When we undertook the excavation in Brixton Deverill in 2018 we discovered and carefully mapped and sectioned key Roman and later features, including a deep trench which contained waterlogged horizons/contexts at its base and what our professional advisers believed to be a gravel platform for some kind of processing. Samples – including snails, pollen, wood, charcoal etc – of each context were taken for further work on specific environmental indicators.
We are amazingly lucky to have on our doorstep in Codford Dr Mike Allen, who owns and runs one of only two privately owned environmental archaeology laboratories in the country. We are also hugely fortunate that he is so keen to support groups like ours and with his abounding energy and enthusiasm help us develop our environmental archaeology skills. So, after some to-ing and fro-ing matching volunteers with Mike’s availability, we settled on the last week in August as our processing period, which as a best guess would be 5 days for 3 volunteers. We were blessed with unusually good weather for the processing, which was lucky since most of the work takes place outside! Sat in the sun around one of Mike’s outdoor seating/BBQ spots, we began with a morning’s introduction to the tasks ahead. He introduced us to the principles of sample preparation by flotation, which has a number of stages: soaking the samples; large tank primary flotation (outside); laboratory based second flotation; separation/fractionation of residue by wet sieving; drying and sorting of residue; examination and assessment of the flots collected by flotation. All the while accurate labelling of EVERYTHING is essential!
Prior to work beginning, several of the samples had been subdivided into 10-litre sub-samples and, in readiness for flotation, pre-soaked in large plastic wine/beer fermentation tubs with a quantity of hydrogen peroxide to break up aggregates.
The first stage of processing takes place using a large stainless steel flotation tank into which the sample ‘soup’ is poured. Water is drawn by a submersible pump from a lower reservoir and forced through a series of tubes within the flotation tank, which have holes in them thus creating a constant flow of agitated water. As the tank fills, the water flows over the lip of a weir and into a fine mesh (300 micron or 0.3mm) sieve. Anything which floats over the weir (eg charcoal, wood, shell fragments, plant fibres, grains and husks) is therefore collected as flots in the sieve and the water drops into the reservoir tank and is re-circulated. The mesh size allows suspended sediment (clays and silts) to pass through. To help the process the operator gently agitates the sediment ‘soup’ by hand, hopefully releasing any flots that may be held down by heavier muck clinging to them. Samples took anywhere from 1 to 3 hours to process (fine in great weather but can’t be much fun in cold and rain!). We quickly learned that opening the water flow valve too much leads to a torrent flooding the sieve and the loss of important flots! Once we (Mike) was satisfied that no more flots were likely, we carefully removed the mesh basket liner and decanted the residue into a bucket for the processing second stage. Meanwhile the flots were set to dry on their labelled sieve. Ready for the next sample!
Mike explained the importance of second stage flotation as many flots are missed in Stage 1. Stage 2 involved working in the lab at a large purpose-built sink. The residual sample (minus its flots from Stage 1 and some of the very fine sediments that went down the waste gate) was now placed in a 5 litre bucket of water. After swirling the sample around with a large spoon or by hand, the liquid was carefully and slowly decanted through a 500 micron sieve; the bucket was then refilled and the procedure repeated until no further flots appeared. The sample was now ready for wet-fractioning through a small nest of sieves, while the flots retained on the sieve were dried for examination and assessment, along with the flots from the primary flotation. Fractioning the remainder of the sample proved tricky, particularly where there was a large proportion of peaty plant remains that clogged up the finest mesh sieve at the bottom of the stack. These samples required heaps of patience as it was very tempting to tip too much sample into the top of the stack, only to find that the bottom sieve was overwhelmed! Eventually there were two sets of flots, material caught on each sieve and a lot of water and silt/clay that had gone down the drain!
Each sieve fraction was now laid out in the sun on a drying tray on newspaper to thoroughly dry. Once dry (anywhere between an hour and two days), the fractions would be examined for any artefacts (which are bagged and labelled individually), and then weighed and the weights recorded on the master chart. The coarsest fractions are discarded and the rest bagged, labelled and stored. Simples, and that’s the story of one 10 litre sample: only another 170 litres to go!
Once the flots have dried they will be assessed by Mike and he will decide whether they have sufficient merit to pass them on to specialists in grain and husk, wood/charcoal etc identification. He will of course do any work on snails himself!
It wasn’t all work! We had some great conversations over coffee and lunch with Mike and his wife about all manner of things from families to gardening and of course archaeology.
This was a very interesting week and we were exceptionally well looked after and trained by Mike, to whom we are very grateful.
Blog by David Croot