IT WAS A TRIP SHROUDED IN MYSTERY from the beginning. Mike Braun, general manager of Emperor Divers in Egypt, had invited me to dive a wreck-site, but the arrangements were cloak-and-dagger. He was unable or unwilling to give me much detail.
He told me it might be something very exciting or it might be nothing. It would occupy a week of my time. Was I prepared to take the risk
An air ticket duly arrived. It brought me, together with wreck-hunter the late Steve Carmichael-Timson, to Hurghada.
It was a long eight hours later, including passing through Egyptian immigration and a less-than-comfortable drive, before we were met in darkness at a flyblown jetty in southern Egypt, and ferried to where mv Emperor Elite waited.
This would be our home for the next week. With so few of us, we were going to rattle around.
Elite is a liveaboard dive-boat that usually carries 20 passengers. Apart from the regular crew, Mike and Steve, there was only senior instructor Ahmed Fadel and dive-guide Cat Parfitt.
We set off south at a great lick on a waxy-calm sea. Elite can do 19 knots since being refitted, and with her new under-hangings.
We travelled overnight to a place that I later discovered was called Habili Saleh. I deduced that it was around 30 miles south of Dangerous Reef, and close to the border with Sudan.
Id like to tell you more, but the information was not made available. What I can tell you is that Ahmed had dived the site before, on an exploration trip, and had stumbled across
something interesting.

CROSSING THE TOP OF THE REEF by inflatable, we dropped onto a sandy-looking seabed dotted with small coral heads. We saw two large anchors, petrified yet strangely upright.
Another large item, which I at first mistook for a third anchor, was later thought by Steve to be the metal part of a rudder assembly from a very old boat.
Then there was what looked very much like a ships bell. All around us, the outlines of pottery showed up through the sand, but this seemed to be from a much earlier era.
There were assorted amphorae, decorative urns and large jugs, strewn across an area of seabed about the size of a football pitch, at 23m deep.
In some cases the pottery looked quite posh, with square handles, while other items were simple conical storage jars.
The problem was that they were not simply lying there, waiting to be picked up by some itinerant diver.
Lifting anything was out of the question, both legally and practically. Everything was well concreted into a rocky substrate that lay only a centimetre or so below a layer of sand.
The big question was whether this was a seriously important ancient wreck site.
My job was to photograph as many of the finds where they lay as I could, so that the pictures could be shown to an expert who could make that judgment.
Steve would check the seabed for evidence of metal structures with his underwater metal-detector, and survey the whole area with his side-scan sonar.
This, in conjunction with his laptop computer, could render a picture of what was down there, covering an area that would have been impossible for a few scuba-divers to survey.
First we dived to see what was worth closer attention, marking likely candidates with ping-pong balls on string. Steve got to work with his detector, and his readings indicated a fair scattering of metal pieces.
We found a few items that looked to us like copper rivets.
Of course, we were equally excited about the bell. It was a simple-looking shape but it was decided that if we could clean and photograph it, it might reveal all the clues we needed to determine what was lying there.

OUR BELL WAS COMPLETELY CONCRETED IN, so we started cleaning it off where it lay. This was hard work.
A depth of 23m may not sound much, but if you want a lot of time down there, decompression stops are a fact of life.
We had to use our dive time carefully, so we decided to work a rota, each diver doing 20 minutes of hard graft, followed by an ascent of 15 or so minutes. This profile allowed us to get in several dives each day.
Mike went in first and got to work. When my turn came it was easy to spot him from the tremendous stream of bubbles he was exhaling, such was the effort he was putting in.
I then did some heart-busting over the relic before Cat took over from me.
Ahmed and Steve were setting up the side-scan sonar. By the next dive we were ready, with a full complement of divers to work on the bell.
We soaked up a lot of nitrogen and expended a lot of effort until we came to do the last dive on the bell. I had given strict instructions that it should not be moved, at least until I had had time to photograph it.
Ahmed went in and got to work, followed by Mike. I passed Ahmed doing a lengthy deco stop as I dropped down, only to be met by Mike, not working around the bell at all, but starting on an amphora nearby.
He motioned me to where the bell lay. I could see that it now had a hole in it.
It wasnt a bell. It was a bell-shaped terracotta pot, and his disappointment was palpable. He signalled to me that we should both ascend.
As we swam towards the reef, we saw several more bells that we simply hadnt noticed before. Funny how the mind plays tricks if youre into wishful thinking.

BACK ON ELITE, we discussed the matter and then put our disappointment aside. We decided to choose some good examples of the amphoras and other pottery, clean them where they lay to reveal them as best we could, and photograph them. Some of the amphoras were disqualified because coral was growing on them.
The others went in first with wire brushes and chipping tools. Steve and I followed later with video, stills photography equipment and scale-bars to lay alongside each item.
The water was thick with sand and detritus liberated by those doing the hard work but the job was quickly accomplished, the pictures digitally processed, and they were sent off to various experts around the world.

ROSS IAIN THOMAS of Southampton University told us that a number of interesting features would help to identify the purpose and date of the mystery wreck.
He started by citing traces of the superstructure, preserved in the form of copper alloy nails and lead sheathing from either the bow or stern end.
Lead sheathing was constructed from large sheets 1-2mm thick that were laid over pitch waterproofing and held in place by copper tacks hammered into the wooden hull, he told us.
This system, used to protect the hull from marine borers, has been found on hulls of 5th century BC to 2nd century AD wrecks in the Mediterranean.
Lead sheeting and copper tacks fitting this description have been found in 1st century BC through to 3rd century AD deposits in the Red Sea Roman ports of Myos Hormos and Berenike (Port Berenice), where such vessels would have anchored and been maintained.
At least two ancient iron anchors can be seen in the photographs, said Ross. Two-armed stock iron anchors occur in the Mediterranean from the Roman Republican period onwards, though the curved arms of these examples are typical of the early Imperial Roman period. Both the hull and anchors suggest that this vessel was built in the traditional Mediterranean fashion.

ROSS DID NOT FIND IT SURPRISING that the vessel should be where it was. Rome controlled much of the Red Sea coast after its annexation of Egypt in 30 BC and Nabataea (modern Jordan) in 106 AD. Greco-Roman merchants took this opportunity to be involved in the Red Sea and Indian Ocean trade of exotic spices and incense with South Arabia, East Africa and India.
So we had been diving an ancient merchant vessel. Ross said that detailed analysis of the cargo amphoras would greatly increase our understanding of such vessels, but that the photos already told him a certain amount.
The bifid handles, bead rim, long cylindrical body and short spike of some of the amphoras told him that they were of a type produced in such locations as Italy, Spain, Greece, Turkey, France, Tunisia and Egypt, from the late 1st century BC to the mid-3rd century AD.
Broad, long amphoras ending in a solid spike with two small looped handles at or under the rim came from Egypt from the same starting date through to the 5th century AD.
Squat amphoras with a flat ring base, short flaring neck and beaded rim came from France and Spain, and similar types from North Africa, from 50-300 AD.
These three amphora forms were specifically used to transport wine, and it is likely that this wreck was on its way to Arabia, East Africa or India to trade wine for spices, incense or other high-value and exotic goods demanded by the wealthy Roman public, Ross told us.
So everything pointed to this Roman merchant ship being wrecked between 50 and 200 AD on its way from Berenike to an unknown destination.
There may be many such wrecks in the Red Sea but, because it is so deep, what is unusual is to actually find one.

Searching the seabed for metal objects.
Surely its a bell
Single-handled jug.
One of the anchors.
Amphoras were scattered everywhere.