The Roman Masterpiece Buried Under A Molehill Time Team Odyssey

This is Withington in the Cotswolds, an area of outstanding beauty with a lot of rich farmland and some very classy houses. In Roman times, it seems to have been pretty much the same because 200 years ago, plowmen somewhere around here discovered a huge villa. Excavations revealed some exquisite mosaics, but no one was able to date the villa and there's no record of exactly where it was. But that's just the beginning of our story. 200 metres below the villa field is this lovely little spring and guess what? Roman mosaic and Roman tile have been found here and in this field over here. No one's dug here, but the two sites are too close together not to be connected. Could the big villa over there have had a huge complex of buildings down here? Could it be a bathhouse or given the proximity of this water, could it even be something sacred like a temple? We've got just three days to find out. The Cotswolds are in the very heart of England. 2,000 years ago, this was the breadbasket of Roman Britain and by the middle of the second century, Cirencester had become the second most important town after London. We were invited to Withington, just a few miles from Cirencester by local archaeologist Roger Box, who's given us a couple of juicy leads before. Rog, you've already located one fantastic villa for us when you found Turkdean a few years ago. Why are you so excited about this site? Well, I don't know, but all I do know is every year, the moles that are in this field, and there are a lot of moles in this field, keep pushing up these small tassery cubes, literally by the hundred. We've got a villa just a few yards up there. Yeah. So could we have another villa here? I think it could, but I'm more interested actually in the spring, which is under the bush over there. This is called Withington by the Wall Well, which rather suggests it was, you know, perhaps named in the Saxon period. And we know that Romans, when they come in, they often take over pre-existing sites like spring, which are religious interest, and put temples and turn them into their own religious sites. Temple? Well, maybe, but this is a very unusual location. It's very wet down here. There's a lot of water about. It's going to flood. I think it's much more likely that the main villa's further up the slope, but perhaps this is the predecessor, the early villa, which was then replaced by a later villa further up the slope. David, what about these bits of mosaic? Do they give us a clue? Well, these could belong to an early villa, but I think it's more likely a bathhouse. So we could have an elaborate bathhouse associated with the villa, either down here or up there, or, and perhaps we could have a swimming pool associated with an early villa. The important thing, which I keep going back to, is what is the relationship between what is happening down here and the really big villa up there? All right, if you were all gambling men, which I know that none of you are, what are you going to say this is? Temple. I'd like a temple. Early villa. Villa. To all. Early villa, bathhouse. Bathhouse, villa, temple, swimming pool. We're not going to know till we start digging, are we? Get going, Phil! Our site is on the slopes of the valley of the River Colne, which flows right past this lower field. Besides mosaics in the molehills, it's full of mysterious lumps and bumps. As you can see, there's huge anticipation as Henry and Geofiz do what they have to do. By 10 o'clock, we've got a massive target. I mean, I see a place there that's crying out for a trench, and it's got to go just in there. Can you locate us on that, then? Of course I can. Can you dig it? So our first trench goes in near the river. No sooner has Phil got the turf off than mosaic pieces start appearing. Got our first one, look. Oh, nice. Tessera. It's out the topsoil. Yeah, wouldn't matter. It proves Roger wouldn't plant in them. Oh, look, there's another one. Popping out all over the place. Yeah. So we're off to a flying start in our first trench. We still don't know, of course, whether there are buildings down here, but we're optimistic because we do know that there was a Roman villa just across the slope in the upper field. It was discovered in 1810, just five years after the Battle of Trafalgar, and some fabulous mosaics were unearthed, which are now preserved in the British Museum. This site's protected by English heritage, but bizarrely, they don't exactly know where the villa is. We've offered to find it, more fool us. The villa was excavated by an eminent antiquarian called Samuel Lysons, and that's the problem, really, because he seems to be more interested in chasing mosaics than he was in working out the extent of the villa, so our first job's gonna have to be to find it. But John, we've got these zonking great pylons here. Aren't they gonna get in your way? It shouldn't be too much of a problem. If we're actually close to the metal pylons themselves, maybe, but the overhead wires shouldn't be. These instruments are designed to filter out that sort of noise. Better get going, then, yeah? Right. Let's have a look at this. Is this a big villa? Well, I don't think it is, Tony, because if you look at the scale, you know, it's only about 120, 130 feet across, which, you know, it's not that big, really, and I've orientated it more or less as we think it is. Do we know if he excavated the whole thing? No, we don't know that. We've got his plan of the walls here, and we've got a nice watercolour he did himself of the excavation, which is quite nice. It's typical