windley 2002; Santosh & Collins 2003, Rakotondrazafy et al. 2008, Feneyrol et al. 2013). Eastern Africa: GT, Garba Tula; M, Mangare; T, Twiga; SN, Si Ndoto; LL, Longido and Lonsogonoi; MM, Mahenge and Morogoro; K, Kalalani and Umba; S, Songea; T, Tunduru. Mozambique: Mz, Montepuez; Madagascar: Ila, Ilakaka; ZS, Zazafotsy and Sahambano; Vo, all the deposits from the Vohibory region; Am, Ambatomena; An, Andranondambo; And, Andilamena; Dy, Didy. South India: My, Mysore; KK, Karur-Kangayam corundum belt; Or: Orissa. Sri Lanka: RE, Ratnapura and Elahera, WC, Wanni complex; HC, Highland complex. Antarctica: NC, Napier complex; RC, Rayner complex; LHC, Lützow-Holm complex; YBC, Yamoto-Belgique complex.

GIULIANI ET AL. �– GEM CORUNDUM

42

bearing rocks Menon & Santosh (1995) and Dissanayake & Chandrajith (1999) proposed the existence of a Neoproterozoic�–early Cambrian gemstone province developed under granulite-facies conditions in East Gondwana. The common presence of ruby and pink sapphire in Kerala and southern Madagascar led Santosh & Collins (2003) to support the view that the Achankovil shear zone in southern India extends into the Ranotsara fault zone (Fig. 2-8) in Madagascar (Collins & Windley 2002). However, few studies have concentrated on the petrology, mineralogy and geochronology of the extensive mineralization associated with the formation of Gondwana. The ages of formation of protolith ruby have been dated by U�–Pb in zircon from the John Saul mine in Kenya, Longido in Tanzania, and Vohibory deposits in Madagascar. The respective ages of 612 ± 6 Ma (Simonet 2000), 610 ±0.6 Ma (Le Goff et al. 2010), and 612 ±5 Ma (Jöns & Schenk 2008) pinpoint a similar period of formation related to the East African orogeny. Moreover, U�–Pb ages of zircon coeval with blue gem sapphire in the Andronandambo skarn deposit range between 523 and 510 Ma (Paquette et al. 1994), and the 40Ar/39Ar ages of biotite syngenetic

with sapphire and ruby from the deposits of Ambatomena, Sahambano and Zazafotsy range between 494 and 487 Ma (Giuliani et al. 2007b). These ages confirm a corundum mineralizing episode during the late Pan-African orogenic cycle (Cambrian), specifically corresponding to the Kuunga orogeny (Meert et al. 1995, Meert 2003). The Cenozoic Himalayan orogeny (45 Ma �– Quaternary) The marble-hosted ruby deposits in central and eastern Asia occur in metamorphic blocks that were affected by major tectonic events during the Cenozoic Indo-Asian collision (Garnier et al. 2006a). The ruby was indirectly dated by 40Ar/39Ar stepwise heating experiments performed on single grains of syngenetic phlogopite (Garnier et al. 2002b, Pêcher et al. 2002, Garnier et al. 2006b), and by ion probe U�–Pb analyses of zircon inclusions in corundum (Garnier et al. 2004b, 2006b). All of the Oligocene to Pliocene ages (40�–5 Ma) are consistent with extensional tectonic events that were active from Afghanistan to Vietnam, in the ruby-bearing metamorphic belt (Fig. 2-9).

FIG. 2-9. Location of marble-hosted ruby deposits from central and southeast Asia. Ar�–Ar dates on

phlogopite syngenetic with ruby are from Garnier et al. (2006b).

GIULIANI ET AL. �– GEM CORUNDUM

43

The Cenozoic alkali basalt extrusions (65 Ma �– Quaternary) Gem corundum occurs worldwide as xenocrysts or megacrysts either in xenoliths or enclaves incorporated in basaltic magmas during their ascent. Such sapphire and ruby deposits occur from Tasmania, Australia, in the south, through eastern Australia, Southeast Asia and eastern China to far eastern Russia (Graham et al. 2008b). They are also found in Nigeria and Cameroon, in the Aïr and Hoggar regions (Wright et al. 1985, Boaka à Koul et al. 2011), the French Massif Central in the Limagne rift (Merle et al. 1998, Gaillou et al. 2010, Giuliani et al. 2009, 2010), and in Northern and Central Madagascar (Rakotosamizanany 2009, Rakoto-samizanany et al. 2009a, 2009b). In general, U�–Pb isotope ages of zircon in sapphire-bearing xenoliths show ages that are similar to K�–Ar and 40Ar/39Ar ages of the basalt (Sutthirat et al. 1994, Sutherland et al. 2002, 2003). The zircon inclusions dated in magmatic sapphire are considered to represent corundum crystallization ages. In the Australian and Asian West Pacific margins intraplate basaltic fields, the U�–Pb ages on zircon are widespread between 60 and 1 Ma (Sutherland et al. 2002, Graham et al. 2008b). Limited U�–Pb data gave ages of 150 and 180 Ma for zircon from Cudgegong (Sutherland et al. 2003), and 400 Ma for those from Tumbarumba (Graham et al. 2004). Generally, the sapphire fields�’ gems produced a simple eruptive event and zircon�–sapphire crystallization age while others had in most two zircon�–sapphire ages and two eruptive events (Graham et al. 2008b). Age correlation of Cenozoic gem corundum occurrences with associated magmatic and tectonic events in central and east Asia is reported in Figure 2-10. It shows that the magmatic�–polygenetic corundum crystals within the basaltic fields largely crystallized in a tectonic regime of distributed E�–W extension (with the opening of the Tasman and Coral Seas) whereas the metamorphic�–metasomatic corundum (ruby in marble) crystallized in a transpressional regime associated with the collision between Indian and Eurasian plates (Graham et al. 2008b). CLASSIFICATION OF CORUNDUM DEPOSITS Several classification schemes have been proposed based on different features such as the morphology of corundum (Ozerov 1945), the geological context of the deposits (Hughes 1990, 1997), the lithology of the host rocks (Schwarz

1998), the genetic processes responsible for corundum formation (Simonet 2000, Simonet et al. 2008), the genetic type of deposits (Kievlenko 2003), the type of deposit and the nature of the corundum host rock (Garnier et al. 2004a; Walton 2004, Giuliani et al. 2007a), and the oxygen isotopic composition of corundum (Giuliani et al. 2005, 2012a). The geological classification schemes group well-known deposits with lithological criteria as a common basis (marble, gneiss, basalt, and others). With this approach, the physical and chemical mechanisms of corundum formation, the source of the fluids, and the origin of the elements are generally not constrained by stable and radiogenic isotopes, fluid inclusions, and geochemistry; and therefore the genesis of the deposits remains uncertain. For example, for marble-hosted ruby deposits, the presence of granite in the proximity of the marble does not imply that the deposit is a skarn type related to the emplacement of the granite, as suggested by Kievlenko (2003). Dating of the mineralization, its host rock, and the intrusion is the first step to take after fieldwork to create a true genetic story. The study of solid and fluid inclusions trapped by corundum during growth sometimes allows assessment of the following conditions of corundum crystallization: temperature, pressure, and composition of the fluids (e.g., Upton et al. 1999, Sutthirat et al. 2001). Muhlmeister et al. (1998) classified ruby deposits into three categories by analyzing the chemical composition of the ruby by X-ray fluorescence: basaltic, marble, and metasomatic deposits (including the deposits of Umba, Tanzania, and Kenya). The concentrations of trace elements (Fe, V, Ga, and Ti) are different among the three categories (Fig. 2-11). Ruby associated with basalt (Thailand and Cambodia) is Fe-rich, and V- and Ga-poor. Ruby in marble (Afghanistan, Myanmar, Nepal, and China) is V-rich, and Ga- and Fe-poor (excepting ruby from Afghanistan and some analyses from China, Nepal and Mogok mines in Myanmar). Metasomatic ruby presents a great variation of trace element concentrations and also high variability for its host rocks. The combination of chemical fingerprinting of corundum with the nature of solid inclusions allows the gemologist to separate great families of ruby, but gemological study should not be disconnected from knowledge of the geological environment and formational processes. Corundum deposits can be classified into primary and secondary deposits. They are considered primary where the corundum is either

GIULIANI ET AL. �– GEM CORUNDUM

44

FIG. 2-10. Age correlation of Cenozoic gem corundum deposits with associated magmatic and tectonic events in central and

east Asia (modified after Graham et al. 2008b). The dates are from: East Tibet�–Indochina igneous rocks (Wang et al. 2001); Himalayan granite bodies (Singh & Jain 2003), tectonic activity (Yin et al. 2002); age of ruby deposits (Garnier et al. 2002b, 2004b, 2005, 2006b, Graham et al. 2006a, 2006b, Yui et al. 2006, Sutherland et al. 2002, Akinin et al. 2004, Graham et al. 2008b).

GIULIANI ET AL. �– GEM CORUNDUM

45

FIG. 2-11. Chemical composition of natural and synthetic rubies (after Muhlmeister et al. 1998). Gallium (Ga)�–vanadium (V)�–

iron (Fe) diagrams (in wt.%) of (A) natural ruby from different deposits, and (B) natural versus synthetic ruby.

hosted in the rock where it crystallized, or in the rock that carried it from the zone of crystallization in the crust or mantle to the Earth�’s surface (as xenolith or enclave). Secondary deposits are sedimentary and of detrital origin. They correspond to the accumulation, in basins of variable extent, of material eroded from primary deposits. This material has been primarily transported by rivers. The primary deposits are subdivided into two sub-groups, according to the magmatic or metamorphic context of the corundum deposit (Fig. 2-12). Magmatic deposits include corundum in alkali basalt, sapphire in lamprophyre, sapphire in syenite, and corundum associated with porphyry copper mineralization and kimberlite (Fig. 2-12a). Metamorphic deposits are divided into metamorphic deposits sensu stricto (marble, mafic and ultramafic rocks, granulite, cordieritite, gneiss, and migmatite), and metamorphic�–metasomatic deposits (Fig. 2-12b) characterized by high fluid�–rock interaction and metasomatism (plumasite [i.e., desilicated pegmatite in mafic and ultramafic rocks] and marble, skarn deposit, shear-zone related deposits in different substrata, mainly corundum-bearing Mg-biotite schist but having experienced granulite-facies metamorphic conditions).

GEOLOGY AND GENESIS OF PRIMARY CORUNDUM DEPOSITS Magmatic deposits Corundum in magmatic deposits is found in plutonic and volcanic rocks. In plutonic rocks, corundum is associated with rocks deficient in silica and their pegmatites, especially syenite and

nepheline syenite. Corundum is formed by direct crystallization from the magmatic melt as an accessory mineral phase. The classical examples are the syenite pluton from Haliburton and Bancroft (Ontario, Canada; Moyd, 1949), and the Kangayan syenite (Coimbatore district, Madras, India, Hughes 1997). In addition, aluminous rocks carried as xenoliths by plutonic rocks can lead to the formation of corundum, as in the xenoliths of biotite schist carried by a monzonite at San Jacinto, in Southern California (Murdoch & Webb 1942). Corundum is found also as an accessory mineral in porphyry Cu deposits (Bottril 1998). In volcanic rocks, sapphire and less commonly ruby are found in continental alkali basalt extrusions (Sutherland et al. 1998a). Corundum is found as xenocrysts in lava flows and plugs of subalkaline olivine basalt, high alumina alkali basalt, and basanite. The sapphire is blue-green-yellow (BGY) and the deposits only have economic importance because advanced weathering in tropical regions concentrates sapphire in eluvial and especially large alluvial placers. Sapphire also occurs in alkaline basic lamprophyre (Brownlow & Komorowski 1988) as mafic dikes of biotite monchiquite (lamprophyre characterized by the abundance of phlogopite and brown amphibole, olivine, clinopyroxene, analcime), ouachitite (a variety of monchiquite) and diamond-bearing kimberlite.

Sapphire in lamprophyric dikes (xenocryst): sapphire was discovered in Montana in alluvial gravel by gold prospectors in 1894 (Voynic 1985, Hughes 1997). The placer deposits occur associated

GIULIANI ET AL. �– GEM CORUNDUM

46

FIG

. 2-1

2. C

lass

ifica

tion

of p

rimar

y co

rund

um d

epos

its b

ased

on

the

litho

logy

of t

he c

orun

dum

-hos

t roc

ks. A

, The

mag

mat

ic d

epos

its w

ith th

eir m

ain

type

s. Fo

r ea

ch ty

pe th

e or

igin

of r

uby

and/

or sa

pphi

re is

pre

cise

d: m

agm

atic

ver

sus m

etam

orph

ic. P

hoto

grap

hs re

spec

tivel

y 1

cour

tesy

of F

. Fon

tan,

2 a

nd 6

of L

.-D. B

ayle

, 3

and

5 G

. Giu

liani

, and

4 o

f S. R

akot

osam

izan

ani.

GIULIANI ET AL. �– GEM CORUNDUM

47

FIG

. 2-1

2. C

lass

ifica

tion

of p

rimar

y co

rund

um d

epos

its b

ased

on

the

litho

logy

of t

he c

orun

dum

-hos

t roc

ks. B

, The

met

amor

phic

dep

osits

with

thei

r mai

n ty

pes.

Phot

ogra

phs r

espe

ctiv

ely

1 to

4, a

nd 6

cou

rtesy

of G

. Giu

liani

, 5 a

nd 6

of L

.-D. B

ayle

.

GIULIANI ET AL. �– GEM CORUNDUM

48

with the Missouri River, Dry Cottonwood Creek, Rock Creek, and Yogo Gulch (Fig. 2-13). The primary deposit of Yogo Gulch was intensely developed from 1897 to 1929. The sapphire is hosted by an Eocene ultramafic lamprophyre and hydrothermal breccia known as the Yogo dike, which is 8 km long and 2 m thick (Brownlow & Komorowski 1988, Meyer & Mitchell 1988). The dike is a member of the Central Montana alkali province (Cade & Groat 2006) and cuts Carboniferous limestone and shale, which underwent contact metamorphism. Underground mining took place at a depth of 100 m from the surface and from 1 to 1.5 km in length. The sapphire grade is highly variable and can reach up to 70 ct per ton, and 5 ct per ton in hydrothermal breccia with limestone fragments (Myachaluk 1995). This sapphire is famous for its cornflower blue coloration, the rarity of inclusions and zoning, and its brilliancy under natural and artificial light (Myachaluk 1995). Among the sapphires produced, 97% are blue, ranging from pale to saturated color; the others are purple to violet. Solid inclusions of pyrite, phlogopite, calcite, rutile, spinel, zircon, and analcime are rare (Gübelin & Koivula 1986). Sapphire crystals are up to 1 ct, but the largest reported raw crystal weighed 19 ct. The crystals (Fig. 2-14) have an uncommon aspect due to the combination of a flattened rhombohedron (k{30.2}, d{01.2}, r{10.1}, b{10.7}) and pinacoid {0001} (Hughes 1990). Nevertheless, the majority of the crystals have not kept their original form because

they were rounded, corroded, and broken by the magma (Clabaugh 1952, Dahy 1988, Gauthier et al. 1995). The sapphire host rock is composed of megacrysts of phlogopite and diopside dispersed in a matrix of phlogopite, clinopyroxene, calcite, analcime, magnetite, and apatite (Brownlow & Komorowski 1988). Accessory minerals include spinel and sapphire crystals with a corona of dark green spinel (Gauthier et al. 1995). Chemical analysis by Meyer & Mitchell (1988) showed that the rock can be classified as a ouachitite, defined by LeMaitre (1989) as an ultramafic lamprophyre containing olivine, phlogopite, amphibole, and/or clinopyroxene phenocrysts within a groundmass of feldspathoid and carbonate minerals, first described from the Ouachita River in Arkansas (Kemp 1890).

FIG. 2-14. Crystal habits of sapphire from Yogo Gulch,

Montana (after Hughes 1997). r, rhombohedron {10.l}; c, pinacoid {00.1}; d, rhombohedron {10.2}; n, second-order hexagonal dipyramid {22.3}.

FIG. 2-13. Map of Montana sapphire localities (after Mychaluk 1995). 1, Yogo Gulch; 2, Missouri River; 3, Cotton Creek; 4,

Dry Cottonwood.

GIULIANI ET AL. �– GEM CORUNDUM

49

Within the lamprophyre Brownlow & Komorowski (1988) described xenoliths of cumulates, composed of megacrysts of olivine, Ti-augite, and hypersthene dispersed in a fine matrix, with a chemical composition between monchiquite and ouachitite. The origin of corundum in lamprophyre bodies is debated on the basis of a magmatic versus metamorphic origin. Clabaugh (1952) suggested an excess of alumina in the magmatic melt by assimilation and fusion of alumina-rich sedimentary or metamorphic rocks that favored the crystallization in situ of the sapphire. Brownlow & Komorowski (1988) proposed the following magmatic genesis: (1) formation of an aluminous magma by the partial fusion of mantle rocks enriched in alumina; (2) magmatic differentiation with crystallization of sapphire and cumulates of olivine, augite, and hypersthene; (3) injection of the magma into the upper crust, with the corrosion of sapphire by the residual liquid magma; (4) fractionation of the residual liquid into a silicate liquid and a gas phase; (5) explosive injection of the dike with brecciation of the limestone and hydrothermal alteration. Berg (2002) observed volcanic glass on two alluvial sapphire crystals from western Montana and suggested dacitic volcanic rocks as a possible source material (in Walton 2004). Garland (2003) studied the trace element chemistry of sapphire and its solid inclusions and concluded that the sapphire might be of metamorphic origin. The xenocrysts were included in the lamprophyric magma and brought up to the surface. The sapphire formed between 580 and 720°C at about 30 km depth. Cade & Groat (2006) showed that reddish-orange garnet inclusions within the Yogo sapphire formed in the mantle in group II eclogite (B), according to the classification of Schultze (2003). They therefore considered the sapphire to occur as xenocrysts in the melt, originating from the mantle. Giuliani et al. (2005) have shown that colored and blue sapphires from Yogo Gulch have 18O-values between 5.4 and 6.8�‰ V�–SMOW (mean 6.0 ±0.4�‰, n = 16). For a mean sapphire 18O of 6�‰, the calculated oxygen isotope composition of water in equilibrium with the corundum as a function of temperature (between 600 and 800°C) indicates an oxygen reservoir characterized by crustal 18O (between 12.6 and 14.5�‰). The genesis of the Montana sapphire is not resolved because the composition of garnet inclusions within the crystals is a strong argument for the hypothesis that sapphire crystals are

xenocrysts probably formed in mantle eclogite (but the 18O values indicate a crustal origin), and subsequently brought up by the lamprophyric magma. At Loch Roag in Scotland sapphire occurs in a Paleozoic monchiquite dike crosscutting Archean pyroxene gneiss (Jackson 1984). The monchiquite contains megacrysts of biotite, sanidine, anorthoclase, apatite, and sapphire, and mantle xenoliths. The corundum crystals are up to 3 cm in diameter and present hexagonal, pyramidal-dipyramidal or bladed habits. The smallest crystals are blue but the biggest are spotty blue and green, some with yellow patches (Kievlenko 2003). Mineral inclusions include hematite and rutile. The Loch Roag monchiquite is similar in geochemical composition to the Yogo Gulch dike. Jackson (1984) noted a similarity in petrographic composition, and the presence of mantle xenoliths, in lamprophyre dikes and sapphire-bearing lavas in Australia and Thailand. He suggested a genetic affinity between the two types of intrusive and extrusive magmatism. Upton et al. (1983) reported some 70 localities in northern Britain and Ireland where deep-seated xenoliths (spinel lherzolite and pyroxenite) and megacrysts have been brought up by basaltic magma. In Scotland, megacrysts are abundant anorthoclase, biotite, clinopyroxene, magnetite, zircon, and apatite. These minerals occur as megacrysts but also as polycrystalline syenite xenoliths. Composite xenoliths provide evidence that the anorthoclasites may occur as pegmatitic veins crosscutting pyroxenitic wall rocks in the shallow mantle (Upton et al. 1999). The geochem-istry and oxygen isotope composition of sapphires (4.6 < 18O < 5.25�‰, n = 4) and feldspar indicate crystallization of these syenitic facies from felsic melts originating from partial melting of meta-somatized mantle rocks (Upton et al. 1999, 2009).

Sapphire in syenite (megacryst): sapphire is a typical mineral of pegmatite in syenitic rocks. The pegmatite contains megacrysts of K-feldspar, titanite, Nb�–Ta minerals, and sapphire (Kievlenko 2003). Corundum is mainly opaque but in some cases it is transparent and translucent with a dominant blue color. Its gemstone production is small, although gems were extracted from syenite pegmatites in Russia (Ilmeny-Urals, Khibiny-Kola Peninsula), Canada (Haliburton, Bancroft, Dungannon, all in Ontario), India (Coimbatore, Madras), and Norway (Seiland, Finmark). Sapphire

GIULIANI ET AL. �– GEM CORUNDUM

50

is also found in a unique example of sapphire bearing-monzonite, the Dusi deposit, known as Garba Tula syenite, in Kenya. The 18O values of sapphire in syenite define a range between 4.4 and 8.3�‰ (n= 25; Giuliani et al. 2010). The Ilmeny-Urals deposit was discovered in the 19th century in the southern Urals. Sapphire-bearing pegmatite occurs as veins some tens of metres long, 1.5 to 2 m thick, and up to 200 to 300 m along strike, in a large biotite-nepheline syenite. It consists dominantly of plagioclase (albite�–oligoclase), with orthoclase, muscovite, biotite, and hornblende (Kievlenko 2003). Corundum is either disseminated or in pockets. It is intergrown with plagioclase and garnet and commonly rimmed by biotite. The crystals have barrel or bladed shapes and are typically 3 to 4 cm long. The corundum is not transparent and is bluish to brownish-gray; less commonly it is pink, yellow or green. The corundum is associated with zircon, ferrocolumbite, aeschinite (Nb�–Y oxide), fluorite, and molybdenite. The same type of deposit is described in the Khibini syenitic massif, Kola Peninsula (Kievlenko 2003). Transparent varieties of sapphire are very rare and are cornflower blue to deep blue. In Ontario, Canada, corundum and limited gem sapphire are found in nepheline syenite pegmatite within an alkaline belt (Field 1951, Kievlenko 2003). Nepheline syenite from Hastings County contains orthoclase and albite feldspar. Nepheline is very abundant and is associated with cancrinite, sodalite, scapolite, calcite, biotite, and hornblende. These silica-undersaturated rocks are formed by low degrees of partial melting in the Earth�’s mantle. Bancroft sapphire crystals are greyish-blue to violet and are associated with microcline, titanite, and biotite. However, the 18O values of respectively 7.6, 7.8, 7.6, and 9.3�‰ for the sapphires of Craigmont, Hastings-New Carlow, Craig Hill, and Dungannon indicate a crustal contamination of the magma oxygen reservoir. The Dusi (Garba Tula) deposit in Central Kenya was discovered by nomads in the 1960s (Fig. 2-15). The deposit is a subvertical dike several metres thick and with an inferred horizontal extension of several kilometres. It is in contact with biotite-hornblende gneiss, but the lack of outcrop does not permit study of the field relations with the host rocks nor appraisal of the continuity of the dike (Simonet et al. 2004). The dike is leucocratic monzonite, and is made of K-feldspar (microcline) and albite (An3�–7) in equal proportions, biotite, zircon and magnetite, and is classified in the QAPF

diagram of LeMaitre (2002) as an alkali-feldspar syenite. Zircon is included in feldspar or in corundum. U�–Pb geochronology on zircon crystals allowed dating of the crystallization of the dike at 579 ±6 Ma (Simonet 2000) which corresponds to one of the tectonic events of the Mozambique metamorphic belt. Sr�–Nd isotopes indicate that the monzonite is of mantle origin, with little or no participation of a crustal component (Simonet et al. 2004). The corundum is barrel-shaped with a size varying from a few millimetres to more than 20 cm. The gem material varies in color from dark blue to golden yellow (Plate 1C), through various shades of blue, green and yellow (Feneyrol et al. 2013). The crystals contain significant amounts of iron (FeO between 0.95 and 1.13 wt.%) and are Ti-poor (TiO2 < 0.04 wt.%). Simonet et al. (2004) compared the Dusi sapphires with those found in the Turkana gem field in northern Kenya, which are associated with Tertiary basalt. Indeed, they have the same range of color (identical to the BGY sapphire suite defined by Sutherland et al. 1998a), the same habit, the same chemical composition, and similar inclusions. Simonet et al. (2004) suggested that the similarities between sapphire crystals and the association with zircon is a strong indication of a similar crystallization in a peraluminous Zr-rich syenitic melt. Yellow and green sapphires from Garba Tula have 18O values of 5.2�‰ (Giuliani et al. 2005) which clearly indicate a mantle derivation ( 18O mantle of 5.5 ± 0.4�‰, Mattey et al. 1994).

FIG. 2-15. Major gem corundum deposits of Kenya

(modified from Simonet 2000).

GIULIANI ET AL. �– GEM CORUNDUM

51

At Urdach and Espéchère in the western Pyrenees, France, dikes of pegmatitic albitite have intruded serpentinized lherzolite bodies. The albitite contains albite, muscovite, and biotite, with chlorite, epidote, zircon, titanite, thorite, pyrochlore, aeschinite, ferrocolumbite, rutile, ilmenite, and magnetite (Monchoux et al. 2006). The corundum-bearing albitite also contains Nb-bearing rutile, samarskite, manganocolumbite, fersmite, and fergusonite. In corundum-free albitite, the paragenesis is composed of allanite, chevkinite, apatite, and monazite. The texture of the rock is pegmatitic and corundum is barrel-shaped, dark blue, pale blue or gray. Geochemical and isotopic data of the albitite rocks document a mantle origin for these rocks (Pin et al. 2006). They have very high contents of Sr (up to 1089 ppm), Ba (up to 860 ppm), Nb (167�–299 ppm), Ta (6�–17 ppm), and an extreme enrichment of the light rare-earth elements (Ce ~ 1020 ppm). Nd values range from 1.9 to 3.4 and the 87Sr/86Sr initial ratios are between 0.7039 and 0.7068, reflecting a mantle source without any major contribution from continental crust. The 18O values of sapphire between 5.1 and 5.4�‰ for the Espéchère deposit confirm the mantle source. Those from Urdach deposits between 6.1 and 6.6�‰, show a 18O increase indicative of a hybridization of the mantle material (Giuliani et al. 2010). Corundum in porphyry Cu deposits (accessory mineral): corundum is an accessory mineral found in several porphyry Cu deposits in Tasmania (Botrill 1998), British Columbia (Simandl et al. 1997), El Salvador and at Los Pelambres deposits in Chile (Singer et al. 2008). At the Empress Cu�–Au�–Mo deposit in southwestern British Columbia corundum is reported in association with an andalusite-pyrophyllite rock free of quartz. Corundum growth has been explained by a rapid heating of the cool meteoric water influx in fractures over the thermal anomaly of the porphyry Cu. The sapphire grains are on a millimetre scale, and can be transparent, colorless, or blue. Although corundum is theoretic-ally incompatible with quartz, a metastable associ-ation quartz�–corundum�–andalusite�–pyrophyllite has been described in a zone of intense hydrothermal pervasive argillic alteration associated with the Bond Range Tasmanian porphyry Cu deposit (Botrill 1998). Corundum in alkali basalt (xenocryst and xenolith): blue-green-yellow (BGY) sapphire makes up most of the gem trade and is concentrated in

huge placers. They include the deposits of the corundum belt of eastern Australia (Fig. 2-16) that extends from Queensland in the north to Tasmania in the south (Oakes et al. 1996, Sutherland 1996, Sutherland & Coenraads 1996, Graham et al. 2008b, Abduriyim et al. 2012); in Asia (Fig. 2-1), the deposits of Mingxi, Fujian Province (Keller & Keller 1986), of Changle, Shandong Province (Guo et al. 1992), and Penglai, Hainan Island in China (Furui 1988); the provinces of Binh Thuan, Lam Dong, Dong Nai, and Dak Lak in southern Vietnam (Smith et al. 1995, Poirot 1997, Garnier et al. 2005); Pailin, Cambodia (Lacombe 1970, Jobbins & Berrangé 1981); the provinces of Chanthaburi-Trat, Kanchanaburi (Vichit et al. 1978, Sutthirat et al. 2001), and Denchai in the province of Phrae in Thailand (Limkatrun et al. 2001); Laos (Sutherland et al. 2002); in central and far eastern Russia with the deposits of Nezametnoye, Podgelbanochny, and Shkotovo Plateau (Graham et al. 2008b); in Africa,

FIG. 2-16. Gem corundum in the basaltic terranes of

Eastern Australia with the main sapphire mines of Australia. Most production is centered on the Lava Plains, Anakie, New England and Oberon fields (modified after Sutherland & Schwarz 2001)

GIULIANI ET AL. �– GEM CORUNDUM

52

the sapphire deposits of Cyangugu in southwest Rwanda (Krzemnicki et al. 1996), of the Kivu region in the Democratic Republic of Congo (Frazier & Frazier 1990), the region of Turkana in the north of Kenya (Barot et al. 1989, Keller 1992, Simonet 2000), the province of Kaduna in Nigeria (Irving & Price 1981, Kiefert & Schmetzer 1987), the regions of Mamfe and Adamawa in Cameroon (Lettermann & Schubnel 1970, Boaka à Koul et al. 2011), the volcanic massif of Atakor in the Algerian Sahara (Conquéré & Girod 1968), and the Aïr massif in Niger (Carbonel & Robin 1972); deposits in the region of Ambondromifehy, Antsiranana Province, and Antsirabe in Madagascar (Lacroix 1922, Schwarz et al. 2000, Rakotosamizanany 2009, Rakotosamizanany et al. 2009a, 2009b); in Europe, the Espaly deposit, near Puy-en-Velay, in France (Carbonel et al. 1973, Forestier 1993, Giuliani et al. 2009, 2010, Gaillou 2003, Gaillou et al. 2010), in the Eifel massif in Germany (Hochleitner 1998), in the alluvial deposits of Jizerská Louka and Trebivlice-Ceské in the Czech Republic (Hughes 1997), the deposit of Wilcza Por ba in Poland (Maliková 1999), in the alluvial deposits of Hajná ka and Gortva in southern Slovakia (Uher et al. 1999, 2012), and in Loch Roag in Scotland (Upton et al. 1983, 1999, 2009), and lastly in South America, the Mercaderes�–Rio Mayo deposit in Colombia (Keller et al. 1985, Sutherland et al. 2008). These deposits have a number of features in common. To begin with, they are generally associated with intra-plate alkali basalts containing ultramafic xenoliths, mainly mantle lherzolite (Coenraads et al. 1990). The corundum is commonly associated with pyrrhotite, clino-pyroxene, zircon, Fe-rich spinel, and in some cases sapphirine (Muhlmeister et al. 1998, Sutthirat et al. 2001, Giuliani et al. 2009, 2010). The xenocrysts of corundum contain numerous mineral inclusions (e.g., Nb-bearing rutile, ilmenite, ferrocolumbite, fersmite, samarskite, zircon, feldspars, apatite and, in some, sulfides (Simonet 2000, Graham et al. 2008a). These mineral inclusions are enriched in Zr, Nb, Hf, Ta, and lithophile elements such as U in betafite and uranopyrochlore, and in Th such as thorite. The corundum commonly exhibits corrosion features on the surface of the crystals due to transport by the magma (Coenraads et al. 1995, Sutherland & Coenraads 1996). In the placers, the crystals are broken but they keep their usual six-sided barrel or "keg of beer" shape (Guo et al.

1996a). Within a single gem province, variations of color observed for corundum found ~10 km apart suggest the existence of multiple sources for the gemstones (Coenraads et al. 1990, Sutherland et al. 1998b). Indeed, Sutherland et al. (1998b) and Sutherland & Schwarz (2001), studying the trace element contents distribution in corundum from the states of New South Wales and Victoria, found evidence for two contrasting geochemical fields, one for magmatic sapphire and one for metamorphic crystals. The same geochemical behavior was confirmed for corundum of the Kanchanaburi-Bo Rai and Nam Yuen deposits in Thailand, and Pailin in Cambodia. The chemical variation diagram Fe/Ti versus Cr/Ga proposed by Sutherland et al. (1998b) is commonly used for the separation of magmatic and metamorphic blue sapphires from several deposits worldwide (Fig. 2-17). On the one hand, the corundum crystals called "metamorphic" have pastel (blue, pink, orange) to ruby color, and are rich in Cr and poor in Ga (Cr2O3/Ga2O3 ratio > 3); solid inclusions are chromiferous spinel, pleonaste, Al-rich diopside and sapphirine. On the other hand, the corundum crystals called "magmatic" are xenocrysts of BGY sapphires with a Cr2O3/Ga2O3 ratio < 1; the solid inclusions are Nb-bearing rutile,

FIG. 2-17. Fe2O3/TiO2 versus Cr2O3/Ga2O3 diagram of

blue sapphire from non-basalt-related and basalt-related origins (after Abduriyim & Kitawaki 2006). The metamorphic blue sapphires originate from the placers of Ilakaka in Madagascar, Ratnapura in Sri Lanka, and Mogok in Myanmar. The magmatic blue sapphires are from the placers of Bo Phloi and Kanchanaburi in Thailand, and Changle in China. The chemical field of the BGY sapphires defined by Sutherland & Schwarz (2001) is also reported.

