Intense hydrothermal alteration has affected at least 8.5 mi² (22 km²) in the Snow Camp-Saxapahaw area, where three different ages of alteration were distinguished. The earliest, and most extensive and intense, was a high-sulfidation event limited to rocks of the intermediate to felsic volcanic complex, mostly within the Major Hill structural block. All of the pyrophyllite-rich deposits and certain types of gold mineralization were formed during this period. The second hydrothermal event was identified only within the Reedy Branch Tuff and has resulted in large areas of weak to moderate quartz-sericitic and propylitic alteration, but not intense high-sulfidation-type alteration nor known metallic mineralization. The third and latest hydrothermal event, the most limited in area of the three, was related to the emplacement of postmetamorphic granitoid porphyritic plutons. The strong hydrothermal effects included extensive potassic and sericitic alteration and the formation of (1) several molybdenum-bearing greisenlike bodies that are associated with the plutons, (2) gold-bearing pyritic zones that are gossans where they crop out within the porphyritic plutons, and (3) probably the remobilization of some of the earlier deposited gold into epithermal veins in favorable sites.

The terminology used to describe alterations systems such as these has varied widely. We follow the general criteria and nomenclature of White and Hedenquist (1990, 1995), who urged the adoption of the descriptive term, "high-sulfidation." Earlier authors used a variety of terms for similar alteration, such as "acid-sulfate," "adularia-sericite," "high-sulfur," and "kaolinite-alunite."

"High-sulfidation alteration," as used hereafter for deposits in the Southeastern United States, can generally be considered equivalent for earlier references to the origin of "Al(umina)-rich assemblages" (Carpenter and Allard, 1982) and to "high-alumina alteration" (Schmidt, 1982). Ririe (1990) was perhaps the first author to use "high-sulfidation" in reference to such deposits in the Carolina slate belt.

High-Sulfidation Mineralization in Other Regions Worldwide

Comparison of these deposits with those in other regions having similar alteration style and zone geometry was facilitated by the recent publication of many new reports describing a variety of mines and prospects. We think the most appropriate deposits to consider are those from Summitville, Colo. (Stoffregen, 1987); Nansatsu district, Kyushu, Japan (Hedenquist and others, 1988); the area around the Hyttsjo intrusion near Persberg, Sweden (Outhuis, 1989); the Temora gold-silver deposit, southeast Australia (Thompson and others, 1986); the Monte Negro deposit, Pueblo Viejo district, Dominican Republic (Kesler and others, 1988); and the Rodalquilar district, Spain (Arribas and others, 1989).

Additional deposits in other regions that are generally similar in scale and alteration zonation, and that contain alteration minerals such as pyrophyllite, andalusite, diaspore, and less commonly, corundum, but clearly are unmetamorphosed, include the Bol'shoy Semiz-Bugu alunite-andalusite-corundum deposit in Kazakhstan (Nakovnik, 1968) and sections of the Comstock lode, Nevada (Hudson, 1984). Several deposits in Fujian and Zhejiang Provinces, southeast People's Republic of China, are also generally similar in scale, alteration mineralogy, and zonation. Some of these were visited by the senior author in 1990, and one deposit, the Emei, has been described by Li (1987). The alteration mineralogy of these and other unmetamorphosed deposits assures us that most of the minerals present in the Snow Camp area deposits could have formed as parts of the original hydrothermal suite. The presence of the minerals themselves did not require a metamorphic event.

Other alteration systems within the Carolina slate belt having zoned high-sulfidation alteration similar to that in the Snow Camp area include the Brewer Mine, Jefferson, S.C. (Butler, 1985; Butler and others, 1988; and Lu and others, 1993), the Nesbit gold mine, Union County, N.C. (McKee, 1985), and the Pilot Mountain-Fox Mountain alteration system, Randolph County, N.C. (Klein and Schmidt, 1985; and Klein and Criss, 1988). Pipe-like breccia zones have been described at the Brewer Mine (Cherrywell and Butler, 1984; Klein and Schmidt, 1985; and Butler and others, 1988), where some of the mineralized breccias are funnel-shaped and flare upward, providing excellent fluid conduits (Butler and others, 1988; and Scheetz and others, 1991). Breccia pipes have been described at similar deposits in other regions, such as the Motomboto prospect, Tombulilato district, North Sulawesi, Indonesia (Perello, 1994), and Nalesbitan, Luzon, Philippines (Sillitoe and others, 1990).

