102. Dairsie

Now this is one for old times sake in so many ways. Dairsie was the first bridge on which I was consulted when I got into arches in the 1980s. I have learnt so much since then and there has been so much water…..

A PDF version is available here.

The area around St Andrews was a very important monastic establishment, with a large cathedral (long since destroyed) in St Andrews itself. Dairsie bridge dates back to that time, though it has obviously had a lot of modifications since then. My friend (and ex student) Frances Ratcliffe asked whether I could show some of what we do with modelling to her son, and I used the opportunity of a trip to Aberdeenshire to stop by in Dairsie and Leuchars (Which also has a couple of bridges of a similar age). Time was limited and access difficult so the model is far short of perfect but it does allow us to examine the bridge in some detail.


This view cannot be achieved in normal photography. The missing bit of the right-hand arch is obscured by a large buddleia, that scourge of bridges but this does give a good general view. It is taken from Downstream. The land at the river bank is roughly level across at the crown of the pointed arch but it rises away from the river at the right. At some stage, the slope was smoothed by carrying it right across the bridge and raising all levels accordingly. Even at this resolution we can see the horizontal bedding in the wall above the central pier base. It is worth having a look at a bit of cathedral to put that in context.


Notice here how the beds are different thicknesses but each bed continues for a considerable distance. Many years ago, at a lecture in St Andrews, I heard that St Andrews and Norwich Cathedrals were supplied as kits from quarries in Caen at the same time. The beds in Caen extend for a considerable distance so it was possible to make a whole layer of one thickness but then use a different bed of stone for the next layer. One layer was shipped at a time.

Zooming in on the middle span already shows some interesting features.


The beds in the left-hand pier are of constant thickness as at the cathedral, though the constancy doesn’t reach over to the right.

Why is there a cutwater on the right-hand pier and not on the left? This is downstream and some medieval bridge builders didn’t see a need for cutwaters there but it seems certain that:

·        The right hand (pointed) span is the original

·        The horizontal line at the top of the cutwater originally continued right across

·        The two arches which now cross the river were originally flatter and pointed like the third

·        And that surely implies the pier was washed away in flood and two spans replaced.

Were the arches raised to increase the waterway?

Zoom in closer and there is more:

The ribs spring from the same level each side of the pier but at different angles.

There is a thin bed of stone at springing level of the arch proper but at the right hand side there is a proper springing while at the left the arch comes right down to land on a thinned part of the thin bed and the stone behind it is a very irregular shape to bring things to a proper level above.

The crest is believed to be that of the Bishop sponsor and sits in a course of badly eroded stone. Look back to the picture above and you can see that this aligns with the top of the cutwater at the right. Is it a hangover from a previous incarnation or did that damage accrue later?


Despite these obvious changes, striking horizontal planes through the model and examining both sides of the left had span we find a degree of correlation in the layers.


Looking closer again at the left abutment there is a lot to see, even though the model is far less than perfect here.


First, it is worth noting that the little corbel is obviously part of the original bridge. There is a marked break in the stonework above the arch which probably coincides with the original to the left and replacement to the right. The original tilts a little so it probably dropped slightly with the arch. There is also a horizontal line above the “original” stone, which may mark the top of the original parapet.

Oh, but there is something else rather strange. That little corbel is noticeably lower than at the other end. Was there a slope in the structure from the start? The difference is 5ft, 1.5m

If we look at the upstream face of that span.


Here there is good correspondence between the courses in the pier and the left hand side of the spandrel but at the right the courses are different. There is a link from abutment to spandrel but then it breaks up. Is it possible that the left hand half span stood and just half the span fell? In England, there were a lot of arches were dropped deliberately during the civil war. Does this damage date from the risings?

Looking at this fresh in the morning makes me think that just this upstream side of the pier may have gone in a flood. Here’s my guess at a break line.


I need to stop soon but I want to look up from below first. Here is the low end span:


And the middle span

The first thing to note is the wavy ribs. At first I thought they were the result of movement but the edge chamfers on the barrel seem straighter so perhaps they were built that way. It is very easy not to think (or to forget) that very slight rotation of the voussoirs will produce curves like this. I think it might be the result of a relatively hasty rebuild by people who had never built a vault rib. (and no, I didn’t think of that myself, it came from a cathedral mason, my friend Peter Dare.)

Though a closer look says there is a slight wave in the top edge of these views right of centre at the top and left at the lower picture.

Note also that the middle span is much wetter than the downhill one.

Also that the spandrel walls are visible in these upward views showing how they slope outwards.

A look at the cross section shows that too. Interesting that the parapets are thinner at the base.


And just before I leave, though there is much more to see, Look at the up-hill end downstream side abutment.

Looks a mess, doesn’t it. If you follow the crease in the wall to the right of the wet patch you will reach a drain pipe.

Tilting the view makes the position clearer.


And taking a section through:


Here, I have added three arrows.

·        Top right marks the gully

·        Centre is the drain from it

·        And below that a step in the wall with water leaking below.

Proper diagnosis would need a little more work but I have seen something like this before. The gully is an attempt to catch water running down the hill before it crosses the bridge. The trouble is, that drain through the spandrel is a very unforgiving thing. The least settlement of the gully will crack it off. The outcome then is that the water is channelled in behind the abutment. Not a happy scenario at all. Part of the problem is that the gully is directly in the wheel path of any heavy vehicles that cross.

There is a 10T weight limit but a very large tractor and trailer crossed while I was there. His tyres were squeezing both curbs all the way across.

Once again, there has been far more to see from the model than we would ever have found on site. We were there about 2 hours, with only two cameras. Access wasn’t easy and many of the photos were taken from far away, especially on the soffits.

In the pointed span I was not able to get far enough away and certainly unable, through the vegetation, to get any sort of link under the bridge. Had this been more than just a passing visit I would have cut down that butterfly bush and perhaps even braved the bulls in the field to get better photos of the uphill, upstream spandrel and the link under the arch. As it was, the combination of a very wet boggy patch a couple of gates blocking the opening kept me out.

Oh, and one last thought. I tried a circle in both spans.


The curve is obviously the same but the left span is shorter.

A closer look shows that left span doesn’t sit happily at the end. I guess the centre wasn’t set up quite right.


And that really will have to do. Back to paid work.