early 19th century. They followed the walls, and they've opened up areas where the mosaics are, but you can see here that they've actually left the areas of soil in the rooms. They haven't actually dug those. So, you know, we don't know whether he dug the whole thing. We don't know whether he dug everything, even in the areas that he did dig. So, realistically, we could end up digging something that someone's already dug. Yeah, but I don't think that's a problem, you see, because archaeologists moved on in the last 200 years. There's lots of things we would want to know that he wasn't interested in. He doesn't, for example, seem to have kept any of the finds, any of the pottery, any of the coins. He wasn't interested in that, whereas if we could find those in the backfill, when, you know, over the site, that would at least give us some broad date ranges of, you know, when the thing was occupied. We don't know that at all. There's already a race on between Stuart and Geophys to find Lyson's villa. Stuart's armed only with Lyson's watercolor. Doesn't seem much to go on. It was painted 200 years ago. Geophys have been out searching for two and a half hours already, with two different sorts of equipment. And still, not much is showing up. The pylons could be their problem. Curiously, it looks as if Stuart's using the pylons to his advantage. How have you done? I think we have a modicum of success, actually. I'm pretty confident this has been drawn from anywhere between where that pylon is there and where we're standing now, looking up this slope. So if it's where the pylon is, then it's on the other side of the road in the next field. Yeah, it's got to remember, though, when Lyson's drew this, this road hadn't been built then, so it wouldn't appear in this drawing. But look, you've got a very, very strong correlation between these trees up here on this ridge. Those are those trees up there. You've got the hill to the right. That's where the pylons go over. Drops down, comes up a notch here, which you can see over there. And this band of trees is that far band of trees on the ridge over there. And in between, you've got this crest line, which is the top of this slope, showing the thing has been drawn somewhere on this slope up above us. But one of the things I was always taught about Roman villas is that they have a spectacular view. You wouldn't really call that view spectacular, would you? Well, it depends how you define spectacular, of course, because if you turn round, what this villa does is looks towards the south. And a lot of Roman villas do face the south because you get the sun all the year round. There are so many tesserae, or pieces of mosaic, coming out in the lower field that Phil's put the machine away to avoid any damage to the archaeology. In front of your spade, that one? Yeah. Another one there, nice one. That's loads, you've found loads. And all I've done is the top soil from there to there. I've found about 10. But what we want is to get them in a floor, man. That's what we want. In the lunch hour, Roger's got some volunteers together to search the stream near the spring. This is where he's previously found Roman roof tile. Oh, here we go. These are bits of sandstone. They're not Cotswold natural stone. They probably come from the Forest of Dean or something. This is sort of typical roof tile, which seems to be predominant going up through here. We've barely been going five hours. Phil says he's found something to bring me out in goose pimples. What you got? Oh, look, this is what we got. We got a mosaic. Look, there it is. Isn't that absolutely incredible? And look, it goes right the way across there. It's still going there. Look, it's still going. Look, there it goes. So, there it is. Let's have a go so we can find it. Go on. On the other tree. There it is. Blimey, O'Reilly. It's still here. Two square meters already. Wow. I know. And of course, it's going back in through there. And the further it goes back in there, the better protected I think it's going to be. Because I think here, we're right on the edge of where it's been taken away by the plowing. When we say a Roman mosaic, this is just white with the occasional bit of red in, isn't it? It's not pretty. Oh, come on, Tony. What do you mean it's not pretty? Look, it's a Roman mosaic. How often do you find a Roman mosaic just like that on day one? Very, very rarely. Did you think that there might be a flaw as close to the spring as this? Well, this is where the mosaic's been coming from. Yeah. You know, this is where the mould's been pushing it up and it's a bang on cue. Well, that's a good start, isn't it? That's incredible. Absolutely incredible. All right, show us the rest of it. Geophys have got their first results from the upper field. There's occupation here, but it's not all Roman. Look, the first thing to note is the pylons have had a very minor effect at the top, but don't worry about that. What we appear to have is a series of curving ditches. I mean, I'm sure that's probably prehistoric. Yeah, they don't look Roman, do they? But we do have an area of noise, and that's what we're hoping we might see that would go with the Roman brick and tile. And then beyond that, a series of rectilinear, regular sort of ditch systems. Well, I think, you know, we can be pretty clear. You found the villa, haven't you, John? I mean, that's, you know, there's your rectangular range. It's about the right size of what Lyson's found. And then look, you know, we've got the courtyard out in front. So it's a classic Romano-British villa with a main range and a front courtyard. Yes, but