Printer Friendly

Halls Gap Lincoln County, Kentucky.

Since 1964, a roadout in central Kentucky has been producing thousands of quartz geodes containing a variety of minerals including extraordinary specimens of millerite, jamborite (?) and unusual forms of pyrite.


The Halls Gap geode locality is located along U.S. Highway 27, 7 km south of Stanford, Kentucky, near the community of Halls Gap [ILLUSTRATION FOR FIGURE 1 OMITTED]. It lies near the top of an escarpment known as Muldraugh Hill. In the development of the improved highway, road engineers serendipitously uncovered what is probably the best millerite geode horizon known in the world. Geodes from this roadcut have been a favorite among midwest collectors for years.

Discovered in 1964 by members of the Cincinnati Mineral Society (Medici, 1981), the locality has produced thousands of fine display specimens and many more study specimens. This article will examine the geology, stratigraphy, origins, and mineralogy of the locality as it relates to the formation of the geodes.


The Halls Gap millerite-bearing horizons are situated near the contact of the Halls Gap and Muldraugh members of the Borden Formation in east-central Kentucky. The units are exposed near the summit, known to geologists as Muldraugh Hill, which borders a section of Kentucky's Bluegrass region to the south and west. The Borden is apparently the result of deltaic conditions similar to those of the Borden siltstone in Illinois, as described by Swann et al. (1965) and Lineback (1966).

The Halls Gap occurrence lies near the apex of the Cincinnati Arch, a large anticlinal fold that runs generally in a north-south direction. It dips toward the east and west. The crest of the Arch, called the Jessamine Dome, is located about 40 km north. The Cincinnati Arch trends southeast to another structural high called the Nashville Dome. Mississippi Valley-type deposits associated with the Cincinnati Arch include the central Kentucky fluorspar district and the zinc deposits in Smith County, Tennessee.

In Kentucky, the Borden Formation consists of, from the base upward, a sequence of shales, siltstones and silificied limestones which differ locally in proportion (Weir, 1970). In the Halls Gap area, it is conformably underlain by New Albany Shale of Late Devonian or Early Mississippian age, and overlain by the Salem Limestone of Middle Mississippian age. These units are quite evident in the roadcuts exposed on U.S. Highway 27 near the town of Halls Gap. Near the top of the hill the contact between the Halls Gap and Muldraugh member is exposed. It is marked by a thin seam of dark green glauconitic shale (Weir, 1970) which is persistent throughout the Midwest. It was deposited contemporaneously over a vast area, extending from Ohio to Alabama and from Virginia to Kansas: it may record a period of extremely slow deposition caused by a sudden rise in sea level (Whitehead, 1976).

The overhanging ledge in the road cut represents the basal unit of the Muldraugh Member, whereas the face underneath, which has been driven back by years of collecting, represents the top-most units of the Halls Gap Member. Millerite-bearing geodes occur most abundantly in the 1.4-meter-thick unit, which is third below the contact. Rarely, millerite-bearing geodes are found in the overhang (Muldraugh); infrequently they are found in the first two units below the contact, and just slightly more frequently in the fourth and fifth units below the contact. There are two areas within the 1.4-meter unit (what is called the "millerite zone" throughout this article) that are unusually rich in geodes, including larger geodes up to 15 centimeters in diameter. One is located about 30 centimeters below the top of this unit, the other about 30 centimeters above the bottom.

Geodes are distributed throughout the strata. The density of distribution varies for each stratum, but within each unit is relatively uniform. Clustering of geodes has been observed, especially in the overlying Muldraugh member. Clustering may be lateral or vertical. Vertical clusters may resemble carrots or chimneys. Some geodes share common walls. Clusters of two to four geodes with shared walls are common, especially in the glauconitic layer. In larger geode clusters (i.e. four or more) the contents of one is a reliable guide to the mineralogy of an adjacent geode.

The millerite-bearing portion of the outcrop is characterized by alternating layers of siltstone of two lithologies: shaley and massive. The lowest exposed interval is a shaley siltstone containing very few geodes. Immediately above this is a massive siltstone ledge, which is probably dolomitic. The unit contains geodes, predominantly carbonate-filled (calcite, dolomite and ferroan dolomite or dolomite with oxidized pyrite). Much of this layer is covered with talus from collectors working the layers above.

The "millerite zone" is capped above and below by a shaley siltstone. The lower siltstone is rarely exposed. The eastern side of the roadcut is kept covered with debris from collectors working higher layers. The western side is heavily overgrown, but is relatively accessible during winter months.

The upper shaley siltstone is 25 to 30 cm thick and is very glauconitic. The rock shows extensive fossilized worm burrows filled with glauconite. The entire shaley siltstone horizon is a deep, bright green. Geodes collected from this layer are easily recognized by their distinctly green exterior.

Above the "millerite zone" is the massive ledge, almost 1.5 meters in thickness, which forms a precarious overhang where collectors have mined geodes. The geodes from this massive siltstone are frequently of large diameter, and there is an overall greater volume of geodes in this layer as compared to the "millerite zone." Solid quartz nodules and geodes to 15 cm in diameter are found in equal abundance. Marcasite nodules from 1 to 15 cm are especially common in this layer; reaction halos due to weathered marcasite form multicolored rings on the massive siltstone. (Such nodules are not found in the "millerite zone.")

