Genesis of thunder eggs

[Admin]

indexarticles.com/reference/rocks-minerals/occurrence-and-genesis-of-thunder-eggs-containing-plume-and-moss-agate-from-the-del-norte-area-saguache-county-colorado/

Occurrence and genesis of thunder eggs containing plume and moss agate from the Del Norte Area, Saguache County, Colorado – Statistical Data Included

Daniel E. Kile

Known since the late 1800s, the thunder-egg beds near Del Norte, Colorado, have provided the collector in recent times with world-class plume and moss agate. Plume agate represents the ultimate in aesthetically arranged featherlike assemblages, whereas moss agate can manifest a bewildering variety of forms, including tubes as well as randomly intertwined fibers. These inclusions exhibit a wide range of color, from white to black and from brown to orange and red. Described as “plant-like” by Brown (1957), they are typically enclosed in chalcedony, which comprises the interior of spherical nodules commonly known as thunder eggs. Plume and moss structures are three dimensional in form, in contrast to the principally two-dimensional dendritic growths that occur between bands of chalcedony, such as those found in Montana “moss” agate.

Thunder eggs are generally defined as nodular structures (Staples [1965] emphasized that they should be considered structures, not “rocks”) that are formed within high-silica extrusive volcanic rocks or welded tuffs, with the silica content of the rock ranging from 75 to 80 percent (Dake 1951; Renton 1951; Staples 1965). Accordingly, thunder eggs are not found in comparatively silica-poor, basaltic rocks such as those capping the Table Mountains near Golden, Colorado. The nodules are approximately spherical in shape and have an exterior shell of rhyolite that is more silicified than the host rock (Renton 1951). Staples (1965) further refines this definition by describing surface textures ranging from smooth to wartlike, with intersecting ribs that encircle the nodule. Ross and Smith (1961) simply define thunder eggs as uncommonly large lithophysae (i.e., hollow gas cavities; Greek derivation = “stone bubble”) that develop in a welded tuff (welded tuffs are composed of a “plastic” volcanic ash that becomes indurated, or fused, by the internal heat that is inherent in the erupted rock [Thrush 1968]).

The thunder egg is composed of a shell of rhyolite that has devitrified (a process of crystallization of the volcanic glass) and in which spherulites appearing as radiating growths of acicular crystals and/or lithophysae occasionally develop. By traditional definition it has a central cavity that is partly to completely filled with chalcedony, opal, or quartz and that may exhibit a geometrical, star-shaped form. The chalcedony may be banded or it may contain “pseudo-algal” structures that are of special interest to collectors and lapidaries. The cavity may also be completely hollow or it may be absent altogether, leaving a solid rhyolite thunder egg. In the interest of simplicity, the term thunder egg as used in this article pertains to any of the above-described nodular structures, including those with no cavities, because all these forms are related in terms of genesis.

The Oregon state legislature designated the thunder egg as the state rock, using the term “thunderegg”; however, use of this term as two words has priority in the literature (Renton 1936; see also Staples 1965). Although thunder egg seems rather informal, efforts to apply a more scientific nomenclature, such as nodule, lithophysae, “rhyolite bomb,” or spherulite, have been fruitless because these terms are either too generic or they imply an inappropriate mode of origin. For example, although spherulite may be descriptive of some occurrences, such as those near Deming, New Mexico, and Silver Cliff, Colorado, it seems less appropriate for the structures at Del Norte because these nodules do not show a fibrous, radiating structure as stipulated by the traditional definition of a spherulite (e.g., see Thrush 1968). Thus, in the absence of a more technical word, numerous substitutes, some bordering on the whimsical, have been interjected. These include buffalo chips, blunder eggs, coconuts, and goof balls, among many others, one of which implies anatomical parts of elephants best not described here (Frazier, Frazier, and Mustart 1993). Amidst this plethora of terms and in the interest of historical precedent, the term thunder egg is used here (except it is thunder-egg when used as a modifier).

Thunder eggs are not uncommon in western North America, with the thunder-egg beds in the John Day Formation in central Oregon being the most prominently known. Other well-known deposits are documented at Dugway, Utah, and near Rockhound State Park near Deming, New Mexico. Although the Oregon thunder eggs have traditionally attracted attention for their spectacular plume and moss agate, the thunder-egg beds near Del Notre have through the years provided plume and moss agate equal to, or better than, that found at other worldwide localities.

Historical Background

Renton (1936) gives credit for the term thunder egg to the Warm Springs Indians in Oregon, whose legend recounts the angry spirits of Mount Hood and Mount Jefferson hurling spherical rocks at one another, accompanied by thunder and lightning; stray rocks became known as “thunder eggs,” which, as noted by Brown (1957), were ostensibly robbed from nests of “thunder birds.”

Nodules occurring in rhyolite from Colorado were documented relatively early, having been described as spherulites in 1891 from the Silver Cliff area in Custer County (Cross 1891) and from the Ute Creek area in Hinsdale County (Patton 1896). Hollow nodules from Specimen Mountain near Estes Park (now part of Rocky Mountain National Park) were also known by this time and were described by Patton in 1896. By contrast, the now widely known Oregon thunder-egg beds, although noted in the late 1920s, remained largely unknown until the mid-1930s (Renton 1951; Schaub 1979a), when articles appearing in The Mineralogist drew attention to them (Renton 1936; Dake 1940; Buddhue 1941).

The Del Norte thunder eggs also seem to have attracted attention at a comparatively early date, with Patton (1896) describing them as being similar to the Ute Creek nodules except for differences in their external color. Plume agate from Del Norte, however, was not documented until much later by Fahl (1948), who relates the “discovery” of the locality in 1944 by Ralph Dabney and Harry Simcox, both of southern Colorado.

Location

The Del Norte thunder-egg beds are located approximately 7.5 miles (by air) northwest of the town of Del Norte in southcentral Colorado. The locality is at the southern end of the La Garita Mountain range and along Old Woman Creek, at the western edge of the San Luis Valley. Although Del Norte is in Rio Grande County, the thunder-egg beds lie just across the county line, in Saguache (pronounced sawatch, or preferably, sa-wash’) County, a Ute Indian word for “blue earth” (Dawson 1954). The deposits in this area are somewhat limited in extent, being confined to the southeastern slopes of Twin Mountains and to areas extending northward for approximately 1.5 miles. Access over the last few miles is by a poorly maintained dirt road; a four-wheel-drive vehicle is recommended, especially in wet weather, although a high-clearance vehicle should suffice in dry conditions. Maps and/or mileage logs to the locality are given by Pearl (1965), Mitchell (1992), Voynick (1994), and Kappele (1995). The site is on the U.S. Geological Survey Twin Mountains 7.5′ quadrangle (1967).

Much of the area is held under nonpatented mining claim. Pearl (1965) reported that claims were located as early as 1958. Juanita Davis of Del Norte was one of the early claim holders who had properties on the southeastern slope of the southernmost of the Twin Mountains; these claims are now owned by parties in California (B. Morley, pers. com., 2001). Dick Siebenaler of Estes Park, Colorado, acquired a nonpatented claim (located somewhat northeast of the Twin Mountains) from Davis in the early 1970s; this property is currently maintained by his daughters, Siebenaler having passed away in 2001 (Morley, pers. com., 2001). Collecting of surface material is permitted on the Siebenaler claim; digging is not allowed. Other areas in the Baughman Creek volcanic center, as shown on a geologic map of the area (Lipman 1976), show evidence of thunder eggs and are good candidates for further prospecting.

Regional Geology

The thunder-egg beds occur within Oligocene rocks of the Baughman Creek volcanic center of the Conejos Formation. The Baughman Creek center is adjacent to the Summer Coon volcanic center, the two volcanoes apparently being concurrently active, as evidenced by similarities in their age and composition (Lipman 1976). The Baughman Creek flow is described by Lipman as a porphyritic, silica-rich rhyolite that consists of a single, pale gray lava flow with a well-developed flow layering; this flow, on the southeast flank of the Baughman Creek volcano, is up to 200 meters thick and now caps the northernmost peak of Twin Mountains. It is composed of about 70 percent Si[O.sub.2] and 5-10 percent phenocrysts, mainly plagioclase and biotite (Lipman 1976). The adjacent Summer Coon center, located northeast of the thunder-egg beds, is composed of a crudely bedded andesite breccia and local flows of alkali olivine andesite, which formed the main cone of the Summer Coon volcano. This flow consists of about 30 percent plagioclase, clinopyroxene, and olivine, as small phenocrysts.

Description of the Del Norte Thunder Eggs

External Structure

Thunder eggs from Del Norte are more or less spherical to oblate, ranging in size from less than 0.4 cm to 3 feet in diameter (or as much as 300 pounds [Fahl 1948]); most are reported to be 4-6 inches in diameter (Eckel 1997; Kile, Modreski, and Kile 1991) or 8-14 inches in diameter (Fahl 1948). By comparison, Oregon thunder eggs are reported to be up to 4 feet across (Renton 1951), but in some localities they average only 2.5-3 inches (Fahl 1948; Dake 1951).

Many of the Del Norte nodules show faint striations on their exteriors that presumably parallel the bedding plane of the host rock, although details of their mode of formation are unknown; they typically also parallel the direction of the long axis of the agate lens that formed within the nodule. Superimposed on these striations are external ribs, which give an indication of the internal agate structure. Thunder eggs from Del Norte can be easily recognized based on their exterior color and textural characteristics; thus nodules from Oregon, and even those from the nearby Houselog Creek locality in Saguache County (see Baldwin 1979; Kile, Modreski, and Kile 1991), can be readily differentiated.

A thin zone of porous material surrounds in situ thunder eggs; it is presumably derived from a low-temperature hydrothermal alteration of the host rhyolite that produces clay and/or zeolite minerals (Zarins 1977; Pabian and Zarins 1994; Smith, Tremallo, and Lofgren 2001). This layer was found by Smith, Tremallo, and Lofgren to be composed of mordenite and montmorillonite surrounding the large spherulites at Silver Cliff, whereas Renton (1951) presumed the layer encircling the thunder eggs at the Oregon localities to be composed of bentonite. X-ray diffraction (XRD) analysis of a chalky off-white material partly surrounding a Del Norte nodule showed the presence of a smectite-group mineral that was computer matched (by Jade software) to montmorillonite; a plagioclase feldspar and phlogopite were also detected in this sample. *

Shell

The thunder-egg shell is composed of what hits been presumed to be a devitrified and silicified rhyolite (e.g., Renton 1936); it is considerably denser and more weather-resistant than the host rhyolite, and in contrast to the gray color of the host rhyolite, the shell is a mottled brown color with varying tints of green.

Examination of thin sections of the thunder-egg shell by polarized-light microscopy shows phenocrysts of plagioclase (often fractured) that are as much as 7 mm in length and that are distinctly calcic- to sodic-zoned (2[V.sub.X] ~65[degrees], biaxial negative, polysynthetically twinned, with the Michel-Levy extinction angle varying from approximately 7[degrees] to 12[degrees], but ranging to 22[degrees], corresponding to an oligoclase-andesine composition); other phenocrysts identified in thin section include biotite/phlogopite (strongly pleochroic, 2[V.sub.X] varying from 30[degrees] to 43[degrees]) and magnetite. The groundmass, composed of minute feldspar crystals, is indicative of devitrification (i.e., crystallization of a glassy rhyolite as a result of falling temperature); the groundmass and phenocrysts are similar in size and density to that of the host rhyolite. All thin sections examined showed a zone of apparent alteration (or possibly a higher degree of silicification) of the rhyolite shell at its contact with the agate center.

Examination under plane-polarized light at high magnifications (i.e., greater than 100x) shows that there is no visual distinction between the thunder-egg shell and the host rock; only at very low magnification does the darker and slightly denser matrix of the thunder egg become evident. This may suggest a degree of silicification of the thunder egg, as stated by Renton (1936) and supported, in part, by data from Smith, Tremallo, and Lofgren (2001) that are based on a study of spherulitic nodules at the Silver Cliff locality in Colorado. Their study of whole-rock chemistry has shown somewhat higher silica contents of the spherulitic structures as compared to the host rhyolite glass.

Center

A thunder egg may have a hollow center or it may have a center that is partly or completely filled by agate and/or opal. The form of the center may range from simple lensoidal to a less commonly seen and more complex star pattern, the cross-sectional shape depending, in part, on the cutting orientation. The border of the cavity in all thin sections examined shows a thin layer of lathlike, tabular crystals. The individual crystals vary in length from an average of [less than or equal to] 0.06 mm to a maximum size of approximately 0.19 mm. They are predominantly uniaxial positive, but some crystals range to biaxial positive with a small optic angle that varies from 13[degrees] to 18[degrees]; they have a low birefringence and an index of refraction higher than that of opal (i.e., about 1.44, see below). Examination by scanning electron microscope (SEM) showed the presence of only Si and O *; single-crystal XRD identified this phase as quartz (H. Graetsch and O. Medenbach, written com., 2002). Although the optical, SEM, and XRD data are consistent with quartz, considering the incongruous crystal symmetry, it is likely that this mineral is actually a quartz pseudomorph after tridymite. The observed biaxial character could result from incomplete replacement in some crystals; tridymite can be metastable at ordinary temperatures (Deer, Howie, and Zussman 1963), and the inversion of tridymite from a hexagonal, uniaxial, high-temperature form to a biaxial, orthorhombic form is reported to be exceedingly slow, taking “millions of years” (R. M. Hutchinson, written com., 1994). The possible existence of a high-temperature form of tridymite may have important implications for thunder-egg genesis (see below). These crystals are in turn usually covered by a thin veneer of parallel fibrous crystals that show optical properties consistent with a weakly anisotropic form of opal (i.e., opal-CT, length slow, index of refraction < chalcedony; see below); occasionally an Fe-bearing phase is also present with this layer, as shown by SEM. Pabian and Zarins (1994), in a discussion of the cavity margin of basaltic agates, referred to "pre-banding" inclusions of celadonite or zeolite minerals, which they presumed to be derived from alteration of the host rock; similarly, Cross (1996) noted an agate ,border zone" composed of celadonite, delessite (chamosite), and "membranous" cristobalite, and Bishop and Rolfe (1989) documented celadonite, chlorite, and saponite as the first-deposited minerals in a gas cavity.

Agate often completely or partly fills the cavities and typically shows little pattern. Less often it is distinctly concentrically banded; occasionally it appears as “Uruguay bands” (sometimes referred to as an “onyx layer”), which are characterized by plane-parallel bands. Drusy quartz crystals sometimes line a hollow central cavity within the chalcedony. The agate centers are often fractured. In thin section, chalcedony commonly appears as radiating fibers that originate from a single point or as multiple bands of discrete, radiating, fan-shaped aggregates. The fibers are generally length fast, indicating that the c-axis of the quartz comprising the fibers is at right angles to the direction of elongation. Periodic, oscillatory extinction bands seen throughout radiating chalcedony aggregates are manifestations of changes in orientation due to periodic twisting of the crystallite fibers around the [11 [bar]2 0] or [1 [bar]1 00] directions (Merino and Wang 2001), resulting in a cyclic rotation of the c-axis around the axis of the fibers (see also Frondel 1978; Heaney 1993) and consequent extinction when the c-axis becomes perpendicular to the stage of the microscope. In contrast, examination of thin sections of Uruguay bands shows granular layers that are composed of randomly oriented equigranular chalcedony (possibly intergrown with opal-C [see Gaines et al. 1997]), the individual bands being characterized by variations in grain size and density.

A small amount of moganite (also known as lutecite [see Gaines et al. 1997]) was identified by X-ray diffraction in chalcedony in Del Norte thunder eggs (see also Heaney and Post 1992; Graetsch 1994). Although the term lutecite has priority, it is seldom used in recent literature, possibly because of its similarity to lussatite and lussatine, anisotropic variants of opal (see below). Moganite is proposed to be a monoclinic phase of Si[O.sub.2] that is distinct from quartz; it is evidenced by three minor peaks seen in X-ray diffraction patterns (e.g., see Florke et al. 1991; Heaney and Post 1992), although a much more elaborate Rietveld refinement technique must be used to unambiguously confirm its presence. The occurrence of this mineral as an intergrowth with chalcedony appears to be widespread in nature, as documented by Heaney and Post (1992) and Godovikov et al. (1993). Its validity as a distinct mineral species has been a subject of controversy, although it was approved by the International Mineralogical Association Commission on New Minerals and Mineral Names in 1999 (P. Heaney, pers. com., 2002). Its presence as a visibly distinct phase was not noted in Del Norte thin sections (i.e., length-slow fibers, inclined extinction, index of refraction ~1.524-1.531 [see Gaines et al. 1997]), although “quartzine,” a length-slow chalcedony with parallel extinction, was noted in one sample (see Godovikov et al. 1993; Florke et al. 1991).

