I have field-walked, observed and written about the Roman Trimontium for over sixty years and feel that I have inspected/known every stone and stain on the ground. However while having another look at the Trimontium section of Richard Strathie’s web site Borders Archaeology,* I noticed a faint but definite rectangular shape with rounded corners on the side of the road leading westwards from the Fort’s south annexe. 

It was in an unusual place but well worth investigating. A divining rod survey confirmed it as a Roman fortlet with a 2.5 m rampart and two 3 m wide ditches. Internally it measures approximately 47m by 28m with four barrack blocks 10m by 4m on the western side. There is only one gateway 3m wide facing on to the modern farm lane/minor Roman road.
As far as I know, this has not been noted before.

* is a great medium to spread and receive archaeological or historical information. Not all archaeologists will acknowledge a sent letter or respond to an e-mail but by placing information in this resource, it allows people who want the information to find it. The Trimontium to Lauder section of Dere Street is superb.

‘Gyrus’ at NT 57328 34433.
When the Department of Archaeological Studies from Bradford University came to excavate at Trimontium in 1987, one of the first tasks was to draw out a plan of the site using the Royal Commission’s air photographs. It was fascinating to watch the story of the complex being built up from the marks left on the ground by the ramparts and ditches.
On some of the black and white photos, there was a large white circle to the east of the Main Fort which did not fit in with the right angles and rounded corners of Roman forts. As it was deemed to be ‘something agricultural’, it was kept out of the plans.
Being brought up on farms where the main motive source was the horse, I knew it was more likely to be ‘something equine’ but at this period, I thought that archaeologists were all-knowing and the new science of geophysics was all-showing, so I kept my views quiet. However I did make a sketch-plan of the circle using divining rods and marked it on the general plan with red ink.

Although I have lost my original sketches, I remembered that the circle was over 20m in diameter with a gateway on the south side. I have some experience of horses by building paddocks and looking on while someone else trained the horse. With this, I decided that the circle had to be a horse enclosure or exercise yard.
As this is located in the field where a spread of Roman glass bracelets, beads and melted glass, suggested a Roman glass-works (NT 57234 34443), I ran divining rods over the circle occasionally and eventually drew out the plan shown below.

This piece of modern art represents the post-wall with a further series of individual posts set at regular intervals 2m from the wall. I visualise that these would be used to support an external platform for the convenience of onlookers.

During research over the years, I have found only one similar structure with a Roman connection. This was at the Lunt near Coventry where a circular stockade has been officially identified as a gyrus, a training ground for horses and cavalry recruits. It is claimed as being ‘unparalleled in Britain’ but this is an over-ambitious claim as there are bound to be more of these training schools not yet identified.

A good case can be made for the existence of a gyrus at Trimontium.
Trimontium is the centre and pivotal point of the Roman occupation of S.E. Scotland and well situated for an attached cavalry unit to nip trouble in the bud before it became a major uprising.
But it was also be good location for a cavalry training school. The country around the fort was ideal for training horses and men, with steep slopes, deep gullies, rivers to cross and enough arable ground in the river haughs to provide grazing, hay and grain to supply the horses.

From the number of pieces of horse harness and skulls found during the excavations at Trimontium, there has been a sizable contingent of cavalry in the garrison – more than would be expected on a fort of this size, even on a frontier post.

To explain the white circle which first drew my attention.
In black and white air photos, light does not penetrate sand as well as it does normal soil and so appears whiter on the photo. This phenomenon is not so apparent on colour photos.
Roman cavalry horses were never shod and so their training was done on a surface which did not damage their hooves; sand made an ideal surface and at Trimontium, this is what is likely to have made a circular shape which corresponds with the pattern traced by my divining rods.

In my search for information, I asked a friend, a skilled horseman, if the 24m diameter circle would be suitable for training horses. He said that it would ideal for a single horse but equally good for training the two horse team required for a chariot or wheeled vehicle; and there is plenty of evidence of these from the Curle excavations.

A final point for my proposition that the light circle is indeed a gyrus. In 1793 a Roman altar was ploughed up in the immediate vicinity. The altar was dedicated to the Campestres, the deities who presided over cavalry parades or training grounds and this was dedicated by Aelius Marcus, a decurion in a cavalry unit, the Ala Augusta Vocontiorum.
I don’t think this was a purely coincidence and there could be more in the same area.

This is as far as I can take my theory. I hope that someone can take it forward by putting trial trenches to establish definite proof.

7 May 2016.                                                                                                 Walter Elliot.  

Associated Pages