Trench L was a 2.3 metre deep lift-shaft of an excavation dug in just under two weeks by Alex, Cate, Josh, Carol and Tony. Dolines, natural vertical shafts in the limestone, are very common on and around New Laund Farm. They are likely to contain vital archaeological and environmental evidence to help us reconstruct the prehistoric landscape. It is easy to see why. A doline acts as a natural trap for anything in the locality inert enough to be washed in (pollen for example), dozy enough to fall in (sheep or people, as it might be) or exciting enough to be deliberately thrown in (probably sheep or people too, who knows in later prehistory…). Unfortunately, because of their ability to swallow everything in their vicinity, they fill up quickly and any prehistoric layers are likely to be very deep. For example, the Early Bronze Age burials and artefacts at Charterhouse Warren Farm Swallett in Somerset were about 17 metres from the modern surface.
So, while we are keen to sample as many dolines as possible, we want to make sure that we know just what we will be getting ourselves into. If any of them are as deep as Charterhouse Warren Farm then it is going to be a serious engineering job even to get near the prehistoric archaeology.
Trench L was designed to explore a relatively small doline. I hoped this would give us some idea of what the sediment sequence was like, and so how far down the archaeology was likely to be. Just as importantly, I hoped it would give us a feel for how effective our excavation techniques were to deal with these deep holes.
This photo shows the most important features of the sequence in trench L. Reading from the top down you can see the modern topsoil slumping into the depression at the top of the doline. Beneath this is a layer of limestone rubble. This is likely to be quite recent too, as a late 18th or early 19th century iron hammer head was found under this rubble. Then there is a thick layer of brown silt and at the base of that a thin charcoal-rich layer. This also follows the curving shape of a depression in the top of the doline and we think that all of these layers may be related to some digging in the top of the doline for walling stone in the last few hundred years.
However, beneath that charcoal rich layer is a thick deposit of brown silty clay which does not seem to have been disturbed. This goes down as far as we were able to dig in the two weeks we spent on this site. There was no bone, animal or human, in this layer, nor any worked stone. Hopefully, there is preserved fossil pollen. Cate, Carol and Josh took a series of four 0.5 metre long soil samples through the best preserved part of this deposit. This should allow a reconstruction of the changing environment around the doline and give us a hint about how deep the sequence has to go before we are down to the Early Bronze Age and Late Neolithic.
We have also learnt a lot about the practicalities of digging these sites. Here we had two sides of the trench with a lot of potentially loose rubble. We shored these early and kept adding new acrow props as we went deeper, trusting that this would keep the whole trench stable. This worked fine in the dry but when we got some heavy overnight downpours towards the end of the dig we found that we were getting many minor landslips in the spaces between our shoring boards. Tony and Carol described an alternative solution that they have used on caving digs. They use a rectangular scaffolding frame, which descends into the hole as you dig, with many more, closely spaced, boards behind the frame as shoring. The disadvantage of this solution for us is that we like to be able to see the sequence in the sides of the trench as we dig.
That said, commercial archaeologists working in advance of construction often excavate in deep, narrow, closely shored foundations and they manage to make sense of what they find without looking in the sections. Maybe we just need to raise our game a little bit. My preferred option would be to find a doline that we could access with a digger, or at least with the digger bucket on a tractor, and make a bigger hole in the upper layers. We could then ‘step’ in the sides of the excavation as we went down. So, an 5 x 5 metre trench would become a 3 x 3 and then a 1 x 1, a bit like an inverted ziggurat, before any shoring had to be attempted at all. Yet another option would be to use a powered auger (basically a giant drill) to extract a soil sample over the full depth of a doline without digging it at all. This should work well for the pollen but, unless we were very lucky, wouldn’t help much with understanding the archaeology.