Ore petrography is the description and study of opaque mineral associations and the textural relationships found within them. These days mineralogists and geologists use ore petrography and especially its textural aspects for a number of purposes: to establish paragenetic sequences (albeit a wonderfully subjective use); to determine equilibrium or non-equilibrium conditions; or simply as an aid in mineral identification, for example some characteristic intergrowths between marcasite and pyrite betray the former presence of pyrrhotite. But there is a further use, one that is so familiar that it has been neglected. The assemblage (minerals and textures) is a record of a set of unique events and, with luck - if the assemblage has lived through interesting times-, this recording has sufficient detail and clarity to become a petrographical fingerprint. In practice, for many assemblages, there is little variety and the fingerprint is feint but for fine- to medium-grained basic to intermediate igneous rocks, and especially those that have suffered alteration ± metamorphism, their petrography is recognisably individual enough to have potential as a provenancing tool.

Provenancing, the matching of artifacts to their geological source areas, has been a major occupation of archaeologists since the inception of the discipline. In 'scientific provenancing' (as opposed to stylistic provenancing) they have followed geologists in terms of methods and approach, so that for lithics and ceramics transmitted light petrography was followed by major and minor element geochemistry, followed by increasingly sophisticated (and expensive) geochemistry as the most favoured method of sourcing rocks. Metalwork is different, here trace element geochemistry, on both metals and ores has been used almost exclusively; for example the number of provenancing studies based on stable lead isotope signatures of metal and ores is greater than those using metallography or ore petrography. As each new method gains popularity there is a gradual tendency to abandon earlier methods (or indeed a belief that provenancing studies have any great value).

However, the interest and attendance at the Mineralogical Society-sponsored Geology and Geochemistry in Archaeology Conference held at the Open University (April, 1996) demonstrated that a major area of interplay between geologists and archaeologists remains the provenancing of lithics (from axe-heads to columns), ceramics, and metals. Of these, lithics and ceramics are still provenanced routinely using transmitted light petrography, lithogeochemistry, or sometimes by a combination of the two. Normally, however, although the lithology may be specified the source area has been quite broad and only a regional provenance has been given, for example Sardinia, Southwest England.

But, as became clear at the conference, there is now a move away from regional towards fine-scale provenancing - the matching of specific artifacts (normally monumental) to individual outcrops within a given lithology or even to within specific areas in an outcrop. This fine scale literally adds a new dimension, namely time, to provenance studies, it now becomes possible to determine the sequence in which a resource (outcrop/quarry) was exploited. Examples of this type of provenancing include the use of stable carbon and oxygen to try to match building stones from the Palace of Knossos on Crete to nearby limestone quarries (Durkin, 1996) or the matching of Roman columns made of granodiorite to specific areas within single quarries at Mons Claudianus in eastern Egypt using magnetic susceptibility (Williams-Thorpe et al. 1996); and the use of lithogeochemistry (Williams-Thorpe and Thorpe, 1990) and ore petrography moderated by lithogeochemistry (Ixer, 1994) to provenance millstones from a Greek shipwreck to individual lava flows on Pantelleria. It is noticeable that in Europe these and other examples are restricted to the Mediterranean Basin and confined to classical (literate) civilisations. In archaeometallurgy it is now being realized that ore deposits are heterogeneous and that ore from different areas of a mine may have different trace element and/or isotopic signatures that can be preserved and passed into the metal (Ixer, 1999). Preliminary work on the ores from the Early Bronze Age copper mine at Ross Island, Killarney, Ireland and Early Bronze Age copper and bronze artifacts within the British Isles suggests that it is possible to match an ore source (even from within specific parts of a mine) to artifacts using standard and stable isotope geochemistry moderated by detailed ore petrography (Ixer and Budd, 1998).

Although in Britain there is a long tradition of provenancing small lithic objects, most notably axe-heads (Clough and Cummins, 1979; 1988), there have been few attempts to do the same for prehistoric monumental objects and even fewer attempts at fine-scale provenancing. However, amongst the best known are the efforts to match the Stonehenge bluestone monoliths to their outcrops using standard transmitted light petrography (Thomas, 1923), lithogeochemistry (Thorpe et al., 1991) and reflected light petrography (Ixer, 1996).

It was H.H. Thomas who in 1923 first proposed that the bluestones were identical to rocks cropping out within the Preseli Hills of Dyfed, southwest Wales and that they had been transported from there to Salisbury Plain by prehistoric peoples. These bluestones, which have been erected as smaller monoliths within the sarsen circle at Stonehenge, comprise 32 spotted or unspotted dolerites (distinctive enough to be called preselite), five altered volcanic ashes, five rhyolite-ignimbrites (tuffs) rhyolites, two micaceous sandstones and a green sandstone, the 'altar stone' (Thorpe et al., 1991). Although Thomas's suggestion of anthropogenic selection and transport of the bluestones found and still finds favour with the majority of archaeologists the hetereogeneity of the 45 bluestones has to be explained. This problem was addressed by the Open University archaeo-geology team, led by Richard Thorpe and Olwen Williams-Thorpe and Graham Jenkins, who in the late 1980s, were allowed by English Heritage to sample 11 dolerites and four rhyolite/tuffs from in situ bluestone monoliths at Stonehenge for geochemical and petrographical analyses. This material, together with museum samples of dolerites and rhyolites collected from past excavations at Stonehenge (but not directly from the bluestones), were compared with newly collected outcrop material from the Preseli Hills and other areas in South Wales, plus the well characterized suites of rocks collected and described by Richard Bevins (Bevins, 1982, 1989). The resulting long, cohesive and comprehensive paper used lithogeochemistry, together with a little petrography, to confirm Thomas's regional scale provenancing, namely that the Preseli Hills were the source of most of the bluestones. They were able to name three sources in east Preseli for the dolerites and four sources in the north Preseli Hills and north Pembrokeshire for the rhyolites, but recognised that others came from much further afield including the green sandstone, the 'altar stone', which is probably from the Palaeozoic of southwest Wales. Although they were able to suggest that, for a few of the monoliths it was possible to find a lithogeochemical match between an individual bluestone and a named outcrop, for the majority the match was with a number of specified outcrops within a broad area of a couple of square kilometres (Figs. 1 and 2) . The number of outcrops needed to encompass all of the rock types within the bluestones, and the distances between these outcrops, led Thorpe et al. to suggest that the bluestones "were derived from a glacially mixed deposit rather than careful in situ collection", so that it was nature not mankind who transported the exotic bluestones to Salisbury Plain. Few archaeologists like this suggestion and the paucity of glacial material on Salisbury Plain is a major difficulty with the proposal. Even if all of the bluestones (and two thirds are yet to be sampled) were matched to their outcrops this still would not resolve the problem of their selection and transport. However, if it were possible to provenance to within an outcrop and show that significant numbers of the bluestones come from within metres or tens of metres of each other, this might suggest purposeful on-site selection (quarrying) and that the transport agency, for those stones at least, becomes obvious.

Initial reflected light petrography on the monolith and outcrop specimens showed that, although the opaque mineral assemblage was not unusual, there was sufficient textural variation to suggest that ore petrography alone, or better, together with lithogeochemistry, might further refine the sources of the monoliths. Since ultimately the aim is to match rocks with identical histories, in this case, from their initial crystallization from the melt, subsequent cooling histories, prograde and retrograde metamorphism, hydrothermal alteration and final weathering, there is a need to establish a ranking of significance between individual components of their petrography. In an earlier study using unmetamorphosed, unaltered basalts (Ixer 1994), where more artifact material was available, a detailed ranking could be established, namely presence/absence of primary igneous phases > primary textures between primary phases > primary textures within a single primary phase > alteration textures > presence/absence of secondary minerals > weathering features. A similar ranking was established for the bluestones and Preseli outcrop material although it was found that the petrographic effects of the low grade greenschist metamorphism were uniform and so of little use as a discriminatory tool.

As figure 2 shows opaque petrography is as sensitive as lithogeochemistry in subdividing samples within a single lithology; in this case altered ophitic dolerite, and just as effective as a provenancing tool. Provenancing by both methods showed such a very high degree of correspondence that combining the results of the two methods to further refine the provenance seems justifiable (column 3, figure 2). Of course neither of these two results is unexpected but the lack of petrographical and geochemical variation between seven of the monoliths (SH33-SH61) is striking, so much so that it is tempting to believe that they all come from within metres or tens of metres of each other. The number of monoliths that belong to this tight lithological-geochemical group is not known; at present the group represents 21% of all the dolerite monoliths and 16% of all the bluestones. Even if these percentages are a maximum (and two thirds of the dolerites and 70% of all the monoliths are yet to be sampled) they represent a significant number of the monoliths.

However, even if many of the dolerite bluestones come from a single outcrop and perhaps even from a small area within that outcrop, suggesting that they were intentionally removed, there remains the degree of lithological randomness inherent in the presence of the other bluestones including the rhyolites/tuffs and sandstones. Could it be that the monoliths are a mixture of quarried stone augmented by randomly chosen rocks there to make up the numbers, so making Stonehenge amongst England's earliest jerry-built public buildings. That is a question for archaeologists not geologists.

What should be said, however, is that detailed (Victorian-style) petrography, allied to standard lithogeochemistry, is as good and effective a method of detailed provenancing lithics as more novel approaches. For the right rocks it really is a case of back to basics.

Bevins, R.E. 1982. Petrology and geochemistry of the Fishguard Volcanic Complex, Wales. Geological Journal 17, 1-21.

Bevins, R.E., Lees, G.J. and Roach, R.A. 1989. Ordovician intrusions of the Strumble Head-Mynydd Preseli region Wales: lateral extension of the Fishguard Volcanic Complex. Journal of the Geological Society of London 146, 113-123.

Clough, T.H. McK and Cummins, W.A. (eds) 1979. Stone Axe Studies. London, Council for British Archaeology. Research Report No. 2.

Clough, T.H. McK and Cummins, W.A. (eds), 1988. Stone Axe Studies (Volume 2) London, Council for British Archaeology, Research Report No. 23.

Durkin, M.K., 1996. Stable isotopes as an aid to the provenance of building stones in the Bronze Age Palace of Minos, Knossos (Crete) (abstract). Geology and Geochemistry in Archaeology. Mineralogical Society Spring Meeting, Open University.

Ixer, R., 1994. Does ore petrography have a practical role in the finger-printing of rocks? In Stories in Stone. Ashton, N. and David, A. (eds). Lithics Studies Society Occasional Papers 4, 10-23.

Ixer, R.A., 1996. Detailed provenancing of the Stonehenge Dolerites using reflected light petrography : A return to the light. In: Archaeological Sciences 1995 (ed A. Sinclair). Liverpool Meeting. Oxbow Archaeological Monograph Series 11-17.

Ixer, R.A. 1999. The Role of Ore Geology and Ores in the Archaeological Provenancing of Metals. In: 'Metals in Antiquity'. S.M.Young, M.Pollard, P.Budd and R.A.Ixer (eds). British Archaeological Reports. 792, 43-52.

Ixer, R.A. and Budd, P. 1998. The mineralogy of Bronze Age copper ores from the British Isles: implications for the composition of early metalwork. Oxford Journal of Archaeology, 17, 15-41.

Thomas, H.H. 1923. The source of the stones of Stonehenge. Antiquaries Journal 3, 239-260.

Thorpe, R.S., Williams-Thorpe, O., Jenkins, D.G., Watson, J.S. with Ixer, R.A. and Thomas, R.G., 1991. The geological sources and transport of the Bluestones of Stonehenge, Wiltshire, U.K. Proceedings of the Prehistoric Society 57, 103-157.

Williams-Thorpe, O., Jones, M.C., Tindle, A.G. and Thorpe, R.S., 1996. Magnetic susceptibility variations at Mons Claudianus and in Roman columns : a method of provenancing to within a single quarry. Archaeometry 38, 15-41.

Williams-Thorpe, O. and Thorpe, R.S., 1990. Millstone provenancing used in tracing the route of a fourth century BC Greek merchant ship. Archaeometry 32, 115-137.