that is very, very noisy-looking splodges. You can't see any clearly defined walls there. No, but it has been partially excavated, hasn't it? So we could be seeing the backfill of the excavations as well as the Roman material. So at four o'clock, we open up the first trench to be dug in this field for nearly 200 years. We're expecting to find Roman walls, but we might only find rubbish from Lyson's spoil heap. Oh, yes. Ooh. Okay. Long time. There we go. Well, that fits in nicely with the other finds that we're getting. Ooh, it does. A lot of bits of brick and things like that. I mean, really, this is just what you would expect the Victorians to be throwing back into the hole, really, because, you know, they didn't, they just weren't interested in this kind of stuff. So, you know, we are just digging through a spoil heap, effectively. In the lower field, the archeology is just getting better and better. Phil's uncovered loads more of the tessellated pavement I was scraping earlier. At the other end of the trench, Matt's beginning to find the walls of a structure in amongst a lot of rubble. Look at it. There it is. More of it. Geophys, who've split into two teams, have a hunch there's also a villa down here by the river. Their first magnetic results told us where to dig. Now, by using resistance, they should be able to pinpoint the walls and the outline of any building that may be here. Our first trench in the upper field has delivered good archeology, but we can't begin any evaluation work on the scheduled villa until we're sure that we've found it. Rich, where are we? We're on a Roman wall. You can see it's completely running off in that direction. And we've got two faces, one running up here and one on this side here. It's measuring about 50 centimetres across. But the problem is, what's going on on either side of it? We've got this really dense material here. We've got various things coming out of it, this compacted material. Look, we've got here tesserae coming up. And we've also got a few bits of this greyware Roman pottery. Oh, look, even licences be a brothel. I don't think that's Roman. Yeah, I mean, it's very striking, isn't it? This is very hard and compact, and that's very brown and soily. I mean, I would lay some good money if that's licences chafed there. This is intact Roman demolition on this side. But I guess the key is, we're only gonna find out if we extend the trench. Exactly. The ideal thing would be to try and define at least one room. By the end of our first day, we found more archaeology than we often see in three days. There's even a second mosaic at the other end of Phil's trench. The problem, as always, is understanding what's going on. Phil, you see this side of the mosaic here? Well, earlier on, you were saying you thought it was plough damage, but isn't it an edge? Well, it is an edge, and the thing is that it is not caused by plough damage. I admit that now. But now we've got the whole trench open, I really can tell you what's going on. What we've actually got at this end of the trench, I think we're outside the building, and you can see we've got the collapsed roof, which has come sliding down out here. Here's the main back wall or front wall of the building. This, we now know, is a corridor going through. So your straight edge becomes part of the next wall, which has been robbed out. And now inside here, we've got the main rooms of the building with a collapsed mosaic. Where's the mosaic, Matt? We've got one piece there, that bit there, and that long piece going across there. So it's quite extensive, isn't it? So what are we going to do here tomorrow? Well, we've got John's new geophysics. This has got the resistance on it, hasn't it? Yeah, and I think what's exciting about the results is when we'd only got the magnetic, we thought we might have just a simple, maybe three-roomed villa. Now we've got the resistance, it clearly shows that the rooms extend in this direction. It's much bigger and more complex than we thought. So we're going to do a bit more with this area tomorrow and then target a trench somewhere in the middle there, probably a 10 by 10, something like that. And what about our villa up there? Well, the big news, Tony, is we've actually found Vycens' villa, but the problem is we don't actually know where we are in it. So we've got to get a trench across it tomorrow. Let's work out where we are on his plan. And I've been looking at the spring over here, which is very fishy. We need to look at that again tomorrow. So we've got a fishy spring and we've got a Roman building that no one ever heard of before. That's not bad for day one, is it? But how does this building link to our villa up there? Let's hope we find out tomorrow. Beginning of day two, and yesterday we found two Roman mosaics in that trench there where we didn't expect to find any. But over on that hill where we hoped we might find a mosaic, we still haven't found one. But of more immediate interest to me is this spring here. Last thing yesterday evening, Mick, come out over there a second. Mick said to me that he thought this spring was fishy. Well, I've been poking around it all morning, Mick, and I haven't found even one stickleback yet. So what do you mean? Well, when I came down here yesterday evening, I cleared all these nettles and grass away. I expected to see the water oozing out of the ground, you know, as you do in the spring. And instead of that, I found this big stone basin. Here, look. And these couple of stone channels, you see how that's been cut through the middle there? Yeah. Is this Roman stuff? Well, I