Opal occurs in several forms at Del Notre and is typically intergrown with chalcedony; it can comprise nearly all of the cavity or only a small part of its volume, with chalcedony constituting the balance. The opal is invariably an opaque to translucent white, with textures in hand samples ranging from vitreous with a conchoidal fracture to chalky; all samples examined show a distinct pale bluish- to greenish-white fluorescence under shortwave ultraviolet radiation. Although opal is commonly thought of as an isotropic mineral, it actually is a complex intergrowth of several phases of microcrystalline silica, including cristobalite and tridymite, with varying states of disorder with respect to the stacking of the tetrahedral silica structures, which imparts distinctive optical properties and X-ray patterns. Thus, opal can vary from a uniformly isotropic, noncrystalline, and vitreous texture, to a weakly anisotropic, fine-grained material known as opal-C[T.sub.M], to an anisotropic mineral with a distinctly fibrous texture corresponding to opal-C[T.sub.LS] and known as lussatite (also designated as a disordered low cristobalite, or “opaline cristobalite,” by Frondel [1962]), or to a phase with a granular appearance described as opal-C (also called opal-[C.sub.P]), which is a platy anisotropic form known as lussatine (see Frondel 1962; Florke et al. 1991; Graetsch 1994; Gaines et al. 1997).

Optical and SEM studies of the Del Norte opal indicate that most of the phases described above are present. In all forms observed, the material shows an index of refraction substantially less than that of chalcedony (i.e., less than 1.54); one sample measured (using calibrated index media) showed n = 1.444 [+ or -] 0.004. The opal is heavily included with an unknown substance, the magnitude of which accounts for its white color and degree of opacity in hand samples, although in thin sections under plane-polarized light the opal has an overall tan to pale brown color. A form of opal-C[T.sub.M] to -C[T.sub.LS] likely constitutes the first thin film of semifibrous material that covers the tabular crystals (described above) that line the cavity walls; this material is mostly isotropic with faint birefringence in crossed-polarized light and is pale brown in plane-polarized light. Examination of this layer by SEM showed trace Al, the structural incorporation of which is consistent with a disordered silica phase. This layer is occasionally overlain by a distinctive and much thicker layer of comparatively anisotropic and distinctly fibrous, length-slow opal-C[T.sub.LS]. A granular, finely textured opal was found in the cavity interior, enclosed by chalcedony. Optical properties (i.e., low index of refraction, uniaxial negative) and XRD data show this material to correspond to opal-C as described by Florke et al. (1991) and to a disordered opal-cristobalite as described by Frondel (1962); furthermore, SEM analysis of this material showed only Si and O present. More work is necessary to completely characterize the nature of the complex silica microcrystalline phases present in the Del Notre thunder eggs.

Other late-formed minerals found in small amounts within hollow quartz-crystal-lined cavities are magnesian-calcite and gypsum, based on XRD patterns.

Plume and Moss Inclusions

Plume is considered to be a rhythmic structure that is organized into arborescent or feathery patterns enclosed by translucent to transparent chalcedony within the thunder-egg cavity. Plume structures are typically 1-2 inches long but can be as much as 5 inches. Black to brown colors predominate, whereas orange-gold colors are common and reds are comparatively rare (Fahl 1948; Kile, Modreski, and Kile 1991; Eckel 1997). Plume inclusions are uncommon at Del Norte. By contrast, moss is much more commonplace, occurring as irregular tubular filaments showing random branching and intertwined structures of infinite variety. Plume and moss inclusions generally originate from the cavity wall of the thunder egg, typically from an apex of the agate near the periphery of the thunder egg. A thin layer of translucent chalcedony, popularly called “Liesegang” rings, commonly encloses the plume and moss structures; sometimes these layers are prominent enough to partly obscure the enclosed plume or moss. Under plane-polarized-light microscopy, the plume structures range from diffuse wispy forms to more discrete arborescent forms, with translucency and color varying from nearly opaque and black to a transparent pale orange.

Plume and moss inclusions are often assumed to be composed of amorphous Fe- and Mn-hydrous oxides and hydrox-ides (Brown 1957; Staples 1965), although there has been little in the way of analytical work done on this material. In the present study, examination with polarized-light microscopy and XRD has shown the orange plume to be a noncrystalline material (SEM showed only the presence of Fe and O), although goethite, the presence of which was proposed by E. S. Larsen, as cited in Dake (1951) and Renton (1936, 1951), was detected in a trace amount in one sample by X-ray diffraction; this may correspond to a weak pleochroism that was noted in some thin sections of plume structures, which is suggestive of a crystalline, anisotropic mineral with a symmetry lower than isometric. Density of the orange plume structures ranged from very thin and filmy to comparatively dense with a low incident-light reflectivity of about 12 percent.

Black plume inclusions were also found to be composed of predominantly noncrystalline material, much of which proved to be Fe by SEM, but analysis of several samples of black or variegated orange-to-black plume structures by XRD and SEM, as well as determination of optical properties based on incident polarized-light microscopy and measurement of Vickers micro-hardness, indicated a variety of Mn-bearing minerals, many of which are collectively grouped under the term psilomelane. Accordingly, ramsdellite (an orthorhombic Mn[O.sub.2]) and possibly romanechite (a Ba-Mn-oxide) were noted, in addition to a trace amount of an “unspecified” Mn[O.sub.2] mineral (identified as such by Jade software) that was detected by XRD. The SEM also showed the presence of a Pb-bearing phase of Mn-oxide, as Mn-Pb-O (with a trace of Fe); based on optical properties (strongly anisotropic; percent reflectance at 540 nm = 24) and Vickers microhardness values (VH[N.sub.50] = 343), this phase is consistent with coronadite. These crystalline minerals likely comprise dark brown to black inclusions within the plume; such inclusions have a metallic to submetallic luster and can themselves vary from plumose textures to discrete granules. In addition to the Mn-bearing minerals, several samples were intergrown with barite, as evidenced by SEM detection of Ba-S-O. A rare-earth-bearing phosphate (i.e., Ce, La, and Ca) was also identified as minute (about 5 [micro]m diameter) inclusions within the Mn-beating phases. Similarly, a botryoidal inclusion of a metallic-appearing black mineral that was enclosed by chalcedony proved by XRD analysis to be intergrowths of cryptomelane (a K-Mn-oxide), hollandite (a Ba-Mn-oxide), and vernadite (a Mn-hydroxide). Thus, some opaque black areas of plume that show a more metallic luster are also likely composed of these or similar crystallized Mn-bearing minerals.

White plume structures were identified as a phase of opal (possibly a mixture of opal-[C.sub.M] and opal-C[T.sub.LS]) by polarized-light microscopy; examination by SEM of one such structure showed only the presence of Si and O.

The presence of the above-mentioned Mn-bearing minerals is consistent with results given by Potter and Rossman (1979), who reported coronadite, cryptomelane, hollandite, romanechite, and todorokite in Mn-bearing dendrites, and by Mitchell, Giannini, and Fordham (1988), who noted romanechite (one of the mineral species formerly called psilomelane) comprising plume structures from an occurrence in Virginia. More detailed work is needed tin the Del Norte material to better characterize and confirm the various mineral phases that constitute the plume structures.

Theories of Thunder-Egg Genesis

The only consistent aspect of the various theories for thunder-egg formation is a relative lack of either consensus or consistency. Part of this perplexing situation arises from the sometimes interchangeable use of the terms thunder egg and spherulite. Although a definition of the thunder-egg structure is relatively straightforward, that for a spherulite is less so. Cross (1891) commented that “the term spherulite is one of many in the nomenclature of petrography to which no satisfactory and consistent definition has as yet been given….” The traditional definition of a spherulite describes radiating bundles of intergrown cristobalite and feldspar fibers that are found in welded tuffs (Thrush 1968; Smith, Tremallo, and Lofgren 2001). They result from nonequilibrium crystallization from the devitrification of volcanic glass (Lofgren 1974), which is postulated to occur at relatively high temperatures following compaction and welding of a hot rhyolitic ash flow (Ross and Smith 1961; Jacobs et al. 1992). Photographic examples of spherulitic textures as seen in thunder eggs and volcanic glass are provided by Bryan (1963), Lofgren (1974), McLemore and Dunbar (2000), and Smith, Tremallo, and Lofgren (2001).

The textures of thunder eggs can vary from distinctly radiating fibrous textures to cryptocrystalline aggregates that preserve the structure of the original rhyolite (Kay 1981). The thunder eggs at Del None, however, seldom show the radial fibrous textures characteristic of spherulitic crystallization. Accordingly, the term spherulite (e.g., as used by Patton 1896) seems not to adequately describe the structure of the Del Norte nodules. However, this term may be more appropriate as applied to other related occurrences (e.g., the Silver Cliff, Colorado [Cross 1891; Patton 1896; Smith, Tremallo, and Lofgren 2001]) and possibly the Deming, New Mexico (McLemore and Dunbar 2000) localities. Bryan (1941) adopted the term spheruloid for nodules without radial crystallization, but it has not gained wide acceptance. Consequently, there seems to be no “scientific” term that satisfactorily describes the Del Notre thunder eggs as far as structure and mode of formation are concerned.

In brief, the genesis of thunder eggs commences in a silicarich rhyolite (which is relatively high in water content) by a process of anhydrous crystallization of the glass that results in the formation of cavities due to exsolution of water vapor. Silica, derived from alteration of the rhyolite, later infiltrates the cavity at a lower temperature and eventually crystallizes into chalcedony, completing the genesis of a thunder egg. However, details of its development remain elusive. Although the general tenets of thunder-egg formation recounted above are mostly accepted, textural features and variations of the Del Norte thunder eggs need to be reconciled with this model. Moreover, a model for the genesis of the thunder-egg shell remains speculative at best. Various hypotheses for thunder-egg development are described below under five general phases. The attributions given are not intended to be comprehensive but rather to provide selected references as a basis for further reading.

1. Rhyolite emplacement. This phase of thunder-egg genesis is among the few that are relatively uncontested. The temperature of the rhyolite extrusion is uncertain; however, based on experimental evidence and the measured temperatures of magma extrusion (see Ross and Smith 1961 and references therein), emplacement of host rhyolite is presumed to be at temperatures less than 1,000[degrees]C. Laboratory research with glass melts has affirmed the mineral assemblages seen in the rhyolite following cooling (e.g., Dunbar, Jacobs, and Naney 1995). Thunder eggs are generally confined to a single stratographic horizon in ash-flow tuffs and are most common in Tertiary or younger rocks (Pabian and Zarins 1994).

2. Formation of rhyolite shell. This phase of genesis seems to be the one most lacking in consensus. Three general hypotheses are described below; unfortunately, there are no quantitative analytical data that directly support the common presumption that the thunder-egg shell bears a higher silica content than does the host rhyolite.

* Hypothesis 1. Genesis of the shell was from an immiscible silica (Si[O.sub.2] * n[H.sub.2]O) at magmatic temperatures (within a “red-hot magma”), which coalesced into a plastic sphere at temperatures above 500[degrees]C, incorporating rhyolite within the silica (Schaub 1979a,b; 1989). This model (and the one described below) does account, in part, for the formation of a thunder-egg shell. However, although supersaturated silica solutions have been hypothesized to arise from the reaction of hot water with volcanic glass (Fournier 1985), there is no proven mechanism that can explain the process of coalescence or accretion of a silica or colloidal phase within a hot volcanic glass.

* Hypothesis 2. Bryan (1954) and Kay (1981) attributed development of a shell to the formation of “colloidal substances” around a nucleus of rhyolite phenocrysts or vapor bubbles. It has been proposed that the latter formed either as residual gas bubbles in the rhyolite or from coalesced volatiles derived by the devitrification process in the still-hot lava. This sequence is illustrated in Bryan (1954), Kay (1981), and Godovikov, Ripinen, and Motorin (1987).

* Hypothesis 3. Intergrown spherulites produced a hardened, weather-resistant outer shell of the thunder egg (Staples 1965). Formation of spherulitic shells for the Deming, New Mexico, thunder eggs has been postulated by Dunbar and McLemore (2000) to occur at relatively high temperatures (e.g., 700[degrees]-1,100[degrees]C) following rhyolite extrusion, based on crystallization studies in artificial melts. However, as there is very little evidence of a spherulitic structure in the Del Norte thunder eggs, this theory seems not to account for the genesis of either the shell or the central cavity at this locality.

3. Genesis of the central cavity. Three hypotheses are presented below, of which only the first is widely accepted today.

* Expansion theory. Genesis of the central cavity commenced with an initial vapor bubble in the host rhyolite, to which devitrification of the host rhyolite or crystallization of spherulites within the developing spherical thunder-egg structure contributed. This was accomplished through an increasing vapor pressure as a result of the formation of anhydrous minerals such as cristobalite and feldspar; as a result of this crystallization, the water in the host rhyolite became immiscible and was exsolved as an aqueous vapor (i.e., steam) that coalesced and formed a cavity (sometimes referred to as a lithophysa) by expansion due to increasing vapor pressure within the thunder-egg structure. A secondary contribution to cavity formation may have been from shrinkage of the cooling host rock. Various aspects of this theory were proposed by Wright (1915), Ross (1941), Renton (1951), Dake (1951), Bryan (1954), Ross and Smith (1961), Staples (1965), Kay (1981), Pabian and Zarins (1994), Cross (1996), Dunbar and McLemore (2000), and McLemore and Dunbar (2000).

The role of devitrification of water-rich rhyolite in the formation of spherulites and thunder eggs was recognized early. Wright (1915) credited von Richthofen for having proposed, in 1860, a model for the formation of lithophysae by the expansion of gas bubbles that were liberated by the crystallization of spherulites, and Patton (1896) hypothesized that this mechanism of formation applied to thunder eggs. Moreover, the apparent early crystallization of tridymite on cavity walls may infer relatively high temperatures during the initial stage of thunder-egg-cavity formation, as the high-temperature [beta] (hexagonal) form crystallizes (at equilibrium) from a vapor phase at temperatures above 870[degrees]C (noting, however, that tridymite could form at much lower temperatures during metastable crystallization). Calculations have shown that the process of devitrification can provide sufficient water vapor to account for the volume in the thunder-egg cavities (McLemore and Dunbar 2000). However, this theory does not explain why all lithophysae within the host rhyolite should be restricted to the siliceous nodular structures, nor does it completely account for thunder-egg structures that have no interior cavity.

* Shrinkage/infiltration theory. Early cavities were formed by uniform contraction in a cooling rhyolite. Alternatively, Reed (1940) proposed that cavities formed as a result of bubbles rising within the host rhyolite (this model does not account for the absence of small cavities throughout the rhyolite). Formation of these initial cavities was followed by crystallization (devitrification) of the rhyolite and consequent migration of aqueous vapor into the cavities, with subsequent cavity filling by a rhyolite “mud” and shrinkage of the mud due to drying and cracking, resulting in a star-shaped cavity (Reed 1940; Dake 1940, 1951). A shrinkage model was also proposed by E. S. Larsen (as related by Renton 1936, 1951; Schaub 1979a).

The shrinkage theory does not account for an absence of a layered mud within the thunder-egg structure (as might be an expected result of a progressive cavity infilling), although it does attempt to explain the formation of a thunder-egg shell. Moreover, transport of a mud through a viscous rhyolite seems rather untenable. Ross (1941) also attributed a minor degree of cavity formation to tensional shrinkage due to cooling of the host rhyolite.

* Deformation theory. Cavity formation within a silica sphere is attributed to tensional deformation and stress at high temperature, creating shear planes, with infiltration of colloidal silica along these shear zones resulting in cavity expansion (Schaub 1979a,b; 1989). Duds (i.e., thunder eggs lacking an agate interior) are explained as having been formed too late in the rhyolite cooling history, in the absence of a silica-rich colloidal phase, to incorporate an agate core. This theory requires a mechanism for developing tensional stress; generation of such stress in many different worldwide occurrences at exactly the same time in the cooling history of the host rock necessitates a rather convoluted reasoning. It also implies a relatively high temperature for the emplacement of the agate (see below).

4. Infiltration of silica; formation of chalcedony in the central cavity. In addition to the references pertaining to agate genesis cited in the section below, Moxon (1996) provides a concise overview of the numerous theories of agate formation. Two models are given, the first of which is not widely accepted.

* Model 1. Agate formation resulted from silica migration into the cavity along tensional shear planes and a rhythmic fractional crystallization on the interior of the thunder egg at high temperatures (i.e., “as temperatures cool below magmatic”) (Schaub 1979a,b; 1989).

* Model 2. Agate formation occurred by infiltration of fluids supersaturated in silica, followed by crystallization under comparatively low-temperature (e.g., less than 300[degrees]C) and low-pressure conditions (see Lund 1960; White and Corwin 1961; Fournier 1985; Taijing and Sunagawa 1994; Landmesser 1984, 1988, 1995, and references therein); the silica is presumed to be in a monomeric form, such as Si[(OH).sub.4] (Landmesser 1984, 1995). Infiltration of these fluids is presumed to be through cracks and microscopic pores in the thunder-egg shell. The silica-rich cavity infilling originates from late-stage hydrothermal fluids that are derived from the host rock and local groundwater; the silica is thus a secondary alteration and dissolution product of the enclosing rhyolite (Ross 1941; Staples 1965; Dunbar and McLemore 2000; McLemore and Dunbar 2000), which produces a siliceous gel, clay, and zeolites (Zarins 1977; Pabian and Zarins 1994).

Although chalcedony is now generally presumed to form under relatively low temperatures, the controversy regarding its temperature of formation is far from over. For example, Merino and Wang (2001) have proposed a model for high-temperature agate formation (E. Merino, pers. com., 2002) that is based on both new (unpublished) isotopic data and on an earlier model of self-organizational crystallization and trace element data (Wang and Merino 1990, 1995). The high-temperature model also postulates chalcedony formation in a closed system from “lumps of silica gels,” in contrast to open-system (and lower-temperature) crystallization proposed by Landmesser (1998) and Heaney and Davis (1995), where the silica is derived from alteration. Cross (1996) summarizes observations that seem to contradict a high-temperature hypothesis (i.e., inclusions of low-temperature minerals in agate; the common occurrence of amethyst [the color of which is not stable above about 349[degrees]C] at many agate geode localities; and numerous sedimentary occurrences of agate, including that of silicified wood).

Crystallization of agate has often been assumed to proceed from a gel matrix (e.g., see Harder 1993; Wang and Merino 1995); layers and bands are presumed to be a manifestation of slight changes in composition of the aqueous gel that result from multiple cycles of hydrothermal fluid infiltration. However, Landmesser (1988) proposed a model of agate formation from both a colloidal gel (i.e., a three-dimensional network of silica), which forms a more or less concentrically banded agate, and from a low-viscosity aqueous silica fluid (i.e., a sol, in which silica colloids are dispersed in a liquid), from which horizontally banded agate is formed. His proposed sequence commenced with an initial infiltration of aqueous silica that was followed by deposition of a layer of a silica gel (or more correctly stilted, a “gel-like” material, as the colloidal silica particles are presumed not to have linked together in a three-dimensional framework [see Landmesser 1998]) on the cavity wall; the deposition of this silica resulted in a layer of banded chalcedony, If the silica was sufficiently concentrated, the sol could continue to form a gel and additional banded agate. Conversely, in areas where the silica was less concentrated, a gel would not form; rather horizontal Uruguay bands would form by precipitation and settling out of silica crystallites from the aqueous sol. Similarly, Uruguay bands have been proposed to form by the gravitational settling of large colloidal silica particles in a gel (Bishop and Rolfe 1989; Landmesser 1998). Both gel and sol phases are hypothesized to be present in a given cavity at the same time. The plane-parallel layers of Uruguay bands are presumed to have been deposited horizontal to the land surface (i.e., in a geopetal orientation) during deposition (see Renton 1951; Brown 1957; McLemore and Dunbar 2000).

An alternative model accounts for chalcedony crystallization by a mechanism of crystallite precipitation from low-viscosity aqueous fluids (Heaney 1993). This concept does not entail the simultaneous existence of a silica gel and a sol, but rather a coexistence of short-chain silica polymers and monosilicic acid that condense to form quartz fibers (P. Heaney, pers. com., 2002).

The final step in this sequence is the continuing crystallization of chalcedony and concomitant depletion of silica from the aqueous colloidal solution, resulting in the formation, within residual open cavities, of minute quartz crystals predominantly composed of rhombohedral faces (Landmesser 1988). Fournier (1985) presumed that the formation of euhedral quartz occurred under conditions of slow crystallization from supersaturated silica fluids.

In summary, the sequence of quartz crystallization is: (1) a layer of chalcedony lining the cavity wall, (2) Uruguay bands, (3) irregular and concentric bands of chalcedony, and (4) euhedral quartz crystals; not all phases are necessarily present in a given agate.

Thin halos of transparent chalcedony are often noted surrounding other inclusions such as plume or moss; they are noted to have eradicated earlier banding (e.g., Uruguay bands) in the agate (Staples 1965). The halos can produce a concentric zoning that replicates a pattern similar to that described by Liesegang (e.g., 1915; see also Brown 1957); consequently, they are often popularly called “Liesegang” rings. Although this term has often been used in the literature, strictly speaking, the halos are not Liesegang rings at all, as they do not show a periodic precipitation sequence that is characteristic of the rings as originally described by Liesegang. In fact, true Liesegang rings were observed in only two samples in an extensive study of agates by Landmesser (1984, 1988). However, this term is occasionally used in the present article in accordance with literature precedent.

5. Formation of plume and moss structures. The formation of plume and moss structures within agate is widely assumed to be a result of a rhythmic, chemical precipitation of the salts. oxides, or hydrated oxides of Fe and Mn by diffusion and osmosis (e.g., see Dake 1951; Renton 1951; Brown 1957; Staples 1965; Cross 1996), which proceeds from cavity walls into a silica-gel matrix prior to its crystallization into chalcedony, a process that is analogous to chemical precipitation in sodium silicate gels (Farrington 1927; Brown 1957). Although XRD, SEM, and polarized-light microscopy analyses reported herein mostly support these compositional assumptions, further analytical data are necessary to completely characterize these structures from Del Norte as well as other worldwide thunder-egg localities, There is also controversy regarding the silica medium in which plume and moss structures formed; whereas many writers have proposed that plume structures formed in a gel matrix (e.g., Dake 1951; Brown 1957; Staples 1965), others have postulated that plume formation occurred in a nonviscous silica solution or sol (Sinkankas 1966; Cross 1996).

Some plume structures from Del Norte occur as “free-standing” entities (i.e., they occur in a cavity that is not filled with chalcedony), surrounded only by thin halos of clear chalcedony. This configuration has also been noted in Oregon thunder eggs (see Dake 1954; Pabian and Zarins 1994). Brown (1957) accounts for these structures by postulating an initial growth in an aqueous solution that subsequently “leaked out,” which suggests a relatively early formation of the plume. This hypothesis is in accordance with plume formation in a sol, as it seems unlikely that a gel matrix could be so easily dispersed.

Thunder-Egg Genesis Theories and Their Relevance to the Del Norte Occurrence

There are elements of many of the above theories that seem pertinent to the Del Norte occurrence. The consensus of most writers is that cavities form due to expansion and that the agate was deposited at relatively low temperatures. However, difficulties arise in applying these concepts to specific occurrences because of considerable variation in thunder-egg development from various localities. Much of the above theory was developed using the Oregon nodules, some of which show a distinct spherulitic structure (reportedly observed in approximately 20 percent of the total thunder eggs [D. Rigel, pers. com., 2002]), as a basis for textural interpretation (Staples 1965; Pabian and Zarins 1994; Renton 1951). Thunder eggs from near Deming, New Mexico, are also documented to show a well-developed spherulitic structure characterized by a radiating fibrous habit that is seen in both hand samples and thin sections (Dunbar and McLemore 2000; McLemore and Dunbar 2000). By contrast, thunder eggs from Del Norte seldom show this feature, leading to a contradiction in deducing cavity formation based on a spherulitic origin of gas.

Of the many Del Norte thunder eggs examined, only three showed an indistinct concentric pattern (but without a radial texture) that constituted an integral component of the rhyolite shell. These forms appear macroscopically as concentrically zoned structures, ranging in size to 1.2 inches in diameter. Although they lack the fibrous, radiating crystals that conform to the traditional definition of a spherulite, their cores show a crystal size and density somewhat higher than that of the surrounding rhyolite shell. These structures may be evidence of early nucleation and devitrification (with corresponding vapor formation) that preceded that in the surrounding host rhyolite. It is thus conceivable that these centers of crystallization served as nuclei for further devitrification and consolidation that ultimately resulted in the final thunder-egg structure. Con, sequently, these concentric structures may be evidence of a key element that contributes to an early phase of formation of the Del Norte thunder eggs, although their apparent absence in most of the material examined is troublesome (noting, however, that parallel slices taken through a three-dimensional thunder egg will only by chance encounter a hypothetical concentric nucleation center). It is also possible that early formed, radially fibrous, spherulitic structures in Del Norte thunder eggs were by some process largely obliterated during later lithification or silicification; evidence for this is sometimes noted in Oregon thunder eggs, which may show a distinct central button that is partly surrounded by a very subtle and diffuse radial structure.

In addition to the three samples described above, a single agate core, weathered from a Del Norte thunder egg, shows a distinct positive and negative spherulite “cast,” or button, such as described and illustrated by Renton (1951) and Pabian and Zarins (1994). Thin-section study of the bulk of the material comprising the Del Norte thunder-egg shell shows almost no development of spherulitic textures based on the traditional definition. Moreover, the thin sections examined show a uniform devitrification of the thunder-egg shell and surrounding rhyolite, with equal development of euhedral phenocrysts, excepting the concentric structures discussed above. In some samples, small areas of minute subparallel fibers with a convoluted texture are evident within the devitrified thunder-egg shell; it is possible that these fibrous structures could represent remnants of early spherulitic growth, although the magnitude of such crystallization on a microscopic scale does not seem adequate to generate sufficient aqueous vapor to support cavity formation. Thus, if the crystallization of megascopic spherulites does not provide an adequate means of aqueous vapor generation in the Del Norte thunder eggs, then another mechanism must be found to account for cavities in these structures.

Based on the above observations of hand samples and thin sections, a model that pertains to the Del Norte occurrence can be proposed. Although this model, which is largely conjecture, incorporates elements from a number of the works discussed above, the general sequence proposed by Bryan (1944, 1954) and Kay (1981) and illustrated by Godovikov, Ripinen, and Motorin (1987) seems to most closely fit the physical structures seen at Del None. Any model pertaining to the Del Norte occurrence needs to account for the formation of a resistant spherical shell structure both with and without an agate center (the latter form being relatively common at this locality) and for cavity formation in the absence of a conspicuous spherulitic structure. Much of this model is predicated on a low temperature of agate formation; needless to say, this model would need a complete overhaul if a high temperature of formation is ultimately proven. Of the six phases outlined below, the second and third are the most tenuous, an indication that a considerable amount of field and analytical work is still needed.

1. Extrusive volcanic activity results in a plastic flow of rhyolite.

2. Silica accretes or radially crystallizes in a rhyolite that is supersaturated in silica (by means of cooling). There appears to be some process by which silica becomes enriched to form a nodular structure, as evidenced by numerous solid or near-solid thunder eggs (i.e., those without internal vesicles). A comparatively high silica content in the thunder-egg shell was proposed by Renton (1936); however, as discussed earlier, there are no chemical whole-rock analyses to confirm this supposition. Thus, for lack of a better term, this process is herein referred to as accretion, with the understanding that there is no specific known mechanism whereby this can happen. This mechanism may be analogous to the formation of carbonate concretions in sediments or chert nodules in limestones (e.g., Knauth 1994). Nucleation of silica may have commenced around a spherulite that formed in the earliest stages of thunder-egg genesis, prior to extensive generation of aqueous vapor. Thus, the concentrically zoned structures discussed above may have served as nucleation centers for silica consolidation or radial crystallization. Alternatively, nucleation could have transpired around a small vapor bubble that originated either from the early devitrification of the still-hot rhyolite or possibly from a reduction in pressure upon extrusion of the host rhyolite. The silica-enriched zone thus formed could have assumed a spherical profile that gave rise to the weather-resistant thunder-egg shell.

3. A cavity forms by coalescence and expansion of vapors derived from devitrification within the still-plastic silica-rich shell; the process of silica accretion may have initiated a localized devitrification prior to that in the surrounding host rock. Exsolution of aqueous vapors would have transpired primarily by crystallization of minute, randomly oriented anhydrous minerals throughout the thunder-egg shell and perhaps also by a secondary contribution due to crystallization of macroscopic spherulitic structures (note that the Del None thunder-egg shells seldom show spherulitic textures, but they do show extensive devitrification). Alternatively, accretion of colloidal silica per se could be hypothesized to directly induce exsolution of an aqueous vapor within the silica structure, with or without concomitant vapor formation via devitrification. Entrapment of a vapor phase within the thunder-egg structure may have been possible as a result of early consolidation of the exterior of the shell. Any of the aforementioned mechanisms, or combinations thereof, would account for cavities that are restricted to thunder-egg structures. Moreover, the presence of tridymite may be suggestive of a comparatively high temperature during the initial phase of cavity formation, which is consistent with the findings of Dunbar and McLemore (2000), based on the temperature of formation of spherulites from experimental melts. Nodules showing no central cavity may have formed by silica accretion/crystallization in the absence of the development of an obvious spherulite morphology, as per the above discussion, or after the rhyolite had become mostly devitrified. Solidification of the shell by continued cooling of the rhyolite fixed the overall structure of the thunder egg.

4. The cavity is infilled by a colloidal silica suspension (i.e., a sol); plume and moss structures form by rhythmic precipitation of Fe and Mn salts in a gel or sol matrix.

5. Silica crystallizes initially as chalcedony (from a gel) that forms the cavity lining, subsequently from a sol that forms Uruguay bands, and finally from a gel that forms concentrically banded chalcedony. Plume and moss structures are now completely enclosed.

6. Euhedral quartz crystallizes from a relatively low-concentration silica solution in a residual cavity, forming drusy quartz that lines the cavity walls.

This model retains elements of many of the previously documented theories for thunder-egg genesis but differs from some in that il combines the hypothetical generation of a silica “concretion” (at high temperatures) around an isolated and small spherulitic or vapor bubble nucleus, Crystallization of anhydrous minerals sufficient to generate the vapor phase required to form an extensive, geometrical agate center within the shell is still speculative. Other details of thunder-egg genesis remain conjecture as well and are still largely unexplained. For example, this model does not explain the absence of numerous small lithophysae outside of the thunder-egg structure, where devitrification of the host rhyolite should also have led to generation of a vapor phase. Conversely, there seems to be no plausible mechanism whereby vapor generated in the host rhyolite should have migrated to areas within the thunder-egg structure. In short, there seems to presently be no adequate theory that comprehensively describes all features of thunder-egg genesis at Del Norte!

Lapidary Treatment of Thunder Eggs (With Some Hints on Field Collecting)

Few thunder eggs contain much of lapidary interest. Although the roughly spherical nodules can be cut in half with the hope of exposing an aesthetic pattern of agate or opal, thunder eggs with moss or plume inclusions are relatively uncommon. There are, however, several external features of thunder eggs that give some indication of their content. For example, the external ribs on a nodule give an indication of its internal agate structure. A thunder egg with a single, central rib circumscribing its equator indicates a single lensoidal agate interior, whereas nodules with multiple external ribs suggest a more three-dimensional agate center that may appear as a geometric star when cut in appropriate orientation. Thunder eggs with minimal or no external ribs and of a comparatively uniform spherical shape (locally known as “cannon balls”) are composed only of siliceous, devitrified rhyolite without a central cavity. Plume agate is said to occur in fewer than 1 percent of the thunder eggs mined (B. Morley, pers. com., 2001). By comparison, Dake (1954) estimated that 80 percent of Oregon thunder eggs were of average-to-poor grade, whereas the remaining 20 percent were fine or better grades; no mention is made of the relative abundance of plume agate at this locality, although it is said that the most productive thunder-egg beds with plume agate were worked out many years ago (D. Rigel, C. Rose, and B. Warrington, pers. com., 2002).

Because plume and moss inclusions often start at peripheral cavity margins, the most expedient way to find a thunder egg with these inclusions is to cob a corner off the thunder egg (generally at a point where the agate is in close proximity to the surface [i.e., at an external rib]) and look for a trace of moss or plume within the exposed agate. The presence of external warts or protuberances, for reasons unknown, is often an indicator of moss or plume within the thunder egg. Thunder eggs tend to exhibit similar characteristics in the field within a given area; for example, if a solid nodule (i.e., without an agate center) is found in situ, chances are that other nodules in the vicinity will be of a similar form. Conversely, plume-agate-bearing nodules often occur in “nests”; thus, if a plume-bearing thunder egg is found, others nearby will be more to contain plume as well. Dake (1954) noted this same trend in thunder eggs from the Oregon deposits.

Once found, the thunder egg needs to be cut with a diamond saw to reveal the (hopefully plume- or moss-bearing) agate core. Because plumes are three dimensional, usually only one or two good slabs can be obtained from a given thunder egg, the others showing only plume tips. Needless to say, those unlucky enough to cut across the top of a plume structure will find only material with a nondescript mosslike appearance, an event that is usually accompanied by vociferous expletives! Accordingly, almost everyone who has cut a number of these nodules has developed a pet theory as to proper orientation. Some lapidaries cut off the ends of a round nodule in an attempt to locate and properly orient a plume, whereas others will saw the thunder egg parallel to its external striations because this direction often gives the maximum exposure of a lenticular agate core. In contrast, Oregon thunder eggs are usually oriented based on the pattern of the external ropes, revealing a geometric star. Regardless of the method, it remains a matter of luck to find the plume-bearing thunder egg in the first place, much less to somehow make the right decision to cut it in an ideal direction.

Occasionally, the plume or moss displays well as an intact thunder-egg half, complete with its external shell. However, most thunder eggs are either found broken in the field or show other deformities or such a degree of asymmetry that the plume needs lapidary treatment for an overall balanced and aesthetic appearance. A detailed discussion of this technique is beyond the intent of this article, but, in brief, use of diamond abrasives is the only means of sanding both the hard agate and the comparatively soft rhyolite and plume without severe undercutting. The thunder-egg shell takes a good polish (presumably a result of its high silica content) and thus constitutes an aesthetic matrix on which a plume structure can be framed.

Conclusions

Over the years the Del Norte thunder-egg beds have provided exquisite plume and moss agate, some of which, when accompanied by the thunder-egg matrix, are aesthetic specimens that can surpass those from the far-better-known thunder-egg deposits in Oregon.

Considerably more work will be necessary to further elucidate details of the genesis of the Del Norte thunder eggs (as well as thunder eggs occurring elsewhere). Whole-rock analysis and microprobe data of both the host rock and the thunder-egg shell, in addition to systematic field sampling and extensive X-ray and thin-section analyses, are needed to investigate the various hypotheses described above. Our understanding of thunder-egg genesis in particular, as well as the broader issue of agate genesis, has changed little since 1966, when Sinkankas stated, regarding the formation of chalcedony and its inclusions: “It is regrettable that more qualified mineralogists have not taken up the challenges offered….”

[FIGURES 1-29 OMITTED]

ACKNOWLEDGMENTS

Particular acknowledgment is given to Tom Michalski (U.S. Geological Survey), who provided literature resources and valuable discussions regarding agate genesis. Peter Modreski (U.S. Geological Survey) reviewed the manuscript, providing important details of igneous petrology in addition to editorial suggestions, which greatly improved the final version. Enrique Merino (Indiana University) and Peter Heaney (Pennsylvania State University) provided thoughtful comments and/or literature on the subject of agate genesis. Special thanks are due Olaf Medenbach and Heribert Graetsch (Ruhr University, Bochum) who provided timely sample preparation and single-crystal X-ray analysis, B. F. Leonard (U.S. Geological Survey, retired) kindly translated selected Russian text from Godovikov, Ripinen, and Motorin (1987); Isabelle Brownfield (U.S. Geological Survey) directed work on the SEM; Susann Powers (library, U.S. Geological Survey, Denver) provided, on short notice, many critical references necessary to complete this work; and Dianne Kile gave helpful editorial suggestions. Joe Taubr, many years ago, shared his knowledge about this locality; Barbara Morley (Dick’s Rock Museum, Estes Park, Colorado) afforded the opportunity to examine and photograph Del Norte plume agate and provided information regarding the mining claims in the area; and William Besse prepared the location map. Use of trade and product names in this paper is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey.

* X-ray diffraction was done with either a Siemens D-500 or a Nicolet diffractometer using a Cu-radiation source (operating at 40 kV, 30 mA), scintillation detector, and a graphite monochromator. Jade Materials Data software (version 5, MDI, Livermore, California) was used for computer analysis of the X-ray patterns.

* JEOL scanning electron microscope, operated at 15 kV, with Oxford EDS and Isis software.

REFERENCES

Baldwin, C. E. 1979. Colorado gem and mineral collecting localities. Golden, Colo.: Noble Gems.

Bishop, A. C., and W. D. I. Rolfe. 1989. Agates. London: British Museum (Natural History); Edinburgh: The National Museums of Scotland.

Brown, R. W. 1957. Plantlike features in thunder-eggs and geodes. In Annual report of the Board of Regents of the Smithsonian Institution for 1956, 329-39. Publication 4272.

Bryan, W. H. 1941. Spherulites and allied structures, part 1. Proceedings of the Royal Society of Queensland 52:41-53.

–. 1954. Spherulites and allied structures, part 2. Proceedings of the Royal Society of Queensland 65:51-69.

–. 1963. The later history of expanded spherulites. Journal of the Geological Society of Australia 10:141-49.

Buddhue, J. D. 1941. A possible explanation of the formation of “thunder eggs.” The Mineralogist 9:338.

Cross, B. L, 1996. The agates of Northern Mexico. Edina, Minn.: Burgess Publishing.

Cross, W. 1891. Constitution and origin of spherulites in acid eruptive rocks. Philosophical Society of Washington 11:411-43.

Dake, H. C. 1940. Trailing the “thunder egg.” The Mineralogist 8:335, 349.

–. 1951. The agate book. Portland, Ore.: Mineralogist Publishing Company.

–. 1954. Agate filled thunder egg genesis. The Mineralogist 22:8-14.

Dawson, J. F. 1954. Place names in Colorado. Denver: Golden Bell Press.

Deer, W. A., R. A. Howie, and J. Zussman. 1963. Rock-forming minerals. Vol. 4, Framework silicates. New York: John Wiley and Sons.

Dunbar, N. W., G. K. Jacobs, and M. T. Naney. 1995. Crystallization processes in an artificial magma: Variations in crystal shape, growth rate and composition with melt cooling history. Contributions to Mineralogy and Petrology 129:412-25.

Dunbar, N. W., and V. T. McLemore. 2000. The origin of rhyolitic spherulites at Rockhound State Park. New Mexico. In Abstracts, New Mexico Mineral Symposium, November, 2000, Socorro, New Mexico, 7-8.

Eckel, E. B. 1997. Minerals of Colorado. Updated and revised by D. S. Collins, R. R. Cobban, E. E. Foord, D. E. Kile, P. J. Modreski, and J. A. Murphy. Golden, Colo.: Fulcrum Press.

Fahl, R. 1948. Colorado plume agate. The Mineralogist 16:34-37.

Farrington, O. C. 1927. Agate: Physical properties and origin. Field Museum of Natural History geology leaflet 8.

Florke, O. W., H. Graetsch, B. Martin, K. Roller, and R. Wirth. 1991. Nomenclature of micro- and non-crystalline silica minerals, based on structure and microstructure. Neues Jahrbuch fur Mineralogie, Abhandlungen 163:19-42.

Fournier, R. O. 1985. The behavior of silica in hydrothermal solutions. In Geology and geochemistry of epithermal systems, ed. B. R. Berger and P. M. Bethke, 45-61. Vol. 2 of Reviews in economic geology. El Paso, Tex.: Society of Economic Geologists.

Frazier, S., A. Frazier, and D. Mustart. 1993. Thunder showers. Lapidary Journal 47:149-72.

Frondel, C. 1962. The system of mineralogy of,lames Dwight Dana and Edward Salisbury Dana, 7th ed. Vol. 3, Silica minerals. New York: John Wiley & Sons.

–. 1978. Characters of quartz fibers. American Mineralogist 63:17-27.

Gaines, R. V., H. C. W. Skinner, E. E. Foord, B. Mason, and A. Rosenzweig. 1997. Dana’s new mineralogy: The system of mineralogy of James Dwight Dana and Edward Salisbury Dana, 8th ed. New York: John Wiley & Sons.

Godovikov, A. A., O. I. Ripinen, and S. G. Motorin. 1987. Agates. Moscow: “Nedra” Publishing House. [In Russian]

Godovikov, A. A,. S. N. Nenasheva, V. S. Pavlyuchenko, and O. I. Ripinen. 1993. New lutecite finds. Transactions (Doklady) of the USSR Academy of Sciences, Earth Sciences Sections 320:169-74.

Graetsch, H. 1994. Structural characteristics of opaline and microcrystalline silica minerals. In Silica: Physical behavior, geochemistry and materials applications, ed. P. J. Heaney, C. T. Prewitt, and G. V. Gibbs, 209-32. Vol. 29 of Reviews in mineralogy. Washington, D.C.: Mineralogical Society of America.

Harder, H. 1993. Agates–Formation as a multi component colloid chemical precipitation at low temperatures. Neues Jahrbuch fur Mineralogie, Abhandlungen 165:31-48.

Heaney, P. J. 1993. A proposed mechanism for the growth of chalcedony. Contributions to Mineralogy and Petrology 115:66-74.

Heaney, P. J., and A. M. Davis. 1995. Observations and origin of self-organizational textures in agates. Science 269:1562-65.

Heaney, P. J., and J. E. Post. 1992. The widespread distribution of a novel silica polymorph in microcrystalline quartz varieties. Science 255:441-43.

Jacobs, G. K., N. W. Dunbar, M. T. Naney, and R. T. Williams. 1992. Petrologic and geophysical studies of an artificial magma. Eos, Transactions, American Geophysical Union 73:411-12.

Kappele, W. A. 1995. The rockhound’s guide to Colorado. Helena, Mont.: Falcon Press Publishing.

Kay, J. R. 1981. Thunder egg deposits of the Wycarbah district, central Queensland. Queensland Government Mining Journal 82:566-79.

Kile, D. E., P. J. Modreski, and D. L. Kile. 1991. Colorado quartz: Occurrence and discovery. Rocks & Minerals 66:374-406.

Knauth, L. P. 1994. Petrogenesis of chert. In Silica: Physical behavior, geochemistry and materials applications, ed. P. J. Heaney, C. T. Prewitt, and G. V. Gibbs, 233-58. Vol. 29 of Reviews in mineralogy. Washington, D.C.: Mineralogical Society of America.

Landmesser, M. 1984. Das Problem der Achatgenese. Mitteilungen Pollichia 72:5-137.

–. 1988. Structural characteristics of agates and their genetic significance. Neues Jahrbuch fur Mineralogie, Abhandlungen 159:223-35.

–. 1995. Mobility by metastability: Silica transport and accumulation at low temperatures. Chemic der Erde 55:149-76.

–. 1998. “Mobility by metastability” in sedimentary and agate petrology: Applications. Chemic der Erde 58:1-22.

Liesegang, R. E. 1915. Die Achate. Leipzig: Steinkopff.

Lipman, P. W. 1976. Geologic map of the Del Norte area, eastern San Juan Mountains, Colorado. Map I-952, U.S. Geological Survey miscellaneous investigations series.

Lofgren, G. 1974. Spherulitic textures in glassy and crystalline rocks. Journal of Geophysical Research 76:5635-48.

Lund, E. H. 1960. Chalcedony and quartz crystals in silicified corals. American Mineralogist 45:1304-7.

McLemore, V., and N. Dunbar. 2000. Rockhound State Park and Spring Canyon Recreation Area. New Mexico Geology 22:66-71, 86.

Merino, E., and Y. Wang. 2001. Geochemical self-organization in rocks: Occurrences, observations, modeling, testing–With emphasis on agate genesis. In Non-equilibrium processes and dissipative structures in geoscience, ed. H.-J. Krug and J. Kruhl, 13-45. Vol. 11 of Yearbook “self-organization.” Berlin: Duncker & Humblot.

Mitchell, J. R. 1992. Gem trails of Colorado. Baldwin Park, Calif.: Gem Guides Book.

Mitchell, R. S., W. F. Giannini, and O. M. Fordham. 1988. Mineral update: Plume agate from Warren County, Virginia. Virginia Minerals 34:5-6.

Moxon, T. 1996, Agate: Microstructure and possible origin. Doncaster, S. Yorks: Terra Publications.

Pabian, R. K., and A. Zarins. 1994. Banded agates: Origins and inclusions. University of Nebraska, Lincoln, educational circular 12.

Patton, H. B. 1896. Spherulites containing chalcedony and opal in Colorado. The Proceedings of the Colorado Scientific Society 5:165-74,

Pearl, R. M. 1965. Colorado gem trails and mineral guide, 2d ed. Chicago: Swallow Press.

Potter, R. M., and G. R. Rossman. 1979. Mineralogy of manganese dendrites and coatings. American Mineralogist 64:1219-26.

Reed, M. F. 1940. Genesis of the thunder egg. The Mineralogist 8:368.

Renton, J. L. 1936. Opal or agate filled “thunder eggs.” The Mineralogist 4:12-52.

–. 1951. Some notes on thunder eggs. The Mineralogist 19:171-77.

Ross, C. S. 1941. Origin and geometric form of chalcedony-filled spherulites from Oregon. American Mineralogist 26:727-32.

Ross, C. S., and R. L. Smith. 1961. Ash-flow tuffs: Their origin, geologic relations, and identification. U.S. Geological Survey professional paper 366.

Schaub, B. M. 1979a. Genesis of thundereggs, geodes, and agates of igneous origin, part 1. Lapidary Journal 32:2340-54.

–. 1979b. Genesis of thundereggs, geodes, and agates of igneous origin, part 2. Lapidary Journal 32:2548-66.

–. 1989. The origin of agates, thundereggs and other nodular structures. Northampton, Mass.: Agate Publishing.

Sinkankas, J. 1966. What do we really know about the formation of agate and chalcedony? In The agates of North America, ed. H. Leiper, 6-10. San Diego: Lapidary Journal.

Smith, R. K., R. L. Tremallo, and G. E. Lofgren. 2001. Growth of megaspherulites in a rhyolitic vitrophyre. American Mineralogist 86:589-600.

Staples, L. W. 1965. Origin and history of the thunder egg. The Ore Bin 27:5-11.

Taijing, L., and I. Sunagawa. 1994. Texture formation of agate in geode. Mineralogical Journal 17:53-76.

Thrush, P. W. 1968. A dictionary of mining, mineral, and related terms. U.S. Department of the Interior, Bureau of Mines. Washington, D.C.: U.S. Government Printing Office.

Twin Mountains quadrangle, Saguache County, Colorado. 1967, U.S. Geological Survey topographic map, 7.5-minute series. U.S. Department of Interior, Geological Survey.

Voynick, S. M. 1994. Colorado rockhounding: A guide to minerals, gemstones, and fossils. Missoula, Mont.: Mountain Press Publishing.

Wang, Y., and E. Merino. 1990. Self-organizational origin of agates: Banding, fiber twisting, composition, and dynamic crystallization model. Geochimica et Cosmochimica Acta 54:1627-38.

–. 1995. Origin of fibrosity and banding in agates from flood basalts. American Journal of Science 295:49-77.

White, J. F., and J. F. Corwin. 1961. Synthesis and origin of chalcedony. American Mineralogist 46:12-119.

Wright, F. E. 1915. Obsidian from Hrafntinnuhryggur, Iceland: Its lithophysae and surface markings. Geological Society of America Bulletin 26:255-86.

Zarins, A. 1977. Origin and geologic history of siliceous metacolloidal deposits, Cathedral Mountain quadrangle, Brewster County, West Texas. Unpublished M.S. thesis, University of Nebraska.

DANIEL E. KILE

U.S. Geological Survey

3215 Marine Street, Suite

E-127

Boulder, Colorado 80303

[e-mail]

Daniel E. Kile is a research geochemist with the U.S. Geological Survey in Boulder. His most recent article for Rocks & Minerals was coauthored with Peter J. Modreski and Dianne L. Kile and was titled “Colorado Quartz: Occurrence and Discovery”; it appeared in the September/October 1991 issue (pages 374-406).

COPYRIGHT 2002 Heldref Publications

COPYRIGHT 2002 Gale Group
Posted in:Rocks & Minerals