Water at Avebury

by Edward Pegler on 10 February, 2013

Was the water supply of Avebury as patchy in the Neolithic and Chalcolithic as it is now?

It’s been a wet winter at Avebury. My recent attempts to visit Windmill Mill and West Kennet Long Barrow have resulted in wading through rivers where normally there’s no water at all. And it got me thinking maybe I could write about this. Some of the following is based on my understanding of geology and maps, some is based on an English Heritage Report.

Avebury and chalk

Avebury, and the surrounding Neolithic monuments, are all built over chalk bedrock. Chalk is a highly permeable rock due to being full of tiny holes and, more importantly, bigger cracks. It allows water to flow through it very easily.

The permeablility of chalk has some interesting effects on the landscape. This is because rain, when falling on unsaturated chalk (i.e. not already full of water), tends to disappear into the ground without ever flowing across the surface. Valleys in the chalk often contain no streams (why the valleys are there is another story). Instead, the water flows relatively slowly underground.

The upper surface of this underground water ‘reservoir’ is known as the water table. Due to changes in rainfall, temperature and vegetation, the level of the water table varies annually, going down in summer and up in winter, generally reaching its highest around February.

Occasionally, such as in wet winters like this one, the chalk will become so full of water that the water table will reach the surface to emerge as puddles or flowing streams. In larger valleys these streams can become frequent enough to be called rivers.

An example winterbourne, north of Aldbourne

An example ‘winterbourne’, north of Aldbourne

The Winterbourne & Kennet rivers

One such established, intermittent river is called the Winterbourne. This river is so named because its course is usually completely dry, but the river reappears in wet winters. Normally the maximum depth of the Winterbourne is about half a metre. This year the rain has made the Winterbourne almost 2 metres deep in places where it is confined. Where unconfined, it is much shallower but can be tens of metres wide. The evidence of the geology in the area (e.g. extensive alluvial soils) suggests that the Winterbourne may have been a more substantial stream in the past (perhaps in the early Holocene) than it is now.

The Winterbourne flows from north to south. At its southern end it is met by another ‘winterbourne’ just to the west of Avebury. A little further south it probably was once met by an even smaller and more intermittent stream coming in from Beckhampton to the west. From this point downstream the river is called the Kennet. Here, the course of the river turns east, ultimately joining the Thames.

Even the more substantial Kennet, which is not seasonal, is normally less than half a metre deep near Avebury. Sometimes, it too can reduce to a trickle or even dry up (and this doesn’t appear to be much affected by modern water pumping).

This means that, due to their unreliability, sources of water have historically been a problem for those living around Avebury. In summer people have either had to rely on the trickle of the Kennet, or on springs such as Swallowhead, or on wells which have been dug down into the chalk to reach the water table several metres below. This probably explains why settlements have always been limited on the chalk downland.

Avebury, the Sanctuary, the Avenues and water

Avebury sits on a section of dry ground just to the East of the Winterbourne. In normal years the water table wouldn’t even reach the ditch around the ancient stone circle, although it may well have filled the bottom of the ditch as it was when first excavated in the Neolithic.

The Sanctuary sits on a hill approximately 25 metres above the Kennet. It is unlikely ever to have had its own water supply and, as such, makes a poor location for a house. However, it’s still no more than 300 metres to the north of the Kennet, so you can always go down there with bucket.

Avebury’s eastern Avenue of stones ran along a valley from Avebury either down to the river or up to the Sanctuary. It’s southwestern end appears to run alongside a small, intermittent river course, although this hasn’t flowed this year and I’d guess didn’t during the Neolithic. For comparison, the now lost western Avenue from Avebury to Beckhampton takes a course across the Winterbourne and doesn’t appear to follow any intermittent stream. I therefore suspect that the Avenues have little to do with river courses.

