The microscopists of last century were therefore able, with the patience and thoroughness of their ilk, to discover many phenomena which we may not have thought possible. We have a good example of this in the work of the famous Scottish scientist, Sir David Brewster, who was an authority on the phenomena associated with light. In the 1840's, Brewster became interested in precious opal, particularly the reason for the development of the colour.

Brewster examined precious opal at high magnifications in transmitted light and discerned a regular pattern on a very fine scale. He realised that these represented a very regular pattern of fine pores in three dimensions, and as these were at the limits of possible resolution of the optical system, he deduced that they were the cause of the colour phenomenon; the colours were caused by diffraction of light from this structure. Details of this phenomenon are given in the section Electron microscopy: Microstructures in common opals, which deals with the causes of the colour in precious opal.

Because of the limits placed on optical resolution, however, he was not able to see the actual structure which caused the pores. In 1845 he published a short paper on the subject in the Edinburgh Philosophical Journal.

The final elucidation of the structure, vindicating Brewster's observations, had to wait until the early 1960's.

In general, precious opal shows little structure in the optical microscope, apart from the type of structure found by Brewster, and this is difficult to detect. However, as described earlier, some precious opal, especially if viewed in sections thicker than standard, shows what appears to be a more or less strong (anomalous) birefringence when viewed between crossed polarisers.

Most precious opals and associated potch are virtually amorphous to X-ray diffraction, as is the glass clear hyalite. The latter material differs from the former in its lower water content (about 3% as opposed to 5%-8%), and is formed under different conditions. Under the microscope it usually shows an entire absence of structure, appearing like a glass.

A special case of amorphous opal is that formed in plants, or as the skeletons of minute organisms such as diatoms and radiolaria (biogenic silica). These are large enough to be examined in some detail in the optical microscope, and some of these organisms were used in earlier times as a measure of the resolution capability of a microscope. There are many forms of these tiny animals, and their remains form very large deposits in many parts of the world, both on the floors of deep oceans, and as fossil deposits on land. These latter deposits are of considerable economic importance. Diatomite for example, is a material of low density with both good thermal insulating and refractory properties as well as being a good absorbent.

Plant opal, such as the phytoliths introduced above, has been studied intensively; it is a common constituent of soil, remaining therein after the decomposition of the plant. In some plants, such as the equisitales, or horse tails, in some grasses, in cereal straw, and in some trees, especially the Myrtacae, it may occur in substantial amounts.

Evidence of this is sometimes seen in the form of 'haystack glass', a slaggy material formed by fusion of the silica and other inorganic constituents when haystacks are burnt. Under the microscope, the latter material is seen to be composed mostly of glass, with inclusions of quartz and other impurities from dust in the hay.

The phytoliths can frequently be isolated from the plant material itself, or from soils. They commonly form microscopic rods and needles, or small particles with serrated edges. The rods and needles may be of the order of five micrometres (0.005 mm) in width and 0.1 mm in length. They are generally colourless and transparent.

G. Baker illustrates these phytoliths well in his work on wheat and sugar cane material (R0386). He and his associates even found opal phytoliths in the rumen of sheep, and as the major constituents of uroliths obtained from a ram.

One of the earliest mineralogists to examine common opals using the optical microscope was the Frenchman E. Mallard, in a paper entitled Sur la Lussatite, nouvelle varieté minérale cristallisée de silice, published in 1890 (R1589). He found that his material had a fine, fibrous structure which exhibited a distinct birefringence when viewed between crossed polarisers.

He also found that the mineral had refractive indices and specific gravity consistent with that of opal, and it also contained 8.3% water. He realised that the material was a form of opal despite its microstructure and birefringence, which clearly indicated that this mineral was crystalline. As it appeared to be a distinct mineral, he gave it the name 'lussatite'.

Mallard's work has, in recent years, been validated. X-ray diffraction has proved that most common opal has a crystalline structure, highly disordered, which is now called opal-CT. Much common opal shows the fine, fibrous structure observed by Mallard; whether it can or cannot be seen depends mainly on the crystallite size.

Many samples have crystallites too small to be resolved optically, so that the material appears amorphous and isotropic. It may be noted in passing that some years ago an attempt was made to validate the name 'lussatite' for this mineral, but this was not accepted by the International Mineralogical Association. However the term is seen from time to time in modern literature.

Other microstructures seen in common opal at the optical level tend to be related to impurities and other phenomena in the particular sample. Iron-rich opals, for example, may show the presence of the inhomogeneous distribution of iron oxides through the sample.

A wide variety of fine structure inherent in the opal itself is seen at the higher resolution available by electron microscopy.

The most striking discovery with regard to the microstructure of precious opal was made more or less simultaneously in Germany by Pense (R1502), and in Australia (Jones, Sanders and the author, R0362; Sanders, R0360) in 1963.

Using the replica technique, fracture surfaces of the opal were examined in an electron microscope, when a remarkable pattern of regularly stacked minute spheres was revealed. This structure, at long last, proved the accuracy of Brewster's observations with the optical microscope in the first half of the 19th century.

Pense's photographs showed the regular structure inherent in the precious opal, although it is more clearly shown in early photographs taken in the TEM by Sanders and SEM photographs by Anderson, Jones and Segnit (R0250). The structures are usually more clearly seen if the fracture surface of the opal is lightly etched with hydrofluoric acid.

Detailed analyses of such photographs showed a variety of features analogous with those of atomic structures. Regular and irregular packing, dislocation, stacking faults and other phenomena were described in detail by Sanders (R0119) and later by Monroe and his colleagues (R1356).

The structure is shown equally well in precious opals both from sedimentary environments and from volcanic rocks. Remarkably good packing is shown in a glass clear opal from the cavity of an altered volcanic rock from Maleny, a locality some 100 km north of Brisbane in Queensland.

Small cavities in the trachyte were filled with colourless, transparent opal, which showed one continuous red flash across the sample (about 4.0 mm diameter). When etched with hydrofluoric acid, sections of the surface either cracked randomly or split along 'structural' planes, and curled into curious shapes.

The close packed sphere structure is confined to silica in the mineral world. It can however, be reproduced by other materials which can be formed into regular spheres of the appropriate size. This is the basis for the manufacture of synthetic or, strictly, artificial, opals, a subject dealt with in section, The Fakes.

Synthetic resins can also be prepared to produce the same phenomenon, and is the basis for the artificial plastic 'opals' produced in Japan. This is a scanning electron micrograph of a synthetic resin (a special sample of known regular particle size used formerly by electron microscopists as a magnification standard) which had dried out over a long period of time in its container. Mr J. Farrant, doyen of Australian electron microscopists, noticed a red 'flash' in the material and, suspecting the reason, gave it to the author for examination, with the fortuitous result illustrated.

In natural opal, the spheres, which have a diameter ranging from about 150 nm to 400 nm, have a detailed internal structure. This can be seen in many of the electron micrographs of etched surfaces in Sanders' papers, and is well shown in a paper by Jones and the author (R0266).

This is caused in most cases by the concentric deposition of very small more or less spherical particles (10-20 nm in diameter) around a central nucleus, thus building the larger spherical particles. The manner in which this happens in nature and in the laboratory is dealt with in section Formation of opal.

There are small amounts of 'impurities' in most opals. These may occur either as discrete grains of a specific phase, or may be material adsorbed onto the surfaces of the silica. Sanders (R0130) noted minute particles or crystallites in precious opal or potch when highly magnified in the transmission electron microscope. These particles often gave a distinct electron diffraction pattern, indicating their crystallinity. Some were claimed to be crystallites of either tridymite or cristobalite. Specific crystal shapes were not usually seen.

Hyalite, the glass clear form of opal, does not show the sphere structure. Its fracture surfaces appear very much like the fracture surface of glass. No texture is exhibited at the resolution obtained in the SEM, except for the curved bands of conchoidal structure formed on breaking the sample.

One curious example of amorphous opal was encountered in Recent sediments on the banks of the River Murray, west of Wentworth in the far south west of New South Wales (Australia). This was a white, typically opaline material, which had been broken and re-cemented (brecciated). SEM examination of the fracture surface of this material revealed masses of small, cavities or tubes of somewhat irregular shape, but packed together. The origin of this structure is obscure; it may be due to rapid precipitation of a gel around solution drops, as may be seen, for example, forming around of sodium hydroxide solution when dropped into a solution of aluminium sulphate.

An unusual structure is described by Sinkankis (R0335) in a hyalite, termed an iris-opal, from Mexico:

Cut stones show flashes of colour:

Electron micrographs showed that the latter opal was formed of layers about 2 µm (0.002 mm) thick, the layers themselves being composed of particles about 25 nm (0.00025 mm) in diameter, with even smaller particles at the junctions between the layers. It was therefore clear that the colour was developed in a different manner to normal precious opal, where the diffracting particles are several hundred nanometers in diameter.

Biogenic silicas are probably the most abundant form of amorphous silica (opal-A) to be found in nature. The greatest deposits are formed in the oceans, and many are recorded in detail in recent years in numerous papers published under the heading of 'Deep Sea Drilling Project'.

Minute marine organisms such as diatoms and radiolaria are the main source of this silica. They construct beautiful skeletons of opal-A which are easily resolved in the optical microscope. Small diatoms were a challenge to the 19th century microscopist. However the advent of electron microscopy has made the reproduction of images of such small creatures simple.

Because of their very high surface area, the skeletons are readily subject to diagenetic processes, and are frequently found to be recrystallised to opal-CT and possibly other forms of silica. The minute rosettes, or lepispheres, are one of the resulting forms.

Large deposits of these organisms are also preserved on land; the most common are composed of diatoms, often little altered. Their microstructure is that of a mass of the skeletons of these organisms.

When alteration has taken place, either by diagenesis when on the sea or lake floor, or later alteration when raised above sea level, the skeletons of the organisms may be largely destroyed, leaving a mass of very fine silica. This is sometimes opal-A, sometimes opal-CT, showing few, if any, skeletal remains. On the other hand these remains can sometimes be remarkably persistent.

Larger opaline organic particles may be well preserved; notable amongst these are sponge spicules. These are needle-like skeletal remains from sponges, and appear to be able to last for very long periods without alteration. A typical example, from an opal claystone, is shown in (pj48, pj49).

These silicas are sometimes mixed with clay minerals; they are variously known as diatomite, diatomaceous earth, kieselguhr, tripoli, gaize, opoka or opal claystone.

Perhaps the earliest attempt to explain the colour play of precious opal was that by Sir David Brewster in 1845. Brewster was famous for his knowledge of the properties of light. While he did not give a complete explanation of the colour, he came remarkably close to the truth, although most later attempts by other authors neglected his findings. Brewster's note is reproduced here: footnote

All of the essential phenomena are here recorded, a remarkable achievement for those times. Brewster was, however, wildly astray as regards the formation of the opal.

Probably because this publication was short and in a journal seldom used by mineralogists and gemologists, it was largely overlooked. A similar conclusion regarding the internal structure appears to have been reached by the French mineralogist, A. Des Cloizeaux, who is quoted by Haill (R1491) as recording that:

Later attempts to explain the colour were usually based upon layers of different refractive index, or minute cracks in the opal, causing interference effects of the type one sees with an oil film on water.

However some time later an apparently authoritative paper by an Austrian scientist, Behrends, described a different structure. Until the 1960's this was regarded as the correct basis for the colour of opal. In Dana's Textbook of Mineralogy (1932 edition), it is stated:

Such an explanation involving layers of different refractive index, or thin parallel cracks, has frequently been quoted in later literature. Raman and Jayaraman (R971) who carried out X-ray diffraction work on opal, put forward a variation on this theme by suggesting that the colour might be caused by interstratification of layers of high and low cristobalite.

However, the discovery of the true microstructure of opal, almost simultaneously in Australia and Germany, showed that Brewster was close to the truth. The Australian workers immediately saw the significance of the close packed spheres of silica in relation to the colours of precious opal (R1633, R0360, R0362) footnote.

The spheres of which precious opal is made are of uniform size and are close packed. The spheres, or, more strictly, the cavities between the spheres, form a three dimensional diffraction grating which has the property of 'reflecting', or, more truly, diffracting one specific wavelength of light if certain conditions are fulfilled.

The conditions are those imposed by Bragg's Law, formulated originally for the diffraction of X-rays by crystal (atomic) structures. In crystalline structures, the atoms are arranged in an orderly manner in parallel planes which are uniformly spaced apart.

The spacings between parallel sets of planes of atoms are of the order of magnitude of the wavelength of X-rays; when an X-ray beam of appropriate wavelength enters the structure at an appropriate angle, an X-ray beam of specific wavelength is diffracted. From such a reflected beam can be calculated the distance between the atomic planes causing the phenomenon.

Bragg's law defines the conditions under which this can take place:

where l = the wavelength of the diffracted beam, d is the spacing between the atomic planes, and q is the angle of incidence of the X-ray beam. The same formula applies to the effects found in precious opal, visible light being a part of the electromagnetic spectrum to which, of course, X-rays belong.

In the case of opal, l = wave length of the diffracted light (and hence the colour), d = the distance between the diffracting planes as defined by the distance between the parallel lines of voids (and, hence, the diameter of the silica spheres), and q the angle of incidence of the light beam.

A beam of white light falls upon the opal surface, and is refracted into the structure. Part of the refracted beam is then affected by the planes of voids according to Bragg's Law, and a beam is diffracted at an angle corresponding to that of the incident beam.

However, the diffracted beam, in contrast to the incident beam, is monochromatic; that is, it is of one colour. This colour will change if the angle of incidence changes; in other words, if the opal is rotated, the colour seen may change. The colour will also depend on the size of the spheres, and hence the spacing between the planes of voids.

If the spheres are about 350 nm (0.00035 mm) in diameter, which is about half the wavelength of red light, a complete spectrum of colours should be visible as the opal is rotated. This, however, is limited by the refractive index of the opal, which as illustrated, limits the angles at which the diffracted colour can be seen.

If the opal is immersed in water, a greater range of the spectrum can be observed on rotating the opal. If the diameter of the spheres is small, only colours of the shorter wavelengths can be seen. When the spheres are only about 200 nm in diameter, only blue colours will be observed.

The patches of colour are due to discrete sections of the opal with uniform spheres in a given orientation, although in any particular piece of the gemstone, the spheres will be of more or less uniform size. If, for example, red colours are seen, it is probable that the whole stone will be composed of spheres of about 350 nm diameter.

In addition to the diffracted beams, the greater part of the incident white light will pass through the stone, or be randomly scattered by imperfections in the sphere structure, by minute inclusions, or by microscopic cracks or other discontinuities.

In the latter cases, the opal may take on a milky appearance, with the diffracted colours showing against a white background. With black opals, the agent, probably finely divided carbon of organic origin, causing the 'blackness' absorbs the scattered white light, allowing the diffracted colours to appear against the dark background. This effect is artificially produced in some opal (so-called 'opal matrix') by depositing carbon in the pores.

As this type of opaline silica, usually that called opal-CT, occurs under a wide variety of conditions in nature, it can be expected that a wide variety of microstructures may be encountered. The fine, fibrous structures (lussatite) sometimes seen in the optical microscope are more clearly defined in the SEM, where masses of fibres can be seen clearly.

Most common opals have a very fine grained structure. Little can be seen in the optical microscope, but more detail is revealed in the SEM. One sample of opal, for example, appears to have been derived from the bulk replacement of wood, but without the preservation of the woody structure. The interior of the trunk was translucent, near-transparent opal, while the outer parts were opaque and white to pink in colour.

Immersion of the latter material in water rendered it, too, transparent; in small flakes, the water could be seen penetrating the opal. The SEM pictures show the reason for this; the milky opal was criss-crossed by a network of fine cracks which caused the opacity of this part of the sample. The cracks were able to take up water rapidly by capillary action, thus rendering the milky material transparent.

A notable structure which has only been seen in recent years is the presence of minute crystals of opal-CT. While these can just be seen at times in the optical microscope, the resolution is not sufficient to define their character.

One of the earliest depictions (R0192) of these crystals, platy in shape, was from samples of common opal, again from Recent sediments west of Wentworth in south western New South Wales. Small cavities in the opal were coated with these tiny platelets.

Subsequently rosettes of such crystals were found to be an important constituent of deep sea siliceous sediments which had undergone diagenesis. Such rosettes are well illustrated, for example by von Rad and Rösch (R0153) and by Weaver and Wise (R1368). Further enlargement of the surfaces of the rosettes shows platy crystallites similar to those seen in the cavities of common opal. These rosettes have been termed 'lepispheres'.

In some cases these minerals have been called cristobalite, but X-ray diffraction patterns have not been published. There is some confusion in the literature regarding the interpretation of opal patterns. For example, Peterson and von der Borch (R1603) describe "chert precipitating as gelatinous opal-cristobalite in lakes associated with the Coorong Lagoon of South Australia". However, the XRD pattern published in this paper is clearly that of opal-A, (amorphous), not opal-CT ('opal-cristobalite').

Many variations in morphology in common opal are to be found.

Many of them look as if they have been derived from deposition as gels. Others show different microstructures as magnification is increased. A picture series show a smooth texture on the outside surface, but with increasing magnification, minute spheres can be seen making