GIULIANI ET AL. �– GEM CORUNDUM

53

ilmenite, ferrocolumbite, pyrochlore, fersmite, samarskite, hercynite, magnetite, zircon and Fe oxides. Over the last decade use of laser ablation�–inductively coupled plasma�–mass spectrometry (LA�–ICP�–MS) analysis applied to trace elements in gemstones (Günter & Kane 1999, Guilliong & Günther 2001) has brought new insights on geographic typing of gem corundum in basaltic environments (Abduriyim & Kitawaki 2006, Peucat et al. 2007, Sutherland & Abduriyim 2009). Generally, Fe and Ga reveal higher concentrations in magmatic sapphire ( 1800 to 11000 ppm Fe and

70 to 570 ppm Ga) than in metamorphic corundum, which have Fe and Ga contents less than 3000 and 75 ppm, respectively (Sutherland et al. 2002, 2009a, Khin Zaw et al. 2006, Peucat et al. 2007). Low Cr and Mg contents are typical for magmatic sapphire (both usually < 40 ppm), whereas gem corundum of metamorphic origin is enriched in these elements (both generally > 60 ppm). Therefore, the Ga/Mg ratio is usually > 6 in magmatic sapphire and < 3 in that of metamorphic origin. Conversely, the Cr/Ga ratio is < 0.1 for magmatic and > 1 for metamorphic sapphire. The Fe/Ti ratio is generally higher in magmatic than metamorphic sapphires. The Fe/Mg ratio is also significantly different in magmatic and metamorphic sapphires (Peucat et al. 2007), very high for magmatic (Fe/Mg >> 100) and low for metamorphic and metasomatic sapphires (Fe/Mg < 100). The use of the different discrimin-ating binary and ternary diagrams and chemical ratios (Fe/Ti, Ga/Mg, Fe/Mg) is somewhat debatable and the origin of sapphire crystals from the Hajná ka placer and Gortva syenite in southern Slovakia is a good example (Uher et al. 2012). The different ratios of Fe/Ti, Ga/Mg, and Fe/Mg for blue sapphires indicate bimodal origins whereas the mineral inclusions were magmatic. They signify that the final composition of the magmatic sapphires depends of the behavior of the parent magma during crystallization which can sometimes generate discrepancies with the use of the chemical diagrams for corundum classification. The deposits from Nigeria and Cameroon are linked to a mantle plume located beneath the rift of the volcanic province of Guinea that extends northward to the basaltic plateau of Aïr and Hoggar (Wright et al. 1985, Déruelle et al. 2007). The deposits in Rwanda, Democratic Republic of Congo, and of Baringo in Kenya (Simonet, 2000) are associated with the East African rift. The deposits in the French Massif Central are located on

the southern extension of the Limagne rift (Merle et al. 1998), also associated with a mantle plume. In Australia, it is generally accepted that the central volcanoes became younger towards the south (Sutherland 1996). This variation, developed during the last 35 Ma, suggests that the migration towards the north of the Indo-Australian plate occurred above a mantle plume. The huge basaltic provinces (70�–14 Ma) resulted from the rise from the mantle of a region of high geothermal gradient, related to the extension provoked by the opening of the Tasman and Coral Seas (Graham et al. 2008b) and/or the separation of Gondwana at the end of the Mesozoic (O'Reilly & Zhang 1995). The geological context of formation of the basalts in Southeast Asia is debatable. Levinson & Cook (1994) thought that the alkali basalts in Thailand, Cambodia, and Vietnam were associated with the subduction of the Indian plate under the Eurasian plate. Barr & McDonald (1979) proposed crustal pinching with a mantle rise that would explain the volcanic eruptions in Thailand. Barr & McDonald (1981), Whitford-Stark (1987), and Hoang & Flower (1998) considered that the Neogene to Quaternary intraplate volcanism is associated with the formation of "pull-apart" basins and rifts. This magmatic activity post-dated the collision between India and Eurasian plates (Fig. 2-9) and is linked to the extrusion of the Indochina Peninsula. Several mechanisms have been proposed for the formation of ruby and sapphire found within alkali basalt. Although the models differ, researchers now agree that the gems formed at great depths, under metamorphic or magmatic conditions; subsequently the parental rocks were intruded by the alkali basalt magma, and mineralized xenoliths or sapphire xenocrysts were transported upwards to the surface (Coenraads et al. 1990; Levinson & Cook 1994, Sutherland & Coeenrads 1996, Guo et al. 1996a, Upton et al. 1999). Experimental studies have shown that corundum and mineral inclusions such as zircon, ferrotantalite, pyrochlore-group minerals, ferro-columbite, and samarskite cannot crystallize from a basaltic magma (Levinson & Cook 1994; Guo et al. 1996a). The sapphires must have grown in an environment rich in Th, Zr, Nb, Ta, V, alkali elements such as Na and K, and Fe, Al, and volatile components. The etching on the surface of corundum suggests that crystals were partially absorbed into the entraining magma. Nevertheless, Sutherland et al. (1998a) thought that such etching is not sufficient to conclude that crystals are

GIULIANI ET AL. �– GEM CORUNDUM

54

xenocrysts; indeed, such etched surfaces are sometimes observed on olivine, pyroxene, and plagioclase phenocrysts in the alkaline olivine basalt, as a result of modification of pressure and temperature. Xenocrysts in alkali basalt may have several origins. Levinson & Cook (1994) proposed a two-stage model of genesis. In the first stage, corundum formed in the crust, at depths of 20 to 50 km, by metamorphism of aluminous rocks (such as shale or mudstone) or by metamorphism of gibbsite. In the second stage the corundum was transported to the surface. While ascending to the surface, the magma incorporated xenoliths of corundum-bearing rocks. These xenoliths were dissolved and the refractory minerals (corundum and zircon) were brought to the surface. This model leads to the problem of the existence of an aluminous protolith of bauxitic or lateritic composition, commonly found at the surface, but not on subduction-prone ocean floors. A subduction zone may have been active before the crustal extension in Australia, but this is not the case either in eastern Africa or in Southeast Asia (Saminpanya 2000), and the model seems dubious here. The origin of ruby has long been debated but its metamorphic origin is now accepted. Study of the trace element contents and mineral inclusions (sapphirine, spinel) in Barrington rubies from Australia (Sutherland et al. 1998a) indicated a metamorphic origin and similarity to some Cambodian�–Thai rubies from Southeast Asian basaltic gem placers. Sutthirat et al. (2001) reported temperatures of crystallization for ruby between 800 and 1150 ±100°C and pressures between 10 to 25 kbar, corresponding to depths of 35 to 90 km, for ruby-bearing clinopyroxene xenocrysts with garnet and sapphirine inclusions from eastern Thailand. They proposed a garnet clinopyroxenite and garnet granulite source in the lower crust or upper mantle (ultramafic rocks), as previously suggested by Sutherland et al. (1998a). The nature of the solid inclusions trapped by ruby is important for P�–T thermobarometry determinations. Sutherland et al. (2003) found Al-rich diopside, meionite and anatase in a xenolith from the Cudgegong-Macquarie River in New South Wales. Clinopyroxene�–corundum thermometry suggested T > 1000�–1300°C, but the presence of meionite (silvialite variety) indicated a probable high-T skarn type source. Rakotosamizanany (2009) found ruby in pyroxenite and metagabbro enclaves in alkali basalt from the primary Soamiakatra deposit in the Ankaratra

massif of Madagascar. Three stages were character-ized by: (1) the association of clinopyroxene 1 + pyrope + scapolite + ruby 1, and spinel + ruby 1 + pyrope, indicated a P�–T formation of 1100°C and 20 kbar, at the limit of the eclogite domain; (2) the destabilization of scapolite in anorthite + calcite related to the transformation of diopside in pargasite (at 1100°C and P < 20 kbar), and by spinel + ruby 1 + pyrope in sapphirine (at 1100°C and P ~ 15 kbar); and (3) the association phlogopite + ruby 2 related to the destabilization of spinel + ruby 1 + pyrope during decompression. Metagabbro went from eclogite to granulite facies. Among magmatic models, Irving (1986) argued that corundum crystallized from undersaturated felsic magma at high pressure, and Aspen et al. (1990) concurred that corundum originated from the crystallization of syenite generated in the crust or the upper mantle. Coenraads et al. (1990) proposed that corundum formed in magmas, enriched in volatile and incompatible elements, derived from the fusion of mantle but with a low rate. This may explain why corundum is not in equilibrium with magmas generated from partial fusion, and why the surfaces of the crystals are commonly etched. Coenraads et al. (1995) proposed that sapphire from Australia and Thailand crystallized from phonolitic magma at pressures equivalent to those found at the mantle�–crust boundary. The thermal events that generated such magma bodies are directly linked to the processes that have generated and caused eruption of the corundum-bearing alkali basalt magma. Guo et al. (1996a) studied mineral inclusions in corundum from Australia, China, Thailand, United States, and Kenya placer deposits. They suggested that corundum forms by the mixing and/or interaction between carbonatite and a magmatic system of alkaline or hyperalkaline magmas (Fig. 2-18), in the liquid or solid state (alkaline granite or syenitic pegmatite). During an extensional event linked to the rise of the asthenosphere, carbonatite bodies crystallized in the crust between 10 to 20 km depth; these carbonatite bodies are locally crosscut by syenite and granitoid plutons. The mechanism of hybridization between these two systems can provoke the rapid crystallization of corundum, and the formation of lenses of rock carrying corundum with its characteristic mineral association. Later, a rapid basaltic volcanic event occurred with included fragments of rock and lenses with corundum. This genetic model is complex because it requires the coincidence of successive magmatic

GIULIANI ET AL. �– GEM CORUNDUM

55

FIG. 2-18. Genetic model for the origin of

corundum xenocrysts in alkali basalts based on a crustal growth model for eastern Australia (after Guo et al. 1996a). 1- Crystallization of carbonatite in crustal extensional zone, and locally crosscut by granitoid rocks. 2- Hybridization process between these two systems leads to the formation of corundum lenses. 3- Later in the history, alkali basalts intrude the corundum lenses and corundum is carried up to the surface.

events. It is not a satisfactory model that can be applied to all the deposits found elsewhere on the Earth. Sutherland et al. (1998a) rejected the model because, in the African rift, corundum is generally found in alkali basalt located at the periphery of the central uplift and not in association with ultramafic rocks and carbonatite, as suggested by the model. Sutherland (1996) proposed that corundum formed during the crystallization of felsic alkali magma produced by the fusion of metasomatized lithosphere. Sutherland & Coenraads (1996), studying the assemblage ruby-sapphirine-spinel found in the placers from Barrington in Australia, proposed that the fusion of ultramafic rocks at high temperatures and pressures by basanite can produce limited volumes of anatectic magma able to crystallize these assemblages with aluminous and refractory minerals. The fusion of the accessory phases rich in Cr, such as chromiferous spinel or chromite, would contribute to the local concentration of Cr, allowing the crystallization of ruby and sapphire in millimetre-size crystals. The discovery of ruby-bearing garnet pyroxenite and metagabbro enclaves in the Soamiakatra alkali basalt from Madagascar (Rakotosamizanany 2009) renders the hypothesis of Sutherland & Coenraads

(1996) unlikely. Rubies originate from mafic and ultramafic rocks in the subcontinental lithosphere (Kornprobst et al. 1990). The range of 18O values defined for rubies found in placers from basaltic environments is between 1.3 and 5.9�‰ (mean 18O = 3.1 ±1.1�‰, n = 80, Giuliani et al. 2007a). The oxygen isotopic range fits within the range of rubies associated with mafic and ultramafic rocks (0.25 <

18O < 6.8�‰, n = 21). The 18O values of ruby lower than 5�‰ indicate a previous high-temperature seawater interaction of the parental rocks (Yui et al. 2006). This hypothesis was already proposed by Pearson et al. (1993) for the low 18O values of clinopyroxene from the ruby-bearing pyroxenite from the Beni Bousera peridotite complex in Morocco; moreover, the 18O value of this ruby is 4.4�‰. Sutherland et al. (1998a) proposed a genetic model in four stages for the formation of 'magmatic' sapphire from eastern Australia (Fig. 2-19). The lithosphere is displaced above a mantle plume. A low rate of initial fusion generated felsic magma enriched in volatile elements in zones where the lithosphere is rich in amphibole, allowing the crystallization of corundum and zircon. This magma can also be derived from a mantle enriched in

GIULIANI ET AL. �– GEM CORUNDUM

56

FIG. 2-19. The successive stages for corundum formation

and eruptive alkali basalt transport for eastern Australia gem fields (after Sutherland et al. 1998a).

amphibole and mica, or from a mantle enriched in felsic components, at between 45 and 90 km depth. When the lithosphere is located above the plume, a high rate of partial fusion produces alkali basaltic magma that extracts and carries assemblages with corundum as xenocrysts and in xenoliths. When the lithosphere moves away from the plume, the rates of fusion decrease and lead again to the crystallization of corundum and zircon. This model explains the enrichment of Hf, Nb, and Ta, generally observed in minerals cogenetic with corundum, and in amphibole veins found in the peridotite xenoliths (Sutherland et al. 1998b). The presence of a mantle plume under the lithosphere is a main geodynamic feature for the

genesis of such "magmatic corundum", but the mechanism of generation of the Al-rich magma is not yet constrained. Pin et al. (2006) have shown that the albitite dikes of Espéchère and Urdach in the western Pyrenees (France), which intruded serpentinized lherzolite, are of mantle origin. The mineral accessory phases of these albitite dikes are roughly similar to those found in xenoliths associated with alkali basalt fields. Such xenoliths are generally composed of alkali feldspar ± zircon ± corundum and megacrysts. Pin et al. (2006) observed that the Si�–Al�–Na-dominated bulk composition is similar to that of certain glass inclusions included in peridotitic xenoliths in alkali basalt. In addition, the extreme enrichment of incompatible elements in the albitite implies premelting metasomatism by a fluid or a melt. These rocks are interpreted to be products of very low degree of partial melting of a harzburgite source previously enriched by carbonatite-related metasomatism. The presence of volatile phases such as H2O and CO2 may account for the variation of the solubility of SiO2 and Al2O3 in mantle fluids and the consequent precipitation of corundum in some batches of felsic magma. The geochemical composition of the albitite bodies plots in the field of syenite in the diagram of Cox et al. (1979) as does that from the sapphire-bearing syenite of Garba Tula (Simonet et al. 2004). A syenitic origin for sapphire is confirmed by the case studies of different examples of sapphire-bearing anorthoclasite from Scotland (Upton et al. 1999), France (Giuliani et al. 2010), Slovakia (Uher et al. 2012) and Madagascar (Giuliani et al. 2009). The sapphires plot in the geochemical field of syenite (Fig. 2-20A) and their solid inclusions are typical of magmatic BGY sapphire (Upton et al. 1999, Gaillou et al. 2010, Uher et al. 2012). The

18O range of sapphires between 4.4 and 6.5�‰ (Fig. 2-20B) fits in the isotopic range of sapphire in syenite (4.4 < 18O < 8.3�‰, n = 25; Giuliani et al. 2010). Nevertheless, the origin of the parent magma has not yet found a consensus: the mantle source model without crustal contamination (Pin et al. 2006) is not in agreement with the models proposed by Guo et al. (1996a) and Sutherland et al. (1998a). We should add that the chemical composition of magmatic glass inclusions trapped in sapphire crystals from Menet, Coupet, and Sioulot in France (Gaillou et al. 2010), are trachytic with high Al2O3 (15 to 22 wt.%), K2O (5 to 6 wt.%) and Na2O (4 to 6 wt.%). Uher et al. (2012) proposed that the anorthoclasite from Gortva in southern Slovakia was

GIULIANI ET AL. �– GEM CORUNDUM

57

FIG. 2-20. The sapphire crystals associated with anorthoclasite xenoliths in alkali basalt (modified from Uher et al. 2012): the

cases of Hajacka (Slovakia), Loch Roag (Scotland), Menet (France) and Kianjanakanga (Madagascar). A, FeO �– Cr2O3 �– MgO �– V2O3 versus FeO + TiO2 + Ga2O3 diagram (in wt.%) used for the geological classification of corundum deposits (Giuliani et al. 2010). The main fields defined for the different types of gem corundum deposit are: for ruby, marble (R1); John Saul Ruby Mine (Kenya) type (R2); mafic and ultramafic rocks (R3); metasomatites (R4); for sapphire, syenitic rocks (S1); metasomatites (S2); xenocrysts in alkali basalt and lamprophyre (S3). The domains of R4 and S2 which correspond to metasomatic�–metamorphic corundum are overlapping. The sapphires from the anorthoclasites of Loch Roag, Kianjanakanga and Menet plot in the field of syenite (S1). The sapphire of Hajnacka plots in the metasomatite-related field (S2). B, Oxygen isotope values ( 18O; V�–SMOW, Vienna Standard Mean Ocean Water) of the same sapphires, with reference to the worldwide 18O database (Giuliani et al. this work). The 18O values are between 4.4 to 5.2�‰ which fall in the range of sapphire crystals in syenite. For comparison the 18O values of the sapphire in syenite from Garba Tula (Kenya), Urdach and Espéchère (France) are reported.

GIULIANI ET AL. �– GEM CORUNDUM

58

derived from an igneous reservoir in the sub-continental spinel lherzolitic mantle ( 18O of sapphire = 5.1 �‰). In the lithospheric mantle, felsic melts crystallized to form anorthoclasite, the most evolved peraluminous variant of the alkaline basaltic melt. The proposed model is consistent with an origin of analogous felsic syenitic xenoliths from southern Slovakia where petrological study indicated a pressure around 600 MPa (6 kbar, ~200 km depth) and liquidus temperature of inclusion melts at ~1080°C (Huraiová et al. 1996). Corundum in kimberlite (accessory mineral, xenocryst and xenolith): corundum has been described as crystal inclusions in diamond, xenocryst or megacryst in xenoliths from kimberlite bodies: (1) ruby and Cr-bearing sapphire crystals have been reported as syngenetic inclusions in type II diamond from the alluvial deposits in Brazil (Watt et al. 1994, Hutchinson et al. 2001, Hutchinson et al. 2004), and in a cut diamond of unknown origin (Meyer & Gübelin, 1981). The size of the crystals in Juina diamonds is less than 200 µm and the color is red or red-brown, blue vitreous or light blue, and white. The Cr2O3 contents of the ruby are between 0.3 and 9.2 wt.%. The association of a ruby crystal with aluminous pyroxene and ferropericlase placed

the origin of ruby-bearing diamonds within the lower mantle at ~770 km (Hutchinson et al. 2004); (2) crystals of ruby and sapphire are found as up to 1.5 cm xenocrysts in the Mbuyi-Mayi kimberlite in the Democratic Republic of Congo (Pivin et al. 2013). The color of ruby varies from red to fuchsia and flashy-pink, and its Cr2O3 content reaches up to 5.6 wt.%. The 18O of 5.55 ± 0.05�‰ (n = 3) indicates for ruby clearly a mantle derivation with no substantial variations ( 18O mantle of 5.5 ±0.4�‰; Mattey et al. 1994). (3) Cr-bearing sapphires of less than 100 µm size were recorded in kimberlite xenoliths in kyanite eclogite from the Roberts Victor mine (Hatton 1978), corganite xenoliths (Mazzone & Haggerty 1989) and corgaspinite xenoliths (Exley et al. 1983; Mazzone & Haggerty 1989) from Jaggersfontein and Bellsbank in South Africa. Padovany & Tracey (1981) recorded a 150 µm long Cr-rich ruby (Cr2O3 content up to 3.2 wt.%) in a pyrope-spinel xenolith of the kimberlite from Moses Rocks Dike, Utah. Metamorphic deposits Corundum is a high temperature mineral that forms naturally by metamorphism of alumina-rich rocks under amphibolite- and granulite-facies conditions (Fig. 2-21). Most corundum is formed as

FIG. 2-21. The different types of metamorphic corundum deposits in their geological and geodynamical environments such as

plumasite, skarn, desilicated gneiss and granite, cordieritite, biotite schist in shear zone, mafic�–ultramafic rocks, and marble. The geodynamic model shows also the location of the ruby and sapphire deposits associated with alkali basalt in continental extension domains. The vertical dimension is not to scale. The subcontinental lithosphere is not represented.

GIULIANI ET AL. �– GEM CORUNDUM

59

emery during regional metamorphism of bauxite lenses in limestone, as fine aggregates of corundum with mica, hematite, magnetite, and sericite e.g., Naxos and Samos islands in Greece, and Smyrna in Turkey, or at the contact of aluminous schist and intrusive mafic and ultramafic rocks e.g., Peekskill in Winchester County, New York State, USA. Gem quality corundum is absent in emery deposits. Metamorphic corundum gem deposits are found in complexes formed by mafic and ultramafic rocks, marble, quartzite, gneiss and metapelite where heated either regionally or by the thermal anomaly generated by the intrusion of different plutonic rocks and crosscut by their dikes and intrusive bodies such as granite and pegmatite. The various formation mechanisms depend on either isochemical metamorphism i.e., regional metamorphism or metasomatic metamorphism i.e., contact and/or hydrothermal affecting the contrasting lithologies. Nevertheless, these mechanisms are not well understood for each type of deposit and they remain hypothetical. Moreover, in some cases, several episodes of metamorphism have affected the complexes, as in the Neoproterozoic metamorphic Mozambique Belt that runs from Somalia through Kenya, Tanzania, Malawi, Mozambique, Madagascar, South of India, Sri Lanka, and to East Antarctica (750�–540 Ma). Each episode had different thermobarometric and chemical conditions, and the formation of corundum would have to be defined accurately on a pressure�–temperature�–time scale. Accordingly, the proposed classification for metamorphic corundum relies on the lithology, and not on the chemistry of the corundum or the type of metamorphism. Metamorphic deposits associated with desilicated pegmatite and skarn-related granite or mafic–ultramafic rocks: these corundum deposits formed as the result of hydrothermal metasomatism occurring either at the contact of intrusive aluminosilicate plutonic rocks or mafic�–ultramafic rocks and host rocks depleted in silica (Fig. 2-22) such as marble, amphibolite, and ultramafic rocks, or at the contact of a shear zone. The origin of desilicated pegmatite erroneously called "anorthosite" (Kievlenko 2003) has been debated since the 19th century. Pratt (1906) advanced a magmatic origin. Du Toit (1918), Hall (1923), and Barth (1928) proposed desilication of a granite�–pegmatitic melt at the contact with mafic rocks, leading to enrichment in alumina of the melt and the formation of corundum. Robb & Robb

FIG. 2-22. Types of metasomatic deposits located at the

contact between two chemically different rocks (after Simonet et al. 2008). The metasomatic fluid circulated along the contact between two rocks of contrasting lithology such as ultramafic rock or marble and pegmatite or gneiss. The percolation inside the rocks induced metasomatic desilication of the silica-rich rocks and the formation of metasomatic rocks such as plagioclasite (desilicated pegmatite, i.e., the plumasite), and/or phyllosilicate-rich rocks (phlogopite schist, vermiculite schist, chloritite).

(1986) thought that corundum was the result of the interaction of invading pegmatitic fluids and vapor with the host rock. Today, the hydrothermal�–metasomatic theory proposed by Korzhinsky (1964, 1970), and revised by Kolesnik (1976) for metasomatic corundum, is commonly accepted by geologists (Simonet et al. 2008). Corundum formed during hydrothermal and/or bi-metasomatic processes developed at the contact of two contrasting lithologies, i.e., granite or pegmatite,

GIULIANI ET AL. �– GEM CORUNDUM

60

against mafic and ultramafic rocks, marble or gneiss (Fig. 2-22). The bi-metasomatic reactions are related to the infiltration of post-magmatic solutions, which originated from the same granite or from other magmatic or metamorphic events. The mechanism of metasomatic desilication involves diffusion of silica, from a pegmatite to an ultramafic rock which plays the role of SiO2 sink, at a rate more rapid than the diffusion of alumina. The Al:Si ratio increases in the pegmatite vein, and metasomatic plagioclase�–corundum association in which the mass of alumina per unit volume is much greater than in the initial rock, may develop if the volume of rock decreases at the same time, and underlined by the zone of quartz dissolution. Another case of bi-metasomatic metasomatism by fluid percolation is the intrusion of mafic to ultramafic rocks into marble, such as those described for the ruby from the Mahenge region in Tanzania (Le Goff et al. 2008, Feneyrol et al. 2013). Desilicated pegmatite bodies with the mineral assemblage oligoclase (75%), corundum (23%), and 2% other minerals (margarite, mica, sericite) are called plumasite, a terminology derived from the locality of Plumas in California, where this association was first described (Lawson 1903). Corundum deposits of this type are known from the northeastern Transvaal (Robb & Robb 1986), Yosemite National Park (Rose 1957), Kinyiki Hill and Mangare in Kenya (Simonet 2000), and the upper Allier Valley in France (Forestier & Lasnier 1969). In the Allier Valley plumasite is hosted by spinel-bearing harzburgite. There, corundum is associated (Fig. 2-23) with plagioclase (oligoclase�–andesine), spinel, biotite, apatite, zircon, uraninite,

pyrochlore and lower temperature minerals such as andalusite, talc, diaspore, chlorite, prehnite, and scapolite (Lasnier 1977). The sapphire crystals are barrel-shaped and elongate along the c axis. They are grey to blue, in some cases with a pronounced asterism. World-class gem deposits in desilicated pegmatite bodies are described from the Umba River (Solesbury 1967) and Kalalani (Seifert & Hyrsl 1999) in Tanzania, at the Sumjam deposit in Indian Kashmir (La Touche 1890, Atkinson & Kothavala 1983), in southern Kenya with the deposits of the Mangare area (Key & Ochieng 1991, Mercier et al. 1999a, Simonet 2000), and from the Russian Polar Ural Mountains (Spiridonov 1998, Kievlenko 2003). The famous blue sapphire from the States of Jammu and Kashmir in northern India was discovered in 1881 by peasants after a landslide. This corundum deposit produced the blue sapphires most prized worldwide for their gorgeous blue color and velvety lustre due to the presence of layers of microscopic liquid inclusions. Mallett (1882) and La Touche (1890) described the deposit formed by the small pit called the 'Old Mine'. The sapphire was embedded in clay within an altered pegmatite. The geological structure and history of the deposit is poorly investigated because the area is politically unstable and at high altitude (4600 m above sea level). The deposit is located in metamorphic formations of Cambrian to Silurian age in the Zanskar range, within the Paddar region (Lydekker 1883). The series (Fig. 2-24) comprises a succession of marble, amphibolite, and garnet-graphite-biotite

FIG. 2-23. Mineralogical assemblages in corundum-bearing plumasite (after Lasnier 1977). The sapphire-bearing plumasite of

Treignac (Haut Allier, France) with the different stages of transformation of spinel. Prism of spinel surrounded by a corona of blue sapphire (center) and a final transformation into corundum (left). On the right, presence of corundum altered by a later retrograde phase composed of andalusite, diaspore and chlorite. The plagioclase is transformed in scapolite and prehnite. Drawing from P. Brun (le Règne Minéral).

GIULIANI ET AL. �– GEM CORUNDUM

61

FIG. 2-24. Geological cross-section of the Sumjam sapphire deposit, Kashmir (after Atkinson & Kothavala 1983).

and amphibole-bearing gneiss which has been intruded by pegmatite (Atkinson & Kothavala 1983). The sapphire is associated with pockets or lenses of metamorphic rocks formed of olivine, talc, and spinel. The lenses range from 1 to 100 m long and up to 30 m thick. They are enveloped by a green aureole of tremolite, actinolite, and anthophyllite-bearing rocks. Laterally, the lenses are surrounded by garnet-bearing amphibolite, then amphibolite. Pegmatite bodies are located at the contact between the lenses and the amphibolite. They have been affected by an alkaline metasomatism that provoked dissolution of quartz. The mineralogical association consists of plagioclase, mica, tourmaline, and sapphire. The rims of the desilicated pegmatite are composed of talc-biotite-carbonate and tourmaline-bearing rocks (Peretti et al. 1990). The greatest concentration of sapphire is in plumasite. Near the surface, the veins are altered into kaolinite, which was derived at depth from fresh plagioclase rich-rocks. The sapphire is embedded in the plumasite and sporadically in the tremolite-actinolite zone. The sapphire crystals have a spindle shape, with hexagonal dipyramids and are partially corroded. Some crystals present small pinacoid faces that are flattened following one of the a axes (Cesbron et al. 2002). The common color is pale blue to deep cornflower blue, but other colored sapphires have been observed, such as violet, green, and yellow. Pink sapphire was recovered from the contact of the graphite-biotite gneiss (Atkinson & Kothavala 1983). Hänni (1990) identified rutile, green and red-brown tourmaline, pargasite, plagioclase, uraninite, allanite, diopside, and zircon as inclusions in the sapphire. The velvety luster of the blue sapphire is caused by minute two-phase (liquid and vapor) fluid inclusions. Peretti et al. (1990) studied the primary

carbonic fluid inclusions trapped by the sapphire by microthermometry. The pressure�–temperature of the fluid is estimated to have been 3.7 to 5.6 kbar and 680�–700°C at the center of the crystal, 3.9 to 4.8 kbar and 590�–600°C in the intermediate zone, and 2.9 to 3.1 kbar and around 500°C in the outer zone (Fig. 2-7). This study showed the variability of the P�–T conditions of crystallization during the growth of the crystals, suggesting either a slow growth or three distinct stages of growth. Alluvial ruby and sapphires were discovered associated with the Umba River in Tanzania in 1960, and field exploration exposed the primary deposits in the Umba and Kalalani serpentinite massifs (Solesbury 1967; Seifert & Hyrsl 1999). Mining ceased between about 1970 and 1980, but in 1989 an important discovery of 80 kg of transparent reddish orange (called 'Tanzanian padparadscha') and yellow sapphires stimulated new mining activity. In 1994, a primary tsavorite (V-rich green grossular) deposit was discovered in the surrounding graphitic gneiss, and pyrope-almandine deposits in the serpentinite. The serpentinite massif is about 2 km long and 0.5 to 1 km wide, and surrounded by marble, kyanite-sillimanite gneiss, and quartzite. Gem-quality sapphire and ruby are confined to desilicated pegmatite bodies that crosscut the serpentinite (Fig. 2-25A, B). They form vertical lenses, 5 to 10 m long and 2 to 3 m thick. The pegmatite bodies pinch and swell and the dimensions are quite variable. They have coarse-grained texture and display meta-somatic reaction zones at the contact with the serpentinite. Two types of corundum-bearing veins have been described (Fig. 2-25C): (1) veins with calcic plagioclase, vermiculite-actinolite-chlorite, and colored sapphire; and (2) veins with vermiculite and orange and yellow-brown sapphire. In Kalalani,

GIULIANI ET AL. �– GEM CORUNDUM

62

FIG. 2-25. The location and geology of the Kalalani gem corundum in Tanzania (modified after Seifert & Hyrsl 1999).

A, Location of the Kalalani and Umba primary corundum deposits. B, The gem corundum deposits are associated with desilicated pegmatite bodies (plumasite) crosscutting serpentinite. C, Geological sketch map and cross-section of a gem corundum deposit. The sapphire is contained in argillitic nodules in the desilicated pegmatite.

other major constituents are chlorite, spinel, and kyanite. Tabular ruby was found within a desilicated pegmatite associated with biotite. The inclusions and color range of reddish to orange sapphire from Umba and Kalalani are very similar to those of the orange sapphire from Tunduru (Tanzania) and Chimwadzulu in Malawi (Henn & Milisenda 1997). The sapphire crystals are tabular with pinacoids and prismatic hexagonal faces. Color zoning in the form of a tube parallel to the c axis, colorless in blue

sapphire, and black in the ruby, is a common feature. Seifert & Hyrsl (1999) demonstrated that desilicated pegmatite results from intense alkaline metasomatism of the pegmatite and the serpentinite at high fluid�–rock ratio. Ruby was discovered and mining initiated in the Mangare area of Kenya in 1973, but the mines were closed during most of the 1970s and 1980s due to litigation. Exploitation began anew in 1993�–1994 and since then has provided a regular supply of

GIULIANI ET AL. �– GEM CORUNDUM

63

facetable and cabochon material of high quality. In 2011 the production decreased and in 2012 the Rockland mine (formerly the John Saul mine) was in discontinuance of business, as a consequence of the discovery and marketing of ruby from Mozambique. In Mangare area the ruby contains low iron and consequently exhibits an intense fluorescence. The John Saul, Penny Lane and Hard Rock deposits occur in serpentinite intercalated in graphite-sillimanite gneiss (Mercier et al. 1999a, Simonet 2000). In the Hard Rock mine, the contact zone between the gneiss and the serpentinite is a tectonic shear zone. The serpentinite is considered to be an imbricate structure formed during the main tectonic Pan-African event (700�–600 Ma) of the Mozambique belt. Corundum is found in two types of structures (Fig. 2-26): (1) micaceous lenses and pockets located at the contact between the serpentinite and the gneiss along the shear zone. They are usually composed of ruby ± spinel ± sapphirine and mica, but in some cases of ruby, phlogopite and magnesian chlorite; (2) veins that crosscut only the serpentinite. Corundum is rare and is associated with zoisite, plagioclase ± phlogopite ± muscovite. These veins resemble plumasite. Corundum formed by the following reaction (Mercier et al. 1999a):

sapphirine + water corundum + chlorite + spinel (1)

In the veins that crosscut the ultramafic rocks, the corundum is associated with anorthite and zoisite. The textures and the mineral associations suggested the following reaction:

anorthite + water corundum + zoisite + silica (2)

Thus, two hydrothermal alteration phenomena, hydration and desilication, are superposed on the shear zone system. The conditions of pressure and temperature for ruby formation are between 700�–750°C and 8 to 10.5 kbar (granulite facies). These P�–T conditions contrast with those of the surrounding gneiss (P = 650°C and 4.5 < P < 6.5 kbar, amphibolite facies; Fig. 2-27. In consequence, Mercier et al. (1999a) suggested that the ruby and the ultramafic rocks are exotic. They correspond to fragments of deeper crust brought up by thrust and shear zone tectonics. Comparison of the thermobarometric and mineralogical data of twenty deposits formed in ultramafic rocks in Kenya and Tanzania show that they formed in the granulite facies and that they are associated with shear zones.