Gold, molybdenum, tin, and copper were introduced into core zones at several locations in the Carolina slate belt, such as at the Brewer Mine (A.K. Kinkle, Jr., U.S. Geological Survey, written commun., 1968; Schmidt, 1985a; Butler and others, 1988; and Lu and others, 1993), and at Pilot Mountain (Schmidt, 1982, 1985a). These elements were also enriched in various degrees in the Snow Camp-Saxapahaw area.

Researchers in other regions have shown that high-sulfidation mineral suites formed in two types of volcanic environments, "steam-heated" generally horizontal zones near the surface and "magmatic-hydrothermal" vertically elongated systems extending downward along a variety of solution channelways. When gold is present, the two environments have different styles of mineralization (Iwao and Udagawa, 1969; Heald and others, 1987; Klein and Criss, 1988; Thompson and Peterson, 1991; and White and Hedenquist, 1995) and different types of alunite (Hayba and others, 1985; and Arribas and others, 1989). Alunite is the most common aluminous mineral, but pyrophyllite, andalusite, diaspore, dickite, and corundum may be present as well.

Previous Interpretations of High-Sulfidation Mineralization in the Carolina Slate Belt

The large areas of high-sulfidation alteration in the Snow Camp-Saxapahaw area share many characteristics with over 40 similar large areas in the Carolina slate belt (Schmidt, 1982). These areas of alteration are scattered from Graves Mountain, Ga., to the northern margin of Granville County, N.C., near the Virginia State line, as well as the Brewer gold mine in South Carolina. Detailed descriptions of some of these have been given by Bell and others (1980), Spence and others (1980), Cherrywell and Butler (1984), Hughes (1985), Klein and Schmidt (1985), McKee (1985), Schmidt (1985a,b), Butler and others (1988), Klein and Criss (1988), Ririe (1990), Schmidt and others (1990), Scheetz and others (1991), and Lu and others (1993).

Early descriptions of the hydrothermal alteration centers in the Carolina slate belt dealt mostly with the minerals present, but a few attempted to explain the style of alteration. Espenshade and Potter (1960) noted that similar mineral suites were present in hot spring areas, such as Yellowstone National Park and several other locations, and suggested that "this type of rock alteration may have been active in the formation of some of the Southeastern deposits of aluminous materials." They speculated further that, "some kyanite-quartz deposits may be masses of volcanic rock that were first altered by solfataric activity and then metamorphosed to the present mineral assemblage. Other deposits, such as some andalusite-pyrophyllite-quartz deposits, for example, might represent the deeper zones of extensive solfataric centers."

The interpretive study of selected gold mineralization in the Carolina slate belt by Worthington and Kiff (1970) suggested that the mineralization in the Carolina slate belt was related to volcanic exhalations associated with the enclosing felsic volcanic rocks. This drew further attention to the significant idea for a volcanic origin, an idea that was then carried forward by many others. Worthington and others (1980) and Spence and others (1980) suggested that certain gold deposits found in conjunction with masses of high-alumina minerals were associated with hot springs and fumaroles. Bell and others (1980) recognized the Brewer mine as "perhaps revealing an explosion vent or volcanic center surrounded by hydrothermally altered rocks." Carpenter and Allard (1980) related the aluminosilicate hydrothermal alteration zones near Lincolnton, Ga., and McCormick, S.C., to hydrothermal systems that vented as submarine fumaroles; and Allard and Carpenter (1981) concluded that the various deposits containing abundant aluminosilicate minerals in the Southeastern United States fit a volcanic hydrothermal model and that the Otake geothermal field, Japan, could be considered to be a modern analog. Carpenter and Allard (1980) applied the useful name "quartz-granofels" to the fine-grained, intensely altered, quartz-rich rock.

The size, type, and alteration zonation of the hydrothermal systems led Schmidt (1978, 1982) to regard the Brewer Mine as having a porphyry copper affinity, and later that both the Brewer Mine and Pilot Mountain were porphyry gold systems (Schmidt, 1982). He interpreted the large volumes of altered rocks to be the result of intense solfataric and subvolcanic activity, analogous to such deposits as Mt. Pleasant, New Brunswick, Canada; certain andalusite-pyrophyllite deposits of Japan; and the Kounrad and other deposits of Kazakhstan. The large scale of some of the alteration systems in the Carolina slate belt was emphasized by Ririe (1990), in his comparison of them with certain gold deposits of Archean age in Western Australia. Klein and Criss (1988) interpreted the Pilot Mountain alteration system to have formed at greater depth than that at the Brewer Mine. They drew parallels between the Pilot Mountain system and high-sulfur epithermal systems elsewhere, citing several analogs. Their oxygen-isotope data for Snow Camp and Pilot Mountain supported the formation by subaerial hydrothermal and volcanic processes. Klein and Criss (1988) and Ririe (1990) interpreted the aluminosilicate alteration mineral suite from the Brewer Mine to be evidence for acid-sulfate (high-sulfidation) hydrothermal alteration. It should be pointed out that Schmidt (1982, 1985a) and Páyas and others (1988) used the term "high alumina" minerals when aluminosilicate minerals would have been more appropriate.

Several Carolina slate belt mines have produced significant amounts of gold from high-sulfidation alteration centers and small amounts of gold have been produced from many centers. There was no known gold mineralization, however, at many more well-documented centers, such as at Daniels Mountain northwest of Oxford in Granville County, N.C. (Hughes, 1985). Furthermore, no significant base-metal mineralization has been reported from most. Individual pyrite-rich drill core samples from the Brewer Mine, however, contain as much as 1.0 percent copper, and some tin and bismuth were noted there (Schmidt, 1978). Traces of molybdenum were found in soils at the Pilot Mountain alteration system (Milton and others, 1983). As discussed elsewhere, some high-sulfidation deposits in other regions of the world contain an even more diverse suite of metals, while other nearby intensely altered high-sulfidation centers are completely barren.

There is a common association of deposits of this general type with calderas (Steven and others, 1974; Hayba and others, 1985; and Arribas and others, 1989), but no circular caldera was recognized here. The Major Hill structural block may have behaved much like a caldera having a series of shallow plutons intruded along the bounding faults as the graben formed.

Possible Linear Control of Mineralization on a Regional Scale

High-sulfidation hydrothermal systems in North Carolina are known to be distributed within a linear belt extending from Montgomery County northeastward to Granville County since the publication of a map of pyrophyllite deposits by Stuckey (1967). Several geologists studying the region noted the confinement of most of the deposits to a trend only a few mi (km) wide, with a few additional deposits scattered up to 6 mi (10 km) on either side (for example, Hughes, 1987). No surface manifestations, such as abundant faults or evidence of recurring fault movement, have been recognized along the deposit trend in North Carolina.

High-Sulfidation Alteration Systems in the Study Area

The minerals in the Snow Camp alteration zones, and the concentric configuration of the zones are typical of many large high-sulfidation alteration systems found in other regions in stratovolcanoes, volcanic domes, and in the volcanic rocks associated with diatremes and calderas (Albers and Kleinhampl, 1970; Lipman and Steven, 1970; Steven and others, 1974; and Sillitoe, 1973, 1989). However, the quartz-rich core zones in this area are enclosed in much larger quartz-sericite-paragonite alteration envelopes than those described from all but a few other regions, such as the Freida River prospect, Papua New Guinea (Britten, 1986). In the Snow Camp-Saxapahaw area, three areas of alteration (fig. 2, fig. 3, fig. 4A, and fig. 5), the Sheeprock (sectors J and N), Major Hill (sectors F and G), and Central-Northeast Highland alteration zones (sectors C and G), form a major northeast-trending zone, 8 mi (13 km) long and several miles (kilometers) wide. The smaller Cane Creek (sectors G, H, and K) and Zachary (sector K) alteration zones and other small areas of quartz-sericite altered rock are present eastward of the major trend. These are in a variety of protoliths within rocks of the intermediate to felsic volcanic complex and, except for small patches in sectors N and O, are all in the Major Hill structural block.

The Sheeprock, Major Hill, and Central-Northeast Highland alteration centers have irregularly shaped quartz-rich core zones surrounded by extensive envelopes of quartz-sericite-paragonite altered volcanic and subvolcanic rocks. There is a non-outcropping core zone in the Zachary alteration center as well, as indicated by slabs of quartz-pyrophyllite float rock and andalusite in panned concentrates in the transecting creek. Core zones were not identified in several of the smaller quartz-sericite-paragonite alteration zones mapped. Large pods or lenses of pyrophyllite-rich rock formed within part of the core zones, but mostly near their margins, such as in the Hinshaw pyrophyllite prospect (sector N), Snow Camp south pyrophyllite prospect and Snow Camp pyrophyllite mine (sector J), and the Major Hill pyrophyllite prospect (sector F).

The margins of alteration zones are transitional, the outer edges of quartz-sericite-paragonite zones more so than the inside boundaries against the quartz-rich cores. The total area of quartz-sericite-paragonite zone (7.3 mi² or 19 km²) greatly exceeds the quartz-rich core zone (1.16 mi² or 3 km²), and the pyrophyllite-rich rock probably occupies less than one-tenth of the area of the core zone.

Rocks of the Quartz-Sericite-Paragonite Alteration Zone
The rocks of this zone were formed by strong hydrothermal alteration of mostly pyroclastic rocks, a few shallow intrusive bodies, and perhaps some flows as well, and they generally retain some or many of their original textures. Penetrative deformation is more obvious here than in the core zone, and, locally, these rocks have been intensely sheared. Light-gray and gray-green colors are characteristic as well as a dull to waxy luster. When examined in the laboratory, rocks mapped as being in the quartz-sericite alteration zone were found to contain abundant white mica as the most important secondary alteration product. Part of this mica was found to be paragonite, but it could be distinguished only by X-ray diffraction (Hughes, 1987). The paragonite is assumed to be widespread.

The quartz-sericite-paragonite rocks forming the envelopes enclosing the quartz-rich cores generally crop out poorly. Most rocks in this zone contain a trace to several percent pyrite, which might have augmented the weathering process. Quartz-sericite-paragonite alteration also is stronger or more readily recognized where the rocks are more sheared, such as at the quarries of the Jones pyrophyllite-sericite mine west of the Central Highland (fig. 2, fig. 3, fig. 4A, and fig. 5, sector F), where the quartz-sericite-pyrophyllite assemblage was identified in the soft platy material.

Quartz-Rich Core-Zone Rock and the Pyrophyllite-Rich Bodies
The quartz granofels that makes up most of the core zones is very fine grained and is generally light gray to white, but some is red, purple, or black. It may have faint ghosts of texture remaining from a protolith or may be a sealed breccia of homogeneous siliceous fragments. There is, however, no recognizable trace of a former internal structure. The rock resembles a very fine, massive quartzite, and the descriptive name, "secondary quartzite," was used in some literature.

The quartz granofels of the core zone here contains more than 90 percent SiO₂ and minor amounts of pyrophyllite, sericite, and hematite or, in a few places, magnetite. Other accessory minerals might include chloritoid, tourmaline and sparse lazulite. Small clusters of fine rutile grainlets, probably formed by alteration of ilmenite in the volcanic protolith, were observed at many sites. Topaz is present in several other Carolina slate belt core zones, but it was not recognized here; however, fine disseminations of topaz resemble quartz granofels except they are heavier and are easy to overlook. Tourmaline is present at the Hinshaw pyrophyllite prospect (fig. 2, fig. 3, fig. 4A, and fig. 5, sector N), and it has been noted at several other Carolina slate belt alteration centers.

Like the quartz-sericite-paragonite rock, the quartz-granofels core rocks formed by the hydrothermal alteration of rocks of a variety of textures and compositions. One cannot overemphasize the intensity of the process, stripping the alkalies, K₂O, Na₂O, CaO, and MgO, and leaching rocks as mafic as andesites, to result in the siliceous quartz granofels (table 8). The presence of the trace minerals, topaz, tourmaline, and lazulite, in many of these systems indicates that the reactive elements, fluorine, boron, and phosphorus, may have contributed to the rigorous alteration. Similar alteration is well known in the Circum-Pacific region (Hedenquist, 1992) and in specific districts such as the Frieda River prospect, Papua New Guinea (Britten, 1986), Archean gold deposits in Western Australia (Ririe, 1990), and the Matsukawa geothermal area, northeastern Japan (Sumi, 1968).

The extent to which the altering fluids may have redistributed or possibly introduced alumina is not clear, but much element transport must have occurred in the formation of both the quartz granofels (>90 percent SiO₂) and local paragonite rock (>40 percent Al₂O₃, Schreyer, 1987). The introduction of alumina has been shown to have been an aspect of alteration at the diaspore-pyrophyllite deposits of Fukue Island, Nagasaki Prefecture, Japan, where the replaced host rock was sandstone (Watanabe, 1953). The removal of alkalis by alteration in the Snow Camp area is further confirmed by the striking deficiency of calcium and magnesium in forest soils on several of the siliceous Snow Camp-Saxapahaw core zones (Payás and others, 1993).

Within the core zones of the Central Highland area (fig. 2, fig. 3, fig. 4A, and fig. 5, sectors C and G), Hughes (1987) found that the alteration assemblage ranged from quartz-sericite±chloritoid±hematite to quartz-pyrophyllite-iron oxide rocks as well as much scattered quartz granofels. No quartz-rich core zones were mapped in the Northeast Highland area, but foliated quartz-sericite±paragonite±hematite rocks are well exposed there (Hughes, 1987).

Breccia Zones and Breccia Pipes
Quartz-granofels breccias healed by massive quartz granofels are present in parts of all of the Snow Camp-Saxapahaw core zones and are especially common along the southern slope of the Sheeprock hill (fig. 2, fig. 3, fig. 4A, and fig. 5, sectors J and N). The fragments range in length from less than 0.04 in (1 mm) to perhaps 0.2 in (5 mm), the matrix differing from the fragments only in pigmentation. The breccia fragments have broken with no regard for the original structure. The pigment is mostly finely divided hematite, or magnetite, making the rock dark red, purple, or black, although the amount of iron oxide is deceptively small, probably less than 10 percent. The matrix is generally more pigmented than the fragments, but the reverse was locally seen. In areas with little pigmentation the fragments appear only as faint "ghosts."

The quartz granofels in the study area lacks the abundant voids commonly described in the unmetamorphosed vuggy quartz rock in many Mesozoic and Tertiary deposits, such as Summitville, Colo. (Steven and Ratte, 1960), and Rodalquilar, Spain (Arribas and others, 1989). In the Motomboto deposit in the North Arm of Sulawesi, Indonesia (Carlile and others, 1990; and Perello, 1994), both vuggy silica and massive quartz granofels have formed.

Upward flaring pipe-like breccia zones have been described in the Carolina slate belt at the Brewer Mine (Cherrywell and Butler, 1984; Klein and Schmidt, 1985; Butler and others, 1988; and Scheetz and others, 1991), which would have provided excellent fluid conduits. They are generally interpreted as explosion-breccia pipes. The distribution of quartz-healed breccias in both outcrop and float in the study area suggests generally equidimensional cross-sectional shapes rather than thin tabular bodies. We think that they have had considerable vertical dimension, perhaps many hundreds of feet (hundreds of meters), similar to the quartz-cemented breccias of the North Amethyst vein in the Creede mining district, Colo. (Foley and others, 1993), and the Chinkuashih district, Taiwan (Tan, 1991). Ferruginous-matrix healed breccias in the study area are thought to resemble gold-bearing rocks at the Hickey's Pond showing, Newfoundland, as described by Huard and O'Driscoll (1986).

At many places in the core zones the individual breccia fragments tend to be somewhat flattened or lenticular, having a planar orientation, or they may be smeared out into thin folia. Fragment surfaces may be coated by thin films of pyrophyllite, and perhaps some kaolinite, and are commonly stained with reddish-brown iron oxide.

Shapes and Orientation of the Core Zones and Pyrophyllite-Rich Bodies
The three most significant pyrophyllite-rich bodies, the Snow Camp pyrophyllite mine and the Snow Camp South pyrophyllite prospect at the northern and southern ends of Mine Ridge (fig. 2, fig. 3, fig. 4A, and fig. 5, sector J) and the pyrophyllite prospect on Major Hill (fig. 2, fig. 3, fig. 4A, and fig. 5, sector F), are all within but close to the margins of the larger quartz-rich core zones; however, they are too small to be shown in figure 2. Rock exposures in the Snow Camp pyrophyllite open-pit mine give the impression of a generally equidimensional orebody, but there may be unmined lateral or downward extensions that would define a lenticular or podiform shape. Outside the main pyrophyllite-rich body at the same mine, the quartz-rich core zone included the assemblage quartz-pyrophyllite±kaolinite, andalusite, chloritoid, sericite, paragonite, and iron oxide (Hughes, 1987). At the Snow Camp pyrophyllite mine and the Snow Camp South pyrophyllite prospect (fig. 2, fig. 3, fig. 4A, and fig. 5, sector J), chloritoid-rich patches are present peripheral to the pyrophyllite-rich rock. These observations were made on samples from the oxidized zone, and pyrite is assumed to have been present below the water table.

Origin of the Alteration Systems Summarized

Alteration systems such as these are now generally interpreted to have been formed by vigorously convecting hydrothermal cells adjacent to near-surface magma bodies and volcanic vents (Sillitoe, 1990; and White and Hedenquist, 1990). We think that the zones of quartz-sericite-paragonite altered rocks and their enclosed quartz-granofels and pyrophyllite-rich bodies in the Snow Camp-Saxapahaw district formed when shallow plutons intruded a thick pile of mostly andesitic to rhyodacitic pyroclastic rocks within a calderalike tilted graben, the Major Hill structural block. Hydrothermal convection cells driven by the heat of the intrusive bodies and supplied with strongly reactive magmatic or meteoric-magmatic fluids subjected the invaded rocks to intense high-sulfidation alteration.

The core zones may be broad, as that including Sheeprock (fig. 2, fig. 3, fig. 4A, and fig. 5, sector J) being almost 1 mi (1.6 km) in greatest horizontal dimension, and have one or several roots or pipes possibly extending downward for an unknown distance. The surrounding large quartz-sericite-paragonite alteration zones probably extend and narrow downward.

From core-zone rocks, such as those in Mine Ridge formed high in the original volcanic pile, to progressively deeper parts of the system, such as intrusive rocks to the southeast including the Lindley Farms Quartz Monzonite plutons, hydrothermal fluids present during alteration diminished and the style of alteration changed, resulting in the formation of peripheral bands of hornfels around the intrusives.

Sericitization in the Reedy Branch Tuff

Two types of alteration affected the Reedy Branch Tuff, moderate quartz-sericite alteration (Zra) and strong quartz-sericite alteration (Zrb). The strong alteration includes shear-related sericitization, especially of oriented white mica microzones that tend to wrap around larger plagioclase and quartz crystals. Locally, potassium feldspar has replaced plagioclase along microfractures. In other places, extensive alteration is indicated by development of coarse white mica plates in two cleavages of larger plagioclase crystals and by secondary biotite and epidote in microfractures. The degree of alteration near the base of the unit does not seem to bear any relation to the locations of intense high-sulfidation alteration in the older rocks below.

Epidote-Rich and Greisenlike Rocks

Epidote-rich rocks and epidosite, consisting almost entirely of quartz and epidote, are present at several places in the district. Common float materials of coarse micaceous rocks that resemble classical greisens were studied in four open fields. Some of them were associated with deuterically altered post-metamorphic felsic intrusive bodies (Cg). Neither fluorine nor boron minerals were identified. The greisenlike rocks are composed of muscovite-quartz+epidote+sulfides (fig. 8), and both rocks and adjacent soils contain molybdenum. The compositions of five of these rocks are given in table 9. Some epidote-rich veins have cut the Reedy Branch Tuff, but the origins and age relations of these intensely altered rocks may not all be the same.