Printer Friendly

Andradite form Antetezambato, North madagascar.

In 2009 some of the world's finest specimens of green and yellow andradite crystals on matrix were found in a mangrove swamp near Antetezambato in the extreme northern area of Madagascar. The workings are constantly being flooded by tidal waters, and are inaccessible during the rainy season. Because of the primitive hand-mining methods in use, it may not be possible to re-open the workings.



A new deposit of andradite garnet of the demantoid and topazolite varieties has recently been discovered in Antetezambato, Ambanja area, Antsiranana province, North Madagascar. The mining activity is mostly devoted to the extraction of rough stones for the gem market but, since April-May 2009, a significant quantity of collector-quality crystal specimens, from thumbnails and miniatures to large cabinet specimens, has emerged as well. Some of these specimens, because of their complex crystal shapes and extraordinary color, luster and transparency, are among the finest andradites in the world.


The new demantoid-topazolite deposit in Madagascar is located about 2.5 km west of Antetezambato village, in the Ambanja area, Diana region, Antsiranana province, North Madagascar. The productive area lies mostly in a mangrove swamp about 5 km from the northwestern coast of Madagascar and extends westward to the slope of a gentle hill. The mined area covers a surface of several tens of acres with its center at 13[degrees]30.460' S and 48[degrees]32.652' E.

The town of Ambanja is about 20 km south of Antetezambato, on the road that runs along the northwest coast of the island connecting the large towns of Mahajanga and Antsiranana (known formerly as Diego Suarez). The area is rather sparsely inhabited, with an economy based on local products such as coffee, raffia, seafood, fruit, etc. Foreigners occasionally come to visit nearby Nosy Be island or the d'Ambre Massif Natural Park located between the towns of Ambilobe and Antsiranana.

The climate is characterized by a winter season from May to October which is rather dry and windy with mild temperatures, and a summer season from November to April which is typically hot and humid. From December to February, strong rains characterize the so-called "rainy season" or "cyclonic (hurricane) season." Vegetation is typical of wet-subtropical areas, with mangrove forests along the shore and thick primary forest covering the coastal plains. Steep mountains occur as isolated massifs and as a wide mountainous belt to the east.


The northwestern coast of Madagascar is subjected to sea tides ranging from about 2.5 meters to over 4 meters. This phenomenon explains the extensive occurrence of mangrove forests along the shore and the presence of complex networks of seawater drainage channels cutting the plains along the coast. The Antetezambato garnet deposit is situated mostly within one of these large mangrove swamps, and is invaded by 50 to 80 centimeters of seawater at high tide every day. The tidewaters reach the garnet deposit through a 5-km complex of channels in the mangrove forest. The tidal surge decreases to a minimum only during very short periods of the year.


Lacroix (1922) described garnet occurrences in Madagascar, especially andradite in igneous rocks, and reported the discovery of some crystals of pale green andradite lining cavities in a nepheline-bearing syenite at Mount Bezavona. This intrusive massif of Cretaceous age is located south of Nosy Be island, and forms the Ampasimena Peninsula. Before the discovery at Antetezambato, the information reported in Lacroix (1922) was the only evidence of the occurrence of green andradite in Madagascar. The green gem-grade garnets found in Madagascar at Gogogogo, in the Toliara region (Mercier et al, 1997), and at Itrafo, south of Betafo, in the Antananarivo region (Adamo et al., in preparation), are not andradite but rather grossular with more or less significant vanadium and chromium contents (the tsavorite variety).

In June 2006, news was circulating among local gem and mineral dealers in Antananarivo about the discovery of small fragments of "green tourmaline of good color" in the Ambanja area. However, no samples were available for analyses and the information was considered unreliable. During the following two years, a few local workers at Antetezambato (at the time a small village of about twenty cabins) produced a very limited quantity of what was being called "green zircon" or "green sapphire" in the local market in Ambanja. Nevertheless, no one bothered to analyze this gem rough and it generated no economic interest. The first fragments of these stones were probably found by natives involved in charcoal production, while they were cutting mangrove wood and digging small terraces at the limit of the swamp for burning the wood. This activity, rather common in the area, involved the local removal of surface soil and probably led to the discovery of the first "green stones." Danet (2009), however, reported a different story, suggesting that the first stones were found by crab fishermen in late 2008.

It was only in November 2008 that French gem dealer Jacque le Quire brought some rough and cut gemstones to the gemological school of the University of Nantes, where they were identified as the andradite variety demantoid (Mocquet et a., 2009a). In the meantime, the first small lots of rough demantoid arrived at the gem market in Antananarivo where gemologist and gem dealer Giuseppe Pocobelli informed the author about the new discovery. Nothing significant was produced in Antetezambato during the next few months, probably because of the strong rains of the hurricane season. Nevertheless, beginning in April of 2009, rumors of the new find and of the new gemstone in Madagascar began circulating throughout the country and several hundred miners, buyers and brokers, both Malagasy and foreign, rushed to the site. Danet (2009) reported that in May 2009 about 2,000 miners were digging at the deposit and several thousand people (probably up to 10,000) were living in nearby Antetezambato village. Using hand tools (crowbars, buckets, etc.), the miners were digging pits at low tide up to 18 meters deep in the mangrove swamp, and also in the dry terrain at the foot of a nearby hill to the west. The better organized groups of miners used gasoline-powered pumps to help drain their pits.


