The AVON which passes through our fair city of Bath, is hardly one of the great rivers of the world, barely deserves inclusion in the top twenty in Britain alone as regards length and amount of water it carries to the sea. Surely, therefore, this little river within the unique region where William Smith deduced the first principles of modern geology must have been studied until there is nothing else to learn about it, and there must be volumes of erudite literature on all our library shelves.

However, this is not so, as several members of our Society discovered last Spring during a series of lectures under the auspices of the Extra Mural Department of the University of Bristol, led by Dr Brian Hawkins, BSc, PhD. The Avon is, in fact, a very complex river, with several features for which explanations are at best controversial. The course was therefore lively and very enjoyable, although it was far too short to cover the subject in the depth of our probing, and arrangements had to be made to continue with the same subject in the coming year. Indeed, having learned at its source, wherever that might be, we arrived only at the junction of the Avon and Bybrook before the final lecture, which was followed by a bright but cold field trip to see what we had been studying.

The Avon rises somewhere north of the M4 Motorway in the general vicinity of the Badminton estate, there being between five and ten places that could qualify as its source depending on definition. The source of a river might be any of the following:

1. The point most remote from its estuary from which surface water drains into it,
2. The point most remote from its estuary where a small brook can be seen to flow in wet weather, in a recognisable channel,
3. The place where a trickle of water can always be found, except perhaps in a protracted drought.
4. The most remote spring that flows in wet seasons, ignoring any minor surface drainage towards that point.
5. The most remote spring that flows continuously, other than in exceptional circumstances of drought.
6. The largest spring contributing water to the river.
7. The head of the stream that carries most water at any junction of two streams, even if it is not as long as the other.

With so many definitions it is not surprising that there is controversy as regards the Avon. There is no doubt that the stream carrying most water where two streams join just above Sherston, is the one from Crow Down Springs about a mile above the junction, quoted in many publications as the official source of the Avon. However, above these springs there is a well defined channel extending for a good quarter of a mile to another group of springs that flow only in wet weather, and small surface drainage channels can be seen running still further into adjacent farmland.

The other branch at the junction, although carrying less water, extends for a much greater distance, to within a few hundred yards of the scarp of the Cotswolds not far from Little Sodbury. This branch once received a lot of water now being taken by the water authorities from Hancock’s Well, not far above the junction, also cited occasionally as the source of the Avon, but the brook is well defined beyond this point for at least three further miles. It begins in a marshy patch of land just to the east of the A46, midway between the Old Sodbury crossroads and Petty France, from which it appears to flow through a culvert under the airfield and on through Badminton before it reaches Hancock’s Well, carrying by then quite a lot of water draining from the surrounding countryside. This marshy patch would be my choice as the source of the Avon, although even here there might be dispute, as several ditches that might once have been natural channels drain into it, any one of which might reasonably be considered to be part of the river.

From the head of one of these ditches it is only a very short walk to the edge of the steep scarp of the Cotswolds overlooking, in the distance, Bristol and the Severn Valley. Here another set of streams can be seen, part of another drainage system unrelated to the Avon of our studies. These streams are energetic and fast-flowing down the steep slopes, in comparison with the slow progress of the upper reaches of our river over the shallow dip slopes it crosses. It is easy to imagine them etching and eroding their courses, digging further and further at the edge of the scarp until they capture part of the Avon, although it will take a very long time before this becomes a problem to local landowners.

The geographical dip slope traversed by the Avon as it flows eastwards for its first few miles, is slightly shallower than the geological dip of the strata below, so that the rocks it crosses become younger as it progresses. Until it reaches the Oxford Clay 1½ mile will be destined to become another tributary of the Thames, which rises only a few miles to the North, but instead, just beyond Malmesbury, it turns southwards to follow the strike of the rocks until it reaches Melksham after passing under the motorway and through Chippenham. This long reach might itself be considered the source.

lead to the Thames if elevations were only a few tens of feet lower, either by joining the River Kennet which reaches the Thames at Reading, or by making independent use of the Vale of Pewsey, but paradoxically it turns westwards towards Bradford-on-Avon, thereafter to cut through much higher hills to reach the Severn, through a series of gorges at Limpley Stoke, Bath, and Bristol.

This pattern is not easily understood. Rivers do not flow up and over quite high hills to begin cutting gorges, so that any mature gorge is difficult to account for, particularly when there appears to be an alternative route for the river across much flatter country.

And this, of course, leads to more explanations than the river has sources, a selection of five separate theories being as follows, to illustrate the diversity of opinion that may be encountered:-

No. 1. A long time ago, as this part of the country emerged from the Cretaceous sea, the Avon established its course across a fairly level plain, and as up uplift proceeded it could be developed into a canyon. Much later this part of the country was on the edge of the northern ice-sheet, which either directly or by means of melt water, altered the general topography through which the river was incised, without changing the base level of its course to any great extent. Parts of the original canyon walls are now visible in the present-day gorges.

No. 2. There was a stage at the close of the ice-ages when sea level was much lower than at present, and the Severn Estuary was entirely blocked by packed snow and ice drifting and sliding from the elevated regions that are now Cornwall, Devon, Somerset and South Wales. The blockage is said to have extended as far as the Scilly Isles. Note that at this time the Severn Estuary was really the Wye Estuary, because before the ice-ages the Severn apparently flowed north-eastwards from its source region to join the Dee in Cheshire. As the ice-line began to recede there was abundant melt water along its margins, but in this region it could not reach the sea through the estuary, which remained blocked. Ultimately the upper reaches of the Severn began to flow, but the water could not escape northwards because of ice, so it was obliged to flow southwards through a gorge it was able to cut at Ironbridge, to join with the waters of the Wye. It was able to do so because the Wye could not find a direct outlet to the sea. These combined streams could not discharge overland, cutting valleys, so they were obliged to flow into the Avon. Later, when the ice blockage in the estuary was dissolved by melting and by rising sea level, these gorges were no longer needed by the Wye/Severn, instead becoming discharge channels in the reverse direction to take off the excess water and the streams that fed them, leading to the establishment of the present day course of the Avon.

No. 3. The upper part of the river may well have once joined the Thames, and a small stream along the strike of the Oxford Clays may well have continued through the Vale of Pewsey, but the ice-ages modified these courses. As the ice-line receded debris and ice outliers of ice blocked old channels, but abundant melt water had to find its way to the sea, first forming lakes (e.g., Lake Trowbridge) which emptied when outlets were established along the present course of the river. The lower sea level at that time, combined with vast quantities of melt water, ensured that there was enough energy to very quickly excavate these outlets into the present-day gorges.

No. 4. Before the ice-ages, a fairly modest stream flowed through shallow valleys along the course of the present Bybrook and lower reaches of the Avon, possibly flowing southwards from Bristol to reach the estuary somewhere near Clevedon. Tributary streams flowed into it from the direction of Frome and from the Bristol Downs. Later, with the advent of lower sea levels and abundant melt water these streams became sufficiently energetic to deepen their valleys into gorges. The gorge cut by the stream from Bristol Downs was finally used by the reverse direction of the blockage of ice or debris developed between Bristol and Clevedon, causing a lake to accumulate and ultimately discharge through gorges into the Severn Estuary, which was never entirely blocked. The stream from the Frome direction also deepened its channel, and a minor stream from the Bradford-on-Avon became more energetic, cutting further and further back until it captured and diverted a fairly large stream crossing the Oxford Clays, thus establishing the present course of the Avon. This explanation seems particularly credible from a viewpoint just above Box, from which the valley of the Bybrook can be seen stretching away towards Bristol, with no hint that a major stream enters from Limpley Stoke away to the left.

No. 5. During the long period when the land that is now the Avon drainage basin was uplifted from the sea and eroded into its present shape by water and climate, a good deal of faulting took place that weakened or shattered the rocks involved. Two separate suites of fault have been identified, one with a predominant East-West trend, and the other NE-SW, exactly the directions of the various reaches of the Avon, which seems to follow lines of rock weakened by faulting all along its course. For example, there is certainly a fault running through the gorge at Bath, invisible at the bottom where it is deeply covered by river sediments. Convincingly revealed by the different elevations of the same strata on either side. The various reaches of the Bybrook follow exactly the same directions in a tighter loop inside the upper reaches of the Avon. No other explanation is therefore needed to account for the basic shape and extent of the Avon/Bybrook drainage system, although the topography has been locally much modified by the ice-ages and subsequent melt water which removed far more material than would be possible by the present day streams, giving rise to the wide valleys and deep gorges with which we are familiar.

These five basically different opinions are each quite plausible, but none is entirely satisfying by itself. It therefore seems to me that perhaps a combination of these ideas is required. Imagine, for example, the pre-ice age Bybrook, a gentle stream rising somewhere to the west of Corsham, finding its way to the Wiltshire estuary along natural fault lines; while well to the north of Corsham two large brooks also made use of faults in their progress towards the Thames. Then the climate changed, For a while the country might have been deeply covered by crisp compacted snow, but most of the time it was on the southern fringe of the ice-cap, where rain, snow and summer melt water would give rise to almost permanent miserable soggy conditions. Bybrook and its tributaries would then collect vast quantities of water, which would only erode their valleys to present-day dimensions but would not cause them to cut back rapidly from their origins but would also capture other swollen streams whose natural courses would have been impeded. Conditions like those whose actual course might well have conditions in Patagonia near sea level, rather than to the fringes of ice-fields in the mountains of Europe or N. America, and would offer a useful explanation of the Avon valley.

Incidentally, the same conditions might explain other problems in our region, such as for example, the huge cave systems of the Mendips and Chepstow region which could hardly have been excavated by the streams now flowing through them, even at their maximum winter flow rates, and Cheddar Gorge which surely needed more water for its construction than could nowadays be collected from the top of the Mendips.

Clearly the subject is far from exhausted, but perhaps in next years Journal it will be possible to report a clarification of ideas, following the continuation of the Extra Mural Course next Spring.

N.B. Dr Hawkins provided the spur to write this article during his Extra Mural Lectures, but he should not be blamed for some of the more fanciful opinions expressed, which arose in subsequent private discussion.

---

References