Before provenance studies were undertaken it was necessary to visit the circle site and make detailed notes on the nature of the stones. Existing literature was utilised to locate potential sites of provenance.

The study at the Stanton Drew site centred firstly on the southwest circle where there is a selection of the different types of lithology.

Five main lithologies were found.

The stones have been documented many times over the last three hundred years. They were first mapped and documented by John Aubrey in 1664. Aubrey included the first crude sketch of part of the complex.

It is uncertain which author first suggested the Harptree area as the source of the stones. East and West Harptree are two small villages approximately five miles south of the Stanton Drew site. Dymond (1896) cites a number of authors who, previous to his work, suggested the area as a possible source for the cherty magnesian limestone.

Dymond not only counts and labels the stones but also offers names and origins for most of the different lithologies found at the site. He identifies the limestone as the Inferior Oolite of Dundry, a site a few miles north west of Stanton Drew. Grinsell (1918) also describes some of the stones as being local conglomerate, and also identifies oolite from Dundry.

Morgan (1887) and Phelps (1836) both mention Oak[e]y Hole as a possible source for the extremely cherty stone found in the southwest circle. Oaky Hole is not found on any map of the area and its exactly location is not given in any literature. However it is possible that I have found this location using a description from Green and Welch (1965). This stone is an oddity and will be discussed in more detail. Green and Welch pinpoint East Harptree as a possible area for the more cherty conglomerate.

Green and Welch (1965) and Woodward (1876) were used to locate the position of various quarries and pits, outcrops and cutting in the area that would expose the bedrock to view. Old issue, six inch Ordnance Survey maps were used to look for extra sites of this nature that may not have been included in the memoirs and newer O.S. maps which may show more recent workings on them.

Field Work

The most interesting and potentially recognisable lithology proved to be the cherty stone. The origin of this stone remained unclear at this time. Its extreme orange colour and lack of clasts means that it easily spotted when it was seen in the walls of many houses in East Harptree. The fact the stone has been used in masonry could indicate that it is local in origin (see fig 5.)

The first task was to visit the localities of interest deemed so by the O.S. and geological maps and the memoirs. If the areas proved interesting then the surrounding area was also scoured for outcrops. The nature of these areas meant fieldwork has been difficult. There are few outcrops and foliage is omnipresent.

Appendix 2 is a map of the area showing the areas worked and the locations at which samples were collected.

A study of each individual stone at Stanton Drew and the Cove was carried out, noting any distinctive features and in particular, missing or recently revealed stones.

Geochemistry

The composition of sedimentary rock preserves a record of the provenance and allows comparisons to be made (McLennan, 1989). Rare Earth Elements (REE) have very similar geochemical properties including large atomic radii, which means they will only enter minerals with large atomic sites. REEs are not easily fractionated during sedimentation and will not be effected to any great extent during a silicification episode (McLennan, 1989). REE contents of sediments and sedimentary rocks reflect the mineral content and the process by which the deposits were formed. REE abundances in shales are taken to represent the REE contents of the upper continental crust. The upper crust generally contains REEs in the same abundances and they fractionate relative to each other (Henderson, 1984). The absolute abundances of these elements decrease with respect to shales in his order: shales, greywackes, sandstones and limestones. This probably reflects the decrease in clay mineral content (Haskin et al, 1966).

REEs tend to remain unaffected by any of these changes. Precambrian sediments which have also undergone metamorphism have REE abundances effectively the same as present day sediments (Wildeman and Condie, 1973). The aim of this analysis is to assess whether the silicified conglomerate at the Stanton Drew site is the same as the samples taken from the East Harptree area. The objective is to see if a match exists between the REE abundances. REE changes due to silicification will not be relevant as both samples have undergone these alterations.

Concentrations of the REE are normalised by dividing their abundances by a reference standard, in this case 44Ru and 75Re. The normalised abundances of each REE are plotted on a logarithmic scale against their ionic radius and the results form curves with distinctive forms and slopes relating to the relative enrichment of the light or heavy REE (McBirney, 1984). Light Rare Earth Elements (LREE) include 57La to 63Eu and Heavy REE 64Gd to &1Lu. HREEs and LREEs will differentiate differently due to their masses. Trends should be seen in the quantitative analysis results.

A semi-quantitative analysis will be undertaken to give a rough idea (+/- 20%accuracy) of what elements are contained in the samples. It is necessary to compare the samples taken from the field on their hand specimen qualities, to a set of control samples. These control pieces are representative of the unsilicified dolomitic conglomerate, a rhyolite and a piece of the megalith itself

It was not deemed worthwhile or economical to take thin sections. With the advanced state of silicification of the conglomerate the sections would have been quite featureless. It would also have been difficult to take representative sections of the rock owing to the large size of many of the component clasts.

The control sample from the Stanton Drew circle site has been in the University of Bristol Earth Sciences department's rock collection for many years. With permission a small sample has been taken from it to be the comparison sample for the rocks from the field. The main difficulty is when sampling conglomerates there will be a bias due to the distribution of clasts.

Crushing samples

The first task was to reduce the samples to fine powders, accomplished with a jaw clamp, then coarse grain and fine grain crushers and gyro-mill. The samples were dried at 110°C for 24 hours.

Acid Digestion

This process involves the dissolution of the silicates in the rocks.

200mg of each sample were taken (within 0.003mg) and placed in a P.T.F.E. beaker. 10 ml of 40% v/v hydrofluoric acid was added with 2.5ml of concentrated nitric acid. This was heated to incipient dryness at 100°C. 20ml of 1%v/v nitric acid was added and reduced for 5 minutes at 230°C. The sample was cooled and diluted to 100ml with 1% nitric acid. A reagent blank is made to contain the elements used in acid digestion but none of the sample elements. The solution to be used is 555-1 dilution the original rock and is used in ICP-MS analysis.

Hydrofluoric acid is used to breakdown silicates. Concentrated nitric acid is for intense oxidation and the lower concentration is for complete oxidation.

Semi-Quantitative Analysis

5ml of the 500-1 solution are mixed with 5ml of 49In which is the standard solution. An induced coupled plasma mass spectrometer was used to gain semi-quantitative analysis of the samples (see appendix 4). Method of acquisition: 1 minute wash, 1 minute to allow sample to reach plasma. Acquisition: 400 short sweeps.

Quantitative Analysis

5ml of the 500-1 solution is mixed with 5ml of the 200ppb Re-Ru standard solution. The ICP-MS is calibrated using different concentrations of REE solution and a reagent blank. (See appendices 5 and 6). Method of acquisition: 2 minute was, 1 minute to allow sample to reach plasma. Acquisition: 3 x 100 long sweeps.