Why does it look so strange?

This article examines one geometric motif from The Hunting Dogs mosaic currently in Corinium Museum, Cirencester. Discovered in 1849 this is a good example of the Romano British style combining both figures and geometric patterns. I want to look at just one such pattern shown here;

The reason for selecting this particular shape has less to do with aesthetic concerns but more about how these elements were set by ancient craftsmen – why was it curved rather than squared?

The Romans were famous for their geometric mosaics, which are seen as aesthetically pleasing and well executed when viewed from afar. But once you get closer to them the truth is revealed: a number of these patterns seem quite rough in appearance with misplaced lines and colours that go out of sync with one another.

These days we expect everything to be neatly done with every detail accounted for. I feel that during Rome’s time period people did not mind so much about getting these details exactly right, as if “close enough” were good enough. There could have been significant variation depending upon the standard of the workers involved and also the individual commissioning them.

Hunting Dogs

The motif is called a ‘Solomon’s knot with interwoven square’. Solomon’s knot isn’t really a knot but two strands in chain links that are interwoven through each other. This pattern then has a square through all of it. This is difficult to see in the original one (left) as the strands change colour so the line drawing in the middle shows how it should be. The image below shows the underlying geometry.

You need to understand that there is the pattern, in this case, Solomon’s knot with interwoven square, and then there is the underlying geometry (left). You can see how the pattern is made up of four lines and four circles, the pattern is then created on top of this. You do not need to know the geometry to be able to set the pattern, once you have the outer shape, (square, circle etc) then you just need to know the width of each strand and the position of the single, white centre tesserae (CT). In this motif you can see that there are 13 CT.

Now let us look back at the original work, below. Each strand, in the centre section at least, is made up of a line of black tesserae on either side of six lines of colour. Obviously, they have had to curve along each side to account for the whole pattern so this will compress the strands of the interwoven square, which results in them having to reduce the number of lines in each section as it curves along the middle. They also have the same effect on the four outer ‘hooks’ of Solomon’s knot so you have to wonder why the mosaicist didn’t just make it smaller overall?

By reducing the lines of colour tesserae in the strands from six to four or three then the pattern would have sat well in the space they had.
One answer may be how the mosaic is set and I will look at the layers of the floor suitable for a mosaic and then two methods that we have evidence for, sinopia and etched lines.

Floor Layers
The image shows how a mosaic floor could be set with the different layers. In red I have shown where the two methods I am about to describe are in relation to the floor surface. The nucleus is the dry surface of the floor and the setting bed is a layer of cement (‘pozzolan’) spread onto the nucleus.

Method 1 – Etched Lines

Etched lines are where the pattern is scratched onto the dry layer, the nucleus. You then need to apply the cement a bit at a time over the pattern and push the tesserae into it. Obviously, the disadvantage here is that the cement obscures the pattern as your work.

Below, my copy of a section of mosaic floor where the tesserae and mortar had broken up showing the etched guidelines underneath.

Method 2 – SinopiaSinopia or ‘painted lines’.

This is where a layer of wet cement is laid onto the floor. Then, while it is still wet basic guidelines are painted onto it using just one colour. The tesserae are then pushed down directly into the cement following the lines. This example is from the Lod mosaic.
This is the setting bed and the tesserae have been removed leaving their imprints in the mortar.
Most likely it seems, in my opinion, that this motif at Corinium was set freehand. Working by eye the mosaicist would have set the single white tesserae in the centre and then set the others working out from that. Once the centre tesserae are in position then the strands are easy to place. I feel we need to see Roman floor mosaics, not as an art form, but as a craft, a job.

When you consider that with the average size of tesserae in a Roman mosaic being 8mm — 12mm square (3/8″) then in one square metre (3′ x 3′) you will have 8,000–10,000 tesserae. Very quickly you will develop your eye but you will also just get them down! It is estimated, and this is borne out by my own work rate, that they could lay about 0.8–1 sqm in a day and this is just one person, with his cutters and other assistants. Even in a smallish room, you could have two people doing the setting and as a craft, it would have been easy to train people up to do the job.
Mosaic work is very simple, how good you get is down to how many you do, your experience. There are no advanced techniques and I am sure that they had a very different attitude to the one we have now where we want everything to be neat and in the right place!

Lawrence Payne

More Posts
Marble surface finishes, what do you get?