Of the geodes found in this layer about a third contain quartz only. Another third are calcite-filled or contain calcite only. The remaining third contains sulfides (pyrite dominating) and may be partially or wholly filled with calcite. One geode in 50 may contain dolomite. Millerite is uncommon in this layer. Millerite epitaxially growing on chalcopyrite or (rarely) sphalerite is more common than as free-standing filaments. Sphalerite crystals to 1 cm occur with smaller pyrite and chalcopyrite crystals encrusting them. Pyrite cubes modified by the octahedron are predominant, though octahedral and pyritohedral forms may be found. Cubes elongated in one or two directions may also be found. Some of the pyrite is beautifully iridescent, much more so than in the "millerite zone." Among geodes in the Muldraugh and upper Halls Gap members, chalcedony is more abundant than drusy quartz (about 10 to 1).

The highest units in the sequence contain fewer geodes. The lithology of these layers consists of massive siltstones in horizontal layers, along with cut-and-fill channel deposits. They are largely inaccessible, and horizontal exposures are relatively small.

The concentration of sulfide-bearing geodes from this locality is similar to a low-grade Mississippi Valley-type deposit. By low-grade, we mean a similar mode of development but low in sulfide minerals. The usual minerals in Mississippi Valley-type deposits are galena, sphalerite, chalcopyrite and pyrite (all present here in varying amounts). Mississippi Valley-type deposits are not associated with igneous rocks, nor are the minerals thought to have been deposited with the enclosing rocks. Instead, this type of ore deposit is believed to originate by gradual leaching of metals from basinal sediments by hot brines moving upslope through the sediments. The original metal content may be very low - the lead for galena, the most important Mississippi Valley-type ore, is derived from trace amounts of lead in potassium feldspar! The metals are deposited where they encounter sulfide ions. In the case of geode minerals, the sulfide may be derived from the reduction of sulfate from precursor minerals such as gypsum and anhydrite (D. Coskren, personal communication, 1991). The full lateral extent of sulfide-bearing geodes on the Muldraugh escarpment is not known.

For comparison, another geode-bearing deposit located on a structural high is found on the Muldraugh Dome in Meade County, Kentucky (Goldstein, 1989, 1990). All sulfide minerals (primarily pyrite) have altered to an unknown, bladed mineral associated with powdered goethite. Curiously, fluorite crystals are found in the geodes on this structural feature, although they are almost unknown in other occurrences.

There are several geode-bearing outcrops farther down the road from the main Halls Gap exposures and lower in the stratigraphic section. However, only the millerite-bearing roadcut is dealt with here in detail. The next roadcut downhill is also in the Halls Gap member, and contains sulfides and abundant dolomite. The sulfides, which are not as spectacular as in the "millerite zone," consist primarily of pyrite and chalcopyrite. Some of the chalcopyrite has been altered to malachite.

Halls Gap is only one of many millerite geode occurrences across the Midwest. Other well-known localities include those near Keokuk, Iowa (perhaps the "old classic" locality); Hamilton, Illinois; Bedford, Indiana; and St. Louis, Missouri. Most of these millerite occurrences appear to have, in part, a stratigraphic control. The common thread that connects most of them is this: they are at or near the top of a thick pile of deltaic sediments, deposited in late Osagean (Middle Mississippian) time.


The geodes at Halls Gap range from spherical to flattened and cylindrical in shape. Some geodes, especially in the Muldraugh Member, are associated with marcasite nodules. These geodes present storage problems, because the marcasite deteriorates. The chalcedony rind prevents the decomposing marcasite from attacking the geode's interior, but the acidic byproduct of the decomposition is deleterious to storage containers made of paper or cardboard.

Some geodes have a double wall. The outer wall may contain pyrite, frequently associated with red chalcedony. The gap between the two walls is seldom more than 3 mm. The interior of the inner geode is frequently solid quartz.

No systematic attempt has been made to plot the lateral distribution of minerals, however more millerite has apparently been found in geodes on the east side of the roadcut (as indicated by the heaviest collecting efforts). The western roadcut has not been as extensively collected; there is one relatively small "mined" area.

One of us (B.M.) has collected sufficiently to gain an understanding of some lateral mineralogical variation. Three occurrences are noteworthy. The north end of the eastern side of the roadcut has an odd form of calcite resembling blisters. The second area lies about 15 meters south, at the comer where collectors have removed the most rock. Here, large octahedral pyrite crystals are abundant. The northern 15 meters of the eastern roadcut has produced 75% of the jamborite. In the same area is a fairly persistent layer of large (to 5 cm) calcite rhombs and calcite fills with some millerite and other sulfides without calcite. This layer is in the lowest 25 cm above the lower shale unit. No other significant mineralogical differences have been noted within the outcrop.

The source of nickel in the local mineralogy is thought to be iron-nickel meteorites. Since meteoritic material continually "rains" down to earth as microscopic dust grains and as occasional meteorites, this source seems at least plausible. There are large cryptoexplosive structures in Shelby and Jessamine Counties, north of Halls Gap, 88 and 69 km respectively. The Middlesboro structure lies 125 km southeast, and the Serpent Mound structure in Adams County, Ohio, and several similar structures in Tennessee are located within a 160 km radius. Any or all of these probable meteorite impacts could have contributed quantities of nickel to the surrounding rock layers through groundwater over tens of millions of years.

D. Coskren (personal communication, 1991) proposes that the glauconite could be a reservoir for nickel. As an iron-magnesium sheet silicate, it could accommodate nickel in substitution for the iron and magnesium. It has been shown to act as a "scavenger" for heavy metals.

Carbonate minerals were deposited from meteoric water acting on limestone deposits. They are widespread in geodes throughout the region.

The origin of the geodes has been a subject of speculation for many years. One popular theory was developed by Chowns and Elkins (1974). Their studies indicate that geodes were preceded by anhydrite or gypsum nodules. The source was thought to be hypersaline water formed in an enclosed unreplenished basin. (Such an area is called a sabkha.) However, the nodular gypsum or anhydrite that preceded the geodes could not have formed in a restricted basin because the depositional characteristics of the rock that contains these geodes does not match the type found in a sabkha. Many geode-bearing rocks in the area (though not at Halls Gap) are richly fossiliferous. Hypersaline seas are not conducive to an abundance of life. The Borden delta, the depositional feature in which the Halls Gap geodes occur, has been studied by numerous geologists. None describe the formation as being a shallow, hypersaline basin.

Where, then, was the source of the anhydrite/gypsum? Dodd et al. (1987) review previous papers on the subject which were not seen - Maliva (1985, 1986) and Dodd et al. (1984). They theorize that the source of the anhydrite or gypsum was most likely the overlying St. Louis Limestone. That formation has been eroded away at Halls Gap. Throughout much of Kentucky and Indiana, the St. Louis Limestone has features characteristic of sabkha deposition. At localities in north-central Kentucky and south-central Indiana massive beds of gypsum with lesser amounts of anhydrite can be found at depth. The brines generated during the time of the deposition of the St. Louis Limestone are thought to have leached down from the overlying evaporites to form the nodules that would eventually become geodes. The brines would have been saturated with gypsum or anhydrite and would have had a high magnesium/sulfate ratio. This would lead to dolomitization (which has been observed in strata of comparable age). As dolomite is formed, calcium is released to combine with the sulfate to form gypsum or anhydrite. As sulfate ions are removed (combined with calcium), additional dolomitization would occur, which would continue the cycle.

At some point after the rock was formed, the anhydrite or gypsum was replaced by chalcedony on the outer edges, forming a geode shell. Further sulfate dissolution may have created a void in which addition mineral deposition could occur. It is possible that the brecciation observed in the inner layers of chalcedony, found in many Halls Gap geodes, was the result of the sulfate being removed, combined with lithostatic pressure compressing the geodes. Slumping and fracturing of the chalcedony, possibly from dissolution, occurred before additional minerals were deposited.


Anatase Ti[O.sub.2]

Anatase has been reported as red to butterscotch-yellow blades on sugary quartz (P. Smith, personal communication, 1991). It was identified by Dr. Harvey Belkin (U.S. Geological Survey) by electron microprobe analysis. Crystals are very tiny, requiring 30 to 60x magnification to be seen well. Anatase-bearing geodes appear to be restricted to the glauconitic layer at the contact of the Halls Gap and Muldraugh Members.

Calcite CaC[O.sub.3]

Calcite is a common mineral which often completely fills the interior of geodes. Crystals form low-angle rhombohedra that range from less than 1 mm to greater than 5 cm across. The larger crystals show internal zoning due to inclusions of a dark mineral or carbonaceous substance. Crystals can be single or intergrown in flower bud-like forms. Calcite crystals are typically translucent white and do not fluoresce. Crystals have occasionally been found impaled on millerite. Overlapping relationships show that calcite is usually one of the last minerals to form. Sulfides, particularly chalcopyrite are found suspended inside calcite. Some calcite may have minute pyrite crystals sprinkled on the faces. The most attractive calcite is found contrasting sharply against bluish black chalcedony.

Calcite may be locally colored green by jamborite. It is unknown whether the color is a stain or a result of microscopic acicular crystals penetrating the calcite.

Rarely, calcite occurs as a crust of minute, intergrown, rhombic crystals. This crust may be observed directly on chalcedony and occasionally forms hemispherical bubbles. The fragile nature of this material may lead to its partial destruction when the geode is cracked open with a hammer. The finest examples of these calcite "blisters" occur on chalcedony with pyrite cubes sharply modified by the octahedron. Associated with this calcite is another kind that may have formed by dissolution. Although 10x is necessary to see it well, this bizarre calcite resembles colorless to white flos-ferri aragonite.

Geodes collected at Halls Gap that are completely calcite-filled may be carefully etched with weak hydrochloric acid, as sulfide minerals may then be exposed. In addition to "frozen" sulfides previously mentioned, sprays of millerite on chalcopyrite, individual millerite crystals, marcasite and pyrite may be found in otherwise bland-looking geodes. Microcrystals (single and intergrown) may be collected by the dozens in some of these geodes.

Chalcocite (?) [Cu.sub.2]S

Chalcocite was described from the Halls Gap area by Chromy (1972) as occurring in geodes. Chromy's report was not seen by the writers. However, one of us (B.M.) has a specimen which, based on visual identification, appears to contain chalcocite. This identification requires confirmation.

Chalcopyrite CuFe[S.sub.2]

Chalcopyrite is an abundant sulfide mineral at Halls Gap, particularly in the "millerite zone." Crystals range from microscopic to 3 mm on edge. The habit is a simple disphenoid, although some crystals appear to be etched or to have unusual modifications. Some crystals form "flags on a pole," elongated in one direction and serrated along the edges. Intergrown chalcopyrite crystals are often found. Chalcopyrite is usually a brass-yellow color. Some of the most spectacular crystals show a range of iridescent colors, from red to magenta, green, blue and purple. Some are even a lustrous black.

The most curious chalcopyrite associations are with millerite. Chalcopyrite-millerite associations tend to be complex. Chalcopyrite probably formed in several episodes, at least one generation having formed after the millerite. Tiny crystals are found rarely in the midst of millerite strands or perched at the end of a single acicular crystal. Small disphenoids may be skewered by several millerite crystals. A 2-mm millerite crystal is known with more than a dozen very minute chalcopyrite crystals skewered across its length. Despite these rare examples, the bulk of the observed chalcopyrite appears to be among the earliest sulfides in the paragenetic sequence. Chalcopyrite may be coated by millerite, and some examples show a few millerite crystals which appear to protrude from the chalcopyrite. In other examples, millerite growth all but obscures the nucleating crystal. Some forms show a preferred orientation of millerite with respect to the chalcopyrite crystal faces. Ruez (1973) indicates that millerite occurs near the edges of chalcopyrite crystals. Polished sections showed substantial corrosion, probably prior to millerite deposition. Etched and corroded chalcopyrite crystals without millerite are common.