Up to the end of May much of the work was in the superficial weathered zone, producing a large quantity (over 20 kilos of rough per week) of broken or loose crystals of value primarily as faceting rough. But in June the deepening pits began to enter unaltered rock containing increasingly fine mineral specimens. The author began visiting the Antetezambato mine in July, when the production of mineral specimens was probably at its peak, and at that time was able to see the extraction of some of the best mineral specimens ever found there.

Unfortunately, because of the unstable political situation in the country, the local authorities subsequently lost control of the area. The social situation degenerated, and much of the gem and specimen trade as well as the supplying of food and other needs of the miners and their families were taken over by organized crime, under the protection of corrupt politicians in Ambanja, Ambilobe, and various other regions in Madagascar. The area became extremely dangerous and several people, Malagasy and foreigners alike, including the author, were victims of armed attacks and robberies. Since October 2009, thanks to the bravery of a few people in the local police station, gendarmerie and tribunal, control has been regained to some extent and crime has been reduced, but the area must still be considered dangerous.

Since last October the production of garnet has significantly diminished, mostly because of the extremely difficult and dangerous mining conditions. Depending on the timing of the tide, the miner's work begins when the seawater level starts going down. At three or four o'clock in the morning, in the dark, the miners walk from the village to the mine bringing their equipment, including electric generators and pumps. They reach their pit and begin pumping out the water, trying not to flood adjacent pits being worked by other groups of miners. Small fish and crabs are everywhere. Pumping out a pit can require up to five hours; then an hour or two of work is required to muck out the mud brought in by the tidal waters. If the generator runs out of gas, or if there is an equipment breakdown requiring several hours for repairs, the day of work is lost. If all goes well, however, mining can then begin at the bottom of the pits and inside the small galleries up to 18 meters below the sea level. Fortunately the inflow of groundwater is very low.

But mining can go on for no more than three or four hours before the tide comes in again. Miners are accustomed to working right up to the moment when the tidal surge reaches the mouth of a pit and begins cascading in, bringing with it mud and rubble. The pits become filled with water and within a few minutes the entire mine area is flooded, with the exception of the small workings at the base of the hill to the west. During these few minutes all of the miners evacuate their pits with their equipment and, running between the other pits, make their way northward to an area where a small market has been established.

It must be mentioned that removal of the rocks and the mud in the course of mining is a huge problem for the miners because there is nowhere to dump it. The waste material is impossible to haul very far away, so the miners pile it between the pits in small ridges or in storage bunkers reinforced with wood piles and ropes. Such accumulations, continuously wet and washed by the movements of the seawater, are unstable and accidents are common.



With the arrival of the summer rainy season in December 2009, many of the workers left the area. At present (February 2010), the garnet deposit is flooded by seawater and by rainwater runoff coming from the nearby hills, and moreover it is totally buried in mud. It is unclear whether the local miners will return following the rainy season to begin work again and produce more mineral specimens and gem material.

Regarding ownership of the mine, some Malagasy individuals registered mining claims on the Antetezambato area with the Ministry of Energy and Mines in 2009. As typically happens in Madagascar when a new deposit is discovered, the claims were granted by the Ministry regardless of whether the boundaries of the claims as described actually included the deposits. But the owners of the claims use their authorization only for trading and exporting the products of the mines, and not for mining, so the discrepancies don't seem to matter. In any case, until the political situation in Madagascar stabilizes it is unlikely that any well-organized mining will take place at Antetezambato.


Madagascar geology can be divided into two major domains: (1) the upper Proterozoic crystalline basement, characterized by high-grade metamorphic rocks of igneous and sedimentary origin, and by large gabbroic to granitic and syenitic plutons representing the roots of the axial zone of the East African orogenic belt; and (2) a Permian-Mesozoic to Tertiary sedimentary sequence, formed during more or less continuous subsidence of the Mahajanga, Morondava and Toliara basins, corresponding to the paleo-Mozambique channel. The crystalline basement is exposed over approximately two-thirds of the surface area of the island, and the sedimentary sequence crops out on the remaining third all along the western coast.


The famous mineral and gem localities in Madagascar, including all of the pegmatitic occurrences, were emplaced within the upper Proterozoic crystalline basement. Weathering and erosion (mostly during upper Permian to Triassic time) of these gemstone-rich crystalline rocks deposited coarse-grained sediments to form the Ilakaka gemstone deposits in southeastern Madagascar.

A renewed period of magmatism affected Madagascar during the Cretaceous period with the emplacement of doleritic dikes and the eruption of basaltic flows along the east coast and in the sedimentary basins of Mahajanga, Morondava and Toliara. Significant quantities of xenolithic sapphires were brought to the surface by the eruption of the basalts in north Madagascar, forming some significant sapphire deposits, including the locality of Ambondromifehy. Zoned ring plutons were emplaced north of Mahajanga (in the cape St-Andre area) and important sub-volcanic and volcanic alkaline massifs were emplaced in the Isalo sedimentary Permian-Triassic formation, in the area of the Ampasmdava promontory, in the northwestern part of the island (BRGM, 1985). The Antetezambato garnet deposit occurs within the latter geological environment and is the first important mineral and gemstone deposit formed by contact metamorphism of the Permian-Mesozoic sediments to be discovered in Madagascar.




In the area of Antetezambato, the 1:100,000 geologic map (Besairie and BRP-IFP, 1962) shows the distribution of the sediments of the Isalo Formation (mostly sandstones). The nearest igneous intrusion shown on the map is the Ambato pluton, an alkaline-granite massif located about 5 km northeast of the garnet deposit.


Field mapping at the mine and observations made by the author inside the diggings indicate that the Antetezambato garnet deposit is characterized, below the superficial alteration layer, by a large block of layered sediments intruded by a network of lamprophyric dikes. The entire block of sediments is tilted, dipping to the south at an angle of 45 to 60 degrees, in contrast to the sub-horizontal regional attitude of the nearby sedimentary formations. The gray lamprophyric rock is rich in phenocrysts and igneous xenoliths, and appears to have a composition similar to that of a trachyte. These igneous rocks are connected to a significant mass of lamprophyric rocks (not shown on the above-mentioned geologic map) which border the garnet deposit to the north and to the west. The sediments, originally composed of interlayered fine to medium-grained fossiliferous sandstones and silica-rich limestones, locally with a nodular structure, have been affected by metasomatic alteration, concentrated along structural discontinuities such as layer boundaries, fractures and contacts with the lamprophyric intrusions. The metasomatic alteration resulted in a network of veins where the sedimentary rocks are transformed into a very fine-grained white to pale green skarn rock, locally rich in cavities containing crystals of demantoid garnet. The skarn rock consists of an aggregate of microcrystals of garnet, with or without quartz. Calcite occurs only in a limited area in the southwestern part of the deposit. Microprobe analyses and preliminary petrographic observations in thin sections indicate that the microcrystals of garnet constituting the massive rock are zoned, with compositions ranging from intermediate between grossular and andradite to nearly pure andradite.

The metasomatic process allowed the preservation of much of the original structure of the sedimentary rock, and even fossils have been replaced Locally by the fine-grained skarn minerals. In a few cases, recognizable fossil shells, corals and even ammonites have been replaced by garnet and quartz, and form the matrix for demantoid crystals.

Genetic Model

It appears that the Antetezambato garnet deposit formed inside the sedimentary sequence during the sub-voicanic intrusion of lamprophyric igneous rocks. This intrusion phase activated a high-temperature pneumatolytic to hydrothermal circulation of fluids that was concentrated along rock discontinuities, metasomatically altering the sedimentary rock into a skarn and depositing andradite in cavities. As indicated by Fehr (2008), at atmospheric pressure, andradite is the most thermally stable species in the garnet group. For this reason it is possible that very shallow depth and correspondingly very low lithostatic pressure during garnet formation at Antetezambato accounts for the near end-member andradite composition of the garnet. The absence of titanium in the sedimentary rocks accounts for the crystallization of the demantoid and topazolite varieties rather than the more common titanium-rich melanite variety of andradite.


The 1:100,000 geologic map indicates a Post-Liassic age for the Ambato pluton, corresponding to the Cretaceous magmatism reported by the BRGM (1985). Nevertheless, significant evidence of Pleistocene magmatism exists in the form of extensive outcrops of volcanic rocks in the region. Moreover, in the center of the garnet deposit, the author observed a small basaltic dike containing vugs rich in zeolites (probably scolecite) similar to those observed in recent basalts on the nearby islands of Nosy Faly and Nosy Be. To establish definitively the age of the Antetezambato lamprophyres and thereby the age of the formation of the andradite crystals, the author is currently conducting isotope-geochronological analyses on selected specimens, the results of which should be available soon.

The Antetezambato demantoid-topazolite skarn is genetically similar to the Namibian deposits and is significantly different from the serpentinite-hosted demantoid deposits at localities such as the Ural Mountains of Russia and Val Malenco, Italy.


Non-gem andradite is typically found in skarns, is opaque and is commonly a more or less brownish color. It has not been given a varietal name. However, the transparent, colored varieties of andradite are of importance in the field of gemology, and consequently the perpetuation of varietal terms for marketing purposes (onerous as that is to many mineralogists) is inevitable, and thoroughly entrenched in the literature. Thus there will be no attempt here to avoid the use of varietal names. The two andradite varieties of principal interest are demantoid and topazolite.


Demantoid is a variety of andradite[[Ca.sub.3][Fe.sub.2][(Si[O.sub.4]).sub.3]] characterized by a deep, bright green to yellowish green color. It is the most famous and most precious of all the garnet varieties, sometimes selling for thousands of dollars per carat. Demantoid was first discovered in 1864 in the central Ural Mountains about 110 km north-northwest of Ekaterinburg. The name demantoid ("diamond-like") was proposed in 1878, in allusion to its high refractive index and its exceptional dispersion which is stronger than that of diamond. Because of its rarity and its brilliance when faceted, demantoid is among the most appreciated and highly priced of the garnet gemstones. (See Anderson, 1975; Stockon and Manson, 1983; Neuendorf et al., 2005; Stephenson and Kouznetsov, 2009; Rouse, 1986; Kievlenko, 2003; Phillips and Talantsev, 1996; and Gill, 1978).

The most famous classic locality for collector-quality demantoid crystals is Val Malenco (Malenco Valley) in the Italian province of Sondrio, in the central Alps. Cossa (1880) provided the first description of the Val Malenco "green garnets." Several localities in the valley have yielded significant specimens in the 20th century, but the best of them came from the now-abandoned Sferlun asbestos mine. In 1947 and again during the second half of the 1960s, well-formed demantoid crystals were collected there from asbestos-filled fractures cutting across serpentinite outcrops (Sigismund, 1948; Bedogne and Pagano, 1972: Amthauer et al., 1974; Bedogne et al., 1993; Bedogne et al., 1999).

Demantoid has always been admired and sought after by mineral collectors. Demantoid gem rough is known from the famous deposits in the Ural Mountains, in Russia (Phillips and Talantsev, 1996), the area west of the Erongo mountains in Namibia (Johnson and Koivula, 1997; Von Bezing et al., 2007), and the Soghan area, Kerman Province, southern Iran (Laurs, 2002; Karampelas et al., 2007), but these localities have only rarely produced collectible crystals. Several other localities in the Italian Alps, San Benito County (California), the Stanley Buttes area (Arizona), Korea, the Democratic Republic of the Congo, and Armenia have in the past produced demantoid gem rough and crystal specimens of minor interest (Rouse, 1986; Kievlenko, 2003). During the last decade the Lac d'Amiante mine in the Thetford Mines area of Quebec has occasionally produced attractive, translucent, slightly grayish green demantoid crystals including a few very rare specimens of world-class quality (Amabili et al., 2009). Despite all these occurrences, Val Malenco has remained the most important source of attractive collectible demantoid crystals until the recent discovery of the Madagascar occurrence.




Topazolite is another variety of andradite, characterized by a brownish orange color, sometimes with a greenish tint which grades into yellow-green demantoid (Gramaccioli, 1975; Rouse, 1986). Topazolite was initially described as a new species by Bonvoisin (1806), the name alluding to the topaz-like golden color and brilliance of crystals from the Ala Valley in Italy; it was later found to be identical to andradite.


Before the discovery of the Madagascar occurrence, topazolite was only very rarely found as attractive, collectible crystal specimens; moreover, this garnet variety has generally been considered to be of no gem interest (Anderson, 1975) and information in the literature is rather scarce. Rouse (1986) reports that the known demantoid occurrences have commonly produced much more yellow to greenish yellow andradite than green crystals that would qualify as demantoid. However, fine, pure yellow andradite crystals without any tint of green or brown are unusually rare. Rouse (1986) also reported topazolite from San Benito County, California.


The classic occurrence in Val d' Ala (the Ala Valley) in Piedmont, Italy remains one of the most significant sources of topazolite crystal specimens; Ala Valley topazolite has been found in crystals occasionally exceeding 1 cm in diameter, but only the smallest crystals are of gemmy quality. The best recent discovery consists of translucent yellow-brown crystals of topazolite to nearly 2 cm, found a few years ago by some local collectors at Monte Civrari in the Susa Valley, Piedmont. The composition of these garnets seems to be intermediate between andradite and grossular (Piccoli et al., 2007). Some significant gemstones of topazolite were cut from rough material found after 1997 in the demantoid deposits discovered west of the Erongo Mountains, in Namibia (Gernot Smolle, personal communication).


Black andradite, though not a gem material, has long been known among mineral collectors by the varietal name of melanite; it occurs in sharp, lustrous crystals in metamorphic rocks and in undersaturated alkaline igneous rocks. A significant titanium content ([TiO.sub.2], exceeding 5%, see discussion in Meagher, 1980) is responsible for the black color.

Gem varieties of andradite generally have compositions close to the ideal end-member composition (Adamo et al., 2009; Adamo et al. in preparation); their color is controlled by stoichiometric [Fe.sup.3+] (green-yellow) and by traces of [Cr.sup.3+] (green) and probably [Ti.sup.3+] and [Ti.sup.4+] (brown color, see discussion in Meagher, 1980, for the presence of [Ti.sup.3+] and [Ti.sup.4+] in andradite).


Concerning the yellow and yellow-brown color varieties of gem andradite, there is general agreement among gemological writers on the varietal name topazolite, regardless of the causes of the color.

Rondeau et al (2009) emphasized that there is no agreement in the literature about the definition of the varietal name demantoid. Some authors, such as Webster (1975), Bariand and Poirot (2004), and Rossman (2009), consider demantoid to be an andradite colored green by traces of chromium; other authors such as Anderson (1975), the GIA's Gem Reference Guide (1988), and the AGI's Glossary of Geology by Neuendorf et al. (2005a) define demantoid as a green to yellow-green andradite without requiring the presence of trace amounts of chromium as the chromophore.

Microchemical and spectroscopic studies (Rondeau et al., 2009) of gem andradite crystals from Antetezambato, Madagascar--with color ranging from vivid green to yellow green to brown--attribute the green color to the presence of [Fe.sup.3+], and the brownish color to the charge transfer effect between iron and titanium (Fritsch and Rossman, 1993). On the basis of these results, Rondeau et al. (2009) suggest that the name demantoid be used for the green variety of andradite regardless of the chromophores(s) involved. The present author is in agreement with this suggestion.


It is interesting to note that demantoid from Russia can be heat-treated at temperatures between 640[degrees] and 720[degrees]C, in a graphite-powder-reducing atmosphere, to enhance the color (Stephenson and Kouznetsov, 2009). This treatment is able to minimize, if not totally eliminate, the yellow component of the color, yielding gemstones of a pure green color. Studies regarding the effect of such a treatment on Madagascar demantoid are in progress.


Specimen Types and Sizes

Andradite [[Ca.sub.3][Fe.sub.2.sup.3+]([SiO.sub.4]).sub.3]] crystals of the demantoid and topazolite varieties are found at Antetezambato as detached crystals and crystal groups, but most commonly as matrix specimens in ail sizes from thumbnails to large cabinet specimens. Crystal size ranges from a few millimeters to nearly 3 cm in diameter, topazolite crystals being generally of the largest size. The luster is normally good and some varieties are characterized by crystals with extraordinarily bright faces.


The garnet crystals occur in cavities of various types, depending on the size of the vein and the structure, texture and mineralogy of the altered sedimentary rock. Each vein or vein-system produces distinctive mineral specimens with garnets having their own particular crystal morphology, color, brightness, and associations.

In massive impure sandstone, garnet veins contain fissure-shaped cavities 2 to 30 mm wide and up to a meter or more in length. The best garnet crystallizations occur at the intersections of two or more of the fissure-like cavities. Such fissures are generally lined with microcrystals of pale green garnet, on which are isolated large crystals and clusters of gem varieties of andradite.

In the more structurally and compositionally heterogeneous sediments, garnet veining is more complex and large volumes of rock can be altered. At Antetezambato, quartz and chalcedony together with very fine garnet aggregates of whitish color are abundant, and large cavities up to several cubic meters in volume can occur. Such cavities, known among the miners as "caves," are partially filled by fine-grained corroded quartz mixed with clay minerals. This pocket fill typically contains large quantities of irregularly shaped fragments of snow-white matrix carrying crystals of gem varieties of andradite and also crystals of quartz.

A wide variety of other types have also been found, including specimens with (1) spongy grey matrix, (2) brecciated matrix, (3) spongy matrix composed of fine to medium-grained green crystals of garnet, (4) spongy matrix composed of partially altered fossiliferous sandstone, (5) massive matrix composed of fine-grained aggregates of garnet and calcite, crosscut by coarse-grained veins of spathic calcite, and so on. Isolated crystals or crystal groups of gem varieties of andradite are abundant in all of these occurrences.


Preliminary data have been published by researchers at the gemological school of the University of Nantes (France) (Mocquet et al., 2009a and 2009b; Rondeau et al., 2009a and 2009b). The analytical results reported in these publications, obtained using an EDS microprobe, and those made by the author, obtained using both EDS and WDS microprobes, indicate that the garnet crystals from Antetezambato are always very pure andradite, regardless of their colors. For the detection and quantification of trace elements, new investigations using the laser ablation technique are in progress by the author in collaboration with the researchers of the CNR of Milano. Thus far, the only available data regarding trace elements in gem crystals of andradite come from one EDS microprobe analyses (made by Giancarlo Parodi of the Museum de Histoire Naturelle, Paris) on one inclusion inside a garnet crystal, containing significant quantities of chromium. Some EDS microprobe spectra (requiring confirmation) indicate very minimal traces of titanium in one brownish crystal (topazolite), and titanium, vanadium and copper in one vivid green crystal with a bluish tint (demantoid). Small and opaque crystals of pale green, pale gray, pale yellow and pale pink color covering the matrix of some specimens were shown to be andradite with a significant content of aluminum (the grossular component).



Gem andradite crystals from Antetezambato can be divided into (1) the demantoid variety of yellow-green to army-green, grass-green, emerald-green, and bluish green color, often of low to moderate saturation, and (2) the topazolite variety of honey-yellow, amber-yellow, brownish yellow and brown color. Some rare crystals of brownish red color are known, occasionally even in gem quality, and must be referred to simply as andradite.

Transparent, color-zoned crystals are rather common, typically with a green or greenish core and a more brownish rim. The first generation of garnet growth was greener in color whereas the second generation was more yellow-brown. In some veins it is not uncommon to find crystals with a topazolite zone surrounding a demantoid core. Thick polished sections of such crystals reveal a complex multiple color zoning which indicates inhomogeneities in the distribution of the trace elements (still to be defined, as previously discussed), as described in Meagher (1980). As pointed out by Rondeau et al. (2009a), between crossed polarizers, these garnets show an anomalous double refraction with undulating extinction and the so-called "tatami" features. This phenomenon is typical of andradite and has been described several times in the literature (e.g. Meagher, 1980).

An alexandrite-like color-change resembling that described in demantoid from some other localities (e.g. Amabili et al., 2009) is also seen in some demantoid and topazolite crystals from Antetezambato. This effect is more evident for crystals which are distinctly greenish yellow under daylight and become reddish yellow under incandescent light. The unusual bluish hue of some green crystals is evident only in daylight and not under incandescent light. This effect seems to be rather different from the one described by Ahn and Park (2008) and it is probably the result of a different mechanism.

Crystal Morphology

Andradite crystals from Antetezambato display an exceptionally interesting variety of crystal habits and forms. The most common forms are the dodecahedron {110} and trapezohedron {211} in various combinations, the first being more typical for demantoid and the second for topazolite. Cube faces {100} are very rare and have been observed by the author only in two demantoid crystals. Many crystals, mostly of the topazolite variety, show modifications by several hexoctahedral forms and, in rare cases, the crystals can display only a single hexoctahedron form. Goniometric measurements have yet to be performed, but computer-generated drawings using the KRISTALL 2000 program can closely approximate the Antetezambato crystals with combinations of the {321}, {532} and {954} hexoctahedra. A few topazolite crystals recovered in December 2009 are modified by the trisoctahedron {332} and tetrahexahedron {210}.


Other unusual and interesting morphological variations have been observed in Antetezambato andradites, including:

(1) Transparent trapezohedral topazolite crystals showing at the center a large green demantoid phantom with dodecahedral form.

(2) Large trapezohedral topazolite or army-green demantoid crystals growing on the end of a small grass-green dodecahedral demantoid crystal which is elongated along the [111] axis, resulting in a scepter shape.

(3) Two dodecahedral demantoid crystals growing on the ends of another dodecahedral demantoid crystal elongated along the [111] axis, resulting in a dumbbell shape.

(4) Some crystals of topazolite found in small fissures have developed extremely flattened shapes, with or without skeletal growth. Such crystals, occasionally very gemmy and with very good luster, can reach a diameter of over 3 cm with a thickness of only 1 mm.


Quartz [SiO.sub.2]

Quartz is the most common mineral associated with andradite crystals (mostly of the demantoid variety) at Antetezambato. Where demantoid crystals are of high quality, the quality of the associated quartz tends to be low, and when quartz occurs in beautiful crystals, the quality of the demantoid is low. Quartz crystals are generally elongated, dull-lustered and rather milky. They can reach 10 cm in length and, in a few cases, they display traces of amethyst color. Typically flattened Japan-law twins are found as well.

Chalcedony of a whitish color is present in some garnet veins as a constituent of the matrix hosting the andradite crystals, more or less in association with quartz. In some cases, chalcedony in globular whitish and translucent aggregates also encrusts demantoid and topazolite. Occasionally, what appear lo be chalcedony masses of lower density have been found, but these are actually opal.



Calcite [CaCO.sub.3]

Calcite is common in only a few veins in the deposit. The most typical crystals are milky white rhombohedrons rimmed by a short hexagonal prism. The normal size is around 1 cm in diameter but, one pocket found in November 2009 contained lustrous, transparent calcite crystals to 10 cm.




Stilbite-Ca [NaCa.sub.4][[Al.sub.9][Si.sub.27][O.sub.72]].28[H.sub.2]O

Stilbite-Ca is occasionally present as isolated crystals or as crystal druses up to 5 mm in length encrusting demantoid and topazolite. This association can produce rather aesthetic specimens.


Pyritc [FeS.sub.2]

Pyrite is present inside the sedimentary rock as aggregates and swarms of cubic crystals up to a few millimeters across. Pyrite and its alteration product (goethite) are occasionally found encrusting crystals of andradite in cavities.

Wollastonite (?) [CaSiO.sub.3]

Sprays of whitish needles and fibers visible as inclusions in gem crystals of demantoid and topazolite arc probably wollastonite. But preliminary microprobe analyses and micro-Raman spectroscopy on polished sections have been unable to confirm the presence of wollastonite, though quartz was identified. It is possible that the original fibrous mineral has been replaced by quartz via hydro-thermal alteration.


Sometimes the natives glue garnet crystals onto suitable matrix that may already be carrying crystals of low quality. The specimens can be difficult to recognize, as the natives have developed some sophisticated techniques using different types of glue and cement mixed with dust collected from the pockets. Such fakes circulate both in the area of the mine and in the capital city, and are offered mainly to inexperienced tourists.


The author would like to thank the following people for their assistance in the researching and preparation of this article: Giuseppe Pocobelli (Pocobelli and Son Company, Antananarivo, Madagascar) and his coworkers helped with the logistics of our visit to the Antetezambato mine, obtained official authorization necessary for our visit, and provided information on collecting. Rakotonirina Georges (Direction of Mines for the Ministry of Energy and Mines in Antananarivo, Madagascar) helped establish contacts with the local authorities. Vadim Fedder (Mineral Treasure, Antananarivo, Madagascar) arranged for the exportation of part of the studied samples. Laurent Thomas (Polychrome France, Chambray-Les-Tours, France) provided the first fragments used in our analytical studies. Valeria Diella (CNR, Milano, Italy) provided preliminary information about microprobe analyses. Wendell Wilson reviewed and revised the manuscript, and Anthony Kampf reviewed the text and provided helpful suggestions. Daniel Trinchillo (Fine Minerals International) and Marco Amabili partially supported the study.


ADAMO, I., BOCCHIO, R., DIELLA, V. PAVESE, A., VIGNOLA, P., PROSPER, L., and PALANZA, V. (2009) Demantoid from Val Malenco, Italy: Review and Update. Gems & Gemology, 45 (4), 280-287.

AHN, Y. K., and PARK, J. W. (2008) A Photoluminescence study of the alexandrite effect in gemstones. GIT2008, The 2nd International Gem and Jewelry Conference. 98-100.

AMABILI, M., SPERTINI, F., AUGUSTE, M. B., and BONIN, G. (2009) The Lac d'Amiante mine. Mineralogical Record, 40 (4), 297-304.

ANDERSON, B. W. (1975) Identification des Pierres Precieuses Hugo Buchser Ed., Geneve, 422 p.

AMTHAUER, G., KURTZ, W., ROST, F., and SCHLOEMER, H. (1974) Chemismus und gencses des andradits aus dem serpentinit des Val Malenco (Bernina-Gebiet/Oberitalien). Schweizerische Mineralogische und Petrographische Mitteilungen, 54, 691-706.

BARIAND, P., and POIROT, J. P. (2004) Larousse des Pierres Precieuses. Larousse Ed. Paris, 290 p.

BEDOGNE, E, and PAGANO, R. (1972) Mineral collecting in Val Malenco. Mineralogical Record, 3, 120-123.

BEDOGNE, F., MONTRASIO, A., and SCIESA, E. (1993) I Minerali della Provincia di Sondrio. Valmalenco. Bettini, Sondrio, 275 p.

BEDOGNE, F., SCIESA, E., and VIGNOLA, P. (1999) II "demantoide" della Val Malenco. Rivista Mineralogica Italiana, 23 (4), 207-217.

BESAIRIE, H., and BRP-IFP (1962) Nosy-Be--Ambato. Geologic Map 1/100,000 R33-S33. Service Geologique, Antananarivo, Madagascar.

BRGM (1985) Plan Directeur d'Actions pour la Mise en Valeur des Resources du Sol et du Sous-Sol de Madagascar. Premiere Phase, Premiere Partie. Directions des Mines et de la Geologie, Ministere de 1'Industrie, de 1'Energie et des Mines de Madagascar, 573 p.

BONVOISIN, C. B. (1806) Description du Peridot Idocrase ct de quatre autre substance lithologiques d'espeee nouvelle, nominees: Succinite, Mussite, Alalite, Topazolite, decouvertes dansles valleess de Lans, department du Po, en Piemont. Journal de Physique, 42, 409-429.

COSSA, A. (1880) Sopra un granato verde di Val Malenco. Gazzetta Chimica Italiana, 10, 469.

DANET, F. (2009) New discovery of demantoid from Ambanja, Madagascar. Gems & Gemology, 45 (4), in press.

FEHR, T. (2008) Insert at page 37 in "Garnet, Great Balls of Fire." Lithographie, LLC, Connecticut. 100 p.

FRITSCH, E., and ROSSMAN, G. R. (1993) The causes of color in garnets. The Mineralogical Record, 24, 63.

GEMOLOGICAL INSTITUTE OF AMERICA (1988) Gem Reference Guide, Gemological Institute of America, Santa Monica, CA, 270 p.

GILL, J. O. (1978) Demantoid: the complete story. Lapidary Journal, 32 (7), 1542-1545.

GRAMACCIOLI, C. M. (1975) MineraliAlpine e Prealpini. Istituto Italiano Edizioni Atlas, Bergamo, 473 p.

KARAMPELAS, S., GAILLOU, E., FRITSCH, E., and DOUMAN, M. (2007) Les grenats andradites--demantoides d'Iran: zonages de couleur et inclusions. Revue de Gemmologie a.f.g., 159, 14-20.

KIEVLENKO, E. Y. (2003) Geology of Gems. Arthur Soregaroli Ed., Ocean Pictures Ltd, 432 p.

JOHNSON, M. L., and KOIVULA, J. I. (1997) Gem News: Demantoid garnet from Russia and from Namibia. Gems & Gemology, 33 (3), 222-223.

LACROIX, A. (1922) Mineralogie de Madagascar. Augustin Challamel editeur, tome 1, 624 p.

LAURS, B. (2002) Demantoid garnet from. Iran, in Gem New International. Gems & Gemology, 38 (1), 96.

MEAGHER, E. P. (1980) Silicate garnets. Reviews in Mineralogy (Mineralogical Society of America), 5, 25-66.

MERCIER, A., MOINE, B., DELORME, J., and RAKOTONDRAZAFY, M.A.F. (1997) A note on a new occurrence of vanadian grossular garnet from Madagascar. Journal of Gemmology, 25 (6), 391-393.

MOCQUET, B., LULZAC, Y., and RONDEAU, B. (2009a) Nouveau gisement de andradite demantoide a Madagascar. GEM Nantes, May 30, 2009.

MOCQUET, B., LULZAC, Y, RONDEAU, B., FRITSCH, E., Le QUERE, J., MOHAMADY, B., CRENN, G. LAMIRAUD, C., and SCALIE, S. (2009b) Ambanja, premier gisement d'andradite demantoide gemme a Madagascar. Revue de Gemmologie a.f.g., 169, 4-8.

NEUENDORF, K. K. E., MEHL, J. P. Jr., and JACKSON, J. A. (editors) (2005) Glossary of Geology. American Geological Institute, Alexandria, Virginia, p. 170.

PHILLIPS, W. R., and TALANTSEV, A. (1996) Russian demantoid, Czar of the garnet family. Gems & Gemology, 32 (2), 100-111.

PICCOLI, G. C, MALETTO, G., BOSIO, P., and LOMBARDO, B. (2007) Minerali del Piemonte e della Valle d'Aosta. Associazione Amici del Museo "F. Eusebio"--Alba, 607 p.

PRASZKIER, T. and GAJOWNICZEK, J. (2010) Demantoide aus Antetezambato auf Madagascar. Mineralien Welt, 21, 32-41.

RAZAFINDRAMIADANA, L. (2009) Un gisement de demantoide a Ambanja. Express de Madagascar, Antananarivo, 06-06-2009.

RONDEAU, B., FRITSCH, E., MOQUET, B., and LULZAC, Y (2009a) Ambaja (Madagascar)--A new source of gem demantoid garnet. In Color, 11, 22-24.

RONDEAU, B., MOQUET, B., LULZAC, Y, and FRITSCH, E. (2009b) Les nouveaux grenats demantoides d'Ambanja, Province d'Antsiranana, Madagascar. Le Regne Mineral, 90, 41-46.

ROSSMAN, G. R. (2009) The geochemistry of gems and its relevance to gemology: different traces, different prices. Elements, 5(3), 159-162.

ROUSE, J. D. (1986) Garnet. Butterworths Gem Books, British Library and Co. Ltd. (publisher), 134 p.

SIGISMUND, P. (1948) Granato e vesuvianite di Val Malenco. Atti della Societa Italiana di Scienze Naturali, 87, 145-171.

STEPHENSON, J., and KOUZNETSOV, N. (2009) Major deposits of demantoid around the world. In Color, 11, 16-20.

STOCKON, C. M., and MANSON, V. (1983) Gem andradite garnets. Gems & Gemology, 19 (4), 202-208.

VON BEZING, L, BODE, R., and JAHN, S. (2007) Namibia--Minerals and Localities. Edition Schloss Freudenstein, Bode Verlag GmbH, 856 p.

WEBSTER, R. (1975) Gems: Their Sources, Descriptions and Identification. Butterworth & Co, 938 p.

Federico Pezzotta

Museo di Storia Naturale

Corso Venezia 55, 20121 Milano
COPYRIGHT 2010 The Mineralogical, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Pezzotta, Federico
Publication:The Mineralogical Record
Geographic Code:6MADA
Date:May 1, 2010
Previous Article:Call for papers on the Minerals of California.
Next Article:Antetezambato North Madagascar.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |