Introduction

Agates are microcrystalline quartz nodules found in veins and in cavities in host rock. They occur as fillings within gas cavities and veins in certain lavas or as a replacement mineral within some sedimentary limestones and claystones. They are however most commonly found within volcanic rocks, in particular andesites, some rhyolites and tholeiilitic basalts. The commonly accepted theory is that they crystallise in the vugs found in flood basalts and usually consist of fibrous chalcedony and coarse quartz crystals surrounding a central void. They are more resistant to erosion than their surrounding altered rock and thus agates often survive long after this host rock has weathered away.

Agates present one of the more spectacular examples of autonomous pattern generation in nature. The present thinking would suggest that biology has had no part in the generation of these patterns but the dissolution of silica by fungi has been reported. In contrast to some environments where the dissolution of silica has been seen to occur in hot, mainly hydrothermal environments similar process of dissolution has been seen at low temperatures. In most conditions this dissolution is extremely slow but in the presence of a number of organic acids such as oxalic, ascorbic and citric acid the dissolution of quartz is accelerated and the rate of this is controlled by the presence of certain microorganisms.

These bacteria represent a major environmental force contributing to the weathering of rocks in nature. There are huge numbers of species in this category, but some of the most renown are the sulphur-oxidizing bacteria, such as Thiobacillus and Sulpholobus. They convert reduced sulphur, which can be a component in rock, into sulphuric acid further accelerating the weathering process. These bacteria are collectively known as Lithotrophs and they have been found well over 3 km below the surface of the planet.
The biogenic dissolution of quartz and its depolymerisation are of fundamental significance in the biological silica cycle. Such organisms, including fungi, which degrade silicates, are an important mediator for those organisms such as diatoms, which later assimilate and concentrate silica.

Diatoms are delicate unicellular organisms that have a yellow-brown chloroplast that enables them to photosynthesize. Their cell walls are made of silica almost like a glass house. The construction of the cell wall, called the frustule, consists of two valves that fit into each other like a little pillbox.
The biogenic silica that the cell wall is composed of is synthesized intracellularly by the polymerization of silicic acid monomers. This material is then extruded to the cell exterior and added to the wall.
What has all this therefore got to do with the formation of agates? The simple answer is, probably nothing but is there a possibility that agates are some form of carbon or silicon based biological concretion of some kind or the fossil remnants of such a biological process? 

Despite the worldwide occurrence of agates, numerous investigations and theories the process of formation of agate is not yet completely understood. No evidence has, so far, been found to unambiguously document agate formation in real time and agates have never been successfully replicated in the laboratory.

It can be safely stated that agates form within cavities of volcanic rock when microcrystalline silica fibres form on vug walls and grow inward. The source of the silica, method of deposition, temperature, pressure and final mechanism of this crystallisation are all unknown.

Silicon dioxide in its crystalline form is called quartz with the chemical formula of SiO2. Chalcedony is microcrystalline quartz made up of twisted crystal fibres. Agates commonly display repetitive textures and trace element compositions. From rim to core agates typically consist of three parts: concentric layers of length-fast fibrous chalcedony, an inner layer of coarse length-slow quartz crystals (chalcedony), and a central void. However, many agates lack a central void and instead are filled in completely with silica, while others consist only of banded fibrous chalcedony. Agates vary widely in shape, size and colour but their structural and compositional characteristics are relatively constant.

Ferruginous rings in white sandstone from the River Nith near Dumfries, Scotland. An example of an autonomous pattern generated in nature? In this case caused by small particles of iron that were trapped within the aeolian Permian sandstone.

Formation

Agates are found throughout the world mainly occupying gas cavities in basalt rocks aged between 3,480 million [Pilbara, Western Australia] and 13 million [Yucca Mt] years old. The formation of lava flows and the formation of agates are not contemporaneous or even connected events. Lavas contain gases held under pressure before being erupted on to the earth surface. At the time of eruption as the pressure is reduced the space this gas occupies increases and gas bubbles form. The gases mainly involved include water vapour, carbon dioxide, sulphur dioxide, chlorine and even hydrogen sulphide. Many of these bubbles burst to the surface and the gas is vented but as the outer lava layer cools some of the gas bubbles are trapped. These bubbles are called vesicles when the lava cools and later when filled with celadonite or agate become amygdales. Amygdaloidal [from the Latin “amygdula”, an almond] lava is so called because the original vesicular lava contained almond-shaped cavities. As well as almond shapes these cavities can be totally misshapen, round, oval, almost flat or most commonly bun shaped. The viscosity of the fluid rock through which the bubble is ascending may determine this shape. Heddle described amygdale shapes as round, rod-shaped, pear- or balloon- shaped, as axe-shaped and even wine bottle shaped.