Copiapite (?) [Mathematical Expression Omitted]

Copiapite is suspected as a decomposition product of pyrite and sphalerite (D. Coskren, personal communication), occurring as yellowish scales and crusts with these sulfides. A similar material associated with chalcopyrite could be copiapite or cuprocopiapite. Identification of these minerals requires confirmation.

Dolomite CaMg[(C[O.sub.3]).sub.2]

Dolomite is less common than calcite at Halls Gap. Crystals occur in white to pink saddle-shaped aggregates. The interiors of geodes may be completely filled with pink dolomite. It formed late in the paragenetic sequence, though dolomite may be sprinkled with tiny pyrite cubes and may be stained orange or red by pyrite oxidation products. The best dolomite is found below the "millerite zone."

Galena PbS

Galena is very rare in the sulfide zone at Halls Gap, occurring with less frequency than jamborite. Crystals are typically less than 1 mm across. The cubic form has been observed, though individual crystals may not be equi-dimensional on all axes. The metallic gray color may resemble silvery pyrite, but is usually distinctive. The texture of the crystals is most often granular, whereas pyrite is smooth. Of the half-dozen crystals observed, only one cube has lustrous faces. It is possible that the granular "galena" may be an alteration to anglesite (PbS[O.sub.4]). Steve Garza (personal communication) indicates that a granular texture is common for anglesite after galena. Scanning electron microscopy has shown the granular texture to be from etched faces. The only clue as to its place in the paragenetic sequence is from a single cube found on a cube of pyrite. Cubic pyrite tends to be late in the sequence.

Glauconite (K,Na)[([Fe.sup.3+],Al,Mg).sub.2][(Al,Si).sub.4][O.sub.10][(OH).sub.2]

Glauconite occurs throughout the rock as inclusions in late-forming quartz. The contact between the Halls Gap Member and Muldraugh Formation contains an extremely rich glauconitic zone with layers and pods of bright pine-green silt and clay.

Goethite [Fe.sup.3+]O(OH)

A single black lathe-like crystal of goethite (?) was observed in one geode, but was lost as it was detached. Its appearance resembled other goethite crystals found in geodes from other Kentucky localities. Goethite pseudomorphs after pyrite are common in the higher geode-bearing units. Goethite powder from decomposing pyrite is found sporadically in sulfide-bearing geodes.

Gypsum CaS[O.sub.4][multiplied by]2[H.sub.2]O

Gypsum (visual identification only) occurs rarely at Halls Gap, resulting from the decomposition of sulfides. Crystals tend to be very crudely formed and would be more accurately called a crust. The mineral is usually associated with decomposing pyrite or chalcopyrite. One specimen has very crude, white or colorless 1-mm lathes suspended in a millerite cluster.

Halotrichite [Fe.sup.2+] [Al.sub.2][(S[O.sub.4]).sub.4][multiplied by]22[H.sub.2]O

Halotrichite has been reported by D. Coskren (personal communication) as white, silky, fibrous crystals associated with millerite, pyrite and other sulfates.

Jamborite (?) ([Ni.sup.2+],[Ni.sup.3+],Fe)(O[H.sub.2])(OH,S,[H.sub.2]O)

Jamboritie (?), formerly known from Halls Gap as "honessite," is a rare alteration product of millerite; Morandi and Dalrio (1973) suggested the identification as jamborite but did not confirm it. An estimated one in 300 geodes in the "millerite zone" contains apple-green jamborite (?) pseudomorphs after millerite. There are gradations from greenish millerite with a metallic luster to a submetallic luster to the non-metallic luster that is characteristic for pure jamborite. It is not unusual for a single pseudomorph of jamborite (?) to have differing luster from one end to the other as the degree of alteration changes.

Jamborite (?) is also associated with corroded chalcopyrite crystals, forming the same "star burst" patterns observed with millerite.

SEM photos reveal jamborite (?) as ram's horns and needles similar to gypsum, but on a microscopic scale. One specimen consists of a crystal terminating in a tight curl. Like salt weathering in gypsum (which reduces limestone boulders to gravel) the expansion of jamborite (?) pushes and distorts the original millerite crystal shape.

Malachite [Cu.sub.2]C[O.sub.3][(OH).sub.2]

Malachite has been reported as pseudomorphs after chalcopyrite disphenoids in the stratum beneath the "millerite zone."

Marcasite Fe[S.sub.2]

Marcasite is not very common at Halls Gap. Ruez (1973) analyzed a dark sulfide with worm-like shapes and determined it to be marcasite of probable colloidal origin. Specimens appear darker than polycrystalline pyrite. Marcasite also occurs as bundles of elongated blades of a brassy yellow color. This habit consists of masses of intergrown crystals from 5 to 30 mm across, which may completely fill the void. Ruez describes this type vaguely as a possible pyrite pseudomorph after marcasite. More study is needed.

Isolated doubly terminated blades less than 1 mm across are found with pyrite.

Marcasite nodules are abundant in several layers in the exposure. The largest nodules may be more than 10 cm in diameter.

Melanterite (?) [Fe.sup.2+]S[O.sub.4][multiplied by]7[H.sub.2]O

Melanterite is an alteration product of decomposing pyrite. Whitish fuzz, tentatively identified as melanterite, forming 1-mm ram's horns, has been observed.

Metavoltine [Mathematical Expression Omitted]

Metavoltine occurs in 0.1-mm lemon-yellow flakey crystals associated with other sulfates including halotrichite, melanterite, roemerite and voltaire. A geode containing these minerals was obtained from a dealer by D. Coskrin.

Millerite NiS

Millerite is the mineral which has made Halls Gap famous among collectors. It is distributed widely across the Midwest and is found in vugs and geodes in several states. At no other geode locality does millerite occur in the quantity it does here. The millerite appears to be concentrated in a layer about 1 meter thick, and is especially abundant in the uppermost 15 cm of the so-called "millerite zone." About one in four geodes in this layer contains millerite.

The abundance of millerite crystals is highly variable. Some geodes contain a single filament. Other geodes may be packed so tightly with crystals that a tearing sound can be heard when separating the halves.

Millerite occurs as slender, hair-like, hexagonal crystals, highly elongated on the c-axis. The tenacity of millerite varies according to the length and diameter of the crystal. Short crystals are very brittle. Longer crystals are flexible to slightly elastic, depending on the diameter. Crystal diameters have not been measured, but there is a substantial range in thickness. Millerite crystals when uncoiled have lengths that may exceed the diameter of the geode.

Filamentary crystals may be twisted along the c-axis due to screw dislocations (W. A. Henderson, personal communication, 1992). These so-called "Eshelby" twists show variability in the degree of rotation. Some crystals are more tightly twisted than others. Smaller filaments may have several rotations per millimeter, whereas larger crystals may twist only once per millimeter. Some alternate irregularly. SEM photographs show a blade-like crystal and reveal striations parallel to the c-axis. Rings of millerite to 2 mm have also been found.

Paul Smith (personal communication, 1991) reports that a tiny white fuzz observed rarely coating millerite is actually microscopic millerite. One of us (A.G.) has observed a single millerite specimen exhibiting this same white material.

X-ray fluorescence analysis of various colors of millerite has revealed some interesting compositions (Ruez, 1973). Violet gray-coated millerite has higher than normal amounts of iron and cobalt. Greenish millerite (jamborite?) has unexpectedly high traces of platinum.

There are several unusual associations with millerite which are summarized here: apparent nucleation around chalcopyrite and sphalerite; impaled or suspended calcite, chalcopyrite, gypsum and pyrite; coatings of late-formed microscopic quartz; the alteration to a possible nickel sulfate; and violarite coatings.

Some millerite appears to be corroded or otherwise altered. The color varies from an orange rust (which may stain the interior of the geode) to a straw-yellow color. The luster of both types tend to be submetallic. Whether or not the end result of this alteration is another mineral species has not been determined. Bish and Livingstone (1981) describe from another locality a citrine-yellow encrusting film associated with chromite consisting of a mixture of reevesite, honessite and hydrohonessite. Whether a similar mixture is present at this locality is unknown. Further work on this yellowish millerite alteration is needed.

Paratacamite (?) [Cu.sub.2][(OH).sub.3]Cl

Paratacamite has been reported as green botryoidal hemispheres on chalcopyrite (Medici, 1981; and D. Coskren, personal communication); confirmation of this identification is still pending.

Pyrite FeS2

Pyrite is one of the most fascinating minerals at Halls Gap. Paragenetically it occurs in several generations, each with a different crystal habit. One type is the most unusual: the pyrite is polycrystalline and forms many odd shapes, including worm-like masses, loops, springs, coils and irregular shapes. Another type is a sceptered habit, which may be partially or completely coated with transparent crystals of quartz. Some ring-like shapes occur late in the paragenetic sequence, found on calcite and glauconite-bearing quartz. It is likely that polycrystalline pyrite formed intermittently throughout the mineralization process. Multiple polycrystalline pyrite habits may occur within the same geode.

Another early-generation pyrite habit is octahedral or cuboctahedral. It appears to be later than most polycrystalline pyrite, which it frequently overgrows. The octahedral forms may be quite complex, with successive parallel overgrowth. Crude octahedrons may form from stacks of smaller octahedral crystals. They can be smaller than 1 mm or larger 3 cm in diameter. Larger crystal faces may be slightly concave while others may show a pattern with triangular faces and holes. Perhaps the most intriguing are octahedral pyrite crystals with late-generation cubes in parallel overgrowth. A specimen was found with tabular pyrite capping the apex of an octahedron. One area in the "millerite zone" has a particularly high abundance of large octahedral pyrite.

Cuboctohedral pyrite is very abundant. It was deposited simultaneously with and somewhat later than the octahedral pyrite. The variation of the forms ranges from cubes with very minor octahedral faces to octahedrons with small cubic faces. Both octahedrons and cuboctahedrons may show etched, silky-lustered crystal faces. Cuboctahedral pyrite also occurs intergrown with sphalerite.

"Graphic pyrite" is another unusual habit at Halls Gap. The patterns may create a simple "L" or more complicated crystals with multiple 90 [degrees] bends. There is no end to the variation that this odd crystal habit creates. In both size and thickness, significant variation occurs. Some crystals are filiform with multiple changes in the x, y and z directions. Others are thicker, with fewer kinks that are easier to see. Some writers (i.e. Berndt, 1969) liken this habit to structural steel used in tall buildings. This pyrite appears be related in time to cubic pyrite.

Cubic pyrite is the latest generation, and the most common habit throughout the outcrop. The largest cubes are about 3 cm on edge and may be distorted (Medici, 1981). Most cubes are lustrous and brassy in color. Some have a silvery tinge or bright iridescent colors, blue being the most striking. One oddity that has been observed is etched cube faces with a "+" pattern or black dot at the center of each cube face.

Cubic or cuboctohedral pyrite may be found skewered by millerite; in such cases pyrite has nucleated on the side of a filamentary millerite crystal, and has grown equidimensionally while suspended on the millerite. Very rarely intergrown pyrite crystals will be found on a single millerite crystal. Usually only a single pyrite crystal will grow on a single millerite crystal. Even less common is a string of crystals. When pyrite grows on millerite, the cube will often be centered along the host's crystallographic axis. Pyrite attached to the wall of the geode may enclose a portion of adjacent millerite filaments. Millerite intersects at random angles to the pyrite's crystal faces. Berndt (1969) reports millerite nucleated around pyrite. All specimens analyzed by the writers show nucleation around chalcopyrite and less commonly sphalerite. Examination of hundreds of sulfide-bearing geodes indicates that most pyrite formed after the millerite.

Quartz Si[O.sub.2]

Quartz is the most common mineral at Halls Gap. Them are several generations of crystalline and cryptocrystalline quartz.

Chalcedony is the earliest material in a geode, occurring in a striking variety of colors including deep blue, pea-green and red-orange. Multiple generations of chalcedony may consist of varying colors and show a thin agate-like banding in cross-section.

Chalcedony is followed paragenetically by crystalline quartz, which varies from drusy microcrystals to euhedral crystals exceeding 5 mm. Drusy quartz may be colorless to orange-red. Amethystine quartz occurs in singly and doubly terminated crystal prisms to 5 mm. Greenish late-forming drusy quartz contains glauconite inclusions.

Smoky quartz can be found in what almost appear to be bipyramidal crystals, mimicking the habit of beta (high temperature) quartz. Closer examination of some bipyramidal quartz reveals a minute c-pinacoid face. Colorless quartz may also be found.

A late generation consisting of microscopic euhedral quartz crystals is uncommon, but has been observed "dusting" larger quartz and other sulfides. A single specimen has been found by one of us (A.G.) that has the appearance of a quartz pseudomorph after millerite.

Quartz crystal scepters are found in many geodes. They are small, typically less than 1 cm in length. Some are stalactitic in appearance and can be straight or curved. Scepters tend to be white or colorless.

Roemerite [Mathematical Expression Omitted]

Roemerite occurs in pale beige blocky crystals 0.25 mm across. It is associated with halotrichite, melanterite, metavoltine and voltaire (D. Koskren, pers. comm.).

Sphalerite ZnS

Sphalerite is relatively common, especially in the strata above the "millerite zone." The crystal habit is tetrahedral, usually in intergrown crystals which range from 1 to 10 mm in diameter; the larger crystals usually fill voids in the geode. Well-formed, isolated, 2 to 4 mm tetrahedra are rare. Millerite nucleated around sphalerite is very uncommon. Pyrite crystals are commonly intergrown with sphalerite crystals. Thin, tabular sphalerite occurs rarely, but usually breaks when the geode is opened.

Sulfur S

Sulfur occurs in microscopic crystals and grains as a result of decomposing sulfides. It is commonly associated with polycrystalline pyrite and marcasite and rarely with unknown mineral #1.

Violarite (?) Fe[Ni.sub.2][S.sub.4]

Violarite has been reported as a dark violet-blue coating on millerite (Medici, 1981). It may be present on one specimen seen by the writers, but this observation needs confirmation.

Voltaite [Mathematical Expression Omitted]

Voltaite occurs as complex cuboctahedra and dodecahedra, dark oil-green to black-green in color and 0.25 mm across (D. Coskrin, personal communication). It is associated with pyrite, millerite and sulfates including roemerite.

Unknown #1

Unknown #1 is uncommon. It occurs as bright silvery mats or dustings on quartz and other sulfides. Under magnification it resembles wrinkled aluminum foil. This mineral often forms extremely unusual shapes, including loops, archs, blisters, fans and other complex forms. It seems to crystallize late in the paragenetic sequence because, other than microscopic sulfur grains, there are no later minerals. This mineral is typically deposited preferentially on one side of a geode. At first glance it resembles polycrystalline pyrite, but the silky texture is distinctive. SEM photos reveal elongated fiber-like crystals that may follow the contour of the substrate or matrix. Steve Garza examined several specimens and described its appearance as similar to lepidocrocite. Ruez (1973) calls it "marcasite."

Unknown #2

Unknown #2 (nickel sulfate?) was discussed by Medici (1981), but there was insufficient material for a detailed analysis. This mineral occurs in greenish to blue-green tufts to 1 mm resembling "star-bursts" and as a "bottle brush" pattern on millerite or jamborite. It is usually sparse, but in at least one geode has been found in large quantities.


In order to establish a paragenetic sequence, one of us (A.G.) examined under magnification hundreds of geodes from the "millerite zone" (collected by B.M.) to determine overlapping relationships. Specimens can be made available to mineralogists for further study.

Geodes are surrounded by a rind of chalcedony, which is first in the paragenetic sequence. The outside find probably formed while anhydrite or gypsum was still in place. Chowns and Elkins (1974) found relict anhydrite in thin-section studies of geode rinds. Thin-rimmed gypsum geodes occur in Hardin and Meade County, Kentucky. Geodes over 15 cm in diameter may have chalcedony walls less than 2 mm thick.

Chalcedony deposition at Halls Gap produced multiple layers and colors. Late-forming chalcedony may show desiccation cracks up to 1 mm in width.

Geodes may be nested doubles, consisting of an exterior chalcedony rind separated by a narrow gap of drusy quartz and sulfides (chalcopyrite, pyrite and/or sphalerite), from an inner, very smooth, second chalcedony shell which is often quartz-filled.

Brecciation of chalcedony due to dissolution and slumping occurred before other minerals were deposited. Crystalline quartz follows chalcedony in most instances. Some early quartz may be followed by later chalcedony, visible in the cross-section of thick-walled geodes. Several episodes of crystalline quartz deposition have been observed. Some of the quartz druses are accompanied by a single large quartz crystal.

Polycrystalline pyrite follows and overlaps with early quartz in the paragenesis. However, it occurs intermittently throughout the paragenetic sequence, including after calcite. Most polycrystalline pyrite is firmly imbedded in early quartz. Smaller "wires" occur as inclusions in quartz.

Other sulfides including chalcopyrite, millerite, sphalerite, marcasite and, to a lesser extent, cuboctahedral pyrite occur at the end or slightly before the end of early quartz deposition. Additional study will be required to determined the precise paragenetic sequence of these sulfides. Each formed very close to the other in time, and in many cases are simultaneous.

Calcite, dolomite and sulfate alteration products of sulfides were late in the sequence. Cubic pyrite is consistently found late, including less than 0.5-mm crystals on calcite and dolomite. Larger cubic pyrite preceded these carbonates, but followed other crystal habits of pyrite and other sulfides. Some chalcopyrite, millerite, and cuboctahedral pyrite occurs as "floaters" locked in calcite. The paragenetic relationship between the earliest calcite deposition and latest of the so-called "early" sulfides needs to be studied further. Galena occurs after cubic pyrite, the evidence provided by a single specimen.

The latest generation of quartz has a sugary appearance. It has been observed sprinkled over crystalline quartz and all sulfides, including millerite. Quartz with glauconite inclusions occurs late in the mineral depositional process.

Anatase is associated with microscopic euhedral quartz. It is probably late in the sequence, but its exact position has not been precisely determined.


Collecting at Halls Gap has been complicated by the presence of a dangerous overhanging ledge which has been created by collectors working in the "millerite zone." We estimate that over 300 tons of rock have been removed by collectors in the eastern roadcut alone! Parts of the overhang are highly fractured. Collectors should avoid collecting beneath unstable overhangs.

The so-called "millerite zone," sandwiched between the overhang and the floor, is 1.4 meters thick. Usually, at the actual working face, about 1 meter of this is exposed. The majority is a dense, compact glauconitic siltstone with horizontal grain structure. In lateral extent, this face can have a good deal of unevenness, owing to the recent activity of collectors, especially those who may use power equipment. This creates localized protrusions that can be removed easier than where the face is laterally flat.

A promising-looking face is first selected using the criteria mentioned above. Additionally, cracks (horizontal or vertical) or localized spots of increased weathering (recognized by delamination of the rock) may be good places to begin. There are places that may be less comfortable to work. Much depends on the tools at hand. The eastern roadcut has been most productive for one of us (B.M.) over the years.

The order of the day when collecting at Halls Gap is to move rock. The number of geodes collected is more or less proportional to the volume of rock liberated from the face. Rock can be moved using chisels of 1.75 to 2.5 cm in width and a hand sledge. These are positioned pointing downward approximately 15 to 45 degrees from horizontal and driven with repeated heavy blows using a 3 to 8-pound hand-sledge. This splits sections of rock by a wedging action and normally removes 1 to 3 kg of rock. The rock can then be split further with chisels or a rock splitter. When a section of rock is removed and it has split along the plane of a geode, a portion or the whole may remain in the wall. It can be removed by carefully chiseling a groove in the wall, concentric with the geode, from 2.5 to 5 cm away. Next, the chisel is carefully worked around the groove, weakening the enclosing rock on all sides. The geode will then pop out, with or without matrix.

A much more productive technique requires drilling a series of horizontal holes with a hammer drill. The holes may be from 2 to 2.5 cm in diameter and 10 to 15 cm in depth. The holes should be drilled 15 to 20 cm apart, 3 to 4 at a time. Wedges and shims are then oriented in the holes to split the rock horizontally when hit sequentially with a sledge. This technique can remove some large blocks of rock which can be further broken to free any geodes. It is best to work one section of ledge at a time, moving from low to high.

The wall thickness of the geodes varies greatly. Much care should be taken in opening those geodes that are light in weight in relation to their volume. Most larger geodes have one or more seams surrounding them, reflecting fractures that have been recemented with quartz. These geodes can break quite erratically without sufficient care. If a geode is obviously hollow, it should be opened under the most controlled conditions possible.

Geodes opened at the outcrop, either accidentally or intentionally, should be immediately stored in egg cartons or other suitable protection. Many millerite tufts have been damaged or lost by a gust of wind from a passing truck.

When possible, the exterior of the geode should be cleaned prior to opening. This eliminates excessive loose debris which can become intermingled with acicular millerite or jamborite (?) crystals. Debris can be removed by upending the geode and gently tapping the bottom with a wooden peg. Dental picks can be used with a steady hand under a lens or microscope to remove stubborn debris. Badly dusted specimens can be dipped upright in distilled water. Soak for 30 seconds to 1 minute and then dry in an oven as described below. Jamborite (?) must not be allowed to get wet in any manner, as it will be destroyed.

Many geodes are totally calcite filled. These can be quite spectacular after the calcite has been removed with 3% to 5% HCl. After "etching," the geode should be neutralized in dilute ammonia, thoroughly rinsed in distilled water and heated to dryness in an oven at 200 [degrees] F for 10 to 30 minutes, depending on the mass of the geode. It is surprising to what extent wet, matted millerite can "spring" back out to nearly its original aspect after drying in this way.

These geodes are best kept in a dust-free environment, such as tight mineral cabinets or display cases. In lieu of this, or when transporting them, they can be kept in hinged plastic boxes and secured with "mineral tack." They should never be packed with tissue or cotton, as the fibers will become hopelessly intertwined with millerite. Most of the minerals in the geodes are chemically stable over many years.


The Halls Gap locality has produced millerite and its alteration products and unusual pyrite since its discovery. Continued collecting keeps the market well-stocked with specimens of millerite. Not everyone puts a lens to each Halls Gap geode; consequently, discoveries of additional alteration products and unusual pyrite habits are possible by anyone obtaining specimens. Much scientific work remains to be done. New studies employing microanalytical techniques are necessary because of the microscopic size of most minerals.


The authors would like to thank the following for their assistance in the research that made this paper possible: Chris Anderson for macrophotography; Dr. Dennis Coskren for information on a possible source of millerite, the determination of paratacamite and comments about other minerals; Dr. John H. Eicher for locating a copy of the thesis by Paul Reuz; Steve Garza for discussion of galena and possible lepidocrocite; Claus Hedegaard for discussion on bipyramidal quartz; William A. Henderson, Jr. for discussion on "eshelby" twists of millerite; Dr. Allen V. Heyl for information on other millerite alteration products; Dr. George Lager for SEM photography of the twisted millerite and discussion of other mineral occurrences; Paul E. Smith and Dr. Harvey Belkin for information on the anatase and additional microphotography; and Dr. Rudi Turner for other SEM photography.


BERNDT, D. J. (1969) Three golden rings. Gems and Minerals, no. 386, p. 30-33.

BIDEAUX, R. A. (1970) Mineral rings and cylinders. Mineralogical Record, 1, 105-113.

BISH, D. L., and LIVINGSTONE, A. (1981) The crystal chemistry and paragenesis of honessite and hydrohonessite: the sulphate analogues of reevesite. Mineralogical Magazine, 44, 339-343.

CARBONATE PETROLOGY SEMINAR (1987) Ramp Creek and Harrodsburg Limestones: A shoaling-upward sequence with storm-produced features in southern Indiana, USA. Sedimentary Geology, 52, 207-226.

CHOWNS, T. M., and ELKINS, J. E. (1974) The origin of quartz geodes and cauliflower cherts through the silicification of anhydrite nodules. Journal of Sedimentary Petrology, 44, 885-903.

CHROMY, B. (1971) The wee minerals. Earth Science, 301-312.

CHROMY, B. (1972) The wee minerals. Earth Science, 253-255.

FISHER, I. S. (1977) Distribution of Mississippian geodes and geodal minerals in Kentucky. Economic Geology 73, 864-869.

GOLDSTEIN, A. (1989) Minerals of Kentucky: The species list. Mineral News, 5, no. 5, 1-3.

GOLDSTEIN, A. (1990) Geodes on the Muldraugh dome: A new (temporary) locality. Mineral News, 6, no. 7, 5.

LINEBACK, J. A. (1966) Deepwater sediments adjacent to the Borden Siltstone (Mississippian) delta in southern Illinois, Illinois State Geological Survey Circular 401, 48 p.

MALVIA, R. G. (1985) An early diagenetic model for the origin of geodes in the Ramp Creek Formation and Harrodsburg Limestone (Mississippian), southern Indiana. Geological Society of America Abstracts with Programs, 17, 300.

MALVIA, R. G. (1986) Quartz geodes: Evidence for non-evaporitic anhydrite nodules. Journal of Sedimentary Petrology, 56.

MEDICI, J. C. (1981) The Halls Gap millerite locality. Rocks & Minerals, 56, 104-108.

MORANDI, N., and DALRIO, G. (1973) Jamborite: A new nickel hydroxide mineral from the northern Apennines, Italy, American Mineralogist, 58, 835-839.

NICKEL, E. H., and WILDMAN, J. E. (1981) Hydrohonessite - a new hydrated Ni-Fe hydroxyl-sulphate mineral; its relationship to honessite, carrboydite, and minerals of the pyroaurite group. Mineralogical Magazine, 44, 333-337.

RUEZ, P. H. (1973) A Field and Laboratory Study of Millerite and Related Ni-bearing Minerals. M.S. thesis, Miami University, Oxford, Ohio, 63 p.

SWANN, D. H., LINEBACK, J. A., and FRUND, E. (1965) The Borden Siltstone (Mississippian) delta in southwestern Illinois. Illinois State Geological Survey Circular 386, 20 p.

WEIR, G. W., GUALTIERI, J. L., and SCHLANGER, S. O. (1966) Borden Formation (Mississippian) in South and Southeast-Central Kentucky. U.S. Geological Survey Bulletin 1224-F, 38 p.

WEIR, G. W. (1970) Field Trip no. 3: Borden formation (Mississippian) in southeast-central Kentucky, p. 29-48. tn Geological Society of America Guidebook for Field Trips (18th annual meeting, Southeastern section): Kentucky Geological Survey, 70 p.

WHITEHEAD, N. H. III (1976) The Stratigraphy, Sedimentary, and Conodont Paleoecology of the Floyds Knob Bed and Edwardsville Member of the Muldraugh Formation (Valmeyeran), Southern Indiana and North-central Kentucky. M.S. thesis, University of Illinois at Champaign-Urbana.
COPYRIGHT 1997 The Mineralogical, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:mineral locality
Author:Goldstein, Alan; McKenzie, Bill
Publication:The Mineralogical Record
Date:Sep 1, 1997
Previous Article:Sainte-Marie-aux-Mines Show 1996.
Next Article:Szenicsite: a new copper molybdate from Inca de Oro, Atacama, Chile.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |