Lawrence the mosaic running under the later Roman stone steps traces of pink mortar surviving against the wall.
Detail of mosaic showing pink mortar between and overlying the tesserae
Back in July we uncovered a corridor mosaic at Chedworth Villa, Glos. It was made up of thousands of little stone cubes (of course) and quite a lot of brick ones. What a life, cutting cubes of stone. Perhaps the apprentices did it before they got involved in the more creative side
The mosaic had been worn and broken 1600-1700 years ago. There was obviously not much money around at the time because they just filled in the gaps with loose stone cubes, bits of stone and pot and then covered it with pink mortar.
One thing we notice when cleaning the loose cubes is that you can always see which way up the floor surface was. It is worn by Roman foot fall and cleaning. The bottom is marked by the cream mortar the mosaic maker pushed it into and the fixing grout can be seen as a pink mortar staining the upper edges.
Cut the stone, fix in the cream mortar, assemble the mosiac, finish with pink grout, use clean and wear, erode break and patch. Abandon, rediscover and excavate 1864, rebury for 148 years and place in a finds tray in 2012.
Archaeology National Trust SW 
Could the pink ‘mortar’ have been just a by product of grinding the surface down? Once the mosaic had been set it would have needed to of been ground down. Sand or grit and water with different grades of grinding stones would have produced a paste that would have filled in the gaps in the same way as a grout.
Dear Lawrence
Yes, this seems unusual. I have not come across grout before when reading about the construction of mosaics and stone dust and dirt is bound to have accumulated between the tesserae. However, the displaced tesserae we found nearly all have patches of this pink hard mortar attached to their sides. This is a deliberate preparation that set after application rather than loose sand and grit lost between the cracks. It is either a grout or comes from a lost pink mortar floor that was placed above the mosaic at a later date and almost fully excavated away when James Farrer excavated the villa in 1864.
With best wishes
Martin
Martin, thanks for the reply.
I was thinking that once the mosaic is set then you would have the rough surface (the riven cut) of the tesserae to grind down. If you grouted this before grinding then you will have the grout sitting on top of the tesserae in the small depressions on the surface. This creates more to grind down.
Do you think it possible that they mixed in some of the cement with the sand or grit used in the grinding process? This way a grout would be formed alongside with the grinding and would fill in the interstices.
The pink colour would come from any terracotta or red stones as this is the dominant colour which takes over from all others in grinding.
Laurence
The grinding down of the surface had not occurred to me but I see how it would be needed to create a finished surface for the client. It seems like hard work. Do you know what they used?. The surface of the mosaic did have a lot of pink mortar remaining on it and this could be most clearly seen on the white tesserae so perhaps the pink mortar between the tesserae derives from this. It looks very similar. I will attach a photo.
With best wishes
Martin
Martin,
It’s most probable that they had a range of abrasive stones to use and (in exactly the same way as we work now to polish stone) they would have had the rough grinding piece, with sand and water, to smooth off the surface. Then they could have used a smoother stone with a fine sand (and water to keep it as a paste) to bring the polish up.
I can’t upload an image but here is a post I’ve just put on my blog which has an image of the finish achieved by hand polishing http://romanmosaicist.wordpress.com/2013/03/01/hand-polishing/
Lawrence
A very fine finish. I’ll attach the photo I mentioned a couple of days ago with the pink mortar attached.
Martin