These subsequently become filled with agate forming materials. Agates can also occur in fissures within the rock called veins, or as long filaments, similar to but more numerous than veins, called stringers. Agates can also form within sedimentary rocks as nodules that are the result of the replacement of a former mineral or some organic material such as coral.

Agates do not appear to form in the final cooling phase of volcanic rock. It is only after complete cooling and burial of the flows to depths of 200-100 metres that they form. 

Although the actual formation of agate is complex, not fully understood and the subject of numerous, occasionally conflicting, theories it can be said that in general they form from a amorphous deposit of silica rich material which fills the cavity. It is thought that silica-rich solutions enter the cavity from outside and form gels inside or that solutions form outside the cavity and then move into the cavity. This solution then deposits silica from the outer part of the cavity to the inner part over an indeterminate period of time at a range of temperatures from 40 to 250°C. Whatever the source of the silica it seems that the banding within the vesicle is caused by chemical and physical reactions due to the ambient temperature and pressure within the cavity as well as the amount of bound water held within the silica solution.

Microstructure of Agates

The feature that most agates have in common at least is the region of banded chalcedony. This banding occurs as two distinct and perhaps not connected parts. There are the so-called growth ring bands of chalcedony and those of the coloured bands that are the result of the chemical deposition of mainly iron oxides. To the naked eye the width of these second bands may vary randomly in the mm range. However by examining thin sections of agates with the polarising microscope visual banding is resolved into a concentric succession of zones on a micron to sub-micron scale.

An interesting characteristic that is almost universal within agates is a distinct band of “first generation” chalcedony. This band is in immediate wall contact and may vary from 1-2 mm thick. It usually has a distinctly different morphology to the inner layer.

Small section of the edge of an agate from Ardownie Quarry Monifieth, Tayside [area approx 8 x 8mm]

Amygdale in lava with green Celadonite coating from Ardownie Quarry, near Monifieth, Tayside [Amygdale 60 x 50mm]

The banded region itself consists of layers made of fine crystalline and untwisted quartz fibres that alternate with layers made of even finer and twisted fibres. This twisting can be seen under the polarising microscope by a change of birefringence along any one fibre. The fibres can range in thickness up to about 0.5 microns and can be up to a few centimetres in length. The fibrous chalcedony is intergrown with variable amounts of another form of silica called moganite. In agate, moganite cannot be observed by the optical microscope but can easily be detected by powder X-ray diffraction. The presence of moganite, which rarely occurs as a pure mineral has been confirmed by powder X-ray diffraction within agate samples. The moganite:quartz ratio is often not uniform but shows a cyclical pattern that correlates with the observed cathodoluminescence (colour and intensity) pattern.

Agate in thin section

Much information can be obtained when rocks and minerals are ground to a thin section (standard thickness is 0.03mm) on a microscope slide. The rock slide is viewed between two polaroids using a microscope. Interference colours are generally created and these are diagnostic of the mineral under study.

Two agates are shown in thin section below. The colours shown by agate in thin section are black and white or straw yellow if the section is slightly thick. These images are not micrographs and should be strictly called montage photographs. They were taken with a close-up lens using a light box for illumination. Separate photographs were taken with some overlap and software was used to produce the completed image. This technique allows the creation of the image on the full microscope slide (7.5 x 5 cm).

Burn Anne agate

All agates have a contrasting micro texture over the initial 1 or 2 mm outer region [a)]. This agate then shows a fibrous sheaf-like growth at b). The chalcedony fibres take over and sweep towards the centre in the direction of the arrows.

Macrocrystalline quartz (qz) grows at the centre of the agate. Note that as the centre is approached, the quartz crystals become larger. This is a typical observation from crystals growing in conditions of changing temperature: small crystals develop from a high temperature and larger crystals when the temperature is lower.

[A – The agate is photographed in plane polarized light
B – The same agate with polars in the extinction position
q z – macrocrystalline quartz
a) – The initial growth; the arrow shows direction of growth
b) – Fibrous sheaf-like growth
Scale bar = 10mm]

[Click on the image to enlarge]

Ethiebeaton Quarry agate

This agate shows a very similar growth pattern to the Burn Anne sample. However, this agate shows a different micro texture [a] with crystal growth shown by the direction of the arrows. A final deposit of macrocrystalline quartz (qz) is at the centre of the agate.

[A – The agate is photographed in plane polarized light
B – The same agate with polars in the extinction position
q z – macrocrystalline quartz
a) – The initial growth; the arrow shows direction of growth
b) – Fibrous sheaf-like growth
Scale bar = 10mm]

[Click on the image to enlarge]

These two agate are from different hosts occurring about 412 million years ago. Although the agate hosts are over 100 miles apart, these thin sections show that agate development is similar. Indeed, the ancient agates from the Pilbara, Western Australia (host age 3480 million years) show the same basic growth pattern.

Agate structure can therefore probably be interpreted as alternating formation of fine-grained, highly defective chalcedony inter-grown with moganite, and coarse-grained low defective quartz. It could therefore be hypothesised that cyclical variation in the moganite content must be a general feature that is connected to the mechanism of agate genesis.

Colour in Agates

The coarse visual colour banding seen in agates is an independent feature imposed on the compositional zonation by relatively long-term variations in the deposition of pigmenting impurities.  Although agates are composed almost entirely of SiO2 it is the trace quantities of various other elements that give agates their colours and lead to their characteristic banding. Most agates are red and blue, although in reality the “red” will vary from pale pink through orange to pillar-box red, whilst the “blue” will vary from grey-blue to cornflower blue through to almost black. Other rarer colours include yellow and green, or white bands standing out from the background hues. Dark browns, even blacks and combinations of all of the above can produce a whole range of strong to pastel shades, each agate being either subtly or completely different from its neighbour. Almost all of these great varieties of colour are due to oxidised iron.

Trace element composition of agates varies widely from location to location but some trends are common to almost all agates of igneous origin. Trace element data are similar for agates from both acidic and basic volcanic hosts. Germanium and Boron are the only elements beside Silicon, which is enriched in almost all agates compared to the Clarke values of the lithosphere (1.4ppm for Germanium, 12ppm for Boron). Unusually high concentrations of uranium are also sometimes detected in agates. Agates contain impurities less than p.p.m. [parts per million] level for most of the elements except Sodium, Potassium and Iron. Nonetheless, even red chalcedony bands often have relatively low concentrations of substitutional Iron, indicating that the colour is caused by fine dispersed iron oxides not incorporated into the structure of the fibres. Iron oxides therefore occur in all the colours, which are met in agates including even the rarer green and purple tints.

It is interesting to note that although the agates found on the island of Mull on the west coast of Scotland tend to be white/grey in colour they fluoresce green/yellow in ultraviolet light. This is most likely due to trace amounts of the mineral calcite trapped within the crystalline matrix.

Large crystal of Calcite in agate from the Island of Mull. White/grey banding fluoresces yellow with UV light[18 x 18mm]

Shades of colour may show a preference for certain areas within the agate or even preferentially around some bands, colour by itself has nothing to do with the growth of an agate. It is as the result of iron-rich solutes entering the gel either initially or subsequently after the agate has crystallised. The one exception to this is the formation of some pure whitish-blue bands that are of an amorphous opal material. These attractive intensely white bands are particularly common in vein agate from Burn Anne in Scotland.

Fortification agate from Turnberry beach showing red Haemachatae [Area shown 25 x 25mm]

Problems associated with Agate Formation

Given the structural and compositional similarities between agates from all parts of the world it has been hypothesised that agate formation occurs independently of any outside influences and that the banding as well as the colouring is largely chemically controlled. In other words the variation in the texture and composition must result from the closed internal dynamics of the growth itself rather than any variable conditions outside the agate. It seems that the crystallisation behaviour that controls this compositionally simple mineral system is very complex. It is so complex that despite numerous investigations, theories and arguments about the formation of agates the following factors have not been fully explained:

• Silica source
• Method of deposition
• Temperature of formation

Silica Source

One of the biggest uncertainties in agate genesis is the source of the silica, which the agates are formed from. The lavas, which the agates are formed in, are generally very poor in free silicon dioxide. It has been proposed that the silica source could be from the surroundings, hydrothermal activity, late magma deposition or a silica glass within the magma. However there appears to be two favoured theories with ample evidence supporting each.

The first theory suggests that silica released from volcanic ash devitrifies and eventually releases silica in the form of a watery gel that then permeates through into underlying lavas via meteoric (atmospheric or rain) water. After the initial lava flows are laid down subsequent eruptions covered the basalts with a silica-rich rock called a welded ash-flow tuff. Alkaline or saline lakes later formed on the tuffs and freed silica from the volcanic ash. This silica rich gel then moved downwards in the rock and enters the cavities.

The second theory believe that agates form from a silica lump or gel within the magma that contains trace elements and water.

Although most believe that agate genesis is contemporaneous with the formation of the host rock, some have argued that agate formation occurs up to tens of millions of years after the formation of the host rock.

One other theory suggests that the silica source for the gate genesis would be mobilised from the surrounding wall rocks by hydrothermal activity. It has also been argued that if the cavity contained the silica gel at the beginning of agate genesis then as a result of contraction and loss of volume this amorphous silica deposit would require an estimated 20% extra in order to maintain a full amygdale of Chalcedony. Therefore the agate amygdales would be reliant on a late input of percolating silica rich solution and that formation would not be a “closed system”. Also the similarity in the rare earth elements between agates and the parent lavas suggest that these elements are mobilised during syn- and post-volcanic alteration of the host rock.

It has also been suggested that there is evidence suggesting that crystallisation from fluids may occur with differing degrees of polymerisation and that the observed alternating crystallisation of quartz and chalcedony may be caused by variations in the degree of silica saturation of the silica bearing fluids. Despite the evidence in support of this theory there is ample evidence to the contrary and no clear single theory has be arrived that that takes into account all the data.

Crystallisation of agates from an initial hydrous silica gel or glass lump has also been put forward. These polymerised silica “lumps” would contain trace elements and water and there is some evidence from high silica concentrations found in present-day hot springs that might support this theory.

One of the main problems with the hydrothermal fluid theory is that the initial crystalline deposit of chalcedony on the inside of the amygdale would be expected to block the ingress of the silica rich fluids. This would not be a problem with a  “closed” lump of polymerised silica within the magma as all the required silica would already be present.

Another variation on this theory has suggested that agates in volcanic lavas  are xenoliths of marine chert because of the similarity in the 18O values between the two. These chert lumps would not be melted and reabsorbed by the lava but would be carried within it as melt drops and later transformed into agates.

As can be seen there is no final answer as to where the silica comes from as no one theory fully explains the process.

Deposition  

Agates have not been successfully reproduced under laboratory conditions to date. Many workers believe that agate formation is contemporaneous with the formation of the host rock. However knowing this fact does not help answer the complex question of deposition. Most of the problems relating to deposition relate to the formation of the small scale micro-crystalline quartzine-chalcedony bands and not the colour banding that can be seen with the naked eye.

It is thought that the zoning seen in the micro-crystalline system is the result of a cyclical interplay between growth rate and diffusion rates at the crystal/solution interface. The high defect (Brazil-twin) density found and the impurities within the agates points to a rapid growth of silica from a strongly supersaturated solution probably with a non-crystalline precursor. This type of depositional process would be self-organisational and not dependent on external factors.

Other workers have attempted to explain the deposition using a chemically controlled method that explains both the twisted nature of the quartz fibres and how that process causes the visible banding in agates. This model assumes that agates form from amorphous lumps of silica within the magma. This model predicts that the fibre size changes periodically as seen in agates.

There are a number of other theories that explain deposition using percolating crystallization from fluids in hydrothermal systems.

Each theory appears to explain agate deposition adequately but so far, unfortunately, none of them have resulted in agate-like patterns in the laboratory. Maybe the time frame of formation that is poorly understood may hold the information needed to construct a truly workable model of agate deposition? 

Temperature of Formation

The temperature of formation of agates in igneous environment is not known. It is a key point in understanding formation and studies undertaken have concluded that temperatures between <50ºC to >400ºC. This wide range suggests one of two things: either agates forms under a wide range of temperatures or the conclusions of such studies are inaccurate.

A direct estimate of the formation temperature of agates is difficult and the plethora of results on the topic does nothing to clarify the picture. Nevertheless the majority view suggests that genesis starts with polymerisation of silica rich fluids at temperatures about 100ºC.  

Discussion

I have attempted here to draw together all the current thinking on the formation of agates and present it in as simplified way as I can but without loosing too much of the detail. It can be safely said that there does not appear to be a single accepted detailed explanation of how agates form. As mentioned above nobody has yet managed to start with some basic ingredients and create a whole agate nodule, exhibiting all the familiar agate characteristics, at the end of the experiment. As well as that all the literature would inform us that neither has any evidence been found to date to suggest that agates are forming in real-time in basalts anywhere on the earth. If evidence of this has not been found to date then I would feel that we couldn’t say for sure that they are not forming somewhere. If we had a better idea of exactly how they do form then perhaps we would know where to look.

Assuming no agates are forming on the earth today I guess that the nearest thing that we can observe and study would be where hot silica rich fluids are being brought to the surface and polymerising into silica. This appears to be happening at places like Yellowstone National Park in America and in parts of New Zealand. This would be akin to what took place in Scotland in the Devonian period at Rhynie in Aberdeenshire. Despite the uncertainty of what the pressure and temperatures involved in agate formation at Rhynie we can say that the silicification occurred at atmospheric pressure and temperatures around 100 degrees centigrade.

Having reviewed all the latest theories on agate formation available in the scientific literature I am left with more questions than answers. Maybe agates form under a variety of differing conditions and therefore there is no one single unifying theory of formation? This certainly appears to be the case with the Zeolite group of minerals which also form within cavities in volcanic rock.

 Zeolites form at low temperatures by alteration of volcanic ash and larger pyroclastic material on the land surface, in freshwater lakes, shallow marine seas, saline, alkaline lakes and in deep sea sediments. Very rarely do they form at low temperatures in vesicular volcanic rocks. Zeolites form at high temperature from localised hydrothermal water in hot springs on the continents and “black smoker vents” under the oceans and in broad regions heated by hydrothermal solutions or burial metamorphism. At high temperatures, Zeolites form from the cooling of volcanic flows, late phases in pegmatite’s and miarolitic cavities in plutons; and as phenocrysts in basaltic magma. The final formation process may take place in a variety of the above conditions so that the final Zeolite may have a complex formational history in a variety of geothermal environments. Could the creation of agates also have a complex multi-formational origin?

Occasionally within the basaltic lavas one can find a beautiful solid agate and then within a few inches another cavity filled with calcite, clay like material or even a geode. Could this observation in the field be explained by this multi-formational origin within different and changing geothermal environments?
 
The replacement of once living material held within sedimentary rocks and initially proceeding towards a fossilising process eventually create a fossil formed of beautiful agate with the internal structures perfectly preserved in agate of differing colours depending on the structures present. This is seen all around the world and includes agatised pine cones [Araucaria Mirabilis] from Patagonia, perfectly preserved wood structure from Arizona and even here in Scotland with the Carboniferous coral Lithostrotion preserved in perfect detail from the lower reaches of the Firth of Clyde. How does agate replace dead organic tissue if the theory of formation involves the cavity initially filling with a silica-rich solution? It must require solutions to enter the potential space occupied by this material as it is decaying.

External geothermal processes must be involved that vary with time, perhaps over a very long but variable length of time. This would be further supported by such simple observations as the initial formation of brown Calcite crystals in a lot of the agates from Mull here in Scotland. The ambient external conditions change and then the Calcite is eventually surrounded by agate. 

These are just a few of the questions thrown up by a close look at the theories of agate formation.

Maybe we should all take note of what Prof Heddle said at the end of the 19th century “These materials represent the decomposition-products of the eruptive rocks, which have been carried by means of water into these closed cavities, from the surrounding rock, by endosmose.” and accept this as the mechanism and then simple enjoy these beautiful objects created by nature for what they are?