Silbury Hill and water

Silbury Hill is a bit weird. This is not just because it’s a large earthen mound, seemingly without any point to it. It’s also that it sits on the edge of the Kennet River’s floodplain, just below the point where the Winterbourne becomes the Kennet. The ditch around Silbury Hill, dug out to make Silbury Hill itself, is usually dry in summer. In winter, more often than not, it is completely filled with water due to the higher water table, and this water can become linked to the Kennet.

Silbury Hill is not the only large Neolithic mound to have been constructed in a river valley close to the water table. Marlborough Mound, in Marlborough College, is of much the same age and is similarly positioned, right next to the Kennet but further downriver. To the south, Marden Mound (now sadly flattened), at the head of the Hampshire Avon, was also located in a similar position, right next to the river.

This has led to one interesting theory, posited by one Lothar Respondek, that Silbury Hill was the result of a primitive form of well digging, the primary aim of the Hill actually being the ditch itself. Certainly, the theory has some merit, in that Silbury Hill is strangely positioned to be easily seen, down at the bottom of a valley. Contrary to what Mr Respondek himself thinks, most experts think that Silbury was constructed at a time when the water table was higher than it is now. However, it doesn’t entirely invalidate the idea if the ditch was supposed to give a summer water supply.

Another feature, pointed out by Paul Whitehead and Mike Edmunds, is that Silbury Hill is sited just at the boundary between the Lower And Middle Chalk formations. The significance of this is difficult to fathom but may relate to the different flow patterns of water in the Middle Chalk, which would have been easy, and the Lower Chalk, which is broken up by clay layers, possibly resulting in the emergence of springs at this location.

Either way, the connection between Silbury and water does seem significant.

The West Kennet Enclosures and water

Two large, oak palisaded enclosures (currently known as Enclosure 1 and 2) also sat right by the Kennet River, just to the east of Silbury Hill and a little downstream, during the late Neolithic. These were built around the same time (approximately) as Silbury Hill. Excavations of these structures by Alistair Whittle and his co-workers in the late eighties and early nineties showed that Enclosure 1 even straddled the course of the present day Kennet, although the Enclosure 2 lay on the south bank of the modern river.

Oddly, the major evidence for occupational debris in these apparently short lived enclosures was the abundance of piglet bones in the ditches. More recently, similar piglet debris at Durrington Walls (near Stonehenge to the south) has been strongly argued to be the result of winter feasting.

There has been some suggestion, based on the excavations at Durrington Walls, that such enclosures were occupied only seasonally, probably in winter. Certainly, there would be a more stable water supply at this time of year. However, siting the enclosure on the river might not be the best idea then. This year, the land just to the south of the river, once part of the enclosures, is completely flooded. Even though this is an exceptional year, it certainly isn’t the only time in recent years that this section of the river has flooded. If the experts are right then this problem is likely to have been worse in the late Neolithic. This makes it odd that the people who built these enclosures didn’t site them somewhere drier, but still by the river, for example on the north bank. Maybe we’re missing something here.

Whatever, the problem of water supply should perhaps be taken quite seriously in the study of why the monuments of Avebury are located where they are.


Whitehead, P. and Edmunds, M. 2012 Silbury Hill, Wiltshire: Palaeohydrology of the Kennet, Swallowhead Springs and the siting of Silbury Hill, English Heritage Research Report 12-2012, pp35.

An excellent report on the hydrology of present and past Avebury.

Respondek, L. 2005 The Mystery of Silbury Hill: Why Was it Built? Elar, pp72.

Not a daft book, but sadly without references, so about as valid as my posts.






{ 2 comments… read them below or add one }

Steve Marshall August 22, 2013 at 4:49 pm

I enjoyed this. Did you see my article on the Silbury springs in the last issue of British Archaeology? If not I can send you it.


Edward Pegler September 2, 2013 at 9:49 pm

Dear Steve

I didn’t see the article, no. I’m a bit out of touch at the moment, though I’d love to see it. Thanks. I’ll send you an email if that’s ok.



Leave a Comment

Previous post:

Next post: