1. Structure: partially accommodated channels (0.16 to 0.80mm in width) above the pan; no aggregates; fewer channels below the pan; channel pattern - straight to curved; vughs (spherical to elongate, not normally connected) 0.4 to 0.8mm in diameter and are not orientated and have random pattern: above pan channels and vughs partially connected; pellicular grain microstructure - weakly developed.

2. Mineral components: coarse fraction: quartz grains range in size from silt to fine/medium sand (<lmm). One large quartz (200mm). Dominance of single and compound quartz grains; instances of biotite and plagioclase; quartz grains -angular/subangular and smooth. Fine fraction: brown, optically amorphous iron oxides, isotropic.

Samples were taken from the ditch and various pits within the enclosure and processed foi carbonised remains. All samples were wet sieved using mesh sizes of 1mm and 300ji and the material from the coarse flot (1mm) was sorted. Of the thirty six samples taken, twelve produced small amounts of carbonised plant material. The results from these seven ditch and five pit samples form the basis of this report.

A list of the species identified in each context is given below in Tables 6 and 7. Evidence for cereal grain is scant. Emmer (Triticum dicoccum) and barley (Hordeum sp.) are present. Both are known from the Bronze Age in Britain and in Scotland emmer appears to remain the commonest wheat into the Roman period and beyond (Jones 1981, 106). The evidence for emmer comes from a single grain and fragments of chaff from the ditch (context 1002A) near the entrance of the enclosure. A fragment of hulled barley was also found in the SW section of the ditch (context 1003). This fragment appeared to be twisted which suggests that the six-row variety is present. One other fragment of barley (possibly hulled) is present in context 1103A in Pit 10 (a large pit to the S of the enclosure).

Purpose of the examination

The identification of distinctive suites of pollen types within the cist floor at Loanleven and several other Bronze Age ‘cemeteries’ in Scotland (see Loanleven report below) led to the need to investigate the possibility of similar patterns within the Balneaves complex.

Samples from the basal and most securely stratified fills of five of the pits at Balneaves were examined (F1110A; Pit 3: F1109C and E from Pit 6: F1107A; Pit7: F1114B; Pit 13: F1115A; Pit 14) together with the blackened encrustations from within Urns 1 and 3 (see ‘The Pottery’; Alison Sheridan) were prepared according to conventional chemical techniques (cf. Moore and Webb 1978), including the use of 10|im nylon sieves and hydrofluoric acid to remove siliceous particles, stained with safranin, and examined at mag. x400 on an Olympus BHS microscope.

Bodysherd of pottery from Pit 15

This sherd of pottery is of a thin red brown fabric with small quartz inclusions. Its internal surface has a thin white slip which gives it the impression of being very well made. It does not fît into any medieval fabric type known by the writer. On balance it seems likely that this sherd is from a vessel of eighteenth or nineteenth century date suggesting a fairly modern date for the feature concerned.

Introduction

The bone fragments from Loanleven had several areas of green staining on their surfaces. They were analysed elementally to determine whether copper was present in the stained areas which might suggest that they were due to corrosion of a copper object.

Method

The bone fragments were analysed quantitatively without sampling or surface preparation using energy dispersive X-ray fluorescence (XRF). A Rhodium target X-ray tube run at an accelerating voltage of 46KV was used to produce the primary X-ray beam which was collimated to give an elliptical irradiated area on the surface of the specimen about 1 mm x 1.5 mm.

Four samples were collected from a vertical sequence with No 1 at the base straddling contexts 0089 and 0066, No 2 straddling contexts 0066 and 0021, No 3 in the centre of context 0021, and uppemiost No 4 between contexts 0021 and 0065.

The four slides were described following the international system for thin section description (Bullock et al., 1985). Slides No 1 and 2 were virtually identical whilst slides 3 and 4 were variations on these. It thus seemed best to provide a full description of slide No 1 which is also applicable to No 2 and then only to describe the differences between these two pairs of slides which are of a particular interpretational value.

In order to test the potential of pollen analysis to resolve certain site environmental questions, a total of 5 pollen samples (numbered 1 to 5) were examined. Numbers 1 to 4 refer to series of Kubiena tins which formed a column up through the sub-caim OGS. Number 5 was also from the OGS but from a different profile.

The site country rock is greywacke and shale of Llandovery age (Silurian). There may be dykes of porphyrite in the near vicinity. This stone, with altered country rock, is worked at Tongland Quarry, 3/4 mile due south of the site.

It was not possible to visit the site during the excavation; the identifications are of samples taken from features on site by the excavator. Samples were supplied from:-

a. outcrop to west of the site and recorded by Coles as one of an outer arc of low standing stones

b. outcrop to north of the site and recorded by Coles (1895) as one of an outer arc of low standing stones

c. two samples of the cairn material

d. one sample from the cairn kerb material

All three pots contain quartz (often milky), feldspar and small lumps of grey granite (Biotite Granite) which looks very similar to the Gatehouse-of-Fleet granite and Creetown Granite. This suggests that the source of this material lies to the west of Kirkcudbright rather than to the east. Hence the grits from all three pots, and therefore the manufacture, are likely to have been quite local to the site.