FIG. 2-26. Geological map of the Hard Rock mine in the

Mangare gem corundum mining district, Kenya (after Mercier et al. 1999a). The serpentinite corresponds to an imbricate structure in the graphite-sillimanite-bearing gneiss. Ruby is found either in lenses and pockets at the contact of the shear zone or in veins that crosscut only the serpentinite.

A special case of desilication of pegmatite in mafic and felsic gneiss is described for the ruby deposits from Karelia in Russia (Kievlenko 2003, Terekhov 2007, Bindeman & Serebryakov 2011). The Belomorian polymetamorphic complex located between the Kola and Karelia Archean blocks underwent the 1.9�–1.8 Ga metamorphism and regional intrusion of pegmatite. Ruby is hosted by plagioclase veins crosscutting corundum-bearing kyanite-staurolite-garnet-biotite-hornblende gneiss and the overlying garnet amphibolite. The veins

GIULIANI ET AL. �– GEM CORUNDUM

64

FIG. 2-27. Pressure�–temperature

conditions of the formation of gem corundum and surrounding metamorphic rocks in the Mangare area, Kenya (after Mercier et al. 1999a). An: anorthite, Anth: anthophyllite, Bt: biotite, Cd: corundum, En: enstatite, Gt: garnet, Kf: K-feldspar, Ky: kyanite, Mg-Chl: magnesian chlorite, Sil: sillimanite, Sp: spinel, Spr: serpentinite, Qz: quartz, V: H2O vapor, Zo: zoisite.

range from several centimetres to 10 m thick and up to 60 m long, parallel to the metamorphic foliation. Corundum crystals have prismatic and bladed shapes, 5 cm long and up to 3 cm in diameter; ruby contains up to 9.4 wt.% Cr2O3. Metasomatism and fluid�–rock interaction with the gneiss caused desilication of the pegmatite and transformed it into plumasite with oligoclase-andesine, phlogopite, tourmaline, zoisite, epidote and chlorite. Phlogopitization of the border of the vein is common. Low oxygen isotope ratio of minerals including ruby is characteristic of this deposit with the report of the lowest known 18O (�–16 to �–27�‰) for terrestrial silicate minerals, oxides and rocks (Bindeman et al. 2010). The highly depleted 18O values are due to the circulation of meteoric glacial subwaters in a rift zone at around 2.4 Ga which affected the initial isotopic composition of the protolith prior to ruby formation (Bindeman & Serebryakov 2011).

Now we will consider skarn-type deposits where pegmatite and granite intruded marble, calc-silicate rocks and Ca-bearing gneiss. In the following examples, the mechanisms responsible for corundum formation are quite similar to those proposed for the genesis of corundum within desilicated pegmatite bodies that intruded mafic and ultramafic rocks. Every post-magmatic process linked to the emplacement of granitic intrusions (or equivalents) has a stage at which the solutions became acid and, therefore, constitute a source of strong reactions with their wall rocks. As the activity of K, Na, and Al increases, these elements enter into reactions with marble or dolomitic marble, increasing the concentration of Ca and Mg in solution. Such reactions produce feldspar, biotite, and Ca-bearing minerals. The association of Ca-bearing, usually Fe-rich silicate minerals (including amphibole, pyroxene, garnet, epidote, and zoisite), or Mg-rich silicate minerals (such as phlogopite,

GIULIANI ET AL. �– GEM CORUNDUM

65

diopside, pargasite, and forsterite), defines the so-called skarns. The surrounding marble is transformed by exo-metamorphism i.e., exoskarn, but sometimes complex phenomena of recrystalliz-ation may affect the granite itself by endo-metamorphism i.e., endoskarn. Corundum-bearing desilicated pegmatite in mafic rocks and marble are typical endoskarns. Nevertheless, some authors believe, erroneously, that all the corundum deposits hosted in marble belongs to the endoskarn type (Kievlenko 2003). The metamorphic formations that contain deposits of this type, such as in Myanmar (Iyer 1953, Harlow et al. 2006, Kyaw Thu 2007, Themelis 2008) or Tajikistan (Terekhov et al. 1999), have been subjected to tectonomagmatic activity, but the effective participation of the magma is not constrained by isotopic studies. In consequence, the marble deposits cannot be classified as endoskarn types and further studies are necessary to characterize them completely. One of the more enigmatic regions is the Mogok Stone Tract in Myanmar. The Mogok metamorphic belt was intruded by syenitic to granitic magma in Cretaceous and Eocene times accompanied by subsequent metamorphism. Kyan Thu (2007) defined four ages for the emplacement of magmatic rocks based on radiometric dates of zircon by the U�–Pb method: (1) 129 ±8.2 Ma (Early Cretaceous) for augite-biotite granite; (2) 32 ±1 Ma (Early Oligocene) for leucogranite; (3) 25 Ma (late Oligocene) for foliated syenite; and (4) 16 ±5 Ma (Middle Miocene) for painite-bearing skarn formed at the contact leucogranite�–marble. Khin Zaw et al. (2008) obtained an age of 31�–32 Ma for a zircon inclusion in the Mogok ruby by the LA�–ICP�–MS technique. Garnier et al. (2004b) obtained 40Ar/39Ar ages at 18.7 ±0.2 to 17.1 ±0.2 Ma on phlogopite syngenetic with metamorphic ruby. Themelis (2008) showed that the diversity of the paragenesis associated with primary ruby and sapphire deposits reflected multiple geological origins either with intrusions of granitoid rocks or regional metamorphism. The assemblage of ruby, calcite, graphite, muscovite and pyrite is common, but adjacent silicate minerals typical of skarn and syenitic rocks, such as cancrinite, scapolite, sodalite, nepheline, sodalite, datolite, and painite (CaZrBAl9O18), are also found e.g., such as at Wet Loo (Harlow et al. 2006, Harlow & Bender 2013). The radiometric data suggest that the Oligocene-age ruby grew either as disseminated crystals in marble (Khin Zaw et al. 2008) or as overgrowths on painite

and tourmaline in a skarn-formed marble (Nissimboim & Harlow 2011). Trace element compositions measured by LA�–ICP�–MS or by EPMA on ruby from different types of environment, i.e., in marble, skarn, and alluvial deposits, show no B or Zr enrichment despite the skarn-related paragenesis of some samples (Harlow & Bender 2013). The metamorphic versus skarn-related source of ruby cannot be discriminated and the Mogok Stone Tract rubies record rather metamorphic�–metasomatic processing. The Bakamuna exoskarn skarn deposit at Elahera in Sri Lanka (Fig. 2-28) was described by Silva & Siriwardena (1988). The skarn body is 35 m long, 12 m thick and occurs in a calcite-dolomite

FIG. 2-28. Simplified geological map of Sri Lanka

(modified after Kriegsman 1995) with the corundum localities in their geological formations (after Dissayanake & Chandrajith 1999, Dissanayake et al. 2000). Main mining districts and mines of Sri Lanka: 1, Polonnaruwa; 2, Nalanda; 3, Elahera; 4, Kurunegala; 5, Rangala; 6, Kandy; 7, Hanguranketa; 8, Nilgala; 9, Avissawella; 10, Hatton; 11, Nuwara Eliya; 12, Passara; 13, Panadura-Horana; 14, Ratnapura; 15, Haputale; 16, Buttala; 17, Alutgama; 18, Rakwana; 19, Timbolketiya; 20, Kataragama; 21, Ambalangoda; 22, Morawaka; 23, Ambalantota; 24, Galle; 25, Matara.

GIULIANI ET AL. �– GEM CORUNDUM

66

marble. The skarn mineralization consists of three different mineralogically distinct zones (Fig. 2-29A): (1) the center zone is a coarse-grained orthoclase-quartz-bearing pegmatite; (2) the middle zone contains fine-grained spinel-scapolite aggregates with disseminations of corundum and phlogopite; and (3) the outer zone has a porphyritic texture made of large crystals of corundum and spinel. The substitution of corundum by spinel is common. The intermediate and outer zones are crosscut by a network of veinlets forming lenses up to several centimetres wide filled with spinel, magnesite, and phlogopite. The vein formation is associated with hydraulic fracturing, i.e., a fluid overpressure in the rock. The contact with the marble (Fig. 2-29B) shows a fringe of phlogopite and spinel, 10 to 20 cm thick, with 70% of spinel by volume, and a border where marble recrystallized into coarse-grained calcite. Gem corundum is represented by light greenish blue and blue sapphire, with crystals up to 1 cm. According to Silva & Siriwardena (1988), the Bakamuna skarn deposit formed by the interaction of pegmatitic fluids with dolomitic marble. The pegmatite is not desilicated and quartz is always present. Thus, the sapphire deposit is an exo-skarn deposit. Percolation of the fluid caused the metasomatic alterations. Three stages of fluid�–rock interaction have been proposed:

Fluid 1 + marble scapolite + corundum + magnesite + CO2 (3)

magnesite + corundum spinel + CO2 (4)

Fluid 1 + magnesite + spinel + scapolite phlogopite + Fluid 2 (5)

Fluid 2, enriched in alumina, also reacted with the marble following reaction (3), and resulted in the crystallization of large quantities of corundum. The low Mg content of the marble allowed the preservation of corundum whereas it was transformed into spinel following reaction (4). In 1952, the French geologist Hibon reported small eluvial sapphire grains associated with hibonite south of the village of Andranondambo in the Tranomaro area in Madagascar. The first deposit was discovered at Esiva, near Tranomaro village in 1991 (Fig. 2-30). News of the discovery spread quickly, and by 1996 around 10,000 miners were estimated to be working in pits up to 15 m deep and 50 to 80 cm wide, at Andranondambo, Antirimena, Analalava, and Andranomitrohy (an area of more than 7000 km2).

FIG. 2-29. The Bakamuna exoskarn skarn deposit at

Elahera in Sri Lanka (after Silva & Sirawardena, 1988). The skarn is developed around the pegmatite that intruded the marble. A, Schematic distribution of the different mineral assemblages in the skarn zones: (1) the center zone is a coarse-grained orthoclase-quartz-bearing pegmatite; (2) the middle zone contains fine-grained spinel�–scapolite aggregates with corundum and phlogopite, and (3) the outer zone has a porphyritic texture made of large crystals of corundum and spinel. B, Details of the outer zone with the concentrations of corundum and spinel at the contact of marble.

A number of studies have been published on the origin of the sapphire. Kiefert et al. (1996) com-pared the Andranondambo deposits to the famous Kashmir sapphire desilicated pegmatite deposit, Milisenda & Henn (1996) advocated a high-grade granulite-facies metamorphism for sapphire formation, and Rakotondrazafy et al. (1996) and

GIULIANI ET AL. �– GEM CORUNDUM

67

FIG. 2-30. Geological map of the Tranomaro area with the

location of the corundum-hibonite skarn deposits in Madagascar (after Rakotondrazafy et al. 2008).

Gübelin & Peretti (1997) suggested that the sapphire formed as a result of various metasomatic skarn stages. The sapphire deposits occur in the high-grade granulite facies of the Proterozoic Tranomaro Group, composed of metasedimentary rocks (metapelite, calc-silicate rock, and marble) inter-layered with leucocratic gneiss (Rakotondrazafy et al. 1996). During granulite-facies metamorphism and syn-metamorphic emplacement of the Anosyan charnockite and granite, marble and calc-silicate gneiss were transformed into skarns called the 'calc-magnesian complex' (Moine et al. 1985). This consists of impure calcitic marble, diopsidite with variable amounts of scapolite, spinel, thorianite, and pargasite, and peraluminous rocks made of plagioclase and/or scapolite with spinel, thorianite, hibonite (CaAl12O19), and/or blue to pink corundum. Clinopyroxenite commonly occurs at the contact between marble and granitic or charnockitic

intrusions from the Anosyan magmatism (Fig. 2-30) and a metasomatic skarn origin is proposed by Rakotondrazafy et al. (1996). Three stages of crystallization in the skarns have been defined (Fig. 2-31). In the first stage (metasomatism at T 850°C and P 5 kbar), Ca-rich hyperaluminous pods of meionite, spinel, thorianite, and corundum crystallized in a titanite-bearing matrix consisting of scapolite and aluminous diopside. Corundum is found also in the diopsidite at the contact with the Ca-rich pods. U�–Pb dating on zircon from a diopsidite gave an age of 565 10 Ma (Andriamaro-fahatra & de La Boisse 1986), which is in agreement with Pan-African ages (540�–580 Ma) obtained for the granulite-facies metamorphism and the syn-metamorphic emplacement of the Anosyan charnockite and granite (Paquette et al. 1994). In the second stage (metasomatism T 800°C and P 3�–3.5 kbar), diopside was partially transformed into F-rich pargasite and most calcian scapolite went into anorthite + calcite. Thorianite crystallized with F-rich phlogopite and hibonite at the expense of corundum and spinel. In the third stage (retrograde amphibolite metamorphism at T 500°C and P 2 kbar), lenses of phlogopite associated with calcite, diopside and anhydrite, and late-stage REE-rich calcite veins with zircon, titanite, and urano-thorianite crosscut the 'calc-magnesian complex'. U�–Pb dating of zircon from a metasomatic calcite vein gave ages of 516

10 Ma (Andriamarofahatra & de La Boisse 1986) and 523 5 Ma (Paquette et al. 1994), which correspond to the latest Pan-African event in the area. At this late stage, blue gem sapphire from the Andronandambo area crystallized in K-feldspar veins crosscutting marble and diopsidite at T 500°C and P 2 kb (Rakotondrazafy et al. 2008). The veins are vertical with centimetre to decimetre width. Sapphire is associated with K-feldspar, F-apatite, calcite, and phlogopite. The common crystalline forms of sapphire are the dipyramids n {22.3}and z {22.1} and the basal pinacoid c {00.1}; the second-order prism a {11.0} and the rhombohedron r {10.1} are also present (Schwarz et al. 1996). Color zoning is very pronounced with dark blue or dark brownish blue areas delineated by faces parallel to the basal pinacoid, the dipyramids n and z. At the border of the veins, marble is feldspathized and the stability of the K-feldspar-corundum-calcite assemblage is controlled by the equilibrium equations:

GIULIANI ET AL. �– GEM CORUNDUM

68

FIG. 2-31. The U�–Th-corundum skarn

deposits from the granulite facies of the Tranomaro area (modified after Rakotondrazafy et al. 1996). A, Al�–(Fe+Mg)�–Ca showing the distribution of the different assemblages of the calc-magnesian and skarn rocks from stages I and II. B, The three stages of skarn formation with stage I meta-somatism (leucocratic segregation in clinopyroxenite; T ~ 850°C, P = 5kb), stage II (formation of hibonite and anorthite + calcite assemblages; T ~ 800°C, P = 3.3.5 kb), and stage III (retrograde amphibolite facies and fracturing stage with REE-bearing calcite veins and gem blue sapphire�–K-feldspar bearing veins; T = 500°C, P = 2 kb).

muscovite K-feldspar + corundum + water (6)

anorthite + CO2 calcite + corundum + H2O (7)

Fluid inclusions in different minerals from gneiss, skarn, and sapphire-bearing vein are CO2-rich (Ramambazafy et al. 1998). Fluids with high PCO2 (XCO2 > 0.8 mol.%) and low PH2O are in equilibrium with the mineral assemblages. The C and O isotopic composition of Tranomaro marble has shown the crustal origin of these fluids and their probable relation with granitic magmatism (Boulvais et al. 1998). The infiltrated marble and metasomatic diopsidite do not record any contribution from a crustal C source and the contribution of fluid infiltration is insignificant. The formation of the skarn in marble, however, results from the transport of Si, Mg, Th, U, Zr and REE by the metasomatic fluid. CO2-rich fluids are not a good candidate for the mobilization and transport of such elements which are rather immobile (Rakotondrazafy et al. 2008). At high PCO2, the

solubility of silica is low and a H2O�–NaCl-rich fluid may possibly transport the metals (Ramambazafy et al. 1998). Unmixing would produce a CO2-rich phase in equilibrium with a H2O�–NaCl brine (Gibert et al. 1998). With such a high CO2 concentration in the fluid, the CO2-rich phase (XCO2 > 0.8 mole%; P = 5 kb and T = 600°C) would have to coexist with NaCl salt (Shmulovich & Graham 1999). Sapphire from the Andranondambo district yielded three sets of 18O values (Giuliani et al. 2007b): (1) 3.9�‰ for a pink sapphire hosted in a pyroxenite (diopsidite) from stage 1 of skarn metasomatism; (2) 10.1 < 18O < 10.7�‰ for blue and pink sapphire hosted by scapolitite and marble from stage 1. The important isotopic variation for sapphire of stage 1 is due to the variability in chemistry of the pre-metamorphic host rock. During stage 2, hibonite crystallized at the expense of corundum and its 18O of 10.6�‰ falls within the isotopic range defined for corundum of stage 2. It confirms that in a closed system and at high temperature (T ~ 800°C), the oxygen isotopic composition of a protolith or a mineral phase can

GIULIANI ET AL. �– GEM CORUNDUM

69

control the final 18O of corundum; and (3) 14.0 < 18O < 15.6 (mean 18O = 14.8 0.6�‰; n = 4) for

blue gem sapphire hosted in K-feldspar veins from stage 3. The isotopic range overlaps the range defined worldwide for sapphire hosted in desilicated pegmatite within marble (14.0 < 18O < 16.4 �‰; Giuliani et al. 2009), and the oxygen isotopic composition of sapphire is buffered by the host rock. The important isotopic variation for sapphire between stages 1 and 3 might be related to the large isotopic variation found in 18O of marble between 6.5 and 19�‰ (Boulvais et al. 1998). The lowest

18O-values are found in dolomitic marble whereas calcic variety has higher 18O between 13.5 and 19�‰. Metasomatic alteration by fluid percolation related to the emplacement of granitic magma, including desilication, can also affect rocks such as gneiss or quartzofeldspathic and calc-silicate rocks (Lecheminant et al. 2004, 2005, Simonet et al. 2008, Zwaan & Zoysa 2008, Liu Shang-I & Zoysa 2011, Dharmaratne et al. 2012). The primary sapphire deposit discovered in 2012 at Thammannawa, near Kataragama in southeastern Sri Lanka, is an example of the effect of the circulation of pegmatitic fluid in gneiss (Dharmaratne et al. 2012). The garnetiferous quartz-feldspar gneiss and charnockitic biotite gneiss were intruded by various granitoid and pegmatite bodies. The pegmatite consists of quartz-free coarse-grained feldspar and mica surrounded by a micaceous layer which hosts blue sapphire. The metasomatic process affected both pegmatite and gneiss. The mineralization extended generally in the foliation direction and measured about 60 cm thick. The crystals of sapphire display a combination of hexagonal bipyramids with rhombohedral and basal pinacoids, flat faces, sharp edges, and a vitreous lustre. They have a good transparency and pure color with cut gems up to 20 ct. The inclusions are rutile, uraninite, graphite, zircon, and spinel (Pardieu et al. 2012). Such metasomatic deposits are also described in other areas of Sri Lanka (Hapuarachchi 1989, Fernando et al. 2001). The Kimmirut sapphire occurrence (Fig. 2-1) was discovered by Seemeega and Nowdla Aqpik in 2002 (Lecheminant et al. 2004). The showing is located approximately 1.5 km southwest of Kimmirut (formerly Lake Harbour), near the south coast of Baffin Island, Nunavut, Canada (62°49.7´N, 69°53´W). The area also hosts other gem minerals in complexly deformed, high-grade metamorphic rocks. These include diopside,

pargasite, garnet, spinel, scapolite, tourmaline, apatite, zircon, moonstone, and lapis lazuli. The continental collision setting of southern Baffin is analogous to gem-producing areas within the India�–Asia collision zone, i.e., from Afghanistan to Vietnam. Most of the sapphire crystals in Kimmirut are deep blue in color but yellow, colorless, and light blue crystals have been found. The sapphire is hosted by calc-silicate lenses in a marble unit of the metasedimentary Lake Harbour Group, near a major terrane boundary within the Paleoproterozoic Trans-Hudson Orogen (Lecheminant et al. 2004, 2005). At the main Beluga showing the sapphire crystals occur with plagioclase, clinopyroxene, phlogopite, muscovite, calcite, graphite, nepheline, and scapolite. Apatite, rutile, titanite, and zircon are common in the host rock, and rare phases include chlorite, tourmaline (dravite), monazite, sanbornite (BaSi2O5), thorianite, and uraninite. The rocks in the Bowhead showing, 170 m south-southwest of the Beluga showing, appear to be less altered. Petrographic studies suggest that this diverse mineral suite formed during retrograde metamorphism accompanied by infiltration of CO2-bearing fluids. Syenitic or ijolitic magmas may have played a role in initial formation of the calc-silicate lenses. The ruby and pink sapphire at the Greyson and Kitwalo mines in the Mahenge mining district in Tanzania are found in marble crosscut by mafic dykes (Le Goff et al. 2008 in Feneyrol et al. 2013). The crystallization of corundum is assisted by chemical exchanges through the percolation of fluids along the contact between the dyke and the marble. Ruby and pink sapphire are located in metasomatic pargasite-phlogopite zones formed along the contact zones. Ruby is associated with phlogopite, spinel, pargasite, sapphirine, rutile, Mg-chlorite and pyrite. The formation of sapphirine originates from the destabilization of corundum-spinel-Mg-chlorite following the reaction:

6 corundum + 2 spinel + Mg-chlorite 2 sapphirine + 4H2O (8) Metamorphic deposits associated with biotitite and cordieritite in gneiss: the deposits found in feldspathic gneiss and cordieritite represent new types of metamorphic gem corundum deposits in Madagascar: sapphire-bearing biotite schist in feldspathic gneiss from Sahambano, Zazafotsy, and Ionaivo deposits (Giuliani et al. 2007a, Uher et al.

GIULIANI ET AL. �– GEM CORUNDUM

70

2012), and ruby and polychrome sapphire in cordieritite intercalated within charnockite at Ambatomena and Iankaroka deposits (Giuliani et al. 2007b, Andriamamonjy et al. 2013). Both types of deposits are within shear zones and corundum is linked to fluid circulation through channels such as fractures, foliation planes, and lithological contacts. The Sahambano deposit is located 30 km east of Ihosy (Fig. 2-32). The deposit was discovered in 1999 and exploited sporadically. The sapphire crystals are multicolored, but rarely of gem quality and treatment is necessary to improve color and transparency. On average, 100 kg of corundum picked from washed material contains 24 kg of colored sapphires with 1 kg of translucent crystals, of which 50 g would be of gem quality. The division of the color is 15% brown to orange, 5% orange to pink, 40% pink to purple, 5% purple to fuchsia, and 35% violet to blue. The deposit occurs in the Tranomaro Group, characterized by a high abundance of calcic and magnesian paragneiss and leptynite. The deposit sits in the Ranotsara fault zone, a 30 km wide and 300 km long steep structure that had a long history of deformation and high temperature metamorphism dated between 600 and 500 Ma (Collins & Windley 2002). 40Ar/39Ar dating of biotite from a sapphire-bearing biotite schist gave a minimum age of formation at 492 5 Ma (Giuliani et al. 2007a). Mylonitization and dextral shears are common in the Sahambano area. The sapphire occurs in feldspathic gneiss lenses intercalated within leptynite at Dominique, Nono, Momo, Jeanne d'Arc, and Ambinda Sud pits. Shearing was accompanied by the opening of fissures and fluid circulation, which resulted in the biotitization of the host rock. Sapphire occurs in biotitite, which also contains sillimanite and spinel, and in gneiss composed of K-feldspar, biotite, sillimanite, spinel, sapphirine, garnet, albite, and sapphire. Sapphire formed during the prograde metamorphism at T ~ 650°C and P ~ 5 kbar according to the reaction (Ralantoarison 2006):

3 hercynite + K-feldspar + H2O 3 corundum + annite (9)

The size of the crystals varies from 1 mm to 5 cm. Sapphires display different colors according to their Cr and Fe contents: colorless, grey, greenish grey, orange, blue, dark pink, purple, brown, pink, and red to fuchsia. All crystals are euhedral and exhibit either short or long prismatic habits, which are often associated with the rhombohedron {10.1}, the hexagonal prism {10.0} or {11.0}, and the basal

FIG. 2-32. Structural and lithological sketch map of

southeast Madagascar with the location of the corundum deposits of mixed origins (modified after Martelat et al. 2000). Corundum deposits as 4: Ambinda (Betroka); 12: Miarinarivo, 13: Zazafotsy, 14: Sakalalina, 15: Ambinda (Ihosy), 16: Sahambano, 18: Vohidava, 19: Iankaroka, 20: Ambatomena, 21: Ianapera, 22: Fotadrevo, 23: Anavoha, 24: Maniry, 25: Gogogogo, 26: Vohitany, 27: Ejeda, 28: Ilakaka, 29: Sakaraha, 30: Andranondambo, 31: Sakeny. Major shear zones referred as A: Ampanihy, B: Beraketa, C: Ranotsara fault zone, D: shear zone of the Phanerozoic cover. Minor shear zones referred as (a) to (h). The Pressure (kbar) and Temperature (°C) of metamorphism are from Moine et al. (1985), Ackermand et al. (1989) and Nicollet (1990).

pinacoid {00.1}. Laminar habits are characterized by well-developed pinacoids, and cylinder habits are formed by hexagonal dipyramids associated with pinacoid {00.1} and/or hexagonal prism. Solid inclusions in sapphire are K-feldspar, zircon, barite, spinel, cheralite, sillimanite, diaspore, albite, and pyrite. The sapphire is contained in metasomatic zones

GIULIANI ET AL. �– GEM CORUNDUM

71

and the peraluminous gneiss supplied Al and the chromophore elements (Cr, Fe, and Ti) necessary for sapphire coloration (Fig. 2-33). The distribution of the color of the sapphire is controlled by lithology: colorless to blue sapphires are found in the biotitite zone developed from feldspathic gneiss, green to brown and greenish sapphires are in the sillimanite-bearing feldspathic gneiss, and finally red to fuchsia to pink and pinky orange sapphires are located in the sapphirine-bearing feldspathic gneiss. The other colored crystals are distributed randomly at the interface of biotitite and biotitized sapphirine-bearing feldspathic gneiss. The Zazafotsy deposit, also called "Amboarohy" (Pezzotta 2005), is located 35 km northeast of Ihosy on National Road 7 (Fig. 2-32). The sapphire occurrence was discovered in 1950 and exploitation by local miners began in 1989. The exploitation was sporadic until 2003, when it started again producing very beautiful euhedral sapphire crystals. The majority of the sapphires are not of gem quality and heat treatment is necessary to improve their transparency and color. The deposit is located in the Itremo Group which is mainly composed of garnet-sillimanite-cordierite leptynite and amphibole-clinopyroxene gneiss with minor intercalations of quartzite and impure limestone. The sapphire deposit lies in the Zazafotsy shear zone system, also linked to the Pan-

African tectono-metamorphic event (Martelat et al. 2000). The 40Ar/39Ar age at 494 5 Ma obtained from a biotite crystal associated with sapphire confirmed that the mineralizing episode was the latest Pan-African event in the area (Giuliani et al. 2007b). Like at Sahambano, the mineralization occurs in several lenses of feldspathic gneiss intercalated within garnet-bearing leptynite and affected by fluid circulation in shear zone fractures (Andriamamonjy 2006). The inner zone of the lens consists of almandine and sapphire up to 10 cm associated with biotite, plagioclase, spinel, and K-feldspar formed around sapphire and garnet (Plate 1A). The outer zone is biotite schist with biotite, sapphire, spinel, and very few crystals of grandidierite (boro-aluminosilicate of Fe and Mg, (Mg,Fe2+)Al3(BO3) (SiO4)O2) which passed into biotitized feldspathic gneiss. In one lens, the outer zone consists of fine-grained metasomatic alternation of biotitite and black tourmalinite developed on a 20 cm-wide scale. In the same area, apatite and beryl are found in pegmatite veins. All of the sapphire crystals are euhedral and, as at Sahambano, exhibit either short or long prismatic habits associated with the hexagonal prism and the rhombohedron terminated by basal pinacoid. Mineral inclusions found in sapphire are zircon, K-feldspar, plagioclase, sillimanite, spinel, and biotite. The sapphire occurs in different colors

FIG. 2-33. Schematic geological section

of the Momo trench in the Sahambano sapphire deposit (modified after Ralantoarison 2006). The sapphire mineralization is contained in biotitized feldspathic gneiss and sapphire-bearing feldspathic gneiss. The respective contents in Al, Mg, Fe, Ti (in wt.%) and in Cr (in ppm) are reported for each different protolith and the metasomatite, i.e., biotitite. The colored sapphire distribution is reported for the different zones (br, brown; bl, blue; bg, blue-green; cl, colorless; p, pink; o, orange; f, fuschia; r, red; pu, purple.

GIULIANI ET AL. �– GEM CORUNDUM

72

including dark blue, light blue, grey blue, fuchsia, orange, pink, violet, mauve, and brown. Deposits in cordieritite include the Iankaroka occurrence, which was reported in 1984�–85, and described by Salerno (1992). It is located 35 km south of the city of Betroka (Fig. 2-32) in the province of Toliara. The deposit is characterized by polychrome sapphire displaying distinct color bands. When observed in a plane parallel to the c axis, the crystal is uniformly pinkish to purple, but in a direction perpendicular to the c axis thin layers of green blue, orange, brown, and pink are visible (Koivula et al. 1992). The size of the crystals varies between 1 to 10 mm. Euhedral crystals are elongate to tabular hexagonal prisms and bipyramids. The sapphire is found in a cordieritite lens, ~ 7 m long and 4 m wide, intercalated concordantly within leptynitic biotite-cordierite-bearing gneiss of the Androyan series. The cordieritite is affected by shearing; on its border this results in the formation of a biotitite developed on the gneiss, and in the cordieritite in the development of sapphire-bearing fissures. The cordieritite is composed of phlogopite, cordierite, plagioclase, green tourmaline, chlorite, pyrite, spinel, and sillimanite. The Ambatomena ruby deposit, located 10 km northeast of the city of Isoanala (Fig. 2-32), was exploited by a private company from 2000 to 2001. The rubies were of good quality consisting of euhedral prismatic crystals up to 3 cm long and 1 to 2 cm in diameter. The deposit occurs in the Androyan metamorphic series composed of paragneiss, orthogneiss, marble, granite, clino-pyroxenite, and quartzite (Andriamamonjy 2010). The ruby is contained in cordieritite layers or lenses intercalated within biotite-cordierite-sillimanite-bearing charnockite. The mineralized zone has suffered intense metasomatism, transforming pegmatite veins into plagioclasite, and in the formation of corundum-free phlogopitite and sapphirine-anorthite-phlogopite-bearing rocks. The ruby occurs in a cordieritite composed of cordierite, rutile, K-feldspar, sapphirine, phlogopite ± diopside. Ruby crystals exhibit a spinel and sapphirine coronitic texture. Spinel results from the destabilization of ruby during the retrograde phase, and in some cases ruby is totally pseudomorphosed into spinel.

Metamorphic deposits associated with marble: the ruby and pink sapphire deposits hosted by meta-pure or impure limestone are of two types: (1) in marble where the corundum crystallized as a result

of retrograde isochemical reactions mainly in a closed system such as the ruby occurring as disseminated crystals in marble from central and southeast Asia (Garnier et al. 2008); and (2) in impure marble containing gneiss and silicate layers where ruby crystallized at the peak of prograde metamorphism such as at Revelstoke in Canada (Dzikowski 2013, Dzikowski et al. 2010, 2013) and Morogoro in Tanzania (Le Goff et al. 2008). The first type of marble deposit forms the main sources for excellent quality ruby with intense color ('pigeon blood') and high transparency in marble-hosted ruby deposits from central and southeast Asia (Fig. 2-9). These types of occurrences are also known in North America, Europe, and Africa: in Sussex county, New Jersey, USA (Dunn & Frondel 1990), at Mercus and Arignac in Ariège, France (Lacroix 1890), at the vicinity of Xanthi in Greece (Hughes 1997), at Prilep in Macedonia (Hunstiger 1990), at Campolungo in Switzerland (Hochleitner 1998), and in the Urals (Kissin 1994). These deposits share many common geological, structural and mineralogical features (Garnier et al. 2008). They are hosted by metamorphosed platform carbonate assemblages associated generally with intercalations of garnet-biotite-sillimanite- or biotite-kyanite-bearing schist or gneiss (Kievlenko 2003). The series commonly contain alternations of quartzite and amphibolite. Dikes of granite and/or pegmatite intrude these metamorphosed sedimentary rocks, but the ruby mineralization is not directly linked with granitoid bodies, as observed for desilicated pegmatites in other protoliths. The mineralization is generally stratiform, i.e., conformable with the bedding of the marble, and occasionally disseminated within a particular level in the marble. The mineralized zones are between 0.5 and 10 m thick. A 'ruby zone' is formed by a succession of benches of centimetre- to metre-scale marble, where ruby is located in veinlets, gash-veins, lenses or disseminations in the carbonate gangue. The estimated grade of the Nangimali deposit in Azad Kashmir (exploited by the AKMDC, Azad Kashmir Mineral and Industrial Development Corporation) was estimated at 11 g/m3 from a pilot study on the economic zone carried out over three years (Malik 1994, Pêcher et al. 2002).

In general the ruby-bearing marble is composed of calcite and dolomite. It contains the oxide minerals spinel and corundum, and diopside, phlogopite, garnet, chlorite, margarite, tremolite, pargasite, edenite, uvite, and forsterite. Graphite is sometimes associated with these minerals as in

GIULIANI ET AL. �– GEM CORUNDUM

73

deposits of Afghanistan, Pakistan, Myanmar, Nepal, and Vietnam. Other minerals include anorthite, anhydrite, aspidolite, titanite, fuchsite, scapolite, zoisite, K-feldspar, epidote, pyrite, and pyrrhotite (Iyer 1953, Harding & Scarratt 1986, Kissin 1994, Garnier et al. 2004c).

The most common crystalline forms include the hexagonal prism or the hexagonal dipyramid (Hughes 1990, Peretti et al. 1995, Smith et al. 1997, Smith 1998, Cesbron et al. 2002). Ruby from Jegdalek is generally tabular and formed by the hexagonal prism a {11.0} and the pinacoid c {00.1} and the positive rhombohedron {10.1}. Ruby from Mogok is tabular and prismatic with a development of the rhombohedron faces (Fig. 2-34). Ruby from Mong Hsu has several hexagonal dipyramids associated with the pinacoid and the hexagonal prism a (Fig. 2-34).

In general the color of ruby and sapphire from marble is variable: red to violet for Tanzanian (Hänni & Schmetzer 1991) and Nepali (Smith et al. 1997); red, pink, brown, yellow, blue, violet or colorless for corundum from the Urals (Kissin 1994). Additionally, ruby from Mong Hsu (Myanmar) shows unusual features such as a blue sapphire core and multiple zoning (Fig. 2-35) from ruby to violet sapphire in the outer zones of the crystal (Peretti et al. 1995). The rubies from Mogok appear to glow red or are like 'pigeon blood' and they are among the most expensive in the world for their top quality. The deposits occur (Fig. 2-9) in Afghanistan, Tajikistan, Pakistan, Azad-Kashmir, Myanmar,

Nepal, northern and central Vietnam, and southern China (Hughes 1997). The mines in Afghanistan are among the oldest in the world and were already exploited seven hundred years ago (Rossovskiy 1980, Hughes 1997, Bowersox & Chamberlain 1995, Bowersox et al. 2000, Cesbron et al. 2002). Ruby deposits from the Hunza Valley in Pakistan were discovered in the late 1970s during the building of the Karakorum Highway which links Pakistan with China (Piat 1974, Okrush et al. 1976, Gübelin 1982, Kazmi & O'Donogue 1990, Garnier 2003). The deposit from Nangimali in Azad-Kashmir was discovered in 1979 by the AKMDC mining society (Malik 1994), and the geology and the geochemistry of the deposit was studied by Pêcher et al. (2002). The Tajik ruby deposits remain poorly known and were first reported by Henn & Bank (1990). Kievlenko (2003) reported geological studies carried out by Russian geologists and Smith (1998) described the standard gemological properties and internal features of the Tajik ruby. The first report of marble-hosted ruby deposits from Nepal was published by Bassett (1984), followed by Harding & Scarratt (1986), and the mineralogy was studied by Smith et al. (1997). The Mogok ruby deposits in Myanmar have been exploited since the sixth century (Chhibber 1934, Gübelin 1965, Kane & Kammerling 1992, Kammerling et al. 1994), whereas the Mong Hsu deposits were first reported in 1991 (Hlaing 1991, Peretti et al. 1995). In northern Vietnam, a peasant discovered the first ruby at Luc Yen in 1983 and exploitation of the occurrences began in 1988 (Kane et al. 1991,

FIG. 2-34. Crystal habits of ruby from Mong Hsu and Mogok, Myanmar (after Smith & Surdez 1994, Hughes 1997).

GIULIANI ET AL. �– GEM CORUNDUM

74

FIG. 2-35. Variations of habit and

color during the growth of ruby from Mong Hsu, Myanmar (Peretti et al. 1995). The successive growth zones are red (R, R'), Violet (V, V') and intermediary (I, I') colors which are associated to a coeval variation of chromium (Cr2O3) and titanium (TiO2).

Garnier 2003, Pham et al. 2004, Nguyen et al. 2011, Le et al. 2012). Galibert & Hughes (1995) reported that ruby-bearing marble was discovered in Yunnan province, China, at the beginning of the 1980s. Detailed characteristics of gems from the Ailoshan structural belt were given by Zhang et al. (2003). Marble-hosted ruby deposits at Jegdalek in Afghanistan, at Nangimali in Azad-Kashmir, at the Hunza valley in Pakistan), at Chumar and Ruyil in Nepal, at Mogok and Mong Hsu in Myanmar), and at Luc Yen-Yen Bai, and Quy Chau in Vietnam, share many common structural, mineralogical and geochronological features. To begin with, they all occur in metamorphic blocks that were affected by major tectonic events during the Cenozoic Indo-Asian collision (Fig. 2-9). The Jegdalek deposit is located in the Western Nuristan Block, in the Indus suture zone. The ruby-bearing marble is underlain by metamorphic basement of presumed Precambrian age (Rossovskiy et al. 1982) intruded by Oligocene (34�–26 Ma) plutonic units that define the southern part of an axial batholith. The Nangimali deposit lies in the High Crystalline Himalaya, in the Southern part of Nanga Parbat. It is located within a syncline formed during an upper Cretaceous to Miocene Himalayan tectono-metamorphic event (Pêcher et al. 2002). The Chumar and Ruyil deposits in Nepal are located on the southeastern flank of the Ganesh Himal, in the Nawakot series, just below the Main Central Thrust. These ruby deposits occur within boudinaged marble lenses, 60 to 150 m wide and up to 1 km long, oriented parallel to the Main Central Thrust. The Hunza Valley deposits (Okrush et al. 1976) are located in the Baltit formation of the Karakorum metamorphic complex, north of the Nanga Parbat Massif, between the Main Karakorum Thrust and the intrusive

Karakorum batholith. The famous Mogok ruby deposits occur in the Mogok Metamorphic Belt. This belt accommodated a large part of the Indo-Asian collision and is marked by high temperature ductile Oligocene stretching and post-Miocene brittle right-lateral faults, such as the Shan scarp fault zone and the Sagaing fault (Bertrand & Rangin 2003). The ruby deposits in northern Vietnam are located along the Red River shear zone that was active during Cenozoic times (Schärer et al. 1990, Tapponnier et al. 1990, Garnier et al. 2004b). The deposits from Luc Yen (An Phu, Bai Da Lan-Mong Son, Khoan Thong, Luc Yen, Minh Tien, and Nuoc Ngap mines) occur in weakly deformed marble units from the Lo Gam tectonic zone, located on the eastern flank of the Day Nui Con Voi belt within the Red River shear zone (Leloup et al. 2001). In general these ruby deposits are hosted by metamorphosed platform carbonate units associated with garnet-biotite-sillimanite- or biotite-kyanite-bearing gneiss and granite (Fig. 2-36). The marble units consist of discontinuous horizons up to 300 m in thickness, now oriented parallel to the main regional foliations, thrusts or shear zones related to Cenozoic Himalayan orogenesis. Dikes of granite and/or pegmatite intruded these metamorphosed sedimentary rocks. Ruby mineralization is generally stratiform and distributed within marble layers. In Luc Yen (Vietnam), Nangimali (Azad-Kashmir), and Hunza (Pakistan), amphibolite bodies alternate with calcitic and/or dolomitic marble. The marble parageneses typically consist of calcite, dolomite, spinel, phlogopite, margarite, amphibole, chlorite, forsterite, and titanite ± graphite ± garnet ± pyrite. The surrounding gneiss or schist exhibits assemblages with biotite-garnet-sillimanite ± scapolite ± kyanite ± amphibole ±

GIULIANI ET AL. �– GEM CORUNDUM

75

FIG. 2-36. The Nangimali Formation showing the location

of the ruby-bearing marble in the metamorphic pile (after Pêcher et al. 2000).

clinopyroxene. At the Jegdalek, Nangimali, and Ruyil-Chumar deposits pegmatites are absent (Smith et al. 1997, Pêcher et al. 2002) and ruby occurs in marble lenses parallel to the main foliation and/or within thin fractures (up to 3�–4 cm wide) in marble horizons (Pêcher et al. 2002). The ruby crystals typically occur: (1) disseminated within marble and associated with phlogopite, muscovite, scapolite, margarite, spinel, titanite, pyrite, and graphite, as at Jegdalek, Chumar and Ruyil, Hunza, Mogok, Mong Hsu, Myanmar, and Luc Yen; (2) in veinlets or gash veins, as in some occurrences in northern Vietnam, and associated with phlogopite, margarite, titanite, graphite, and pyrite, and sometimes related to micro-shear zones, as at Nangimali; or (3) in pockets associated with orthoclase, phlogopite, margarite, graphite, and pyrite in some occurrences of northern Vietnam. Ruby hosted in marble was indirectly dated by 40Ar/39Ar stepwise heating performed on single grains of phlogopite syngenetic with ruby (Garnier et al. 2002b, Garnier et al. 2006b), and zircon inclusions in ruby were dated by U�–Pb ion probe techniques (Garnier et al. 2004b). 40Ar/39Ar ages of

phlogopite associated with ruby are Oligocene (24.7 ±0.3 Ma) at Jegdalek, Miocene at Mogok (18.7 ±0.2 to 17.1 ±0.2 Ma), at Hunza (10.8 ±0.3 to 5.4 ±0.3 Ma) and Chumar (5.6 ±0.4 Ma), and Pliocene (4.6 ±0.1 Ma) at Ruyil (Fig. 2-8). In Vietnam, a zircon included in a ruby from Luc Yen yielded a 238U/206Pb age of 38.1 ±0.5 Ma, indicating that ruby formed around 40 Ma when ductile deformation was active under peak metamorphic conditions in the Red River shear zone. All these ages are consistent with extensional tectonic processes that were active from Afghanistan to Vietnam, in the ruby-bearing metamorphic belts, between the Oligocene and the Pliocene. The ruby deposits are contemporaneous with the intrusions spatially associated with them. Only a few detailed geological studies have been devoted to these ruby deposits but several hypotheses on their genesis have been advanced. One hypothesis suggests that ruby formation is a consequence of amphibolite-facies regional meta-morphism in the calcareous rocks enriched in Al relative to Si, Cr and Ti (Okrush et al. 1976). Such original bulk rock chemistry may result from lateritic weathering of an impure limestone, formed in a karst environment, before metamorphism. This model, proposed by Okrush et al. (1976) for deposits in the Hunza Valley, was also used by Rossovskiy et al. (1982) and Bowersox et al. (2000) to explain the genesis of the Jegdalek, Pamirs, and Mogok deposits. Hunstiger (1990) explained that corundum formed from diaspore as a result of intensification of pressure and metamorphism of an initial impure protolith following the prograde transformation from hydrargillite, to boehmite and diaspore, and finally corundum. Kissin (1994) proposed a variant model with metasomatic trans-formations by metamorphic fluids of terrigenous layers in marble.

A second hypothesis suggests that ruby has an indirect relation with granite and its formation is due to contact metamorphism (skarn-type) developed around granitic intrusions, as proposed by Iyer (1953) and Harlow et al. (2006) for some of the Mogok deposits where ruby-painite [CaZrAl9O15 (BO3)] and typical skarn assemblages are found. In that case, the source of heat and some of the chemical elements necessary for ruby formation is the granite. The hypothesis of alkaline meta somatism of marble by the circulation of magmatic and/or pegmatitic fluids is retained by Nguy et al. (1994) for the genesis of Luc Yen deposits. A third model suggests that the ruby has a direct genetic relation with a magma and formed

GIULIANI ET AL. �– GEM CORUNDUM

76

during the emplacement of alkaline magmatism intrusions in marble for the Tajikistan deposits in the Central Pamirs (Terekhov et al. 1999); magmatic solutions carrying Al interacted with marble. The deep-seated source is reinforced by numerical modeling of fluid�–rock interaction between marble and an alkaline fluid, performed by Kolstov (2002). Stability of the corundum-bearing assemblages in the modeling was described by the equilibrium equation:

2 pargasite + corundum + 9 CO2 3 anorthite + 2 albite + calcite + dolomite + 2 H2O (10)

The high concentration of CO2 appeared unusual to Kolstov (2002) for most of the fluids in amphibolite-facies metamorphism and for fluids degassing from magmatic chambers. Thus a deep-seated source which interacted with an initial rock package made of marble and kyanite-garnet-biotite schist was preferred. The metasomatic interaction of the fluid with the composite protolith resulted in the formation of metasomatic columns where corundum formed in Mg or Ca-rich marble zones. To the contrary, based on isotopic and geochemical data on the corundum-bearing metasomatic rocks in the Central Pamirs, Dufour et al. (2007) demonstrated by numerical simulations of the fluid�–rock alteration that ruby was the product of regional alkaline metasomatism during the closing stages of Himalayan metamorphism. The terrigenous material hosted by the Phanerozoic carbonate reservoir was desilicated by the metasomatic fluids and ruby formed at T = 600�–650°C and P = 4�–6 kbar. A fourth model suggests that the ruby formed during high-pressure metamorphism of originally impure limestone in evaporitic series. Spiridonov (1998) proposed this hypothesis for the ruby deposits of Turakuloma, Pamirs, and the Uralian folded areas. Ruby occurs within a sequence of dolomite, calcite and magnesite marble containing intercalated schist. Ruby is associated with carbonate minerals, scapolite, and fuchsite. The model is based on experiments and petrological observations which show that at a high CO2 fugacity spinel breaks down into corundum according to the reaction:

MgAl2O4 + CO2 Al2O3 + MgCO3 (11) spinel + CO2 corundum + magnesite

Spiridonov (1998) thought that both meta-evaporite and meta-ultrabasite rocks are favorable environ-

ments for ruby crystallization because Si, K and Na activities are low, and above 400°C alumina is transported within aqueous fluids. The model requires the presence of Cr and also a fluid-saturated environment. Kissin (1994) observed the formation of ruby in marble in the Urals from the destabilization of spinel at a temperature between 620�–660°C and pressure around 2.5 kbar, according to the reaction:

spinel + calcite + CO2 corundum + dolomite (12)

The formation of corundum by spinel breakdown is the main reaction observed for the ruby deposits from southeast and central Asia (Garnier 2003). In this ruby belt, the experimental reaction proposed by Spirodonov (1998) does not occur and magnesite has never been identified:

spinel + CO2 corundum + magnesite (13)

A fifth model suggests that during the metamorphism of evaporite lenses which reacted as molten salts with the marble and its impurities i.e., organic matter and phengite (Garnier 2003, Garnier et al. 2008) to produce CO2�–H2S�–COS�–S8�–AlO(OH)-bearing fluids subsequently trapped by Vietnamese rubies (Giuliani et al. 2003). Pêcher et al. (2002) have shown that the formation of Nangimali ruby was a good example of fluid transfer during metamorphic processes. Parental fluids of ruby are of metamorphic origin and CO2 is derived from the decarbonation of dolomitic marble. In marble, the presence of aspidolite and anhydrite (Garnier et al. 2004c), together with anhydrite and salts in ruby, implies the presence of evaporite rocks for the genetic formation of ruby. Such sulfates and salts in ruby were also described at Mogok (Smith & Dunaigre 2001, Garnier et al. 2008), Luc Yen (Giuliani et al. 2003), Quy Chau, Afghanistan, Nepal, Tadjikistan, and Campo Lungo in Switzerland (Garnier et al. 2008). Raman and infrared spectroscopy, combined with microthermo-metric investigations, on primary and secondary fluid inclusions in gem ruby from all the ruby deposits from central and southeast Asia provided evidence for CO2�–H2S�–COS�–S8�–AlO(OH)-bearing fluids with native sulfur and diaspore daughter minerals, without visible water (Giuliani et al. 2012b). Crush-leach analyses identified sulfates and chlorides that are attributed to the presence of anhydrite and Na�–Ca�–Cl salts in the ruby crystal (Giuliani et al. 2012b). The thermal reduction of evaporitic sulfates explains the fluid chemistry.

GIULIANI ET AL. �– GEM CORUNDUM

77

As we have seen, marble-hosted ruby deposits from central and southeast Asia share common features: (1) they are hosted by metamorphosed marine carbonate assemblages; (2) salts and sulfate minerals are found as solid inclusions in the ruby samples; (3) minerals associated with ruby are enriched in Mg, Al, F and in some cases Cl; (4) they have trapped CO2�–H2S-bearing fluids with native sulfur and diaspore daughter minerals; and (5) they formed during or directly after upper amphibolite- to granulite-facies metamorphism. (C�–O�–H) isotopic studies have shown that marble units were not infiltrated by externally derived fluids (Garnier et al. 2008). CO2-rich fluids are of metamorphic origin and derived from devolatilization of marble. The second genetic model, based on a possible circulation of magmatic and/or pegmatitic fluids, and the third model, based on a direct genetic relation of ruby with magma or a deep seated source as proposed for the Tajikistan deposits (Terekhov et al. 1999), are not tenable. Nevertheless, ruby is sometimes associated with skarn at the contact granite�–marble in some mines at the Mogok district (Nissimboim & Harlow 2011). The presence of a high CO2 fugacity for ruby formation is in agreement with the five proposed models for ruby formation and experimental results (Kolstov 2002). Such thermodynamic conditions are favorable for spinel breakdown according to the main reaction:

spinel + calcite + CO2 corundum + dolomite (14)

This reaction occurring during the retrograde metamorphic path has been commonly observed in marble from Jegdalek, Afghanistan, Hunza and Luc Yen in Vietnam (Fig. 2-37A; Garnier et al. 2008), and is also described from the Urals (Kissin 1994). In dolomitic marble, the presence of aspidolite and anhydrite together with hypersaline fluids (salt and sulfate) in rubies implies the necessary presence of evaporites for the generation of ruby (Fig. 2-37B, C). The petrographical and geochemical evidence is reinforced by S and B isotopic studies of anhydrite and tourmaline respectively that have pointed to the participation of non-marine and marine evaporites in the formation of ruby (Garnier et al. 2008). At Hunza the presence of inclusions of anhydrite and spinel in ruby (Fig. 2-37B), inclusions of spinel in anhydrite, and restite of anhydrite in marble, suggests that the limestone was deposited in an environment propitious for the deposition of

evaporites. In the Mogok rubies, multi-solid fluid inclusions (Fig. 2-37E) are coexisting with the primary CO2�–H2S�–COS�–S8�–AlO(OH)-bearing fluids (Fig. 2-37D). They are composed of molten salts and solid inclusions which have been studied by micro-Raman spectrometry. The main molten salt and sulfates are dawsonite (NaAl(CO3)(OH)2, barite, unnamed sulfates, Al-nitrates, Na�–Ca�–CO3 mineral like shortite Na2Ca2(CO3)3, and complex carbonate minerals (REE-carbonates?). Halite cubes are sometimes present. The other minerals are fluorite, calcite, dolomite, apatite, boehmite, magnetite, graphite, and rutile. The presence of molten salts corroborates the model of formation for ruby based on the thermal reduction of sulfates of evaporitic origin (Giuliani et al. 2012b). These different features support models four and five proposed for the formation of marble-hosted ruby deposits. Elemental sulfur, hydrogen sulfide and COS in the CO2-inclusions originated from the partial dissolution and reduction of anhydrite by organic carbon (Giuliani et al. 2003). Such chemical association, and especially the presence of COS in geological fluids, was previously described by Grishina et al. (1992) for the Siberian platform, where dolerite intruded carbonate�–anhydrite and halite beds. Furthermore, nitrates as NO3

�– were identified in the crush leach and Giuliani et al. (2012b) proposed a non-marine evaporite origin i.e., evaporitic lakes. The high temperature conditions for ruby formation imply a heat source, i.e., metamorphic or magmatic. Contact metamorphism or regional metamorphism are both suitable. Control of the mineralization is lithologic, with ruby formed in marble intercalated with meta-evaporite lenses (Pêcher et al. 2002, Garnier 2003). As no externally derived fluids are interpreted to have circulated through the marble, the source of Al and Cr must have been deposited coevally with the carbonate rocks on the platform. The Al was probably contained in clay minerals, which are produced in great quantities by weathering of continental rocks and transported by rivers to the sedimentary basins. A contribution from organic matter for Cr and Al is a possibility, as is apparent in bitumen associated with Colombian emeralds in a black shale environment with Cr 3�–87 ppm, Al 700�–2200 ppm (Giuliani et al. 2000), but also from detrital minerals enriched in Cr, and possibly Ti. Thus, ruby is present only in marble units that preferentially contained these Cr�–Al�–bearing minerals, as

GIULIANI ET AL. �– GEM CORUNDUM

78

FIG. 2-37. Petrography and fluid inclusions in ruby from marble (Garnier et al. 2008, Giuliani et al. 2012b). A, Ruby (Crd) in

equilibrium with dolomite (Dol) forms from the reaction of spinel (Sp) with calcite (Cc). Spinel is a pre-ruby phase and it contains high Cr (up to 19 wt.%) and Zn contents (10 wt.%). Spinel contains also anhydrite (Anh). Hunza valley, Pakistan. B, Detail of anhydrite (Anh) and spinel (Sp) inclusions in this ruby. SEM photographs A and B are back-scattered electron (BSE) mode. C, Inclusions of anhydrite (Anh) and phlogopite (Phg) in ruby (Ru) from Nangimali in Azad Kashmir. SEM photograph. D, Ruby from Mong Hsu (Myanmar). Two pseudo-secondary biphase fluid inclusions containing a carbonic liquid (l) and vapor phases (v). The biggest FI contains diaspore crystals (Di) and one globule of native sulfur (S8). E, Polyphase fluid inclusions with molten salts in a Mogok ruby with respectively, calcite (Ca), graphite (C), dawsonite (Dw), apatite (Ap), halite (H), shortite (Sh), fluorite (F), corundum (Crd), and a liquid CO2�–H2S�–S8 fluid (l).

FIG. 2-38. Synthesis presenting the proposed model for genesis of ruby deposits in marble (Garnier 2003). The different stages

of formation of these deposits are illustrated from sedimentation, i.e., Precambrian to Permo-Triassic, to the Himalayan metamorphism. Stage I: sedimentation of the protoliths in a continental platform with a detrital supply, and deposition of carbonate, dolomitic carbonate, organic matter-bearing shale, and marine and non-marine evaporite sediments i.e., lagoon, sabkha paleogeography as found in the representative lithostratigraphic section of Nangimali in Azad Kashmir. Stage 2: Cenozoic Himalayan metamorphism a consequence of the Indo-Asian collision with the metamorphism of the ruby protoliths as shown on the metamorphic unit cross-section of Nangimali. Corundum formed during the prograde metamorphic P�–T path and then during the retrograde stage from the destabilization of different mineral assemblages as drawn in the P�–T sketch: 1, from margarite, then 2, from micas, and 3, from spinel. At this stage, corundum is destabilized at lower temperature and pressure in margarite, and then diaspore. Gem ruby formed at a pressure between 2.6 and 3.3 kbar and temperature between 620 and 670°C (black rectangles). Abbreviations: An: anorthite; And: andalusite; Cc: calcite; Co: corundum; Do: dolomite; Dsp: diaspore; Kfs: K-feldspar; Ky: kyanite; Mrg: margarite; Ms: muscovite; Si: sillimanite; Sp: spinel; W: water.

GIULIANI ET AL. �– GEM CORUNDUM

79

GIULIANI ET AL. �– GEM CORUNDUM

80

proposed by Okrush et al. (1976) and Hunstiger (1990). The Al2O3 content of the marble is not a critical factor for ruby formation (Garnier et al. 2004c). At Nangimali the marble contains 0.1 wt.% Al2O3, and in the marble from the Urals deposits, the Al2O3 content is between 0.08 and 0.13 wt.% (Kissin 1994). The following model (Fig. 2-38) has been proposed for formation of the Indo-Asian ruby deposits (Garnier et al. 2008): (Stage I) deposition of sedimentary protoliths on the Paleotethys platform with local formation of restricted endoheric basins. The ages of sedimentation vary from Precambrian at Jegdalek (Terekhov et al. 1999) to Permo-Triassic at Nangimali (Malik 1994). Restriction of sedimentary basins may be due to tectonism, volcanism (e.g., intrusion of basaltic dikes at Nangimali) or sedimentation dynamics with coral reefs, oolitic units. (Stage II) during the Indo-Asian collision, the sediments were metamorphosed in upper amphibolite to granulite facies, transforming carbonate into marble. Mg, Al, and Na enrichment of the minerals associated with ruby in marble, the presence of tourmaline and anhydrite and salts, and the replacement of some of the hydroxyl groups by F and Cl are evidence for the participation of evaporites in the formation of ruby. Molten salts (NaCl, KCl, CaSO4 and Na2CO3) enriched in F, Cl and B, mobilized (over short distances) Al and transition metal elements contained in the marble. Organic matter played a key role in the mineralizing process. Thermal reduction of evaporitic sulfate, based on an initial assemblage of anhydrite, calcite and graphite, is a convenient explanation for the fluid chemistry found in fluid inclusions but also for the main paragenesis associated with ruby, i.e., carbonate and pyrite, with a reduction of sulfate and oxidation of organic matter. The deposition of pyrite occurs as follows:

7H2S + 4Fe2+ + SO42�– 4FeS2 + 4H2O + 6H+ (15)

Gem ruby formed during the retrograde metamorphism at temperatures of ca. 620�–690 ºC and pressures of approximatively 2.2�–3.3 kbar (Garnier et al. 2008). Evaporites appear to be a key factor in the formation of ruby deposits: (1) Cl and F acted as activators as proposed by Peretti et al. (1996) to mobilize Al in marble, which normally is poorly mobile. F and Cl facilitate the mobilization of Al from mica, formed by the metamorphism of clay minerals, organic matter, and spinel minerals,

leading finally to the formation of ruby; (2) thermochemical reduction of sulfates produced S8, COS and H2S. H2S combined with Fe to produce pyrite; (3) the presence of lenses of evaporites in some specific horizons explains why ruby is sporadic in a single homogeneous marble level, even though Al is ubiquitous; and (4) this genetic model for natural ruby in marble resembles the industrial process that uses molten salts (NaF�–AlF�–Al2O3) to produce Al (Lacassagne et al. 2002). Al fluoride and oxy-fluorine Al are major chemical species involved in dissolution of Al in a F-rich fluid. Guidelines for prospecting for new deposits must take into account (1) the lithological control of mineralization: the protolith deposited in a carbonate platform, in a lagoonal to lacustrine environment, isolated from open sea, and promoting the deposition of marine and non-marine evaporites; and (2) the presence of index minerals such as Na-rich phlogopite (aspidolite) or pargasite and edenite, Mg-tourmaline, Cr-rich titanite (more abundant than ruby itself), and spinel and anhydrite in marble (or in alluvium), are all indicators of possible marble-hosted ruby mineralization in the area. The second type of marble deposit concerns ruby and pink sapphire in marble with intercalated silicate and gneiss layers (Plate 2A, B). It is the case at the Revelstoke occurrence in British Columbia, Canada, discovered in 2005 (Dzikowski et al. 2010, 2013). The occurrence is in the Monashee Complex of the Omineca belt of the Canadian Cordillera. The corundum occurs in thin, folded and stretched layers with a predominant assemblage of green muscovite + Ba-bearing K-feldspar + anorthite (An0.85�–1) ± phlogopite ± Na-poor scapolite. Other silicate layers within the marble are: (1) diopside + tremolite ± quartz and (2) garnet (Alm0.7�–0.5Grs0.2�–0.4) + Na-rich scapolite + diopside + tremolite + (Na,K)-amphibole minerals. Non-silicate layers in marble are either magnetite- or graphite-bearing. The pink sapphire and ruby crystals have elevated Cr2O3 ( 0.21 wt.%) and low amounts of TiO2; rare blue rims contain higher amounts TiO2 ( 0.53 wt.%) and Fe2O3 ( 0.07 wt.%). The associated micas have elevated Cr, V, Ti, and Ba. Petrography of the silicate layers show that corundum formed from muscovite at the peak of metamorphism (T~ 650�–700°C at P ~ 8.5�–9 kbar). The scapolite-bearing assemblages formed during or after decompression of the rock at T~ 650 °C and P~ 4�–6 kbar. Whole-rock geochemical data show that the corundum-bearing silicate (mica�–feldspar) layers

GIULIANI ET AL. �– GEM CORUNDUM

81

formed by mechanical mixing of carbonate with the host gneiss protolith; the bulk composition of the silicate layers was modified by Si and Fe depletion during prograde metamorphism. High element mobility is supported by homogenization of

18O and 13C values in carbonates and silicates for the marble and silicate layers. The silicate layers and the gneiss contain elevated contents of Cr and V due to a volcanoclastic component of their protolith. Such a deposit has been also described in Morogoro, Tanzania (Le Goff et al. 2008). At the mine Ngong'Oro, the ruby-bearing horizon is formed by several centimetre-wide metasomatic zones parallel to the metamorphic foliation, and formed at the contact between marble and intercalated biotite gneiss layers such as at Revelstoke. Ruby is associated with spinel and sapphirine. Hänni & Schmetzer (1991) described ruby from Morogoro. The crystals are composed of the faces of the pinacoid c and the rhombohedron r, but in some crystals the faces a {11.0} are present and thus the crystals look like octahedral spinel (Fig. 2-39). Metamorphic deposits in granulite, charnockite and gneiss: granulite is a metamorphic rock with a high-pressure and high-temperature origin. It is fine-grained with dominant lenticular quartz and feldspar, and hypersthene and garnet, with accessory minerals such as sillimanite, kyanite, rutile, cordierite, and spinel. Granulite is found in Precambrian terranes with other rocks such as leptynite, kinzigite, charnockite, and gneiss. Several gem corundum deposits are hosted by biotite-sillimanite-K-feldspar gneiss in the granulite facies, for example, those from Azov in Russia, Froland in Norway, Mysore in India, Hida in Japan, and southern Madagascar (Schwarz 1998). Other deposits are known from Indaia in Brazil (Epstein et al. 1994) and in the metamorphic Mozambique belt

FIG. 2-39. Crystal habits of ruby from Morogoro,

Tanzania (after Hänni & Schmetzer 1991). a, hexagonal prism of second order {11.0}; r, rhombohedron {10.l}; c, pinacoid {00.1}.

(Simonet 2000). Grew et al. (1989) also described biotite-plagioclase-spinel-corundum in the gneiss of the Aldan craton in eastern Siberia. In Sri Lanka, the rich alluvial gemstone deposits are located in the Precambrian Highland-Southwestern Complex formed of granulitic rocks (Fig. 2-28). Gemstones are rarely discovered in situ and the prospecting methods for primary deposits are based on geology, geochemistry, and mineralogy (Wells 1956, Cooray & Kumarapeli 1960, Katz 1972, Dahanayake 1980, Dissanayake et al. 2000, Walton 2004). The mother rocks of corundum may include charnockite, gneiss with quartz, feldspar, garnet, sillimanite, graphite, and marble and calcitic gneiss (Cooray & Kumarapeli 1960, Munasinghe & Dissanayake 1981; Dissanayake & Chandrajith 1999). The P�–T conditions of formation of gem corundum in the amphibolite facies was approached by De Maesschalk & Oen (1989) studying fluid inclusions trapped by corundum from gem gravels. Several studies have explored the modes of occurrence of gemstones and guides for exploration. Dahanayake (1980) studied gem-bearing gravels and suggested that gem corundum (Plate 1F) and spinel derived from both garnet-bearing gneiss and skarn deposits. Munasinghe & Dissanayake (1981) suggested that corundum formed at the contact between basic charnockitic intrusions and aluminous metasedimentary rocks. Rupasinghe et al. (1984) and Dissanayake & Weerasooriya (1986) used Be and F geochemistry to locate the bedrock sources. Rupasinghe & Dissanayake (1985) proposed that the intrusion of charnockite into aluminous metasedimentary rocks was responsible for desilication in the pelitic sedimentary rocks and the formation of corundum following the reaction:

spinel + sillimanite cordierite + corundum (16)

Mendis et al. (1993) applied structural geology in the exploration for residual gem deposits and found that they are apparently associated with structurally deformed areas where marble and granite occur (skarn-type deposit). Gamage et al. (1992) showed that high Rb/Sr ratios in stream sediments correspond to high gem concentrations. They suggested that depletion of Sr is a function of fractionation of the parental granite generated during granulite metamorphism. Ruby and sapphire associated with granulite-facies metamorphism from Sri Lanka show numerous mineralogical and geological similarities

GIULIANI ET AL. �– GEM CORUNDUM

82

with those from Eastern Africa (Mercier et al. 1999b, Simonet 2000), Madagascar (Lacroix 1922, Rakotondrazafy et al. 1996, 2008), and southern India (Santosh & Collins 2003, Giuliani et al. 2007a, 2007b). Based on the correlation of gem and graphite fields, Menon & Santosh (1995), and Dissanayake & Chandrajith (1999) proposed the existence of a Neoproterozoic�–early Cambrian gemstone province developed in East Gondwana under granulitic conditions (Fig. 2-8). Indeed, the gemstone areas are located in the Precambrian basement, which includes remnants of early crust (de Wit 2003) which were intensely reworked between 950 and 450 Ma, during Pan-African tectonometamorphic events (Kröner 1984). The collision processes between East and West Gondwana resulted in the formation of Neoproterozoic mobile belts (~650 Ma), mostly metamorphosed to high-grade granulite facies. They include similar metamorphic rocks varying from amphibolite, feldspar gneiss, charnockite, granulite, cordieritite, anorthosite, and mafic and ultramafic rocks (Mercier et al. 1999a, 1999b; Feneyrol 2012; Feneyrol et al. 2013). Southern Madagascar is a key area for the characterization of primary corundum deposits formed under granulite-facies conditions (Rakotondrazafy et al. 2008). The metamorphic rocks are high- to medium-pressure granulite and are well exposed throughout southeastern Madagascar (Fig. 2-32). The granulitic terranes are divided into four major lithostratigraphic groups (Besairie 1967, de Wit 2003) corresponding to the juxtaposition of tectonic blocks of different crustal levels (Martelat et al. 2000). This patchwork is due to the relative movements of major ductile shear zones, reflecting a crustal scale strike-slip system. Rocks in all blocks underwent metamorphism around 750°C. The pressure shows an E�–W decrease from 11 to 8 kbar in the west to 5 to 3 kbar in the east (Nicollet 1990). Granitoid rocks are abundant in the eastern part of the country, whereas anorthosite and metabasite are abundant in the west. Corundum deposits are found in all the different tectonic blocks, but are strongly associated with the presence (Fig. 2-32) of major or minor shear zones (Rakotondrazafy et al. 2008). These structures acted as preferential fluid pathways, and the corundum-bearing rocks have suffered intense fluid�–rock interaction resulting in huge amounts of metasomatic alteration. The nature of the parental host rock varies from feldspathic gneiss (Zazafotsy and Sahambano deposits), cordieritite (Iankaroka

and Ambatomena), amphibolite, and anorthosite (Ejeda, Fotadrevo, Vohitany, and Gogogogo) to impure marble (skarns of Andranondambo, Tranomaro). The 'sakenite' described by Lacroix (1941) is found in a metamorphic series made up of paragneiss with intercalations of amphibolite, clinopyroxenite, and impure marble (the Sakeny, Vohidava, Ejeda-Anavoha, and Andranondambo occurrences) and consists of anorthite veins or segregations ± spinel ± corundum ± sapphirine ± phlogopite and (± hibonite). Their original protolith is either metapelite (Lacroix 1941, Raith et al. 2008) or meta-anorthosite (Boulanger 1959, Ohnenstetter et al. 2008). The diversity of bedrock sources of gem corundum explains the incredibly rich variety of sapphires found in the Ilakaka placer, located on the western border of the granulitic domain (Garnier et al. 2004a, Giuliani et al. 2007a, Rakotondrazafy et al. 2008). The sources of ruby and colored sapphire are multiple, and the expression 'gem corundum in granulite' includes a variety of rocks that might contain sapphire and/or ruby. The Malagasy example illustrates the complexity in the search for primary deposits in Sri Lanka. Metamorphic deposits in migmatite: migmatite is a mixture of rocks of granitic and gneiss types. The gneiss is coarse-grained with a diffuse foliation. These rocks formed at the boundary of high-grade metamorphic and magmatic rocks. They formed by partial melting during ultrametamorphism of an initial rock called the paleosome; one part of the rock melts and one part remains solid and is called a restite. The whole range of migmatitic rocks is grouped under the term anatexite. The corundum-bearing anatexite at the vicinity of Morogoro in the Precambrian terranes of Tanzania is an anatectic gneiss (Altherr et al. 1982). It is composed of different assemblages: (1) a medium-grained gneiss with albite + muscovite + phlogopite + corundum and albite + kyanite or sillimanite + phlogopite, and accessory minerals as rutile and baddeleyite; (2) a coarse-grained gneiss with nests of corundum and antiperthite, with minor phases of albite, muscovite, phlogopite, rutile, baddeleyite, and tourmaline. Assemblages with albite + muscovite + phlogopite + corundum are restite, while those with kyanite or sillimanite are recrystallized paleosome. The anatexite formed at PH2O = 7.7 kbar and T = 695°C. Cartwright & Barnicoat (1986) described pale blue sapphire in kyanite-muscovite restite in

GIULIANI ET AL. �– GEM CORUNDUM

83

anatexite from Scotland. The P�–T conditions at the peak of metamorphism were estimated at 11 kbar and 900�–925°C. Metamorphic deposits in amphibolite and gabbro: corundum in amphibolite originating from the metamorphism of gabbroic and dunitic rocks is an important source of ruby. Generally ruby is not of gem quality but the crystals are used for cabochons and ornamental uses, for example the famous 'anyolite' from Longido in Tanzania (Plate 1B, Dirlam et al. 1992, Le Goff et al. 2008, Feneyrol et al. 2013) or the 'smaragdite ruby' from North Carolina in the USA (Hughes 1997). Three recent ruby deposits discovered in the last decade are highly economic for their quality and quantity such as at Winza in Tanzania (Peretti et al. 2008, Schwarz et al. 2008), at Aappaluttoq in Greenland (Rohtert & Ritchie 2006, Fagan et al. 2011, Fagan 2012), and at Montepuez, Ruambeze and M'sawize in Mozambique (Pardieu et al. 2009a, 2009b, 2009c, Pardieu & Chauviré 2013). Such corundum deposits in amphibolite and gabbro are known at Kittila in Finland (Haapala et

al. 1971), at Sittampundi in India (Janardharan & Leake 1974), at Chantel in France (Forestier & Lasnier 1969), at Losongonoi in Tanzania (Dirlam et al. 1992), at Kitui in Kenya (Barot & Harding 1994), at Chimwadzulu in Mozambique (Andreoli 1984, Henn et al. 1990), in the Vohibory formations in Madagascar (Nicollet 1986, Mercier et al. 1999b), in the Harts Range in Australia (McColl & Warren 1980), at Dir in Pakistan (Aboosally 1999), and Hokkaido in Japan (Morishita & Kodera 1998). These deposits display several common features. The corundum-bearing amphibolite is associated with mafic and ultramafic complexes metamorphosed to granulite facies. The common assemblage is composed of corundum, anorthite, gedrite, and margarite (Tenthorey et al. 1996, Nicollet 1986). Others minerals such as sapphirine (Forestier & Lasnier 1969, Herd et al. 1969, Nicollet 1986, Tenthorey et al. 1996), garnet (Nicollet 1986), spinel (Forestier & Lasnier 1969, Morishita & Kodera 1998), kornerupine (Rohtert & Ritchie 2006), phlogopite and zoisite (Dirlam et al. 1992) may be present (Fig. 2-40A, B). The initial basic composition of the protolith includes gabbro

FIG. 2-40. Mineralogical assemblages in corundum-bearing amphibolite and gabbro (after Lasnier 1977, in Giuliani et al.

2010). A, Ruby-bearing amphibolite from Chantel (Haut Allier, France). Successive transformation of spinel in sapphirine, and sapphirine in corundum. B, The ruby-bearing coronitic gabbro from Champtoceaux is amphibolitized. Hornblende (Hb) formed a corona around pargasite (Pr) and symplectites (S) with neoformed Na-rich plagioclase (Pl2) onto plagioclase of the first generation (Pl1). During the amphibolitization ruby is transformed in zoisite (Z). C, Ruby-bearing coronitic gabbro from Champtoceaux (Bretagne, France). The transformation of spinel (sp) in corundum. Spinel formed granules and 'vermicules' onto the ortho (opx)- and clino (cpx) pyroxenes which are transformed in pargasite (Pr). Ruby grows on spinel. Drawings from P. Brun (le Règne Minéral).

GIULIANI ET AL. �– GEM CORUNDUM

84

(Fig. 2-40C, Morishita & Kodera 1998), but also troctolite (Nicollet 1986, 1990; Tenthorey et al. 1996). The conditions of metamorphism are granulite facies; for example P = 9�–11.5 kbar and T = 750�–800°C for the ruby-bearing amphibolite from Vohibory in Madagascar (Nicollet 1986) and P = 7 to 10 kbar and T = 800�–850°C for those from Buck Creek in North Carolina (Tenthorey et al. 1996). At the Winza deposit in central Tanzania (Peretti et al. 2008, Schwarz et al. 2008), gem corundum crystals are hosted by 1800�–2000 Ma basement rocks of the Usagaran Belt (Sommer et al. 2005). The main rock types are migmatitic and well-foliated gneiss, indicative of upper amphibolite- to granulite-facies conditions. Mining of the primary deposits has revealed corundum crystals embedded within dark-colored amphibolite. The corundum is locally associated with areas of brown garnet ± feldspar, with accessory Cr-spinel, mica, kyanite, and allanite. Weathering of the primary deposits

resulted in an overlying soil horizon (eluvial deposit) that contained gem corundum. The ruby and sapphire crystals (Fig. 2-41A, B) are closely associated with cross-cutting dikes of a garnet-bearing pargasitite rock. Macroscopic observations clearly showed that, according to its mineralogy, the dike-like body was mafic in composition before it was metamorphosed. Possible lithologies for the protolith include high-alumina layered gabbro or leucogabbro. The central part of the dike is composed of garnet (pyrope�–almandine) + pargasite ± plagioclase ± corundum ± spinel ± apatite. Metamorphic conditions estimated by using garnet�–amphibole�–corundum equilibria showed that the metamorphic overprint occurred at 800 ±50°C and 8�–10 kbar (Schwarz et al. 2008). Spinel that overgrew or is included in the corundum is Fe- and Mg-rich, whereas spinel included in the host amphibole is Cr-rich with up to 32 wt.% Cr2O3. The presence of two chemically contrasting spinel

FIG. 2-41. The Winza corundum deposit in Tanzania (modified after Schwarz et al. 2008). A, The primary mineralization is

contained in amphibolite dykes. The hand specimen consists of fine-grained amphibolite (Am) margins flanking orangey brown garnetite (Gt) and corundum (Crd). B, Detail of the dike showing the association of sapphire (Sa) and ruby (Ru) in a same crystal. C, Fe2O3 versus Cr2O3 diagrams showing (on the left) the Winza ruby composition compared with those from Mogok and Mong Hsu in Myanmar (marble type) and from Thailand and Cambodia (basaltic type). On the right, chemical comparison with some East African ruby deposits such as Chimwadzulu (Malawi), Mangare (Kenya), and Songea (Tanzania).

GIULIANI ET AL. �– GEM CORUNDUM

85

compositions indicates that two different generations are present. The Cr-spinel is probably relict from the initial magmatic rock before the metamorphism. The Fe- and Mg-rich spinel included in the corundum crystallized during the metamorphic stage. Two types of ruby and sapphire crystals are distinguished by their crystal morphology (Schwarz et al. 2008): (type 1) prismatic-tabular-rhombo-hedral, and (type 2) dipyramidal. In general, medium-red and dark (orangey) red top-quality ruby are type 1 (rhombohedral). Pinkish and purplish red ruby, as well as pinkish, purplish, and bluish (commonly strongly color-zoned) sapphire crystals are type 2 with tabular (very rare) and prismatic forms, or dipyramidal. The inclusions in both types are pargasite, spinel, garnet, and rare apatite, xenotime, and zirconolite (Peretti et al. 2008). The chemical compositions of Winza ruby and sapphire (Fig. 2-41C) indicate moderate contents of Cr (0.1�–0.8 wt.% Cr2O3) and Fe (0.2�–0.8 wt.% Fe2O3), very low to low amounts of Ti (55�–192 ppm TiO2) and V (up to 164 ppm V2O3), and low to moderate Ga (64�–146 ppm Ga2O3). The relatively high Fe content of Winza ruby separates it from most other natural counterparts. The oxygen isotope values of Winza corundum defined a consistent and restricted 18O range of 4.7 ±0.15�‰ (n = 3) which fits with the worldwide range of ruby hosted in mafic and ultramafic rocks (0.25 to 6.8�‰, n = 21). The Winza ruby crystals are of metamorphic origin, with a high-alumina meta-(leuco)gabbro protolith for the host rock. The gem corundum deposits of southwest Greenland are another example of the ruby-hosted type in gabbroic rocks submitted to metamorphic metasomatism (Fagan 2012). The deposits are located within the Fiskenæsset Complex, a layered cumulate igneous complex that occurs in the lower zone of the Bjørnesund structural block (Windley & Smith 1971). It appears today as a sheet-like body, apparently concordant with the adjacent orthogneiss and amphibolite. The complex is mid-Archean (2.86 Ga), and individual layers in the Complex range in width from 2 km to less than a metre. The complex and the gemstone district are very large, extending over an area in excess of 400 km2 (Weston 2009). The intrusive suite comprises layers of gabbro, ultramafic rocks, leucogabbro, and calcic anorthosite. The corundum mineralization is connected with a specific stratigraphic horizon in the complex; this probably represents a unit with high Al content, low silica and a high fluid flux and

it lies between layers of ultramafic rock and leucocratic gabbro (Reggin & Chow 2011, Fagan 2012). In the early 1970s the Danish geological survey located Aappaluttoq and several other ruby showings while conducting regional scale geological mapping in the Fiskenæsset Complex. The Aappaluttoq deposit was subsequently �‘lost�’ in the archives until its rediscovery in 2005 by an exploration team from True North Gems Inc., the Canadian junior mining company who hold the current exploration rights to the deposits. To date the company has discovered a total of more 50 individual corundum occurrences in the gemstone district, three of these show significant potential for development as future gem corundum mines. At this time, the company is focussing on pre-mining development at the Aappaluttoq Deposit, but are expecting to drill and define the resources at the Siggartartulik and Kigutilik occurrences in the upcoming years. The geology of the Aappaluttoq deposit is dominated by a sequence of intrusive gabbro to leucogabbro with significant volumes of ultramafic rocks. This sequence is intruded into and is structurally juxtaposed against the felsic gneiss basement suite (Groves & Banfield 2009). The currently defined dimensions for the deposit are ~160 m long, ~100 m wide and 65 m deep; however corundum has been recovered in drill core below 200 m depth, and the geological structures remain open along strike. The main corundum bearing ore is composed of three main rock types (Plate 2C): sapphirine-gedrite, leucogabbro and a phlogopitite; the latter being the most important of these. The sapphirine-gedrite rocks at Aappaluttoq lie 30 km from the world type locality for the mineral sapphirine ([Mg,Al]8[Al,Si]6O20) located within the village of Qeqertarsuatsiat (formerly known as the village Fiskenaesset). The Fiskenæsset igneous Complex contains multiple areas rich in sapphirine-gedrite; many of these were described in detail by Herd et al. (1969). The presence of the sapphirine-gedrite rock at surface has been used by True North Gems Inc. as an exploration tool, as it acts as a key indicator for potential corundum mineralization. This rock only forms along the boundary between the ultramafic units and the Al-rich rocks of the gabbroic suite in fluid-metasomatized areas. It is believed that this unit represents a SiO2 sink between the ultramafic units and the gabbro, which depleted the availability of SiO2 in the metasomatic fluid, and relatively increased the amount of Al present.

GIULIANI ET AL. �– GEM CORUNDUM

86

The leucogabbro contains a small amount of gem corundum as ruby, but a large amount of pink sapphire. The gabbro is rich in Al, and is believed to be the unit responsible for releasing the Al to form the corundum. Its structural location, distal from the ultramafic, is thought to be responsible for the lack of ruby, as the Cr geochemical gradient was not sufficient to enrich the corundum within the gabbroic units sufficiently to create more than a pink coloration. The phlogopitite is the host for the majority of the ruby held within Aappaluttoq, and has little pink sapphire within it. This is due to the proximity of this unit to the ultramafic Cr source-rock and also reflects the relative mobility of Cr in a fluid-rich environment. This unit is a metasomatic product, and comprises around 90% phlogopite, 5% biotite, and 5% corundum (Plate 2C). The gemstones within the Greenland ruby deposits are highly abundant and of a good quality. In the primary deposits, there is a large variation in the quality of corundum material. It is estimated that over 85% of the corundum is non-gem (opaque) corundum, with around 5% classed as gem grade and 10% of cabochon quality (Mattinson 2007). The color of the stones range from a deep intense red, through to a very light pink, and occasionally white. At Aappaluttoq the majority of the stones are pink to dark red (Plate 2D); they have 0.26�–1.14 wt.% Cr2O3 and 0.21�–0.45 wt.% Fe2O3, and occasional elevated TiO2 (0.34 wt.%). The stones can grow to a substantial size (the largest recovered to date is a 440 ct opaque stone that was carved into the 303 ct Kitaa Ruby). The Aappaluttoq deposit is the world�’s first gem deposit to have its value calculated and defined according to Canadian Securities Regulations (NI 43-101). It alone represents a total of approximately 405 million ct to a depth of 65 m, giving it a minimum mine life of 10 years. Since the mineralization continues outside of this it would be reasonable to conclude that this resource outlines only a portion of the potential at this site. Together, Appaluttoq with other deposits such as Siggartartulik and Kigutilik, represent the three most promising gem concentrations in southwest Greenland; they also represent significant gem deposits that will be capable of producing sizable volumes of good ruby and sapphire gemstones for years to come. The Neoproterozoic metamorphic Mozambique Belt of Mozambique is subdivided in several complexes grouped in four geological entities which are from West to East (Fig. 2-42, Norconsult 2007a,

2007b, Boyd et al. 2010): (1) Ponta Messula complex, (2) Unango and Maruppa complexes, (3) Nampula complex, and (4) the east�–west Pan-African overthrusts (e.g., Nairoto, Meluco, Muaquia, M'sawize, Xixano, Lalamo, and Montepuez) which form the Cabo Delgado complex overlying the Marrupa complex. Recent discovery of ruby in Mozambique within the Cabo Delgado complex concerns the deposits of Ruambeze and M'sawize in the Niassa province, and the deposits of Namahumbire and Namahaca in the Delgado province (Fig. 2-40, Pardieu et al. 2009a, 2009b). The Ruambeze deposit (or Luambeze) is located between Marrupa and Meculo along the Luambeze River (no precise location of the deposit). It was discovered apparently in 1992 (Pardieu et al. 2009a) and it produced dark red to brown cabochon grade material. The deposit is probably located in the Marrupa complex formed of felsic orthogneiss associated with metasedimentary and migmatitic paragneiss. Two episodes of metamorphism affected the series in the amphibolite facies, and dated respectively at 953 Ma and 555 Ma (Norconsult, 2007a, 2007b). The M'sawize deposit (called also Niassa, 12°40'53"S 36°49'20"E) is located in the Niassa National Reserve, about 43 km southeast of M'sawize village (Pardieu et al. 2009a). The M'sawize complex is formed of orthogneiss (granitic to amphibolitic in composition), and paragneiss. The M'sawize ruby deposit is contained in metagabbro and gabbroic gneiss (Pardieu et al. 2009b). Ruby in situ is associated with actinolite, anorthite, scapolite, epidote, and diopside, as well as in eluvial ruby-rich soil located above the primary deposit (Pardieu et al. 2009b). Preliminary examination of ruby revealed the presence of many twinning planes with boehmite and silks formed by rutile-like needles. Chemical analyses of ruby indicate Fe and Cr contents between 0.28�–0.33 and 0.12�–0.48 wt.%, respectively, which fall in the range of the Winza ruby from Tanzania (Pardieu et al. 2009c). In 2009 around 10% of the gem material did not require any type of heat treatment while 85% was treated using lead glass and 5% by borax flux heat treatments (Pardieu et al. 2009a). The Namahaca (13°03'32"S, 39°11'08E) and Namahumbire (13°05'18"S, 39°20'35E) deposits (also called Montepuez), discovered in May 2009, are located at about 30 km from Montepuez in the direction of Pemba. The deposits are hosted in the Pan-African Montepuez overthrust. The Montepuez complex is made of Neoproterozoic orthogneiss

GIULIANI ET AL. �– GEM CORUNDUM

87

FIG. 2-42. Geological map of

northeast Mozambique (after Boyd et al. 2010) with the location of the M'sawize (Niassa), and Namahumbire and Nahamaca (Montepuez) ruby deposits (modified after Pardieu et al. 2009a, 2009b). The deposit of Ruambeze is not reported because the GPS location is not precisely described. The deposit may be located in the Marrupa complex or in the Xixano complex.

Ruby deposits: 1, M'sawize; 2, Montepuez.

(granitic to amphibolitic in composition), and paragneiss mainly formed by quartzite, meta-arkose, marble, quartz-feldspar gneiss, and biotite gneiss (Boyd et al. 2010). The regional deformation appears to be mainly ductile in nature, with large regional folding and localized tight isoclinal folds that make the highly covered sequences difficult to interpret. In general the rocks appear to have experienced regional amphibolite-facies meta-morphism. Paragneiss in the western part of the complex was dated to 942 ±14 Ma with a meta-morphic overprint at 599 ±10 Ma (Boyd et al. 2010). Chemostratigraphic dating of marble in the Complex suggests their deposition between 1100 and 1050 Ma. The ruby area exploited by the Montepuez Ruby Mining Ltd. hosts two distinct geological types of deposit (Pardieu & Chauviré 2013), alluvial and primary. The alluvial sedimentary sequence with a thickness from 0.5 m to over 10 m ranges in grain size from 10 cm cobbles through to fine quartz sand. The grains tend to be equidimensional with significant rounding along the grain edges. No significant quantities of lithic fragments or clay

fractions have been noted in this unit. The layer appears to be grain supported, with minor muddy matrix material and quartz grains in the 5�–10 mm size range making up the bulk of the unit. This unit has extensive artisanal workings cut through it, and a minor amount of corundum does occur within its upper layer: however most of this material appears to be non-gem quality. The primary deposit appears to be a highly metamorphosed or metasomatized rock that has been completely altered from its original primary nature (Plate 2E). The resulting clay appears to be a complex Cr-rich smectite group clay, with at least 5% of the layer made up of corundum, altered amphibole, and yellowish mica (Pardieu et al. 2009b). Large gemmy stones recovered from the artisanal workings have been noted in the Montepuez village (Plate 2F), and these pieces can be of significant gem quality with good color and shape (around 5% of the stones). The majority of the production has a great color but with too many fissures; however, 70% of the production can be improved by Pb glass heat treatment and 25% (such as the milky stones) by borax flux heat treatment (Pardieu et al. 2009a). To date no mining

GIULIANI ET AL. �– GEM CORUNDUM

88

operations have commenced; however commercial bulk sample operations are under way at the Namahumbire site by Gemfields Inc. of London (www.gemfields.co.uk); no record of geological work can be found for the Namahaca Deposit. The total thickness of the corundum-bearing layers, and thus the overall resource potential of the area, is unknown at this time. Metamorphic deposits in ultramafic rocks: the ultramafic-hosted ruby deposits of South Africa, Zimbabwe, and New Zealand have been the subject of several detailed studies. The corundum-fuchsite rocks from South Africa occur in eastern Transvaal (Hall 1923) and in the Barberton greenstone belt (Anhaesseur 1978). In Zimbabwe, the 'boulder corundum' supplied four-fifths of the world production of corundum in the 1970s (Morrison 1972). The ruby-fuchsite rocks located at O'Briens in the Archean greenstone belt of Harare occur in metavolcanic rocks, ferruginous quartzite, banded iron formation, siliceous limestone, and schist. They form lenses in ultramafic rocks (Schreyer et al. 1981, Kerrich et al. 1987). The lenses contain more than 89 wt.% of Al2O3, and Cr2O3 up to 2.8 wt.% (Schreyer et al. 1981). These rocks are so-called 'verdites'; generally the corundum is not of gem quality, but instead is used for carving. In Zimbabwe, the verdites contain andalusite, chlorite, and associations of margarite, tourmaline, diaspore, rutile, and gersdorfitte (NiAsS), and native Bi. In Transvaal, the verdites contain biotite, plagioclase, kyanite, and rutile with exsolution of Cr�–Fe�–Al oxides. The verdites are enriched in Al, Cr, B, V, and As (Schreyer et al. 1981). Schreyer et al. (1981) argued that the verdites from Zimbabwe have been metamorphosed at temperatures around 400°C and pressures lower than 3.5 kbar whereas the verdites from Transvaal at temperatures of 600°C and pressures higher than 5 kbar. A debate on the origin of these rocks ensued (Kerrich et al. 1987, Schreyer 1988). Schreyer et al. (1981) thought that the rocks formed by exhalative alteration of komatiite before metamorphism. The elements B, K, Rb, As, Sb, Bi, and Te were deposited from the solutions, while Al, Cr, Ni, and V were concentrated as immobile remainders from the original rock. The mineralogy of the alteration products was governed by aluminous sulfide minerals such as alunite, which under regional metamorphism formed fuchsite and andalusite (Schreyer et al. 1981). Kerrich et al. (1987) proposed a hydrothermal model for the formation of

corundum and fuchsite in which the minerals are sequential alteration products of ultramafic rocks by high temperature, low pH hydrothermal solutions related to a deep level acid epithermal system. Grapes & Palmer (1996) proposed that the fuchsite, margarite, tourmaline, and ruby-bearing boulders from Westland in New Zealand formed by metasomatism of quartzofeldspathic enclaves located in serpentinite (Fig. 2-43). This rock so called 'goodletite' from Hokitika crystallized at T ~ 450 ±20°C and P ~ 5�–6 kbar, and the ruby contains between 0.5 and 13.0 wt.% Cr2O3. GEOLOGY AND GENESIS OF SECONDARY CORUNDUM DEPOSITS Placer deposits are of major commercial importance and provide most of the gem corundum. Corundum from primary deposits is liberated from their parent rocks by weathering, transported by

FIG. 2-43. The gem corundum deposit of Westland in

New Zealand (after Grapes & Palmer 1996). A, Section of a ruby-bearing pebble. B, Chemical distribution of the ruby (filled circles) and sapphires (unfilled circles) in the (Cr3+)�–(Fe3+)�–(Ti4++Fe2+) triangular diagram.

GIULIANI ET AL. �– GEM CORUNDUM

89

streams and rivers, and concentrated in hydro-graphic traps. These deposits are considered sedimentary and corundum is either found as pebbles in gravel lenses or in lithified conglomerate (paleo-placer). Two types of concentrations are recognized for present day placers (Fig. 2-44): (1) eluvial and colluvial deposits: residual eluvial concentrations are the product of chemical weathering of parent rocks. The altered and decomposed primary deposit settled only vertically. Eluvial-colluvial deposits are on the slopes and in karst cavities (marble type). Colluvial deposits correspond to decomposed primary deposits which have moved vertically and laterally downslope as the hillside eroded away; (2) alluvial deposits result from the erosion of the host rock and transport of corundum by water in streams and rivers. The concentration occurs where water velocity drops at a slope change in the hydrographic profile of the river, at the base of a waterfall, broad gullies, debris cones, meanders, and inflowing streams. The placer deposits of barrier sea beaches are rare and cited examples include the blue sapphire occurrence from the Vendée region in France (Goujou 2002), and the BGY-sapphire deposits from Madagascar at Befotaka in Nosy-Be Island, and in the Andovokonko area in the Ambato Peninsula (Ramdohr & Milisenda 2004). Paleo-placer deposits are characterized by different mineral phases which are cemented in a carbonate or silica-rich matrix. Paleo-placers of corundum are known from alkali basalt deposits in Madagascar: (1) in the deposits of the Antsiranana province, the paleo-placer is formed by a carbonate karst breccia contained in karst cavities in Jurassic limestone (Schwarz et al. 2000, Giuliani et al. 2007b); (2) in the Ilakaka area, gem corundum is found in three gravel levels from alluvial terraces deposited on the Isalo sandstone. The terraces are

weakly consolidated, but correspond to paleo-placers (Garnier et al. 2004a). The erosion and remobilization of these sedimentary rocks by rivers results in the formation of newer present day placers. Climate is a major factor for the formation of secondary deposits. In tropical areas rocks are exposed to meteoric alteration resulting in an assemblage of clay minerals, Fe and Mn oxides, and other supergene phases. Corundum and zircon are resistant minerals and are found in soils, laterite, and in gravelly levels overlying bedrock. In Myanmar, these gem-bearing levels enriched in pebbles, sand, silt and clay, and iron oxides, are called "byon" (Kane & Kammerling 1992). Other examples of such deposits are found in central and southeast Asia where the effect of the monsoon in low altitude environments is an effective mechanism of erosion (Hughes 1997). In high-altitude areas such as in India, Nepal, Pakistan, and Azad-Kashmir the weathering alteration is mechanical and corundum is found in situ in boulders accumulated on the lower part of the slopes (e.g., 'old mine' at the Sumjam sapphire deposit, Kashmir). At Nangimali, a ruby boulder was discovered in a creek in 1979, but due to difficult access and the short working season, the primary occurrence was not found until 1988 (Malik 1994). Australia is an example of sapphire and ruby occurring in alkali basalt that is subject to a tropical climate. The resulting alluvial deposits are in eastern Australia, primarily central Queensland and northern New South Wales (Sutherland & Abduryim 2009, Abduryim et al. 2012). The main ruby production is located in Barrington Tops, Yarrowitch, and Tumbarumba. The sapphire-bearing basaltic sediment, locally called 'wash', occurs in layers 1 to 3 m thick, underneath dark clayey soil some 1 to 3 m below the surface (Abduriyim et al. 2012). In the Inverell district at the Mary Ann Gully mine, the brown

FIG. 2-44. Main types of mechanical and physical hydrographic traps for corundum in placer deposits

(modified after Hughes 1997).

GIULIANI ET AL. �– GEM CORUNDUM

90

basaltic sedimentary soils are extracted from an area of 10 to 40 km2. The basaltic flows came from a large eroded volcano and the fine-grained basalt contains large quantities of corundum, black spinel, zircon, ilmenite, hematite, olivine, pyroxene, feldspar, and amphibole. Many methods are used for the exploitation of placers. In developing countries, the discovery of a new deposit provokes the migration of a huge population of poor people working with rudimentary tools. These independent miners, called garimpeiros in Brazil and guaqueiros in Colombia, are equipped with shovels and pickaxes to dig pits, often tens of metres deep, to reach the mineralized zone. The gem gravels are lifted by bags and buckets with the help of a pulley, washed in sieves at the river, and finally the gravels are hand-picked. Mechanization permits the intensive exploitation of placers, especially those from southeast Asia, Australia, and North America, more rarely in Africa and Madagascar. Barren levels are displaced by mechanical shovel and bulldozer. The gem gravels are brought back to the washing plant either by trucks or pumps after water cannon extraction. A classical washing plant contains one or several trommels, which facilitate a cut-off of the pebbles according to their size, and jigs, which sort the pebbles by their density. Finally, the jigs are cleared out and the pebbles sorted by hand. The exploitation of placer deposits in deep rivers is aided by mining dredges. This type of exploitation is active in the Missouri River in Montana for the recovery of sapphires (Hughes 1997). The washing plant is directly fixed to the barge which dredges the river bottom. Important secondary gem corundum deposits are found in Australia (Queensland and New South Wales), Madagascar (Ilakaka, Ambondromifehy, Vatomandry, Didy), Tanzania (Tunduru, Songea, Umba), Myanmar (Mogok and Mong Hsu districts), Vietnam (Luc Yen, Yen Bai, Da Lat, Ban Me Thuot), the United States (Montana), and Sri Lanka (Ratnapura, Elahera, Kataragama). The historical, geological, and mineralogical features of these deposits are described in details by Heilmann & Henn (1986), Silva & Siriwardena (1988), Hughes (1997), Sutherland et al. (1998a, 1998b), Schwarz et al. (2000), Garnier et al. (2004a), Pham et al. 2004, Themelis (2008), Pardieu et al. (2009a, 2012), Hughes et al. (2011), Abduriyim et al. (2012), Coldham et al. (2012), Pardieu (2012), Pardieu & Chauviré (2013), and Peretti & Hahn (2013). Generally, placers contain gem corundum originating from multiple sources. The geological

origins of these gems are thus difficult to assess, although the problem can be approached by the study of the solid and fluid inclusions trapped in the individual crystals (Hughes 1997), and their chemical (Muhlmeister et al. 1998) and isotopic composition (Giuliani et al. 2005; 2007a; Fig 2-45). Placer deposits in marble environments: the case of Vietnam: the placers formed in marble environ-ments such as those of Luc Yen and Yen Bai in Vietnam, and of Mogok and Mong Hsu in Myan-mar, are closely related to the formation of karst and chemical weathering of marble and associated rocks i.e., schist, gneiss and granitoid rocks. The corundum is trapped in eluvial, colluvial, alluvial, and fracture-filling and cave deposits (Fig. 2-46). In Vietnam, the placers are proximal to the main primary deposits of ruby and colored sapphires in marble. They are located in the Luc Yen and Yen Bai provinces in the Red River shear zone area and in the Nghe An province in the Bu Kang metamorphic dome in central Vietnam (Pham et al. 2004). They are in Luc Yen, An Phu, Khoang Thong, and Nuoc Ngap (Luc Yen district), in Truc Lau, Yen Thai, Tan Huong, Tan Dong, Hoa Cuong, and Tran Yen (Yen Bai district), and Quy Hop and Quy Chau (Nghe An district). The oxygen isotopic study of the ruby and sapphire crystals shows that the majority of these corundum have a 18O range between 15.5 and 22.3�‰ (Fig. 2-47), that overlaps the worldwide oxygen isotopic range defined for ruby hosted in marble (Giuliani et al. 2005; Fig. 2-40). Four sapphires respectively of blue, grey and yellow color are in the field of desilicated and skarn vein type in marble. Placer deposits in alkali basalt fields: the case of the French Massif Central: gem blue-green-yellow sapphire and ruby crystals are mostly recovered from alluvial deposits in continental basaltic fields (Hughes 1997, Sutherland 1996, Abduriyim et al. 2012). The large corundum-bearing regions are located in Australia, the United States, Laos, Thailand, Cambodia and Madagascar. In France, alluvial sapphire is located in the French Massif Central and has been known since the Middle Ages and exploited up to the 19th century (Forestier 1993). The sapphire occurrences are located within the volcanic areas of the Chaîne de la Sioule, Chaîne des Puys, Mont Dore, Cantal, Devès and Velay (Lacroix 1901, Gaillou et al. 2010). The sapphires crystals are either transparent to dominantly pastel blue, pastel lilac and colorless, or

GIULIANI ET AL. �– GEM CORUNDUM

91

FIG. 2-45. Oxygen isotopic ranges of corundum from the giant placer of Ilakaka (Madagascar) with a special focus on the

origin of the blue sapphires. These sapphires are separated in two distinct groups: one group dominated by rutile needle inclusions (reported as needle in the figure), the other by bands and growth structures (reported as band in the figure). The oxygen isotopic composition of blue sapphire from Ilakaka are compared with the oxygen isotopic ranges defined for corundum types of deposit worldwide (Giuliani et al. 2005, 2007a, 2009, 2012a; Yui et al. 2003, 2006, 2008) and for colored sapphires of primary deposits from Madagascar. The oxygen isotopic data are reported in the conventional delta notation ( 18O, expressed in per mil, �‰) relative to V�–SMOW (Vienna Standard Mean Ocean Water). Color in diamonds represent color for ruby (in red) and colored sapphires (others). White diamonds represent colorless sapphires.

blue, deep blue, greenish, greyish, yellowish, pinkish, and bronze colored but milky. The O isotope composition of the sapphires ranges between 4.4 and 13.9�‰ (Fig. 2-48). Two distinct groups have been defined (Giuliani et al. 2009): (1) the first with a restricted isotopic range between 4.4 and 6.8�‰ (n = 22; mean 18O = 5.6 ±0.7�‰) falls within the worldwide range defined for blue-green- yellow sapphires related to basaltic gem fields (3.0 < 18O < 8.2�‰, n = 150), and overlaps the range defined for magmatic sapphires in syenite (4.4 <

18O < 8.3�‰, n = 29; Fig 2-45). The presence of inclusions of columbite-group minerals, pyrochlore, Nb-bearing rutile, and thorite in these sapphires

provides an additional argument for a magmatic origin. (2) A second group, with an isotopic range between 7.6 and 13.9�‰ (n = 9), suggests a metamorphic sapphire source such as biotite schist in gneiss or skarn. They are metamorphic sapphires from granulite which were transported to the surface by the basaltic magma. Placer deposits in sedimentary fields: the case of Ilakaka (Madagascar): the Ilakaka mining district is located in the Isalo massif, between the cities of Sakaraha and Ilakaka (Fig. 2-32). Other districts are found north of Ilakaka and near Bezaha, 120 km southwest of Ilakaka. The alluvial giant deposit was

GIULIANI ET AL. �– GEM CORUNDUM

92

FIG. 2-46. Main types of gem corundum deposits in the Mogok area (modified after Themelis 2008). The model represents the

deposits of Dattaw (left) and Thurein-taung (right). The primary deposits are metamorphic or magmatic. The secondary deposits are residual, colluvial, alluvial, fracture-filling and cave deposits. Drawing courtesy of T. Themelis.

discovered in late 1998 on the border of Route National 7, and the rush of locals and immigrants resulted in the discovery of several mining zones between Ilakaka and Sakaraha, including Sakalama, Ampasimamitaka, Vohimena, Bekily, and Manombo Vaovao. The deposits are generally exploited by illegal miners using 1 m diameter shafts reaching up to 20 m deep with windlass systems, washing gravels in pans and removing the stones from the sieves by hand. The deposits are sometimes exploited in open pits following benches to reach the mineralized zones.

The deposits produce very fine blue, pink, blue-violet, violet, purple, orange, yellow, and translucent sapphire crystals along with zircon, alexandrite, topaz, garnet, spinel, andalusite, and tourmaline. In 2002, test mining analysis by the Gem Mining Resources company during 38 days of production resulted in approximately 43 kg of gemstones, comprising 4% semi-precious stones, volcanic glasses, and ruby, and 96% sapphire comprising 58% pink sapphire, 30% blue sapphire, and 8% other colored sapphire (e.g., yellow, orange, 'padparadscha', green). The operators exploited the alluvial terraces of the Ilakaka and Benahy rivers, which lie on the Isalo sandstone. The terraces resulted from the erosion and stripping of the

Triassic sandstone. The deposit is not well consolidated and is composed of quartziferous sands which contain pebbles of ferruginous laterite, rounded blocks of Isalo sandstone and quartz, and quartzite and schist pebbles (Garnier et al. 2004a). On the Benahy River, two levels of gemmiferous terraces locally called "lalambato" are found (Fig. 2-49). The bottom terrace is the richest in gemstones, with concentrations reaching up to 5 to 7 g/m3 for the mechanized exploitations. The concentration in some terraces exploited by washing and hand recovery reached about 10 g/m3. They correspond to deep potholes or sinuous meanders which are not accessible for mechanical exploitation. River sands, sampled in the stream bed, contain between 0.2 and 2.1 g/m3 of sapphire. This alluvial concentration results from the erosion of old terraces.

The placers of Ilakaka represent accumulation of ruby and colored sapphire from different origins (Fig. 2-45). Microscopic examination of 30 ruby and 58 blue sapphire crystals has shown that all rubies have the same gemological features, i.e., growth, twinning, color bands and solid inclusions, but that blue sapphires defined two different mineralogical groups.

GIULIANI ET AL. �– GEM CORUNDUM

93

FIG. 2-47. Oxygen isotopic composition of ruby and

sapphire originating from placer deposits of the Luc Yen and Yen Bai mining district (north Vietnam). The oxygen isotopic range of these gem corundum samples overlaps that of ruby and sapphire hosted in marble from the same area (Giuliani et al. 2007a). The data are reported in the conventional delta notation ( 18O, expressed in per mil, �‰) relative to SMOW (Vienna Standard Mean Ocean Water).

Ruby crystals are within the same 18O range, between 2.6 and 3.8�‰ ( 18Omean = 3.5 ±0.5�‰, n = 4). The isotopic range corresponds to that defined for primary ruby from mafic and ultramafic rocks from southern Madagascar (range 2.5 to 6.8�‰; Giuliani et al. 2007b). The Phanerozoic sedimentary sequences and the Isalo Group underlie the northern part of the Vohibory unit, where intercalations or complexes of mafic and ultramafic rocks hosting ruby and pink sapphire are common in amphibolitic and aluminous gneiss (Rakotondrazafy et al. 2008). The erosion of the Vohibory unit during the Late Carboniferous to Middle Jurassic contributed to the concentration of ruby in the basin. A proximal

source for ruby in Ilakaka from the Vohibory region is likely.

Inclusions and UV�–Visible spectroscopy allowed for the differentiation of two distinct groups in the blue sapphire crystals from Ilakaka (Dunaigre & Giuliani, unpublished data). One group dominated by rutile needle inclusions, the other by bands and growth structures. Two samples from the "band" group with 18O values of respectively 7.4 and 10.1�‰, and two from the "needle" group with

18O values of respectively 9.1 and 13.6�‰, are compared to previous data obtained on blue sapphire ( 18O value of 3.8�‰; Giuliani et al. 2007b), and considering the different types of primary corundum deposit from southern Madagascar (Rakotondrazafy et al. 2008). The 18O value of 3.8�‰ suggests two possible geological origins for sapphire (Fig 2-45): (1) BGY sapphire in alkali basalt; and (2) biotite schist in gneiss. The chemical composition of the sapphire has shown that it was incompatible with a basaltic origin, a type which has not been reported in southern Madagascar. Therefore the second hypothesis is highly probable as blue sapphire in biotite schist has been described in the Ionaivo deposit (with 18O = 3.3�‰; Uher et al. 2012) located south of Sahambano deposit (Fig. 2-32). The band group contains inclusions of carbonate, phlogopite, graphite, and garnet. The

18O value of 7.4 and 9.1�‰ suggests two possible geological origins for sapphire (Fig. 2-45): biotite schist deposit which results from the metasomatism of feldspathic granulitic gneiss ( 18O between 3.3 and 9.0�‰ for Ionaivo, Sahambano, Ambinda sud, and Zazafotsy deposits; see the section above: Metamorphic deposits, p. 58), and skarn in calc-silicate rocks and marble ( 18O between 7.7 and 10.7�‰ for Andranondambo deposit). Carbonate and phlogopite are found in sapphire from the Andronandambo skarn (Schwarz et al. 1996) but graphite and garnet have never been described. Nevertheless, graphite is a mineral which is often found in marble and gneiss from the Tranomaro Formation which hosts the sapphire and U�–Th mineralization (Pili et al. 1997; Rakotondrazafy et al. 1996). Phlogopite, calcite and garnet are described in the sapphire from the biotite schist but graphite is lacking (Ralantoarison et al. 2006). The needle-dominant group contains rutile and spinel which are also characteristic inclusions of sapphire hosted in schist-type deposits such as Sahambano and Zazafotsy. Besides, the 18O of 9.1�‰ for one sapphire fits with the isotopic range

GIULIANI ET AL. �– GEM CORUNDUM

94

FIG. 2-48. Oxygen isotope values of colored sapphires found in placers within the volcanic fields of the French Massif Central

(Giuliani et al. 2009), with reference to the worldwide oxygen isotopic database of corundum reported by Giuliani et al. (2005, 2007a, 2009). The data are reported in the conventional delta notation ( 18O, expressed in per mil, �‰) relative to V�–SMOW (Vienna Standard Mean Ocean Water).

defined for the Zazafotsy deposit (Fig 2-45). The other crystal with 18O of 13.6�‰ is close to the isotopic range defined for blue to colorless gem sapphires associated with skarn veins in marble from Andronandambo ( 18O between 14.0 to 15.6�‰). Nevertheless, its origin cannot be accurately determined. The oxygen isotopic composition of blue sapphire from Ilakaka deposits shows the complexity of the determination of origin for corundum in placer. Nevertheless, the two gemological groups of blue sapphire from Ilakaka are metamorphic in origin. This result fits with the chemical fingerprinting and the types of solid inclusions found in the sapphire crystals. But at least three geological sources have probably contributed to the blue sapphire concentration in the Ilakaka deposit (biotite schist, skarn in marble or calc-silicate rocks, and skarn vein in marble). However there is currently no correlation between the gemological groups and the isotopic composition and origin of the sapphire crystals.

SUMMARY The primary deposits can be divided into two groups, depending on their geological setting, magmatic versus metamorphic (Fig. 2-12A, B). Corundum in magmatic deposits is found in plutonic and volcanic rocks. In plutonic rocks, corundum is associated with rocks deficient in silica and their pegmatites, especially syenite and nepheline syenite. Corundum is found also as an accessory mineral in porphyry copper and kimberlite deposits. In volcanic rocks, sapphire and ruby are found in continental alkali basalt extrusions. Corundum is found as xenocrysts in lava flows and plugs of subalkaline olivine basalt, high alumina alkali basalt and basanite. The sapphire is blue-green-yellow and the deposits only have economic importance because advanced weathering in tropical regions concentrates sapphire in eluvial and especially large alluvial placers. Sapphires also occur in alkaline basic lamprophyre dikes of biotite monchiquite. The ruby is metamorphic in origin and occurs as xenocrysts or megacrysts in xenoliths in alkali basalt.

GIULIANI ET AL. �– GEM CORUNDUM

95

FIG. 2-49. Lithological section of the gem corundum terraces and aspect of the pebble levels exploited in the Vohimena Talo

placer, Ilakaka district, Madagascar. L: lutite, A: arenite, R: rudite (Garnier et al. 2004a).

The metamorphic deposits are hosted by different kinds of rocks of variable grade of metamorphism. Among these deposits, marble represents the most important source of ruby, such as the famous pigeon blood-colored rubies from Mogok in Myanmar. They are found from Afghanistan in central Asia to northern Vietnam in southeast Asia. The other metamorphic deposits are hosted by gneiss, anatexite, granulite (gneiss, metashale, charnockite, cordieritite), and mafic to ultramafic rocks. New discoveries of ruby hosted in amphibolite (deposits of Greenland and Mozambique) will supply the demand on the market in terms of quantity and quality in the near future. Finally, the formation of ruby and sapphire deposits

is also linked to the intrusion of pegmatites and/or granitoid bodies into marble or mafic to ultramafic rocks, like the famous Kashmir sapphire deposits of Sumjam. Gem corundum forms by fluid�–rock interaction and metasomatism of the host rock (desilicated pegmatite and skarn). The secondary deposits are sedimentary and of detrital origin. Several types are distinguished: colluvial, eluvial, deluvial and alluvial placers, and beach ridges. The minerals contained in eluvial placer deposits are little transported and the parental rocks of the gemstones can thus be constrained. In the other cases, the placer deposits can form by the weathering of paleoplacers and transport of the minerals. In some cases, gemstones coming from

GIULIANI ET AL. �– GEM CORUNDUM

96

several types of deposits are mixed in a single basin (Tunduru, Songea, Ilakaka, Didy, Ratnapura, Elahera, etc.). The study of gemological characteristics, solid and fluid inclusions, as well as the geochemical and isotopic signatures of these gemstones are useful techniques for determining their geological origins. ACKNOWLEDGEMENTS The authors wish to thank the Mineralogical Association of Canada, GET UMR 5563 CNRS-IRD-CNES-UPS in Toulouse, CRPG UMR 7358 CNRS-UL and GéoRessources UMR 7359 CNRS-UL in Nancy, Geological Survey of Pakistan, Institute of Geological Sciences in Hanoi, University of Antananarivo, University of Kathmandu, SUERC Glasgow, School of Earth Sciences Leeds, University of Montpellier 2, MNHN-UMR 7160 Paris, Ministère des Mines de Madagascar, and T. Themelis for the Mogok illustration. The authors thank very much Dr. David Turner for his very carefully and constructive review, and Dr. Rob Raeside and Dr. Lee Groat for final reviews and editorial efforts. REFERENCES ABDURIYIM, A. & KITAWAKI, H. (2006):

Applications of laser ablation-inductively coupled plasma-mass spectrometry (LA�–ICP�–MS) to gemology. Gems & Gemology 42, 98�–118.

ABDURIYIM, A., SUTHERLAND, F.L. & COLDHAM, T. (2012): Past, present and future of Australian gem corundum. Austr. Gemmol. 24, 234�–242.

ABOOSALLY, S. (1999): Update on production in Pakistan, Afghanistan. Jewellery News Asia 1, 60�–64.

ACKERMAND, D., WINDLEY, B.F. & RAZAFINIPARANY, A. (1989): The Precambrian mobile belt of southern Madagascar. In Evolution of Metamorphic Belts (J.S. Daly et al. eds.). Geol. Soc. Spec. Pub. 43, 293�–296.

AKININ, V.V., VYSTOSTSKIY, S.V., MAZDAB, F.K, MILLER, E. & WOODEN, J.L. (2004): SHRIMP dating of zircon from the Podgelbanochny alkali basalt volcano in Primorye, Russian Far-East: application to genesis of megacrysts. In Khanchul, A.I., Gonevchuk, G.A., Mitrokhin, A.N., Simanenko, L.F., Cook, N.J., Seltmann, R. (Eds), Metallogeny of the Pacific Northwest: tectonics, magmatism and metallogeny of active continental margins. Dal'naka, Vladivostok, Russia, 323�–326.

ALTHERR, R., OKRUSCH, M. & BANK, H. (1982): Corundum- and kyanite-bearing anatexites from the Precambrian of Tanzania. Lithos 15, 191�–197.

ANDREOLI, M.A.G. (1984): Petrochemistry, tectonic evolution and metasomatic mineralizations of Mozambique belt granulites from Malawi and Tete (Mozambique). Precamb. Res. 25, 161�–186.

ANDRIAMAMONJY, S.A. (2006): Les corindons associés aux roches métamorphiques du Sud ouest de Madagascar: le gisement de saphir de Zazafotsy. Mémoire de DEA, Université d'Antananarivo, Madagascar.

ANDRIAMAMONJY, S.A. (2010): Importance des fluides sur la métallogénie des gisements de saphir et rubis dans le domaine granulitique de haute température du Sud de Madagascar: cas de Zazafotsy et d'Ambatomena. Mémoire de Doctorat, Université d�’Antananarivo, Mada-gascar.

ANDRIAMAMONJY, A., GIULIANI, G., OHNENSTETTER, D., RAKOTONDRAZAFY, A.F.M., RALANTOARISON, T., RANATSENHO, M.M., RAKOTOSAMIZANANY, S. & FALLICK, A.E. (2013): Les corindons d�’âge néoprotérozoïque de Madagascar. le Règne Minéral, Cahier 2, "Les Gemmes du Gondwana", 109�–117.

ANDRIAMAROFAHATRA, J. & DE LA BOISSE, H. (1986): Premières datations sur zircon du métamorphisme granulitique du sud-est de Madagascar. 11ème Réunion des Sciences de la Terre, Clermont-Ferrand, Résumés.

ANHAESSEUR, C.R. (1978): On an Archean marundite occurrence (corundum margarite rock) in the Barberton Mountain Land, Eastern Transvaal. Trans. Geol. Soc. Afr. 81, 211�–218.

ANTHONY, J.W., BIDEAUX, R.A., BLADH, K.W. & NICHOLS, M.C. (1997): Handbook of Mineralogy, vol. III, Halides, Hydroxides, Oxides. Mineral Data Publishing Tucson, Arizona, U.S.A.

ASPEN, P., UPTON, B.G.J. & DICKIN, A.P. (1990): Anorthoclase, sanidine and associated megacrysts in Scottish alkali basalts: high-pressure syenitic debris from upper mantle sources? Eur. J. Mineral. 2, 503�–517.

ATKINSON, D. & KOTHAVALA, R.Z. (1983): Kashmir sapphire. Gems & Gemology 19, 64�–76.

BAROT, N. & HARDING, RR. (1994): Pink corundum from Kitui, Kenya. J. Gemmol. 24, 165�–172.

BAROT, N., FLAMNIN, A., GRAZIANI, G. & GÜBELIN, E.J. (1989): Star sapphire in Kenya. J. Gemmol.

GIULIANI ET AL. �– GEM CORUNDUM

97

21, 467�–473. BARR, S.M. & MACDONALD, A.S. (1979):

Palaeomagnetism, age and geochemistry of the Denchai basalt, Northern Thailand. Earth Planet. Sci. Lett. 46, 13�–124.

BARR, S.M. & MACDONALD, A.S. (1981): Geochemistry and geochronology of late Cenozoic basalts of southeast Asia. Geol. Soc. Am. Bull. 92, 1069�–1142.

BARTH, T. (1928): Zur genesis der pegmatite im Urgebirge in sudlichen Norwegen. Neues Jarb. Mineral. 58, 385�–432.

BASSETT, A.M. (1984): Rubies in the Himalayas of Nepal. Report submitted to the Nepal Department of Mines and Geology, Katmandu, Nepal.

BERG, R.B. (2002): Probable sources of alluvial sapphires in western Montana. Geological Society of America, Rocky Mountain section. 54th Annual Meeting, Session 20, Gemstone and semiprecious minerals and host rocks in the Western United States, May 7�–9.

BERTRAND, G. & RANGIN, C. (2003): Cenozoic metamorphism along the Shan scarp (Myanmar): evidences for ductile shear along the Sagaing fault or the northward migration of the eastern Himalayan syntaxis. Geophys. Res. Lett. 26, 915�–918.

BESAIRIE, H. (1967): The Precambrian of Madagascar. In The Precambrian (S. Rankama, ed.). Interscience, London, UK, 133�–142.

BINDEMAN, I.N. & SEREBRYAKOV, N.S. (2011): Geology, petrology and O and H isotope geochemistry of remarkably 18O depleted Paleoproterozoic rocks of the Belmorian belt, Karelia, Russia, attributed to global glaciation 2.4 Ga. Earth Planet. Sci. Lett. 306, 163�–174.

BINDEMAN, I.N., SCHMITT, A.K. & EVANS, D.A.D. (2010): Origin of the lowest-known 18O silicate rock on Earth in Paleoproterozoic Karelian rift. Geology 38, 631�–634.

BOAKA À KOUL, M.L., YONGUE-FOUATEU, R. & GIULIANI G. (2011): Mineralogical features and geological origin of the sapphires from the Mayo Kewol paleoplacer in Admawa region (North-Cameroon). In Advances in Materials Science Research (M.C. Whythers, ed.), Nova Science Publishers, 125�–142.

BOTRILL, R.S. (1998): A corundum-quartz assemblage in altered volcanic rocks, Bond Range, Tasmania. Mineral. Mag. 62, 325�–332.

BOULANGER, J. (1959): Les anorthosites de Madagascar. Imprimerie L. Jean.

BOULVAIS, P., FOURCADE, S., GRUAU, G., MOINE, B. & CUNEY, M. (1998): Persistence of pre-metamorphic C and O isotopic signatures in marbles subject to Pan-African granulite-facies metamorphism and U�–Th mineralization (Tranomaro, Southeast Madagascar). Chem. Geol. 150, 247�–262.

BOWERSOX, G.W. & CHAMBERLIN, B.E. (1995): Gemstones of Afghanistan. Geoscience Press, Tucson, U.S.A.

BOWERSOX, G.W., FOORD, E.E., LAURS, B.M., SHIGLEY, J.E. & SMITH, C.P. (2000): Ruby and sapphire from Jegdalek, Afghanistan. Gems & Gemology 36, 110�–126.

BOWLES, J.F.W. (2011): Corundum �–Al2O3. In Rock-Forming Minerals, 2nd edition, Vol. 5A, Non-silicates: Oxides, Hydroxides and Sulphides. Bowles, J.F.W., Howie, R.A., Vaughan, D.J. & Zussman, J. (Eds). The Geological Society, 47�–91.

BOYD, R., NORDGULEN, O., THOMAS, R.J, BINGEN, B., BJERKGARD, T., GRENNE, T., HENDERSON, I. MELEZHIK, V.A, OFTEN, M., SANDSTAD, J.S., SOLLI, A., TVETEN, E., VIOLA, G., KEY, R.M., SMITH, R.A., GONAZALEZ, E., HOLLICK, L.J., JACOBS, J., JAMAL, D., MOTUZA, G., BAURE, W., DAUDI, E., FEITIO, P., MANHICA, V., MONIZ, A. & ROSSE, D. (2010): The geology and geochemistry of the East African Orogen in Northeastern Mozambique. S. Afr. J. Geol. 113, 87�–129.

BRARD, C.P. (1805): Manuel du minéralogiste et du géologue voyageur. de l'Imprimerie de Feuguerry, Paris.

BRISSON, M.J. (1787): Pesanteur spécifique des corps. Paris, L'imprimerie royale, France.

BROWNLOW, A.H. & KOMOROWSKI, J.C. (1988): Geology and origin of the Yogo sapphire deposit, Montana. Econ. Geol. 83, 875�–880.

CADE, A. & GROAT, L.A. (2006): Garnet inclusions in Yogo sapphires. Gems & Gemology 42, 106.

CARBONEL, J. & ROBIN, C. (1972): Les zircons gemmes dans les roches ignées basiques. Le gisement d�’Espaly (Haute Loire, France). Implications génétiques. Rev. Géograph. Phys. Géol. Dynam. 14, 159�–170.

CARBONEL, J., DUPLAIX, S. & SELO, M. (1973): Géochronologie par traces de fission des zircons et par K�–Ar des andésites basaltiques d�’Espaly

GIULIANI ET AL. �– GEM CORUNDUM

98

(Haute Loire). Position du Villafranchien régional et évolution magmatique du Sud-Est du Massif Central français. Contrib. Mineral. Petrol. 40, 215�–224.

CARTWRIGHT, I. & BARNICOAT, A.C. (1986): The generation of quartz-normative melts and corundum-bearing restites by crustal anatexis: petrogenetic modelling based on an example from the Lewisian of North-West Scotland. J. Metam. Geol. 3, 79�–99.

CAUCIA, F. & BOIOCCHI, M. (2005): Corindone. I cromofori nei cristalli policromi di Amboarohy, Ihosy, Madagascar. Risultati degli studi microanalitici. Riv. Mineral. Ital. 2, 126�–129.

CESBRON, F., LEBRUN, P., LE CLÉAC�’H, J.M., NOTARI, F., GROBON, C. & DEVILLE, J. (2002): Corindons et spinelles. Minéraux et Fossiles 15, Paris, France.

CHENEVIX, R. (1802): Analysis of corundum, and some of the substances which accompany it. Philosophical Transactions of the Royal Society of London 22, 327�–338.

CHHIBBER, H.L. (1934): The mineral resources of Burma, Mcmillan & Co., London, U.K.

CLABAUGH, S.E. (1952): Corundum deposits of Montana. U.S. Geological Survey Bulletin, 983, U.S.A.

COENRAADS, R.R., SUTHERLAND, F.L. & KINNY, P.D. (1990): The origin of sapphires: U�–Pb dating of zircon inclusions sheds new light. Mineral. Mag. 54, 113�–122.

COENRAADS, R.R., VICHIT, P. & SUTHERLAND, F.L. (1995): An unusual sapphire-zircon-magnetite xenolith from the Chanthaburi gem province, Thailand. Mineral. Mag. 59, 465�–479.

COLDHAM, T., OGDEN, J. & PARDIEU, V. (2012): A trip around the gem fields of Eastern Australia. InColor 20, 20�–29.

COLLINS, A.S. & WINDLEY, B.F. (2002): The tectonic evolution of central and northern Madagascar and its place in the final assembly of Gondwana. J. Geol. 110, 325�–339.

CONQUÉRÉ, F. & GIROD, M. (1968): Contribution à l�’étude des paragénèses précoces des basaltes alcalins : les spinelles du volcan de l'Oued Temorte (Massif de l�’Atakor, Sahara Algérien). Contrib. Mineral. Petrol. 20, 1�–29.

COORAY, P. & KUMARAPELI, P. (1960): Corundum in biotite-sillimanite gneiss from near Polgahawela, Ceylon. Geol. Mag. 97, 480�–487.

COX, K.G., BELL, J.D. & PANKHURST, R.J. (1979): The Interpretation of Igneous rocks. Allen & Unwin, London, U.K.

DAHANAYAKE, K. (1980): Mode of occurrence and provenance of gemstones of Sri Lanka. Mineral. Deposita 15, 81�–86.

DAHY, J.P. (1988): The geology and igneous rocks of the Yogo sapphire deposit and the surrounding area, Little Belt Mountains, Judith Basin County Montana. MSc Thesis Montana College of Mineral Science and Technology, Butte, U.S.A.

DE MAESSCHALK, A. & OEN, I. (1989): Fluid and mineral inclusions in corundum from gem gravels in Sri Lanka. Mineral. Mag. 53, 539�–545.

DÉRUELLE, B., NGOUNOUNO, I. & DEMAIFFE, D. (2007): The 'Cameroon Hot Line' (CHL): A unique example of active alkaline intraplate structure in both oceanic and continental lithospheres. C. R. Geoscience 339, 589�–600.

DE WIT, M.J. (2003): Madagascar: Heads it�’s a continent tails it�’s an island. Earth Planet. Sci. Lett. 31, 213�–248.

DHARMARATNE, P.G.R., PREMASIRI, H.M.R. & DILLIMUNI, D. (2012): Sapphires from Thammannawa, Katagarama area, Sri Lanka. Gems & Gemology 48, 98�–107.

DIRLAM, D.M., MISIOROWSKI, E.B., TOZER, R., STARK, K.B. & BASSETT, A.M. (1992): Gem wealth of Tanzania. Gems & Gemology 28, 80�–102.

DISSANAYAKE, C.B. & CHANDRAJITH, R. (1999): Sri Lanka-Madagascar Gondwana linkage: evidence for a Pan-African mineralized belt. J. Geol. 197, 223�–235.

DISSANAYAKE, C.B. & WEERASOORIYA, S.V.R. (1986): Fluorine as an indicator of mineralization �– hydrogeochemistry of a Precambrian mineralized belt in Sri Lanka. Chem. Geol. 56, 257�–270.

DISSANAYAKE, C.B., CHANDRAJITH, R. & TOBSCHALL, H.J. (2000): The geology, mineralogy and rare element geochemistry of the gem deposits of Sri Lanka. Bull. Geol. Soc. Finland 72, 5�–20.

DOLLASE, W.A. (1968): Refinement and comparison of the structures of zoisite and clinozoisite. Am. Mineral. 53, 1882�–1898.

DU TOIT, A.L. (1918): Plumasite (corundum aplite) and titaniferous magnetite rocks from Natal. Trans. Geol. Soc. S. Afr. 21, 53�–73.

GIULIANI ET AL. �– GEM CORUNDUM

99

DUFOUR, M.S., KOL'TSOV, A.B., ZOLOTAREV, A.A. & KUZNETSOV, A.B. (2007): Corundum-bearing metasomatic rocks in the central Pamirs. Petrology 15, 160�–177.

DUNN, P. & FRONDEL, C. (1990): An uncommon margarite/corundum assemblage from Sterling Hill, New Jersey. Mineral. Rec. 21, 425�–427.

DZIKOWSKI, T.J. (2013): A comparative study of the origin of carbonate-hosted gem corundum deposits in Canada. Ph.D. Thesis University of British Columbia, Canada.

DZIKOWSKI, T.J., GROAT, L.A., DIPPLE, G.M. & MARSHALL, D. (2010): Origin of the Revelstoke carbonate-hosted gem corundum occurrence, British Columbia, Canada. International Mineralogical Association Conference (August 21�–27), Budapest, Hungary.

DZIKOWSKI, T.J., CEMPIREK: J., GROAT, L.A., DIPPLE, G.M. & GIULIANI, G. (2013): Origin of rubies in marble: The Revelstoke occurrence in the Canadian Cordillera of British Columbia. Lithos submitted.

EICHHOLZ, D.E. (1965): Theophrastus. De Lapidibus. Oxford Clarendon Press, VIII, GB.

EMMETT, J.L., SCARRATT, K., MCCLURE, S.F., MOSES, T., DOUTHIT, T.R., HUGHES, R., NOVAK, S., SHIGLEY, J.E., WANG, W., BORDELON, O. & Kane, R.E. (2003): Beryllium diffusion of ruby and sapphire. Gems & Gemology 39, 84�–135.

EPSTEIN, D.S., BRENNAN, W. & MENDES, J.C. (1994): The Indaia sapphire deposits of Minas Gerais, Brasil. Gems & Gemology 30, 24�–32.

EXLEY, R.A., SMITH, J.V. & DAWSON, J.B. (1983): Alkremite, garnetite and eclogite xenoliths from Bellsbank and Jagersfontein, South Africa. Am. Mineral. 68, 512�–516.

FAGAN, A.J. (2012): Geological Exploration at Tasit marluk and other gemstone deposits, SW Greenland: A field summary report. Internal True North Gems Report, Vancouver, Canada.

FAGAN, A.J., GROVES, I.D. & CARMAN, C. (2011): A field report into geological exploration on the True North Gems Inc. Fiskenaesset Project. Internal True North Gems Report, Vancouver, Canada.

FENEYROL, J. (2012): Pétrologie, géochimie et genèse des gisements de tsavorite associés aux gneiss et roches calco-silicatées graphiteux de Lemshuku et Namalulu (Tanzanie). Thèse de Doctorat, Université de Lorraine, Nancy, France.

FENEYROL, J., GIULIANI, G., OHNENSTETTER, D., LE GOFF, E., SIMONET, C., RAKOTONDRAZAFY, A.F.M., ICHANG�’I, D., MALISA, E., DESCHAMPS, Y., FALLICK, A.E., KHAN, T. & NYAMAI, C. (2013): Les gemmes néoprotérozoïques de la ceinture métamorphique mozambicaine du supercontinent Gondwana (1000�–542 Ma). le Règne Minéral, Cahier 2,13�–50.

FERNANDO, G.W.A.R., HAUZENBERGER, C.A. & HOFMEISTER, W. (2001): Origin of corundums in Sri Lanka: evidence from case studies of in situ deposits. J. Geol. Soc. Sri Lanka 10, 37�–47.

FIELD, D.S. (1951): Ruby and sapphire in Canada. Can. Mining J. 71, 54�–57.

FORESTIER, F. (1993): Histoire de l�’un des gisements de gemmes les plus anciennement connu d�’Europe occidentale : saphirs, grenats et hyacinthes du Puy en Velay. Cahiers de la Haute Loire 43.

FORESTIER, F. & LASNIER, B. (1969): Découverte de niveaux d�’amphibolites à pargasite, anorthite, corindon et saphirine dans les schistes cristallins du Haut-Allier. Contrib. Mineral. Petrol. 23, 194�–235.

FRAZIER, S. & FRAZIER, A. (1990): South of the Equator-The importance of gem mineral deposits in Southern Africa is saluted by this year's Intergem show. Lapidary J. 44, 36�–58.

FRITSCH, E. & ROSSMAN, G.R. (1987): An update on color in gems. Part 1: Introduction and colors caused by dispersed metal ions. Gems & Gemology 23, 126�–139.

FRITSCH, E. & ROSSMAN, G.R. (1988): An update on color in gems. Part 2: Colors involving multiple atoms and colors centers. Gems & Gemology 24, 3�–14.

FRITSCH, E., CHALAIN, J.-P., HÄNNI, H., DEVOUARD, B., CHAZOT, G., GIULIANI, G., SCHWARZ, D., ROLLION-BARD, C., GARNIER, V., BARDA, S., OHNENSTETTER, D., NOTARI, F. & MAITRALLET P. (2003): Le nouveau traitement produisant des couleurs orange à jaune dans les saphirs. Rev. Ass. Fr. Gemmol. AFG 147, 11�–23.

FURUI, W. (1988): The sapphires of Penglai, Hainan Island, China. Gems & Gemology 24, 155�–160.

GAILLOU, E. (2003): Les saphirs du Massif central : Etude minéralogique des saphirs du Sioulot, du Mont Coupet et du Menoyre. Détermination de leur origine. Mémoire de DEA, Université Blaise Pascal, Clermont-Ferrand, France.

GIULIANI ET AL. �– GEM CORUNDUM

100

GAILLOU, E., DEVOUARD, B., VIELZEUF D., BOIVIN, P., ROCHAULT, J. VALLEY, J. & HARRIS, C. (2010): Le Mont Coupet, Menet et le Sioulot : trois gisements de saphirs du Massif central. le Règne Minéral 93, 28�–36.

GALIBERT, O. & HUGHES, R.W. (1995): Chinese ruby and sapphire �– a brief history. J. Gemmol. 24, 467�–473.

GAMAGE, S.J.K., RUPASINGHE, M.S. & DISSANAYAKE, C.B. (1992): Application of Rb�–Sr ratios to gem exploration in the granulite belt of Sri Lanka. J. Geochem. Explor. 43, 281�–292.

GARDE, A. & MARKER, M. (1988): Corundum crystals with blue-red color zoning near Kangerdluarssuk, Sukkertoppen district, West Greenland. Report Gronlands Geologiske Undersokelse 140, 46�–49.

GARLAND, M.I. (2003): The alluvial sapphires of western Montana. Gem Materials and Mineralogy. Special Session, Geological Association of Canada (abstracts), May 26�–28, 2003, Vancouver, Canada.

GARNIER, V. (2003): Les gisements de rubis associés aux marbres de l’Asie Centrale et du Sud-est : genèse et caractérisation isotopique. Thèse de Doctorat INPL, Nancy, France.

GARNIER, V., OHNENSTETTER, D., GIULIANI, G. & SCHWARZ, D. (2002a): Rubis trapiches de Mong Hsu, Myanmar. Rev. Ass. Fr. Gemmol. AFG 144, 5�–12.

GARNIER, V., GIULIANI, G., MALUSKI, H., OHNENSTETTER, D., PHAN, T.T., HOÀNG, Q.V., PHAM, V.L., VU, V.T. & SCHWARZ, D. (2002b): Ar�–Ar ages in phlogopites from marble-hosted ruby deposits in northern Vietnam: evidence for Cenozoic ruby formation. Chem. Geol. 188, 33�–49.

GARNIER, V., GIULIANI, G., OHNENSTETTER, D. & SCHWARZ, D. (2004a): Saphirs et rubis. Classification des gisements de corindon. le Règne Minéral 55, 4�–47.

GARNIER, V., OHNENSTETTER, D., GIULIANI, G., MALUSKI, H., DELOULE, E., PHAN, T.T., PHAM, V.L. & HOANG, Q.V. (2004b): Age and significance of ruby-bearing marbles from the Red River shear zone, northern Vietnam. Can. Mineral. 43, 1315�–1329.

GARNIER, V., OHNENSTETTER, D. & GIULIANI, G. (2004c): L�’aspidolite fluorée : rôle des évaporites dans la genèse du rubis des marbres de Nangimali

(Azad-Kashmir, Pakistan). C.R. Geoscience 336, 1245�–1253.

GARNIER, V., OHNENSTETTER, D., GIULIANI, G., FALLICK, A.E., PHAN, T.T., HOANG, Q.V., PHAM, V.L. & SCHWARZ, D. (2005): Age and genesis of sapphires from the basaltic province of Dak Nong, Southern Vietnam. Mineral. Mag. 69, 21�–38.

GARNIER, V., GIULIANI, G., OHNENSTETTER, D., SCHWARZ, D. & KAUSAR, A.B. (2006a): Les gisements de rubis associés aux marbres de l'Asie centrale et du Sud-est. le Règne Minéral 67, 17�–48.

GARNIER, V., MALUSKI, H., GIULIANI, G., OHNENSTETTER, D. & SCHWARZ, D. (2006b): Ar�–Ar and U�–Pb ages of marble hosted ruby deposits from central and southeast Asia. Can. J. Earth Sci. 43, 1�–23.

GARNIER, V., GIULIANI, G., OHNENSTETTER, D., FALLICK, A.E., DUBESSY, J., BANKS, D., HOÀNG, Q.V., LHOMME, TH., MALUSKI, H., PÊCHER, A., BAKHSH, K.A., PHAM, V.L., PHAN, T.T. & SCHWARZ, D. (2008): Marble-hosted ruby deposits from central and Southeast Asia: towards a new genetic model. Ore Geol. Rev. 34, 169�–191.

GAUTHIER, G., GROAT, L.A., TAYLOR, R.P. & FALLICK, A.E. (1995): Mineralogical, lithogeochemical and stable isotope characteristics of the sapphire-bearing Yogo dyke, Montana. Annual Meeting of the Geological Association of Canada and the Mineralogical Association of Canada, Victoria, British Columbia, Canada, 17�–19 May 1995, Abstracts.

GIBERT, F., GUILLAUME, D. & LAPORTE, D. (1998): Importance of immiscibility in the H2O�–NaCl�–CO2 system and selective CO2 entrapment in granulites: experimental phase diagram at 5�–7 kbar, 900°C and wetting textures. Eur. J. Mineral. 10, 1109�–1123.

GIULIANI, G. (2013): Les gemmes du Gondwana. Cahier N°2 le Règne Minéral. Editions du Piat, Glavenas, France.

GIULIANI, G., FRANCE-LANORD, CH., CHEILLETZ, A., COGET, P., BRANQUET, Y. & LAUMONIER, B. (2000): Sulfate reduction by organic matter in Colombian emerald deposits: chemical and stable isotope (C, O, H) evidence. Econ. Geol. 95, 1129�–1153.

GIULIANI ET AL. �– GEM CORUNDUM

101

GIULIANI, G., DUBESSY, J., BANKS, D., HOANG, Q.V., LHOMME, T., PIRONON, J., GARNIER, V., PHAN, T.T., PHAM, V.L., OHNENSTETTER, D. & SCHWARZ, D. (2003): CO2�–H2S�–COS�–S8�–AlO(OH)-bearing fluid inclusions in ruby from marble-hosted deposits in Luc Yen area, North Vietnam. Chem. Geol. 194, 167�–185.

GIULIANI, G. JARNOT, M., NEUMEIER, G., OTTAWAY, T., SINKANKAS, J. & STAEBLER, G. (2004): Emerald- The most valuable beryl; the most precious gemstone. ExtraLapis English N°2, Lapis International, LLC East Hampton, CT, U.S.A.

GIULIANI, G., FALLICK, A.E., GARNIER, V., FRANCE-LANORD, CH., OHNENSTETTER, D. & SCHWARZ, D. (2005): Oxygen isotope composition as a tracer for the origins of rubies and sapphires. Geology 33, 249�–252.

GIULIANI, G., OHNENSTETTER, D., GARNIER, V, FALLICK, A.E., RAKOTONDRAZAFY, A.F.M. & SCHWARZ D. (2007a): The geology and genesis of gem corundum deposits. In Geology of Gem Deposits (Groat, A. ed.), Mineralogical Association of Canada, Short Course Series 37, Yellowknife, Canada, 23�–78.

GIULIANI, G., FALLICK, A.E., RAKOTONDRAZAFY, A.F.M., OHNENSTETTER, D., ANDRIAMAMONJY, A., RAKOTOSAMIZANANY, S., RALANTOARISON, TH., RAZANATSEHENO, M.M., DUNAIGRE, CH. & SCHWARZ, D. (2007b): Oxygen isotope systematics of gem corundum deposits in Madagascar: relevance for their geological origin. Mineral. Deposita 42, 251�–270.

GIULIANI, G., FALLICK, A.E., OHNENSTETTER, D. & PÉGÈRE, G. (2009): Oxygen isotopes of sapphire from the French Massif Central: implications for the origin of gem corundum in basaltic fields. Mineral. Deposita 44, 221�–231.

GIULIANI, G., LASNIER, B., OHNENSTETTER, D., FALLICK, A.E. & PÉGÈRE, G. (2010): Les gisements de corindon de France. le Règne Minéral 93, 5�–22.

GIULIANI, G., OHNENSTETTER, D., FALLICK, A.E., GROAT, L. & FENEYROL, J. (2012a) Geographic origin of gems linked to their geographical history. InColor 19, 16�–27.

GIULIANI, G., DUBESSY, J., LHOMME, TH. & OHNENSTETTER, D. (2012b): Raman micro-spectrometry applied to the identification of inclusions in ruby from marble deposits in central and south-east Asia. GeoRaman 10, Nancy (June

11�–13), University of Lorraine, France. GOUJOU, J.C. (2002): Les saphirs des plages

vendéennes : un gisement à localisation variable. Le Règne Minéral 46, 19�–27.

GRAHAM, I.T., SUTHERLAND, F.L., WEBB, G.B. & FANNING, C.M. (2004): Polygnetic corundums from New South Wales gemfields. In Metallogeny of the Pacific Northwest: tectonics, magmatism and metallogeny of active continental margins. (A.I.. Khanchuk, G.A., Gonevchuk, A.N. Mitrokhin, I.F. Simanenko, N.J. Cook & R. Seltmann, eds.). Falnauka, Vladivostok, Russia, 336�–339.

GRAHAM, I.T., SUTHERLAND, F.L., KHIN ZAW, NECHAEV, V. & KHANCHUK, A. (2006a): Advances in our understanding of the basalt-derived gem sapphire-ruby deposits of the West Pacific margins. 12th Quadriennal IAGOD Symposium 2006, Moscow, Russia, Extended Abstract CDROM.

GRAHAM, I.T., NECHAEV, V., CHASHCHIN, A., KIKHNEY, E., VYSTOSTSKIY, S. & SUTHERLAND, F.L., (2006b): New insights into the late Cenozoic gem-sapphire-zircon occurrences of southern Primorye, Russia. 12th Quadriennal IAGOD Symposium 2006, Moscow, Russia, Extended Abstract CDROM.

GRAHAM, I.T., KHIN ZAW & COOK, N.J. (2008a): The genesis of gem deposits. Special Issue Ore Geology Reviews, Elsevier B.V., Amsterdam, Netherlands.

GRAHAM, I., SUTHERLAND, L., ZAW, K., NECHAEV, V. & KHANCHUK, A. (2008b): Advances in our understanding of the gem corundum deposits of the West Pacific continental margins intraplate basaltic fields. Ore Geol. Rev. 34, 200�–215.

GRAPES, R. & PALMER, K. (1996): (Ruby-sapphire)-chromian mica-tourmaline rocks from Westland, New Zealand. J. Petrol. 37, 293�–315.

GREW, E.S., DRUGOVA, G.M. & LESKOVA, N.V. (1989): Högbomite from the Aldan shield, Eastern Siberia, USSR. Mineral. Mag. 53, 376�–379.

GRISHINA, S., DUBESSY, J., KONTOROVITCH, A. & PIRONON, J. (1992): Inclusions in salt beds resulting from thermal metamorphism by dolerite sills (eastern Siberia, Russia). Eur. J. Mineral. 4, 1187�–1202.

GROAT, L.A., ed. (2007): Geology of Gem Deposits. Mineralogical Association of Canada, Short Course Series 37, Yellowknife, Canada.

GIULIANI ET AL. �– GEM CORUNDUM

102

GROVES, I. & BANFIELD, C. (2009): The Aappalutoq and Sarfaq Ruby Mineralization – A review of the 2007/2008 drilling programs and geological interpretation. Report True North Gems, Vancouver, Canada.

GÜBELIN, E. (1965): The ruby mines in Mogok in Burma. J. Gemmol. 9, 12, 411�–425.

GÜBELIN, E.J. (1982): Gemstones of Pakistan: emerald, ruby and spinel. Gems & Gemology 28, 123�–139.

GÜBELIN, E.J. & KOIVULA, J.I. (1986): Photoatlas of Inclusions in Gemstones. ABC Edition, Zurich, Switzerland.

GÜBELIN, E.J. & PERETTI, A. (1997): Sapphires from the Andranondambo mine in SE Madagascar: evidence for metasomatic skarn formation. J. Gemmol. 25, 453�–516.

GULLIONG, M. & GÜNTHER, D. (2001): Quasi 'non-destructive' laser ablation�–inductively coupled plasma�–mass spectrometry fingerprinting of sapphire. Spectrochim. Acta B 56, 1219�–1231.

GÜNTHER, D. & KANE, RE. (1999): Laser ablation�–inductively coupled plasma�–mass spectrometry: a new way of analyzing gemstones. Gems & Gemology 35, 160�–166.

GUO, J., WANG, F. & YAKOUMELOS, G. (1992): Sapphires from Changle in Shandong Province, China. Gems & Gemology 28, 255�–260.

GUO, J., O�’REILLY, S.Y. & GRIFFIN, W.L. (1996a): Corundum from basaltic terrains: a mineral inclusion approach to the enigma. Contrib. Mineral. Petrol. 122, 368�–386.

GUO, J., O�’REILLY, S.Y. & GRIFFIN, W.L. (1996b): Zircon inclusions in corundum megacrysts: I. Trace element geochemistry and clues to the origin of corundum megacrysts in alkali basalts. Geochim. Cosmochim. Acta 60, 2347�–2363.

HAAPALA, I., SIIVOLA, J., OJANPERÄ, P. & YLETYUNEN, V. (1971): Red corundum, sapphirine and kornerupine from Kittilä, Finnish Lapland. Bull. Geol. Soc. Finland 43, 221�–231.

HALL, A.L. (1923): On the marundites and allied corundum-bearing rocks in the Leysdorf district of the Eastern Transvaal. Trans. Geol. Soc. S. Afr. 24, 43�–67.

HÄNNI, H.A. (1990): A contribution to the distinguishing characteristics of sapphire from Kashmir. J. Gemmol. 22, 67�–75.

HÄNNI, H.A. & SCHMETZER, K. (1991): New rubies

from the Morogoro Area, Tanzania. Gems & Gemology 27, 156�–167.

HAPUARACHCHI, D.J.A.C. (1989): Some observations on the origin of gem corundum in Sri Lanka. J. Geol Soc. Sri Lanka 2, 5�–9.

HARDING, R.R. & SCARRATT, K. (1986): A description of ruby from Nepal. J. Gemmol. 20, 3�–10.

HARLAN, S.S. (1996): Timing of emplacement of the sapphire-bearing yogo dike, little belt mountains, Montana. Econ. Geol. 91, 1159�–1162.

HARLOW, G.E. & BENDER, W. (2013): A study of ruby (corundum) compositions from the Mogok Belt, Myanmar: searching for chemical fingerprints. Am. Mineral. 98, 1120�–1132.

HARLOW, G.E., PAMUCKU, A., NAUNG, U. S. & THU U.K. (2006): Mineral assemblages and the origin of ruby in the Mogok Stone Tract, Myanmar. Gems & Gemology 42, 147.

HATTON, C.J. (1978): The geochemistry and origin of xenoliths from the Roberts Victor mine. PhD thesis, University of Cape Town, South Africa.

HAÜY, R.J. (1817): Traité des caractères physiques des pierres précieuses. Mme Courcier V., Imprimerie-librairie.

HEILMANN, G. & HENN, U. (1986): On the origin of blue sapphire from Elahera, Sri Lanka. Austr. Mineral. 16, 2�–4.

HENN, U. & BANK, H. (1990): Gem-quality rubies from the Pamir range, USSR. Goldshmiede Zeit. 6, 107�–108.

HENN, U. & MILISENDA, C.C. (1997): Neue Edelsteinvorkommen in Tansania: die region Tunduru-Zongea. Z. D. Gemmol. Ges. 46, 29�–43.

HENN, U., BANK, H. & BANK, F.H. (1990): Red and orange corundum (ruby and padparadscha) from Malawi. J. Gemmol. 22, 83�–89.

HERD, R., WINDLEY, B.F. & GHISLER, M. (1969): The mode of occurrence and petrogenesis of the sapphirine-bearing and associated rocks in West Greenland. Gronlands Geol. Undersogelse, Report 24, 44.

HERNAULT, L. (1890): Marbode, évêque de Rennes. Sa vie et ses oeuvres (1035–1123), Rennes, France.

HLAING, U TIN (1991): A new Myanmar ruby deposit. Austr. Gemmol. 17, 509�–510.

HOANG, N. & FLOWER, M. (1998): Petrogenesis of Cenozoic basalts from Vietnam: implication for

GIULIANI ET AL. �– GEM CORUNDUM

103

origins of a �‘diffuse igneous province�’. J. Petrol. 39, 369�–395.

HOCHLEITNER, R. (1998): Europa: Korunde zum Sammeln. In Rubin, Saphir, Korund: schön, hart, selten, kostbar (C. Weise, ed.). ExtraLapis 15, 76�–85.

HUGHES, R.W. (1990): Corundum. Butterworths Gem Books, London, U.K.

HUGHES, R.W. (1997): Ruby and sapphire. RWH publishing, Boulder, U.S.A.

HUGHES, R.W. & MANOROTKUL, W. (2014): Ruby & Sapphire: A Collector's Guide. Bangkok, Gem and Jewelry Institute of Thailand.

HUGHES, R.W., ISHMAEL, A., ISATELLE, F. & WANG, P. (2011): Laos �– land of a million elephants �…and sapphires. InColor 16, 12�–19.

HUNSTIGER, C. (1990): Darstellung und Vergleich primaerer Rubinvorkommen in metamorphen Muttergesteinen �– Petrographie and Phasenpetro-logie �– Teil III. Z. D. Gemmol. Ges. 39, 121�–145.

HURAIOVÁ, M., KONE NÝ, P., KONE NÝ, V., SIMON, K. & HURAI, V. (1996): Mafic and salic igneous xenoliths in Late Tertiary alkaline basalts: fluid inclusion and mineralogical evidence for a deep-crustal magmatic reservoir in the Western Carpathians. Eur. J. Mineral. 8, 901�–916.

HUTCHINSON, M.T., HURSTHOUSE, M.B. & LIGHT, M.E. (2001): Mineral inclusions in diamonds: associations and chemical distinctions around the 670 km discontinuity. Contrib. Mineral. Petrol. 142, 119�–126.

HUTCHINSON, M.T., NIXON, P.H. & HARLEY, S.L. (2004): Corundum inclusions in diamonds �– discriminatory criteria and a corundum composition dataset. Lithos 77, 273�–286.

IRVING, A.J. (1986): Polybaric magma mixing in alkali basalts and kimberlites; evidence from corundum, zircon and ilmenite megacrysts. Geol. Soc. Austral. 16, 262�–264.

IRVING, A.J. & PRICE, R.C. (1981): Geochemistry and evolution of lherzolite-bearing phonolitic lavas from Nigeria, Australia, East Germany and New Zealand. Geochim. Cosmochim. Acta. 45, 1309�–1320.

IYER, L.A.N. (1953): The geology of gem-stones of the Mogok Stone Tract, Burma. Geological map of the Mogok Stone Tract. Mem. Geol. Survey India 82.

JACKSON, B. (1984): Sapphire from Loch Roag, Isle of Lewis, Scotland. J. Gemmology 19, 336�–342.

JANARDHANAN, A. & LEAKE, B.E. (1974): Sapphirine in the Sittampundi complex, India. Mineral. Mag. 39, 901�–902.

JOBBINS, E.A. & BERRANGÉ, J.P. (1981): The Pailin ruby and sapphire gemfield, Cambodia. J. Gemmol. 27, 555�–567.

JÖNS, N. & SCHENK, V. (2008): Relics of the Mozambique Ocean in the central East African Orogen: evidence from the Vohibory Block of Southern Madagascar. J. Metam. Geol. 26,17�–28.

KAMMERLING, R.C., SCARRATT, K., BOSSHART, G., JOBBINS, E.A., KANE, R.E., GÜBELIN, E.J. & LEVINSON, A.A. (1994): Myanmar and its gems �– an update. J. Gemmol. 24, 3�–40.

KANE, R.E. & KAMMERLING, R.C. (1992): Status of ruby and sapphire mining in the Mogok Stone Tract. Gems & Gemology 28, 152�–174.

KANE, R.E., MCCLURE, S.F., KAMMERLING, R.C., KHOA, N.D., MORA, C., REPETTO, S., KHAI, N.D. & KOIVULA, J.I. (1991): Rubies and fancy sapphires from Vietnam. Gems & Gemology 27, 136�–155.

KATZ, M.B. (1972): On the origin of the Ratnapura-type gem deposits of Ceylon. Econ. Geol. 67, 113�–115.

KAZMI, A.H. & O'DONOGUE, M. (1990): Gemstones of Pakistan. Elite Publishers, Karachi, Pakistan.

KELLER, P. (1992): Gemstones of East Africa. Geoscience Press, Phoenix, Arizona, U.S.A.

KELLER, A.S. & KELLER, P.C. (1986): The sapphires of Mingxi, Fujian Province, China. Gems & Gemology 22, 41�–45.

KELLER, P., KOIVULA, J. & JARA, G. (1985): Sapphire from the Mercaderes�–Rio Mayo area, Cauca, Colombia. Gems & Gemology 21, 20�–25.

KEMP, J.F. (1890): The basic dykes outside of the syenite areas of Arkansas. Ann. Report Geol. Surv. Arkansas 2, 392�–406.

KERRICH, R., FYFE, W.S., BARNETT, R.L., BLAIR, B.B. & WILLMORE, L.M. (1987): Corundum, Cr-muscovite rocks at O�’Briens, Zimbabwe: the conjunction of hydrothermal desilicification and LIL-element enrichment �– geochemical and isotopic evidence. Contrib. Mineral. Petrol. 95, 481�–498.

GIULIANI ET AL. �– GEM CORUNDUM

104

KEY, R.M. & OCHIENG, J.O. (1991): The growth of rubies in south-east Kenya. J. Gemmol. 22, 484�–496.

KHIN ZAW, SUTHERLAND, F.L., DELLAPASQUA, F., RYAN, C.G., YUI, T.Z., MERNAGH, T.P. & DUNCAN, D. (2006): Contrasts in gem corundum characteristics, eastern Australian basaltic fields: trace elements, fluid/melt inclusions and oxygen isotopes. Mineral. Mag. 70, 669�–687.

KHIN ZAW, SUTHERLAND, L., GRAHAM, I., & MCGEE, B. (2008): Dating zircon inclusions in gem corundums from placer deposits �– a guide to their origin. 33rd International Geological congress, Oslo, Norway.

KIEFERT, L. & SCHMETZER, K. (1987): Blue and yellow sapphire from Kaduna Province, Nigeria. J. Gemmology 20, 369�–370.

KIEFERT, L., SCHMETZER, K., KRZEMNICKI, M.S., BERNARDT, H.-J. & HÄNNI, H.A. (1996): Sapphires from the Andranondambo area, Madagascar. J. Gemmol. 25, 185�–209.

KIEVLENKO, E.Y. (2003): Geology of Gems. Ocean Pictures Ltd, Littleton, CO, USA.

KING, C.W. (1865): The Natural History, Ancient and Modern, of Precious Stones and Gems. Orbiting Books, Hereford, London, UK.

KISSIN, A.J. (1994): Ruby and sapphire from the Southern Ural Mountains, Russia. Gems & Gemology 30, 243�–252.

KOIVULA, J.I., KAMMERLING, R.C. & FRITSCH, E. (1992): Sapphires from Madagascar. Gems & Gemology 38, 203�–204.

KOLESNIK, YU.N. (1976): High Temperature Metasomatism in Ultramafic Massifs. Novosibirsk, Nauka. Russia.

KOLSTOV, A.B. (2002): Ruby-bearing metasomatites in marbles: conditions and numerical mode of formation. Experiments Geosci. 10, 94�–96.

KORNPROBST, J., PIBOULE, M., RODEN, M. & TABIT, A. (1990): Corundum-bearing garnet clinopyrox-enites at Beni Bousera (Morocco): original plagioclase-rich gabbros recrystallized at depth within the mantle? J. Petrol. 31, 717�–745.

KORZHINSKII, D.S. (1964): An outline of metasomatic processes. Int. Geol. Rev. 6, 10�–12.

KORZHINSKII, D.S. (1970): Theory of Metasomatic Zoning. Clarendon Press, Oxford, U.K.

KRIEGSMAN, L.M. (1995): The Pan-African event in East Antarctica: a view from Sri Lanka and the

Mozambique Belt. Precambrian Res. 75, 263�–277.

KRÖNER, A. (1984): Late Precambrian plate tectonics and orogeny: a need to redefine the term Pan-African. In Klerkx, J. & Michot, J. (Eds), African Geology. Tervuren, Musée Royal de l'Afrique Centrale, Belgique, 23�–28.

KRZEMNICKI, M., HÄNNI, H.A., GUGGENHEIM, R. & MATHYS, D. (1996): Investigations on sapphires from an alkali basalt, South West Rwanda. J. Gemmol. 25, 90�–106.

KUPFERBURGER, W. (1935): Corundum in the Union of South Africa. Bull. Geol. Surv. South Afr., 6., RSA.

KYAN THU (2007): The igneous rocks of the Mogok Stone Tract: their distributions, petrography, petrochemistry, sequence, geochronology and economic geology. PhD University of Rangoon, Myanmar.

LACASSAGNE, V., BESSADA, C., FLORIAN, P., BOUVET, S., OLLIVER, B., COUTURES, J.P. & MOSSIOT, D. (2002): Structure of high-temperature NaF�–AlF3�–Al2O3 melts: a multinuclear NMR study. J. Phys. Chem. 106, 1862�–1868.

LACOMBE, P. (1970): Le massif basaltique quaternaire à zircons-gemmes de Ratanakiri (Cambodge Nord-oriental). Bull. BRGM 4, 33�–79.

LACROIX, A. (1890): Contribution à l�’étude des roches métamorphiques et éruptives de l�’Ariège. Bull. Serv. Carte Géol. France 11, 1�–49.

LACROIX. A. (1901): Corindon. In Minéralogie de la France et de ses colonies. Librairie polytechnique Baudry et Cie, Paris, vol III, 237�–243.

LACROIX, A. (1922): Minéralogie de Madagascar, tome 1. A. Challamel (Ed.), Paris, France.

LACROIX, A. (1941): Les gisements de phlogopites de Madagascar et les pyroxénites qui les renferment. Ann. Géol. Serv. Min., Imprimerie officielle, Tananarive, Madagascar.

LASNIER, B. (1977): Persistance d’une série granulitique au cœur du Massif Central français (Haut Allier) – Les termes basiques, ultrabasiques et carbonatés. Thèse de Doctorat, Université de Nantes, Nantes, France.

LA TOUCHE, T. (1890): Sapphire mines of Kashmir. Rec. Geol. Surv. India 23, 59�–69.

LAWSON, A.C. (1903): Plumasite, an oligoclase

GIULIANI ET AL. �– GEM CORUNDUM

105

corundum rock, near Spanish Peak, California University of California. Pub. Geol. Sci. 3, 219�–229.

LE, T.-T.H., HÄGER, T., HOFMEISTER, W., HAUZENBERGER C., SCHWARZ, D., PHAM, V.L., WEHMEISTER, U., NGUYEN, N.K. & NGUY, T.N. (2012): Gemstones from Vietnam: an update. Gems & Gemology 48, 158�–177.

LECHEMINANT, A.N., GROAT, L.A., DIPPLE, G.M., MORTENSEN, J.K., GERTZBEIN, P., ROHTERT, W. & QUINN, E.P. (2004): Sapphire from Baffin Island, Canada. Gems & Gemology 40, 344�–345.

LECHEMINANT, A.N., GROAT, L.A., MORTENSEN, J.K., GERTZBEIN, P. & ROHTERT, W. (2005): Sapphires from Kimmirut, Baffin Island, Nunavut, Canada. 15th Annual Goldschmidt Conference, Abstracts A280.

LE GOFF, E., DESCHAMPS, Y., COCHERIE, A., GUERROT, C. & KETTO, D. (2008): Structural, petrological and geochronological constraints of the Tanzanian ruby belt. 22ème RST Meeting Nancy (April 21�–24), University of Nancy, Nancy, France.

LE GOFF, E., DESCHAMPS, Y. & GUERROT, C. (2010): Tectonic implications of new single zircon Pb�–Pb evaporation data in the Lossogonoi and Longido ruby-districts, Mozambican metamorphic Belt of north-eastern Tanzania. C. R. Geoscience 342, 36�–45.

LELOUP, P.H., ARNAUD, N., LACASSIN, R., KIENAST, J.R., HARRISON, T.M., PHAN, T.T., REPLUMAZ, A. & TAPPONNIER, P. (2001): New constraints on the structure, thermochronology, and timing of the Ailao Shan-Red River shear zone, SE Asia. J. Geophys. Res. 106, 6683�–6732.

LE MAITRE, R.W. (2002): Igneous Rocks: a classification and Glossary of Terms. Cambridge University Press, Cambridge, U.K.

LETTERMANN, M. & SCHUBNEL, H.J. (1970): Un nouveau gisement de saphir. Rev. Ass. Fr. Gemmologie AFG 24, 8�–9.

LEVINSON, A.A. & COOK, F.A. (1994): Gem corundum in alkali basalt: origin and occurrence. Gems & Gemology 30, 253�–262.

LIMKATRUN, P., KHIN ZAW, RYAN, C.G. & MERNAGH, T.P. (2001): Formation of the Denchai gem sapphires, northern Thailand: evidence from mineral chemistry and fluid/melt inclusion characteristics. Mineral. Mag. 65, 725�–735.

LITTRÉ, E. (1850): Pline l'Ancien : Histoire naturelle. In Dubochet, JJ, Le Chevallier et al.(Eds), Tome second, Livre XXXVII, Paris, France.

LIU SHANG-I, E. & ZOYSA, G. (2011): The Mirisatahela primary sapphire deposit in Sri Lanka. InColor 18, 34�–39.

LYDEKKER, L. (1883): Geology of Kashmir and Chamba. Mem. Geol. Surv. India 22, 335�–336.

MALIK, R.H. (1994): Geology and resource potential of Kashmir ruby deposits. Distt. Muzaffarabad (ak) Pakistan. Azad-Kashmir Mineral and Industrial Development Corporation. Unpublished report. Pakistan.

MALIKOVÁ, P. (1999): Origin of the alluvial sapphires from the Jizerska Louka alluvial deposit in North Bohemia, Czech Republic, Europe. Austr. Gemmol. 20, 202�–206.

MALLETT, F.R. (1882): On sapphires recently discovered in the northwest Himalaya. Rec. Geol. Surv. India 15, 138�–141.

MARTELAT, J.E., LARDEAUX, J.M., NICOLLET, C. & RAKOTONDRAZAFY, R. (2000): Strain pattern and late Precambrian deformation history in southern Madagascar. Precambrian Res. 102, 1�–20.

MATTEY, D., LOWRY, D. & MACPHERSON, C.G. (1994): Oxygen isotope composition of mantle peridotite. Earth Planet. Sci. Lett. 128, 231�–241.

MATTINSON, J. (2007): Ruby and Pink Sapphire Evaluation Report – Polished Ruby and Pink Sapphire for the Fiskenaesset Ruby Project Greenland. Report True North Gems Inc., Vancouver, Canada.

MAZZONE, P. & HAGGERTY, S.E. (1989): Peraluminous xenoliths in kimberlite: metamorphosed restites produced by partial melting of pelites. Geochim. Cosmochim. Acta 53, 1551�–1561.

MCCLURE, S.F.M., MOSES, T., WANG, W., HALL, M. & KOIVULA, J.I. (2002): Gem news internal special report: A new corundum treatment from Thailand. Gems & Gemology 38, 86�–90.

MCCOLL, D. & WARREN, R.G. (1980): First discovery of ruby in Australia. Mineral. Rec. 11�–12, 371�–375.

MEERT, J.G. (2003): A synopsis of events related to the assembly of eastern Gondwana. Tectonophysics 362, 1�–40.

GIULIANI ET AL. �– GEM CORUNDUM

106

MEERT, J.G., VAN DER VOO, R. & AYUB, S. (1995): Paleomagnetic investigation of the Neoproterozoic Gagwe lavas and Mbozi complex, Tanzania and the assembly of Gondwana. Precambrian Res. 74, 225�–244.

MENDIS, D.P.J., RUPASINGHE, M.S. & DISSANAYAKE, C.B. (1993): Application of structural geology in the exploration for residual gem deposits. Bull. Geol. Soc. Finland 65, 31�–40.

MENON, RD & SANTOSH M (1995): The Pan-African gemstone province of East Gondwana. Geol. Soc. India 34, 357�–371.

MERCIER, A., DEBAT, P. & SAUL J.M. (1999a): Exotic origin of the ruby deposits of the Mangari area in SE Kenya. Ore Geol. Rev. 14, 83�–104.

MERCIER, A., RAKOTONDRAZAFY, A.F.M. & RAVOLOMIANDRINARIVO, B. (1999b): Ruby mineralization in Southwest Madagascar. Gondwana Res. 2, 233�–238.

MERLE, O., MICHON, L., CAMUS, G. & DE GOER, A. (1998): L�’extension oligocène sur la transversale septentrionale du rift du Massif Central. Bull. Soc. Géol. France 169, 615�–626.

MEYER, H.OA. & GÜBELIN, E. (1981): Ruby in diamond. Gems & Gemology 17, 153�–156.

MEYER, H.O.A. & MITCHELL, R.H. (1988): Sapphire-bearing ultramafic lamprophyre from Yogo, Montana: a ouachitite. Can. Mineral. 26, 81�–88.

MILISENDA, C.C. & HENN, U. (1996): Compositional characteristics of sapphires from a new find in Madagascar. J. Gemmol. 25, 177�–184.

MOINE, B., RAKOTONDRATSIMA, C. & CUNEY, M. (1985): Les pyroxénites à urano-thorianite du Sud-Est de Madagascar: conditions physico-chimiques de la métasomatose. Bull. Minéral. 108, 325�–340.

MONCHOUX, P., FONTAN, F., DE PARSEVAL, PH., MARTIN, B. & WANG, R.C. (2006): Igneous albitite dikes in orogenic lherzolites, western Pyrénées, France: a possible source for corundum and alkali feldspar xenocrysts in basaltic terranes. I. Mineralogical associations. Can. Mineral. 44, 811�–836.

MORISHITA, T. & KODERA, T. (1998): Finding of corundum-bearing gabbro boulder possibly derived from the Horoman peridotite complex, Hokkaido, northern Japan. J. Mineral. Petrol. 93, 52�–63.

MORRISON, E.R. (1972): Corundum in Rhodesia. Mineral. Resour. Ser. Rhod. Geol. Surv. 16.

MOYD, L. (1949): Petrology of the nepheline and corundum rocks of southeastern Ontario. Amer. Mineral. 34, 736�–751.

MUHLMEISTER, S., FRITSCH, E., SHIGLEY, J.E., DEVOUARD, B. & LAURS, B.M. (1998): Separating natural and synthetic rubies on the basis of trace-element chemistry. Gems & Gemology 34, 80�–101.

MUNASINGHE, T. & DISSANAYAKE, C.B. (1981): The origin of gemstones of Sri Lanka. Econ. Geol. 76, 1216�–1225.

MURDOCH, J. & WEBB R.W. (1942): Minerals of California, California State. Div. Mines Bull. 136, San Francisco, California, U.S.A.

MYCHALUK, K.A. (1995): The Yogo saphire deposit. Gems & Gemology 31, 28�–41.

NGUY, T.N., NGUYEN, N.K., PHAN, V.L., NGUYEN, N.T. & HOÀNG, T.T. (1994): Dac diem tinh the �– Khoang vat hoc va dieu kien thanh tao cua corindon Viet Nam. Tap chi Dia Chat A 222, 9�–16.

NGUYEN, N.K., SUTTHIRAT, C., DUONG, A., NGUYEN, V.N., NGUYEN, T.M.T; & NGUY, T.N. (2011): Ruby and sapphire from the Tan Huong-Truc Lau area, Yen Bai province, northern Vietnam. Gems & Gemology 47, 182�–195.

NICOLLET, C. (1986): Saphirine et staurotide riche en magnésium et chrome dans les amphibolites et anorthosites à corindon du Vohibory Sud, Madagascar. Bull. Minéral. 109, 599�–612.

NICOLLET, C. (1990): Crustal evolution of the granulites of Madagascar. In Vielzeuf, D. & Vidal Ph. (Eds.), Granulites and Crustal Evolution. Kluwer, Dordrecht, 291�–310.

NISSINBOIM, A. & HARLOW, G.E. (2011): A study of ruby on painite from the Mogok Stone Tract. Gems & Gemology 47, 140�–141.

NORCONSULTCONSORTIUM (2007a): Mineral Resources Management Capacity Building Project, Republic of Mozambique; Component 2: Geological Infrastructure Development Project, Geological Mapping Lot 1; Geophysical interpretation. National Directorate of Geology, Republic of Mozambique. Report no. B4.

NORCONSULTCONSORTIUM (2007b): Mineral Resources Management Capacity Building Project, Republic of Mozambique; Component 2: Geological Infrastructure Development Project,

GIULIANI ET AL. �– GEM CORUNDUM

107

Geological Mapping Lot 1; Sheet Explanation: 32 sheets; Scale: 1: 250000. National Directorate of Geology, Republic of Mozambique, Report no. B6.f.

NOTARI, F. 1997. Le saphir "Padapadradscha". Rev. Ass. Fr. Gemmol. AFG 132, 24�–27.

NOTARI, F. & GROBON, C. (2002): Gemmologie du corindon et du spinelle. In Corindons et Spinelles, Minéraux et Fossiles 15, Paris, France, 48�–59.

OAKES, G., BARRON, L.M. & LISHMUND, S.R. (1996): Alkali basalts and associated volcanoclastic rocks as a source of sapphire in Eastern Australia. Aust. J. Earth Sci. 43, 289�–298.

OHNENSTETTER, D., GIULIANI, G., FALLICK A.E. & RAROTONDRAZAFY, A.F.M. (2008): Are "sakénites" from southern Madagascar a jigsaw puzzle of a metamorphic dismembered anorthositic suite ? GAC�–MAC Quebec, 26�–28 May 2008, p. 125.

OKRUSCH, M., BUNCH, T.E. & BANK, H. (1976): Paragenesis and petrogenesis of a corundum-bearing marble at Hunza (Kashmir). Mineral. Deposita 11, 278�–297.

O�’REILLY, S.Y. & ZHANG, M. (1995): Geochemical characteristics of lava-field basalts from eastern Australia and inferred sources: connections with the subcontinental lithospheric mantle? Contrib. Mineral. Petrol. 121, 148�–170.

OZEROV, K. (1945): Form of corundum crystals as dependent upon chemical composition of medium. Dolk. Akad. Nauk SSSR XLVII, 49�–52.

PADOVANI, E.R. & TRACEY, R. (1981): A pyrope-spinel (alkremite) xenolith from Moses Rock Dike: first known North American occurrence. Am. Mineral. 66, 741�–745.

PAQUETTE, J.L., NÉDELEC, A., MOINE, B. & RAKOTONDRAZAFY, A.F.M. (1994): U�–Pb single zircon Pb-evaporation and Sm�–Nd isotopic study of a granulite domain in SE Madagascar. J. Geol. 102, 523�–538.

PARDIEU, V. (2012): Ruby and Sapphire rush near Didy, Madagascar (April�–June 2012), http:// www.giathai.net/pdf/Didy_Madagascar_TH.pdf

PARDIEU, V. & CHAUVIRÉ, B. (2013): Les rubis du Mozambique. le Règne Minéral, Cahier 2 101�–108.

PARDIEU, V., JACQUAT, S., SENOBLE, J.B., BRYL, L.P., HUGHES, R.W. & SMITH, M. (2009a): Expedition report to the Ruby mining sites in Northern Mozambique (Niassa and Cabo

Delgado provinces), Bangkok, GIA Laboratory, http://www.giathai.net/pdf/Niassa_Mozambique_ Ruby_September13_2009.pdf

PARDIEU, V., JACQUAT, S., BRYL, L.P. & SENOBLE, J.B (2009b): Rubies from northern Mozambique. InColor 12, 32�–36.

PARDIEU, V., THANACHAKAPHAD, J., JACQUAT, S., SENOBLE, J.B., BRYL, L.P. (2009c): Rubies from the Niassa and Cabo Delgado regions of Northern Mozambique. A preliminary examination with an updated field report annex. Bangkok, GIA Laboratory, http://www. giathai.net/pdf/Niassa_ Mozambique_Ruby_September13_2009.pdf

PARDIEU, V., DUBINSKY, E.V., SANGSAWONG, S. & CHAUVIRÉ, B. (2012): Sapphire rush near Kataragama, Sri Lanka (February–March 2012). GIA News from research, www. gia.edu/research-resources/news-from-research/Kataragama_Z_US _version_0503.pdf.

PEARSON, D.G., DAVIES, G.R., NIXON, P.H., GREENWOOD, P.B. & MATTEY, D.P. (1993): Oxygen isotope evidence for the origin of pyroxenites in the Beni Bousera peridotite massif, North Morocco: derivation from subducted oceanic lithosphere. Earth Planet. Sci. Lett. 102, 289�–301.

PÊCHER, A., GIULIANI, G., GARNIER, V., MALUSKI, H., KAUSAR, A.B., MALIK, R.M. & MUNTAZ, H.R. (2002): Geology and Geochemistry of the Nangimali ruby deposit area, Nanga-Parbat Himalaya (Azad Kashmir, Pakistan). J. Asian Earth Sci. 21, 265�–282.

PERETTI, A. & GÜNTHER D. (2002): The color enhancement (E) of fancy sapphires with a new heat-treatment technique (Part A): Inducing color zoning by internal (I) migration (M) and formation of color centers (E(IM). Contributions to Gemology 1, GRS Gemresearch Swisslab AG, Lucerne, Switzerland.

PERETTI, A. & HAHN L. (2013): Record-breaking discovery of ruby and sapphire at the Didy mine in Madagascar: investigating the source. InColor 21, 22�–35.

PERETTI, A., MULLIS, J. & KÜNDIG, R. (1990): Die Kaschmir-Saphire und ihr geologisches Erinnerungsvermögen. Forsch Techn. Neue Zürch Zeit. 187, 59.

PERETTI, A., SCHMETZER, K., BERNHARDT, H.J. & MOUAWAD, F. (1995): Rubies from Mong Hsu. Gems & Gemology 31, 2�–26.

GIULIANI ET AL. �– GEM CORUNDUM

108

PERETTI, A., MULLIS, J. & MOUAWAD, F. (1996): The role of fluoride in the formation of color zoning in rubies from Mong Hsu (Myanmar, Burma). J. Gemmol. 25, 3�–19.

PERETTI, A., PERETTI, F., KANPRAPHAI, A., BIERI, W., HAMETNER, C. & GÜNTHER D. (2008): Winza rubies identified. Contrib. Gemol. 7, 1�–97.

PEUCAT, J.J., RUFFAULT, P., FRITSCH, E., BOUHNIK-LE COZ, M., SIMONET, C. & LASNIER B. (2007): Ga/Mg ratios as a new geochemical tool to differentiate magmatic from metamorphic blue sapphires. Lithos 98, 261�–274.

PEZZOTA, F. (2005): Rubini e Zaffiri. Corindoni policromi di Amboarohy, Ihosy, Madagascar. Riv. Mineral. Ital. 2, 116�–124.

PHAM, VL., HOÀNG, Q.V., GARNIER, V., GIULIANI, G. & OHNENSTETTER, D. (2004): Marble-hosted ruby from Vietnam: Can. Gemmol. 25, 83�–95.

PIAT, D. (1974): Le rubis himalayen d�’Hunza. Rev. Assoc. Fr. Gemmol. AFG 41, 5.

PILI, E., RICARD, Y., LARDEAUX, J.M. & SHEPPARD, S.M.F. (1997): Lithospheric shear zones and mantle-crust connections. Tectonophysics 280, 15�–29.

PIN, CH., MONCHOUX, P., PAQUETTE, J.-L., AZAMBRE, B., WANG, R.C. & MARTIN, B. (2006): Igneous albitite dikes in orogenic lherzolites, western Pyrénées, France: a possible source for corundum and alkali feldspar xenocrysts in basaltic terranes. II. Geochemical and petrogenetic considerations. Can. Mineral. 44, 837�–850.

PIVIN, M., GIULIANI, G., FALLICK, A.E., DEMAIFFE, D. & OHNENSTETTER, D. (2013): Oxygen isotopes in ruby and sapphire megacrysts from the Mbuyi-Mayi kimberlite in the Democratic Republic of Congo. (in preparation).

POIROT J.P. (1997): Rubis et saphirs du Viêt-Nam. Rev. Ass. Fr. Gemmol. AFG 131, 3�–5.

POMIAN-SRZEDNICKI, I.P. (1997): Caractérisation des corindons par des mesures du rapport isotopique de l’oxygène 16O/18O. Mémoire de diplôme de l�’Institut de Minéralogie et Pétrologie de l�’Université de Lausanne, Suisse.

PRATT, J.H. (1906): Corundum and its occurrence and distribution in the United States. US. Geol. Surv. Bull. 269.

RAITH, M.M., RAKOTONDRAZAFY, R. & SENGUPTA, P. (2008): Petrology of corundum-spinel-sapphirine-anorthite rocks (sakenites) from the type

locality in Southern Madagascar. J. Metam. Geol. 26, 647�–667.

RAKOTONDRAZAFY, A.F.M.., MOINE, B. & CUNEY, M. (1996): Mode of formation of hibonite (CaAl12O19) within the U�–Th skarns from the granulites of S-E Madagascar. Contrib. Mineral. Petrol. 123, 190�–201.

RAKOTONDRAZAFY, M., GIULIANI, G., FALLICK, A.E., OHNENSTETTER, D., ANDRIAMAMONJY, A., RAKOTOSAMIZANANY, S., RALANTOARISON, TH., RAZANATSEHENO, M., OFFANT, Y., GARNIER, V., MALUSKI, H., DUNAIGRE, CH., SCHWARZ, D., MERCIER, A., RATRIMO, V. & RALISON, B. (2008): Gem corundum deposits in Madagascar: a review. Ore Geol. Rev. 34, 134�–154.

RAKOTOSAMIZANANY, S. (2009): Les corindons gemmes dans les basaltes alcalins et leurs enclaves à Madagascar: significations pétrologique et métallogénique. Thèse de Doctorat, Université Henri Poincaré, Nancy, France.

RAKOTOSAMIZANANY, S., GIULIANI, G., OHNENSTETTER, D., RAKOTONDRAZAFY, A.F.M. & FALLICK, A.E. (2009a): Les gisements de saphirs et de rubis associés aux basaltes alcalins de Madagascar: caractéristiques géologiques et minéralogiques. 1ère partie: Caractéristiques géologiques des gisements. Rev. Assoc. Fr. Gemmol. AFG 169, 13�–21.

RAKOTOSAMIZANANY, S., GIULIANI, G., OHNENSTETTER, D., RAKOTONDRAZAFY, A.F.M. & FALLICK, A.E. (2009b): Les gisements de saphirs et de rubis associés aux basaltes alcalins de Madagascar: caractéristiques géologiques et minéralogiques. 2ème partie: Caractéristiques minéralogiques. Rev. Assoc. Fr. Gemmol. AFG 170, 9�–18.

RALANTOARISON, L.T. (2006): Les corindons associés aux roches métamorphiques du Sud de Madagascar : le gisement de saphirs de Sahambano (Sud-Est d'Ihosy). Mémoire de DEA, Université d'Antananarivo, Madagascar.

RALANTOARISON, T., OFFANT, Y., ANDRIAMAMONJY, A., GIULIANI, G., RAKOTONDRAZAFY, A.F.M., OHNENSTETTER, D., SCHWARZ, D., FALLICK, A., RAZANATSEHENO, M., RAKOTOSAMIZANANY, S., MOINE, B. & BAILLOT P. (2006): Les saphirs multicolores de Sahambano et Zazafotsy, région granulitique d'Ihosy, Madagascar. Rev. Assoc. Fr. Gemmol. AFG 158, 4�–13.

GIULIANI ET AL. �– GEM CORUNDUM

109

RAMAMBAZAFY, A., MOINE, B., RAKOTONDRAZAFY, A.F.M. & CUNEY, M. (1998): Signification des fluides carboniques dans les granulites et les skarns du Sud-Est de Madagascar. C. R. Acad. Sci. 327, 743�–748.

RAMDOHR, R. & MILISENDA, C.C. (2004): Neue Vorkommen von Saphir-Seifenlagerstätten auf Nosy-Bé, Madagaskar. Z. Dt. Gemmol. Ges. 53, 143�–158.

REGGIN, L. & CHOW, J. (2011): Prefeasibility study on the Aappaluttoq Ruby Project, Greenland, True North Gems Inc. & SEDAR, Vancouver, Canada.

ROBB, L.J. & ROBB, V.M. (1986): Archean pegmatite deposits in the North-Eastern Transvaal. Mineral Deposits S. Afr. 1�–2, 437�–449.

ROHTERT, W. & RITCHIE, M. (2006): Three parageneses of ruby and pink sapphire discovered at Fiskenæsset, Greenland. Gems & Gemology 42, 149�–150.

ROMÉ DE L'ISLE, J.-B.L. (1784): Cristallographie, ou description des formes propres à tous les corps du règne minéral. Paris, 4 volumes.

ROSE, H. (1843): Traité pratique d'analyse chimique. Librairie de l'Académie Royale de Médecine. Ed. Baillière J.-B., Paris.

ROSE R.L. (1957): Andalusite- and corundum-bearing pegmatites in Yosemite National Park, California. Am. Mineral. 42, 635�–647.

ROSSMAN, G.R. (2009): The geochemistry of gems and its relevance to gemology: different traces, different prices. Elements 5, 159�–162.

ROSSOVSKIY, L.N. (1980). Mestorozhdeniya dragotsennykh kamney Afghanista [Gemstone deposits of Afghanistan]. Geol. Rudnykh Mesorozhdenii 22, 74�–88.

ROSSOVSKIY, L.N., KOVALENKO, S.I. & ANANJEV, S.A. (1982): Conditions of ruby formation in marbles. Geol. Rudnykh Mesorozhdenii 24, 57�–66.

RUPASINGHE, M.S. & DISSANAYAKE, C.B. (1985): Charnockites and the genesis of gem minerals. Chem. Geol. 53, 1�–16.

RUPASINGHE, M.S., BANERJEE, A., PENSE, J. & KISSANAYAKE, C.B. (1984): The geochemistry of beryllium and fluorine in the gem fields of Sri Lanka. Mineral. Deposita 19, 86�–93.

SALERNO, S. (1992): Minéraux et pierres de Madagascar. Rev. Assoc. Fr. Gemmol. AFG 111, 9�–10.

SAMINPANYA S. (2000): Mineralogy and origin of gem corundum associated with basalt in Thailand. PhD, University of Manchester, U.K.

SANTOSH, M. & COLLINS, A.S. (2003): Gemstone mineralization in the Palghat-Cauvery shear zone system (Kakur-Kangayam belt), Southern India. Gondwana Res. 6, 911�–918.

SCHÄRER, U., TAPPONNIER, P., LACASSIN, R., LELOUP, P.H., ZHONG, D. & JI, S. (1990): Intraplate tectonics in Asia: a precise age for large-scale movements along the Ailao Shan-Red River shear zone. China: Earth Planet. Sci. Lett. 97, 65�–77.

SCHMETZER, K & BANK, H. (1981): The colour of natural corundum. N. Jahr. für Mineral., Monat. 11, 59�–68.

SCHREYER, W. (1988): A discussion of: �“Corundum, Cr-muscovite rocks at O�’Briens, Zimbabwe: the conjunction of hydrothermal desilicification and LIL-element enrichment �– geochemical and isotopic evidence�”. Contrib. Mineral. Petrol. 100, 552�–554.

SCHREYER, W., WERDING, G. & ABRAHAM, K. (1981): Corundum-fuchsite rocks in greenstone belts of Southern Africa: Petrology, geochemistry and possible origin. J. Petrol. 22, 191�–231.

SCHULZE, D. (2003): A classification scheme for mantle-derived garnets in kimberlite. A tool for investigating the mantle and exploring for diamonds. Lithos 71, 195�–213.

SCHWARZ, D. (1998): Aus Basalten, Marmoren und Pegmatiten. Spezielle Ursachen formten in der Erdkruste edle Rubine und Saphire. In Rubin, Saphir, Korund: schön, hart, selten, kostbar (C. Weise, ed.). ExtraLapis 15, 5�–9.

SCHWARZ, D., PETSCH, E.J. & KANIS, J. (1996): Sapphires from the Andranondambo Region, Madagascar. Gems & Gemology 32, 80�–99.

SCHWARZ, D., KANIS, J. & SCHMETZER, K. (2000): Sapphires from Antsiranana Province, Northern Madagascar. Gems & Gemology 36, 216�–233.

SCHWARZ, D., PARDIEU, V., SAUL, J.M., SCHMETZER, K., LAURS, B.M., GIULIANI, G., KLEMM, L., MALSY, A.K., HAUZENBERGER, C., DU TOIT, G., FALLICK, A.E. & OHNENSTETTER, D. (2008): Ruby and sapphires from Winza (Central Tanzania). Gems & Gemology 44, 322�–347.

GIULIANI ET AL. �– GEM CORUNDUM

110

SEIFERT, A.V. & HYRSL, J. (1999): Sapphire and garnet from Kalalani, Tanga Province, Tanzania. Gems & Gemology 35, 108�–120.

SHIGLEY, J.E., DIRLAM, D.M., LAURS, B.M., BOEHM, E.W., BOSSHART, G. & LARSON, W.F. (2000): Gem localities of the 1990s. Gems & Gemology 36, 292�–335.

SHMULOVICH, K.I. & GRAHAM, C.M. (1999): An experimental study of phase equilibria in the system H2O�–CO2�–NaCl system at 800°C and 9 kbar. Contrib. Mineral. Petrol. 136, 247�–257.

SILVA, K.K.M.W & SIRIWARDENA, C.H.E.R (1988): Geology and the origin of the corundum-bearing skarn at Bakamuna, Sri Lanka. Mineral. Deposita 23, 186�–190.

SIMANDLE, G.J., JONES, P., OSBORNE, J.W., PAYIE, G. & MCLEOD, J. (1997): Use of heavy minerals in exploration for sapphires, Empress Cu–Au–Mo deposit, British Columbia. British Columbia Geological Survey, Canada, Geological Fieldwork 1997, Paper 1998�–1, 1�–12.

SIMONET, C. (2000): Géologie des gisements de saphir et de rubis. L’exemple de la John Saul mine, Mangari, Kenya. Thèse Doctorat, Université de Nantes, Nantes, France.

SIMONET, C., PAQUETTE, J.L., PIN, C., LASNIER, B. & FRITSCH, E. (2004): The Dusi (Garba Tula) sapphire deposit, Central Kenya �– a unique Pan-African corundum-bearing monzonite. J. Afr. Earth Sci. 38, 401�–410.

SIMONET, C., FRITSCH, E. & LASNIER, B. (2008): A classification of gem corundum deposits aimed towards gem exploration. Ore Geol. Rev. 34, 127�–133.

SINGER, D.A., BERGER, V.I. & MORING, B.C. (2008): US Geological Survey Open File Report 2008-1155.

SINGH, S. & JAIN, A.K. (2003): Himalayan Granitoids. In Granitoids of the Himalayan Collisional Belt (S. Singh, S., ed.). J. Virtual Explorer 11, 1�–20.

SMITH, C.P. (1998): Rubies and pink sapphires from the Pamir Mountain Range in Tajikistan, former USSR. J. Gemmol. 26, 103�–109.

SMITH, C.P. & DUNAIGRE, CH. (2001): Gem News: Anhydrite inclusion in a ruby from Myanmar. Gems & Gemology 37, 236.

SMITH, C.P. & SURDEZ, N. (1994): The Monh Hsu ruby: a new type of Burmese ruby. Jewelsiam 5, 82�–98.

SMITH, C.P., KAMMERLING, R.C., KELLER, A.S., PERETTI, A., SCARRATT, K.V., KHOA, N.D. & REPETTO, S. (1995): Sapphires from Southern Vietnam. Gems & Gemology 31, 168�–186.

SMITH, C.P., GÜBELIN, E.J., BASSETT, A.M. & MANANDHAR, M.N. (1997): Rubies and fancy-color sapphires from Nepal. Gems & Gemology 33, 24�–41.

SOLESBURY, F. (1967): Gem corundum pegmatites in NE Tanganyika. Econ. Geol. 62, 983�–991.

SOMMER, H., KRÖNER, A., HAUZENBERGER, C.A. & MUHONGO, S. (2005): Reworking of Archaean and Palaeoproterozoic crust in the Mozambique Belt of central Tanzania as documented by SHRIMP zircon geochronology. J. Afr. Earth Sci. 43, 447�–463.

SPIRIDONOV, E.M. (1998): Gemstone deposits of the former Soviet Union. J. Gemmol. 26, 111�–124.

STERN, R.J., TSUJIMORI, T., HARLOW, G. & GROAT, L. (2013): Plate tectonic gemstones. Geology, online May 2013, doi:10.1130/G34204.1

SUNAGAWA, I., BERNHARDT, H.-J. & SCHMETZER K. (1999): Texture formation and element partitioning in trapiche ruby. J. Cryst. Growth 206, 322�–330.

SUTHERLAND, F.L. (1996): Alkaline rocks and gemstones, Australia. Austr. J. Earth Sci. 43, 323�–343.

SUTHERLAND, F.L. & ABDURIYIM, A. (2009): Geographic typing of gem corundum: a text case from Australia. J. Gemmol. 31, 203�–210.

SUTHERLAND, F.L. & COENRAADS, R.R. (1996): An unusual ruby-sapphire-sapphirine-spinel assem-blage from the Tertiary Barrington volcanic province, New South Wales, Australia. Mineral. Mag. 60, 623�–638.

SUTHERLAND, F.L. & SCHWARZ, D. (2001): Origin of gem corundums from basaltic fields. Austr. Gemmol. 21, 30�–33.

SUTHERLAND, F.L., HOSKIN, P.W.O., FANNING, C.M. & COENRAADS, R.R. (1998a): Models of corundum origin from alkali basaltic terrains: a reappraisal. Contrib. Mineral. Petrol. 133, 356�–372.

SUTHERLAND, F.L., SCHWARZ, D., JOBBINS, E.A., COENRAADS, R.R. & WEBB, G. (1998b): Distinctive gem corundum suites from discrete basalt fields: a comparative study of Barrington, Australia, and West Pailin, Cambodia, gemfields. J. Gemmol. 26, 65�–85.

GIULIANI ET AL. �– GEM CORUNDUM

111

SUTHERLAND, F.L., BOSSHART, G., FANNING, C.M., HOSKIN, P.W.O. & COENRAADS R.R. (2002): Sapphire crystallization, age and origin, Ban Huai Sai, Laos: age based on zircon inclusions. J. Asian Earth Sci. 20, 841�–849.

SUTHERLAND, F.L., COENRAADS, R.R., SCHWARZ, D., RAYNOR, L.R., BARRON, B.J. & WEBB, G.B. (2003): Al-rich diopside in alluvial ruby and corundum-bearing xenoliths, Australian and SE Asian basalt field. Mineral. Mag. 67, 717�–732.

SUTHERLAND, F.L., DUROC-DANNER, J.M. & MEFFRE, S. (2008): Age and origin of gem corundum and zircon megacrysts from the Mercaderes�–Rio Mayo area, South-west Colombia, South America. Ore Geol. Rev. 34, 155�–168.

SUTHERLAND, F.L., GIULIANI, G., FALLICK, A.E., GARLAND, M. & WEBB, G.B. (2009a): Sapphire-ruby characteristics, West Pailin, Cambodia: clues to their orgin based on trace element and O isotope analysis. Austr. Gemmol. 23, 373�–432.

SUTHERLAND, F.L., KHIN ZAW, MEFFRE, S., GIULIANI, G., FALLICK, A.E. & WEBB, G.B. (2009b): Gem-corundum megacrysts from east Australian basalt fields: trace elements, oxygen isotopes and origins. Austr. J. Earth Sci. 56, 1003�–1022.

SUTTHIRAT, C., CHARUSIRI, S., FARRAR, E. & CLARK, A.H. (1994): New 40Ar/39Ar geochron-ology and characteristics of some Cenozoic basalts in Thailand. In: Proceedings of the International Symposium on Stratigraphic correlation of Southeast Asia, 15�–20 November, Bangkok, Thailand, 306�–321.

SUTTHIRAT, C., SAMINPANYA, S., DROOP, G.T.R., HENDERSON, C.M.B. & MANNING, D.A.C. (2001): Clinopyroxene-corundum assemblages from alkali-basalt and alluvium, eastern Thailand: constraints on the origin of Thai rubies. Mineral. Mag. 65, 277�–295.

TAPPONNIER, P., LACASSIN, R., LELOUP, PH. H., SCHÄRER, U., ZHONG DALAI, WU HAIWEI, LIU XIAOHAN, JI SHAOCHENG, ZHANG LIANSHANG & ZHONG JIAYOU (1990): The Ailao Shan/Red River metamorphic belt: Tertiary left-lateral shear between Indochina and South China. Nature 343, 431�–437.

TENTHOREY, E.A., RYAN, J.G. & SNOW, E.A. (1996): Petrogenesis of sapphirine-bearing metatroctolites from the Buck Creek ultramafic body, southern Appalachians. J. Metam. Geol. 14,

103�–114. TEREKHOV, E.N. (2007): REE distribution in

corundum-bearing and other metasomatic rocks during the exhumation of metamorphic rocks of the belmorian belt of the Baltic Shield. Geochem. Int. 45, 364�–380.

TEREKHOV, E.N., KRUGLOV, V.A. & LEVITSKI, V.I. (1999): Rare earth elements in corundum-bearing metasomatic and related rocks of the Eastern Pamirs. Geochem. Int. 37, 202�–212.

THEMELIS, T. (1992): The Heat Treatment of Ruby and Sapphire. Gemlab Inc., U.S.A.

THEMELIS, T. (2008): Gems and Mines of Mogok. Themelis Publisher, Thailand.

UHER, P., SABOL, M., KONE NÝ, P., GREGÁ OVÁ, M., TÁBORSKÝ, Z. & PU�ŠKELOVÁ L. (1999): Sapphire from Hajná ka (Cerová Highland, southern Slovakia). Slovak Geol. Mag. 5, 273�–280.

UHER, P., GIULIANI, G., SZAKALL, S., FALLICK, A.E., STRUNGA, V., VACULOVIC, D., OZDIN, D. & GREGANOVA, M. (2012): Sapphires related to alkali basalts from the Cerová Highlands, Western Carpathians (southern Slovakia): composition and origin. Geologica Carpathica 63, 71�–82.

UPTON, B.G.J., ASPEN, P. & CHAPMAN, N. (1983): The upper mantle and deep crust beneath the British Isles: evidence from inclusions in volcanic rocks. J. Geol. Soc. London 140, 105�–121.

UPTON, B.G.J., HINTON, R.W., ASPEN, P., FINCH, A.A. & VALLEY, J.W. (1999): Megacrysts and associated xenoliths: evidence for migration of geochemically enriched melts in the upper mantle beneath Scotland. J. Petrol. 40, 935�–956.

UPTON, B.G.J., FINCH, A.A. & SLABY, E. (2009): Megacryst and salic xenoliths in Scottish alkali basalts: derivatives of deep crustal intrusions and small-melt fractions from the upper mantle. Mineral. Mag. 73, 943�–956.

VICHIT, P., VUDHICHATIVANICH, S. & HANSAWEK, R. (1978): The distribution and some characteristics of corundum-bearing basalts in Thaïland. J. Geol. Soc. Thailand M4, 1�–27.

VOYNIC, S.M. (1985): Yogo–The Great American Sapphire. Mountain Press Publishing Company, Missoula Montana, U.S.A.

WALTON, L. (2004): Exploration criteria for coloured gemstone deposits in the Yukon. Yukon Geological Survey. Open file 2004-10, May 2004.

GIULIANI ET AL. �– GEM CORUNDUM

112

WANG, J., HARRISON, T.M., GROVE, M., YUQUAN, Z. & GUANGHONG, X. (2001): A tectonic model for Cenozoic igneous activities in the eastern Indo-Asian collision zone. Earth Planet. Sci. Lett. 188, 123�–133.

WARD, F. (1992): Rubies and Sapphires. Gem Book Publishers, Bethesda, U.S.A.

WATT, G., HARRIS, J.W., HARTE, B. & BOYD, S.R. (1994): A high-chromium corundum (ruby) inclusion in diamond from the São Luiz alluvial mine, Brazil. Mineral. Mag. 58, 490�–493.

WEISE, CH. (1998): Rubin, Saphir, Korund: schön, hart, selten, kostbar. Extra-Lapis, 15, Germany.

WELLS, A. (1956): Corundum from Ceylon. Geol. Mag. 93, 25�–31.

WESTON, B. (2009): NI 43-101 Report On Field Activities For The Fiskenæsset Ruby Project, Greenland, Vancouver, Canada.

WHITFORD-STARK, J.L. (1987): A survey of Cenozoic volcanism on mainland Asia. Geol. Soc. Am., Special Paper 213, USA.

WILSON, A.C. (1975): An occurrence of sapphire in the Land's End granite, Cornwall. Bull. Geol. Surv. Gr. Brit. 52, 61�–63.

WINDLEY, B.F. & SMITH, J.V. (1974): The Fiskenaesset complex. Bull. Gronlands Geol. Undersogelse 108, 54.

WINDLEY, B.F., RAZAFINIPARANY, A., RAZAKAMANANA, T. & ACKERMAND, D. (1994): Tectonic framework of the Precambrian of Madagascar and its Gondwana connections: a review and reappraisal. Geol. Rundschau 83, 642�–659.

WRIGHT, J., HASTINGS, D., JONES, W. & WILLIAMS, H. (1985): Geology and Mineral Resources of West Africa. Allen & Unwin (eds.), London, U.K.

YAGER, Y.G., MENZIE, W.D. & OLSON, D.W. (2008): Weight of production of emeralds, rubies, sapphires, and tanzanite from 1995 through 2005. Open File Report 2008-1013. USGS Reston, Virginia, USA.

YIN, A., RUMELHART, P.E., COWGILL, E., BUTTER, R., HARRISON, T.M., INGERSOLL, R.V., COOPER, K., ZHANG, Q. & WANG, X.-F. (2002): Tectonic history of the Altyn Tagh Fault in northern Tibet inferred from Cenozic sedimentation. Geol. Soc. Am. Bull. 114, 1257�–1295.

YUI, T.F., KHIN ZAW & LIMKATRUN, P. (2003): Oxygen isotope composition of the Denchai

sapphire, Thailand; a clue to its enigmatic origin. Lithos 67, 153�–161.

YUI, T.F., WU, C.-M., LIMKATRUN, P., SRICHARN, W. & BOONSOONG A. (2006): Oxygen isotope studies on placer sapphire and ruby in the Chanthaburi-Trat alkali basaltic gemfield, Thailand. Lithos 86, 197�–211

YUI, T.F., KHIN ZAW & WU, C.-M. (2008): A preliminary stable isotope study on Mogok ruby, Myanmar. Ore Geol. Rev. 34, 182�–199.

YURKOVICH S.P. (1985): Ruby mines of North Carolina. Rocks & Mineral. 60, 54�–57.

ZHANG, J., ZHOU, C. & HU, C. (2003): Mineralization characteristics of gems in Ailaoshan structural belt, Yunnan Province. J. Gemmol. 5, 27�–30.

ZWAAN, J.C. & ZOYSA, E.G., (2008): New primary gem occurrences in Sri Lanka. Z. Dt. Gemmol. Ges. 57, 23�–32.

View publication statsView publication stats