think it could be, but rather more interesting is that the water doesn't ooze out of the ground here. It actually comes out of a stone-built channel, look. So it's not a spring at all? Well, the spring must be somewhere back up that way, and it's been channeled to this point here. Hang on a minute. We thought that there was a spring here and that there would probably be a building, a temple or something that respected it, and it might be that building there. Well, it obviously can't be. It's too far away from the spring now. So does that imply there might be another building we haven't found yet up there somewhere? There could be somewhere back up where this is coming from. So I think we need to geophys in this field and see if we've got any structures up there now. Excellent. So, beginning of day two, and we've got yet another target. Not a bad start. The Romans were renowned for their water engineering, but if this is a Roman culvert, it seems odd to go to such trouble to bring fresh running water down to these buildings from a spring further up the valley when the river was so close. Henry's marking out the grid so that geophys can begin searching for the spring head where there might be a temple structure. In the lower field, our Roman experts might have the answer. They've been looking at the finds from this trench, and they're now convinced that the building was a bathhouse. A basic Roman bathhouse had three rooms, a warm room, a hot room to sweat out your dirt, and a cold plunge pool to cool off. They think we're digging in the hot room. Come on, how can you tell that from one trench? Because it's got distinctive types of tile, including this one, which is a fussoir, which is one of the box tiles forming the vault in the ceiling. So you literally get a tube of boxes going around the vault. Allowing the hot air to circulate up the walls and around the ceiling. So this sort of odd shape is deliberate? Oh, yes. That is incredible. So that's it? And we've even got a bit of the vault. This is a chufa, nice light limestone, so you can get the whole of the stone vault, but half the weight. So you'd get a really good stone vault instead of a wooden roof, which would rot in a hot, steamy room. So is this the vaulted sort of roof that has caused our mosaic to collapse in? Highly likely, yeah. And presumably it all belongs to the 4th century. That's what the coins are saying anyway. That does it, because I've got the piece of box tile here from a wall box tile, which has roller stamp decoration and is almost certainly 1st or 2nd century. Implying, therefore, that we may have an earlier bathhouse on the site. So this decoration is literally put on with a roller. Correct. You get a piece of clay and you imprint. As opposed to a comb decoration, freehand comb decoration. Hey, this is really good, isn't it? Because it looks as though we've got not one bathhouse, but two bathhouses. Yes. I need to get my head round this. David and Richard are saying that we've got two phases of bathhouse here, 300 years apart. And as this might be a hot room, could there possibly be a hypercaust, an underfloor heating system, right here under the rubble? The floor certainly looks hollow. At this stage, we don't know if these are air pockets caused by the collapse, or if there is something more engineered, like a hypercaust, running back under our mosaic pavement. Our search in the upper field was supposed to be the easier task. The 19th century antiquarian Samuel Lysens excavated an important villa here during the summers of 1811 and 1812. This is his own record of the walls and mosaics he discovered. But we're still searching for Lysens trenches. And because this is a scheduled site, we can only dig an area of 100 square metres. Each trench has to deliver important information. In the lower field, John Giafiz has just got the latest magnetic results for the area around the bathhouse. Are they impressive, or is it just the colour scheme? Crikey, John, that's good. What have you got there? It's pretty good, isn't it? Yeah. That is your trench, Phil. Right. The mosaic coming through this way. And this is the stream following this edge here. So clearly we've got this sort of big room extending in this direction. So is that edge there the same as this breaker slope behind us here? Yes. Right, so we've only got a day and a half left. Well, I think if we go across this edge and find the limit. Yeah. What do you actually think that is, John? I'd like to think it might be a plunge pool. Really? Hey, that'd be good, wouldn't it? A plunge pool? Well, at least John's theory fits in perfectly with the plan. We'll see if this is a runner once Phil's got through the topsoil. We're almost halfway through day two, and I'm beginning to get really frustrated about this so-called villa, which is up on this hill, because we still haven't been able to find it. And quite frankly, you two are just flashing around, aren't you? I'm a little perplexed, Tony. This should be our Roman villa. Can you see here, very clearly, there's a rectangular structure with a courtyard out in front. On the basis of that, we've put our test pit in there. Problem is, I can't relate what we're finding in our test pit to what licence says we should be finding. We should be in here. But you've got a wall in the trench. Yeah, but it doesn't look like licence has been there before us, and it's untouched. It's virgin ground. And thirdly, I can't relate any of it to the landscape setting. But I talked to you yesterday, and you said, there's no problem. I know where the villa is, because I've worked it out on the picture. I still don't see any problem. I'm not as perplexed as Neil is here, because we've got lots of other evidence, as well as licence stuff. We've got the stone, look, behind you. We've got stones sticking out the ground. Oh, wow. There's a lot more further down there, Tony. Oh, it goes on, doesn't it? It's not the only one, by any means. So this isn't just sort of surface scatter? No, no, I think this would turn out to be part of the villa. And see, it's not the only one. There's more down there. Other evidence, we've got this slope down here in the ground, so we're on top of a terrace. And look at the, we talked yesterday about the view, perfectly oriented towards that view. The courtyard that Neil showed you on there, that's that low area. It's actually cut out of the slope. Well, this is quite interesting. I've got to give Stuart this, because there is this noise on the geophysics. So perhaps licence excavation took place here, not there. Might we possibly be able to put another trench in and check this? I think a test pit in here, say two by two, because if we can find licence to dig, it means we've got one building and a second one he's never seen before. So Neil gets a new test pit underway right on top of Stuart's anomaly. In the lower field, John's getting excited about his plunge pool. Phil's being very patient. I'd like to think you've got a wall somewhere there and a wall somewhere here. We've got a definite change there. There is. The right side of the room here. Yeah, yeah. That's definitely the edge of it. But this is, there's, well, there is a lot more. Look at that. Yeah. You can push your trowel right in there. You can't up here. Well, I reckon between those two walls, we'll have the plunge pool. Well, I think we can clean a lot of this more off with a digger. We'll find your walls and then we'll see what you're right. Oh, I'll come back when you've done it. Shoving my cosy. I do sense a slight reluctance on Phil's part to share in John's enthusiasm. But the simple fact is that we've got great archeology here, which can't be said for the upper field. Bridge has extended our first trench here, but after nearly a full day's digging, it looks like we're still clutching at straws. Where are we then, Bridge? You're in a trench where it's a bit of a confused state of affairs. So we know we've got a Roman wall here behind me, but either side of it, we've got this demolition rubble. It's throwing up loads of late fourth century pottery. So we know it's good, but it's looking very undisturbed at the moment. So this hasn't been dug through by license as far as we can see then. So we're not in the area of his excavations. I think that's clearly in situ, isn't it? That's undisturbed by license. Is there any point in doing any more here, in fact? Well, not really, because all we've got is this demolition rubble and the scheduling of it means that we can't actually go into that. So we can't generate any more information. We don't want to disturb this, do we? Because it isn't the area that we were looking at that was disturbed before. So what, we shut this down? I think the key is to record this, shut it down. I think we should get John back for some detailed resistivity. Let's just see if we can get a clearer plan of the villa, which then relates to license, and then say, okay, that's where we're gonna put our trench. Okay, well, let's do that. And then perhaps put some trenches across the sort of new area of survey. Yeah. Yeah. Okay. Sadly, Stuart's anomaly turned out to be a modern field drain. So we'll continue to search for licensed trenches tomorrow with the help of GFIs. One of the questions we asked ourselves yesterday was how did licensed villa relate to the buildings in the lower field? We still don't know. What's more, Phil's hit his first hurdle. Well, Roger, we resolved John's geophysical anomaly. Oh, yeah. I'm afraid it's not as good as we would hope. One thing, it is not. It is definitely not a plunge pool. Okay. What it is, is an infield quarry. Look. This is all made up ground. Blue stone. I think what they've done is they dug a slot into the side of the hill to take the stone away. And that's what John was seeing on the geophysics. The thing of it is, it's not all doom and gloom, because we do actually have some preserved Roman archeology in here. Oh, yeah. I mean, we've got tassery in there, look. Got it. And animal bone. So I think what's happened is that literally this quarry has chopped away all the Roman archeology. John GFIs is not gonna be happy. His ambitions for a plunge pool have just been reduced to a pile of old stones. But there's plenty more to be getting on with. We should know if we've found a temple or a springhead by the end of the day. There may be a sense of euphoria at the coalface because our finds have been so spectacular. But on site, with so much going on, I'm worried that we're spreading ourselves too thin. We've had a cracking couple of days. We really have. I mean, the quality of the archeology is absolutely amazing. We've got new site areas, haven't we? Not only that, we've got a new mosaic, too. The one that's broken down in, an absolute cracking new mosaic. Yeah, well, we could all pat ourselves on the back. On the other hand, we could say we came here to find a villa. We've got a responsibility to find a villa and we haven't found it at all. And on the lower site, we still don't know what the thing is. We have no idea what to do with it. Oh, we do! We do know what it is. It is a bathhouse. We know how big it is. We know how well-preserved it is. We've done two good days' work. The top site is still a problem, you know, because we haven't really dropped on that yet, have we? Not only that, but that was the reason that we came here. No, it's one half of the reason. We were given specific permission to come here. We had a special piece of paper from someone on high that said that we evaluate that top site in order to find where Lyson's Villa was. And have we found it? We haven't got anywhere near finding it. But we've got a strategy. We might get there yet, you know. Yes, we've done it. Yes, we've got another day left. I mean, what we really need to see is John's resistivity results. What have we got on the geophys? Not a lot. I'm confident that we've got the villa based on the magnetics, but we haven't got a clear wall plan based on the resistance. But if you take us, as you've done, to where we think it is, then we put some big trenches in that point. And you've got the meterage, too? We have, we have. We've done 35% of the area we're allowed. We may have plenty of meterage, but we haven't got plenty of labour. We have, because if we pull most people off the bottom, having done the evaluation there, we can put most people on the top area. Because we know what the lower site's all about. Yeah, then we can dig, what, a couple of trenches? I think two trenches, each long enough to get the whole width of the villa. I'm confident. Yeah. I am confident that by tomorrow night we will have found License's plan. As you were at the beginning of day one. John, all that begs the question, the third potential site, the middle field, the possible springhead. Now, you've jiffed, is that? We certainly have. And? Well... Er... Oh, it's a team, isn't it? Yeah, it's a team. There's the top field. Bottom field, the mosaics. That's where the well is. Yeah. Look at this, look at that. Crikey. It's just an amazing series of responses. You'll presumably finish that bit tomorrow, won't you? Yeah, but I mean, look at this for the moment. Isn't it a pity that Mick's just said that all the labour's got to go to the top field when... Hang on, hang on. No. I think the key is we must get our trenches open first thing tomorrow morning. Labour in there, let's make some progress and then we can free some labour for there. Effectively, we're beginning to think of closing that site down. Yeah, because I think we've answered what we can hear about that. So do I. Complacent archaeologists are archaeologists doing their job properly. I'll leave it entirely to you. But tomorrow we're going to find License's villa and hopefully we'll get a temple, too. Cheers to you. Beginning of day three on our Roman dig site here in the Cotswolds. And the big challenge today is to find the lost Roman villa. Only problem is where to dig and how. It's a big field and not many hours left. Well, that's the problem. We need to hit as many of these walls as possible on the License plant to know that we're in the right place where the villa is. I mean, here's the villa on the geophysics. You know, very clear. It's about the same length as License. But, you know, did License dig here or did he dig here? Yeah, I think we've got to go somewhere across that noisy area, haven't we? Something like that. And I'd say a 20 metre long trench should give us the full width here so we get as many walls as possible. And then we can say, OK, now we know where we are. The trouble with a narrow trench to book it with is you won't know which walls you've got on this plan. I'd like to see us open a bigger area up so we can try and understand this. Yeah, but if we open a bigger area up to start with and we're in the wrong place, then we've used our allowance up. I mean, I'm fairly confident, Mick, that if we open one trench, one and a half metres wide, we'll know whether we're in a disturbed area or not. We'll know whether License has been there. And if he hasn't, we've still got enough allowance for another trench to go somewhere else. Yeah, OK, well, let's start with that. And then if we think we're in the right area, we can open a bigger area up to see more of it. That would be the best way to do it, I think. Reluctant agreement? I'm afraid so. Considering all the extraordinary and unlikely things that time teams discovered over the years, you'd think we'd be able to find a scheduled Roman villa. It comes down to a battle between technology and instinct. John, stung by seeing his plunge pool idea reduced to a pile of rubble, says the upper feels too dry to get really crisp resistance results, which would help us see walls clearly. So we're having to use magnetometry. I don't know what John thinks, but I can't see any walls in there. Stuart's always been tantalised by License watercolour. His instinct tells that the villa should have been built in a rather different place. But for now, we're following the geophys, and it remains to be seen who's right. It looks like we might be having a bit of a wall come up. Oh! There's water spread here. This all looks disturbed round here to me. I think... I feel much happier. Yep. Definitely has a feeling that someone's been here before us. Someone's been here. What's the giveaway? It just comes off so easily, doesn't it? It just literally peels off onto this level, almost as if someone's dug this out before and then just loosely put the earth back on top. Last night, we wanted to close down the lower field. We've solved Roger's molehill mystery. We're convinced it's a Roman bathhouse, complete with collapsed mosaic floors. But every trench seems to be producing more and more stuff. Last night, Phil did open another trench to check out the pile of rubble, but was shocked to find lots more walls. This has made him think again about John Gator's plunge pool idea. And as the rubble comes out... See that? That blue clay, I'd have thought that could be lining or something like that, see? I tell you what, if it is clay lining and it carries on raining like this, it really will be a swimming pool. Cheers, Phil. Plunge pool or not, the experts are already planning the room layout. Of course, you know who's not going to let me forget about this, don't you? Yeah. Gator. I think I'd better prepare me defence now. I think you'd better have. What did John say yesterday? He said if it was a plunge pool, he'd like to see walls at either end. Have you got the other side there then? Well, I ain't sure, but I don't want to risk taking it out if it is. Hmm. Now we can see how the rooms of the Roman bathhouse related to each other in this lower field. Next door to the plunge pool, Matt's beginning to uncover what we suspected all along. David, come over here a minute. There was a hypercaust here, an underfloor heating system, so this must be the hot room. Like an in-situ bit of the bathhouse, if you put it down there. Look at this. Oh, definitely, that's the peely for the floor. This stack here is still in-situ. Stacks of tiles called peely supported the mosaic floor, allowing hot air to circulate underneath and through flues in the walls. We've got, what, one, two, three, four, five, six at least. Five, yeah, and they're still going down as well. Yeah, we're getting down there and there's rubble and stuff. No sign of the floor yet? Well, we've got bits of this collapsed floor, yeah. We've got a bit of boussoir here. Yeah, but the bottom of this, I can't see how far down it is actually going. Bits of tesserae coming out all over the place. You may find another stack somewhere here. They're not so close. Yeah, well, they're sometimes separated by roughly 30 centimetres. That's good. Lovely. Thanks. At last it's coming good in the upper field. After all the hours of geophys and collective head-scratching, they've stumbled on the remains of a mosaic, and alongside is the decisive clue. A mosaic has been lifted from here, and who else would have done that other than our old friend Samuel Lysons. It's great to see this at long last, but we still don't know where we are on the Lysons' plan. OK, Ian. It's 12 o'clock and the hunt's on for the temple in the middle field. We should have started this trench much earlier, but there's a great feeling that we've probably left the best till last. John's chasing one of the biggest responses we've ever seen. We all laughed at him when he first identified the plunge pool, but he was right, and his stock has never been higher. John believes these lines are some sort of watercourse that might somehow be connected to the conduit Mick discovered yesterday. Do you think that building's associated with this wellhead? Yes, I'm sure. Right, yeah, you can see it there on the base. All right, yeah. Quite clearly. Meanwhile, in the lower field, Max uncovered more of the peely stack, an essential part of the hypercourst. In another part of the trench, we can see how the mosaic floor has collapsed into the hypercourst, due almost certainly to the roof falling down on top of it. Stuart's convinced that our three sites make up a classic villa landscape. If that is the bathhouse for our villa, it's a heck of a long way away, isn't it? I don't think it is, actually, Tony, because what we're dealing here with is a concept of views and connection with the earth. You've got the villa here, halfway up the slope you've got the spring, where there's a shrine, and going to the bathhouse is part of a journey. You probably go down to the shrine, pay a bit of homage there, then down to the bathhouse, and the same on the way back. And it's all within that one view down there. And you'd always have the god in your view, wouldn't you? You would. The Romans believed that the gods gave them the fruits of the earth, they gave them water. This is your domain, and within it you've got shrines, bathhouses. You believe the gods are with you on this site. Whether the gods are with John and Kerry remains to be seen as the search goes on in the middle field. So far there's nothing but Cotswold brash, the local stony soil. There's no sign of that springhead, nor any of the features that John's sure must be water channels. With responses like this, John is baffled. I'm baffled. Back in his plunge pool, Phil's eating humble pie. You've dug out your quarry then? Well, yeah, except it's not a quarry. What is it? John was right, it is the plunge pool. You're kidding. No, I'm not. Can I come in? Yeah. In fact, if you come under that step there, I reckon that that is just about the level at which the Roman would have walked across the floor down a series of steps to get into the pool. Wow. This is a Roman step? Yeah. So what's that on the other side? Well, that's the other side of the pool. It's a lovely substantial wall, and then the actual floor of the thing is this clay, look. Just a sealant, really. And then the floor itself, these rather nice red tiles. Turned out pretty well this day, didn't it? It's absolutely incredible, hasn't it? Just alongside the plunge pool in our hot room trench, David is trying to sort out the dating by identifying the different mosaics. He believes this one, the finest in the building, was made in the 2nd century, but was put on top of an earlier floor, presumably when the owner had saved up enough money. Nearly 200 years ago, in the upper field, Samuel Lysons excavated a Roman villa and discovered some exquisite mosaics. In our dreams, we were hoping to find mosaics similar to the fabulous Orpheus mosaic, which can now be seen in the British Museum. So are we anywhere near sorting this trench out? Yeah, you can see there's lots of walls running across it, and we now know where it is in relation to the villa that Lysons found. Oh, wow, look at this. David, tell me all about this mosaic. Well, we've got a series of intersecting circles. You can see the semicircle here and here, and this is the actual edge of the panel. The corner, you can see, is just about here. It's just here. And the opposite corner would be roughly in this position. It's been ploughed out. So you can imagine that to be a square panel. So when you say a square, the pattern is in here with this great big area all the way around it. They're big tesserae, though, aren't they? Yeah, quite coarse. And we're looking at a mosaic, I would say, from around about A.D. 360, probably made by a local man, possibly based somewhere around Cirencester. And that fits the dates of the coins in the trench. We've had coins of the 340s, 50s, 60s and 70s. So we've actually got quite a tight dating now. OK, so we've got the dating, but have we managed to work out what this trench means in relation to what Lysons found? At long last, yes. This is the mosaic. You mean this is this? That's right. The problem we've had for two days is that we've been chasing Lysons. Was his rooms running in this direction or were they going in that direction? So where would the Orpheus mosaic have been? That would have been under the grass at the far end of the trench in that direction. Well, it's been a struggle. Yes. Thank goodness I was able to chivvy you on, otherwise we might never have worked it out. It's all down to you. It's disappointing that we didn't find more of the Lysons villa that we assumed would be easy to locate. The quality of these mosaics alone tells us that this was an impressive 4th century villa occupied most probably by wealthy Yeoman farmers, the early ancestors of England's medieval knights. Yesterday, back in the lower field, the experts thought that there were two bathhouses on the site, one earlier than the other, and today Matt's found the concrete evidence. We've got this huge chunk of opus signinum concrete there. As you can see, it's got this lovely kind of curve on there. Do you know which part...? Oh, yes, I know exactly what this is. This is the first floor for this room and it's part of the same phase as the concrete that you've got, the concrete floor that predates the mosaic. Right, so this is the earliest phase. And in this phase, I suspect the room was being used very much as part of the bathhouse for the use of the bathhouse and it was a hot room, and that when the mosaic was put down, possibly the use of the room had changed somewhat. That there would have formed the base, the floor of the bath, and here's the edge along there. That's the quarter-round moulding, yes. This would have been totally waterproof. And at the bottom of the bath? No, this is not in situ. This has been ripped out and just dumped. But this isn't quite as fancy as a mosaic, though, is it? No, it's just plain... Yes, but it's waterproof, so it's vital. Undeterred by failing to find a springhead, John's at it again. 25 past 5 and still geophysics comes. Worth waiting for, though. Look at this. We've got a whole landscape now, magnetically. Here's the scheduled villa, a series of prehistoric enclosures surrounding it. Then we've got this avenue that we saw yesterday, and that appears to connect the two fields with the possible shrine down here we still don't really know. And then in this field, sort of bathhouse field, as we've been calling it, look at the resistance results. There's the stream, there's the stream, just to orientate. There's the plunge pool, there's the trench with the mosaic. Look, we've got a whole new range. Corridors, rooms, absolutely fantastic. And basically it extends up the hillside there. And even more exciting, look at the scale. It's actually larger than the scheduled villa up there. Is it another villa? It looks like it, doesn't it? It's got to be. 2,000 years ago, there would have been many more people living in this landscape. Based on our finds, it looks as if a modest range of farm buildings, including a bathhouse, was in use by the river just after the Roman occupation in 43 AD. This settlement grew into a larger and more affluent complex with its own piped running water. But it would have been vulnerable to flooding. So perhaps it was the same family that built the villa that Lyson's discovered further up the hill on the site of some other prehistoric farm buildings. This later villa almost eluded us, but no-one could have predicted the quality and the significance of the other discoveries we made. We came here because of a handful of tesserae in this field, but we found this amazing bathhouse. Once Romans would have sat among these beautiful mosaics sweating, then they would have sprinted over to Phil's plunge pool to cool off and exchange the gossip of the day. But this being time, team, in the last few minutes we've been handed this. It's geophys of yet another villa up on this slope here. It's bigger than Lyson's original discovery and part of a huge Roman complex. Of course, we haven't got time to answer all the questions it raises. Did it once look out onto this stream? Was there a temple in that field there? Did all the buildings in all three fields exist at the same time? These will have to remain some of Withington's many secrets.