Rescue archaeologists at work: the mechanical digger, digging a trench for a new sea wall, reveals a prehistoric house. Just to the left of the bucket can be seen the headscarf of one of the excavators, Mysie Macrae.

deposit.

It seemed that the material on the beach, indicative of the Beaker period, was redeposited, perhaps by the action of the Allt Garbh. We were encouraged to follow the stratum containing the midden deposits. The aim was to recover any artefacts present before the deposit was swept away by the chance combination of northwesterly gales, a Spring high tide and a stream in flood after a storm; and to keep watch for any in situ material. At this stage, meticulous recording of 232

position was not necessary, due to the re-deposited nature of the site and the disturbance caused by the excavator during the sea-wall construction.

From June 1979 until August 1982, our work at Dalmore was undertaken on a small scale, often for short spells during family visits to the beach. The first bone awl caused excitement and the first stone arrowhead was an even greater thrill: we little realised how large the finds collection would become!

By January 1981, the collection of bone tools, decorated pottery fragments and stone tools was already extensive enough to justify another week's visit to Lewis, for cataloguing purposes, by Trevor Cowie, by then on the staff of the National Museum of Antiquities in Edinburgh.

But the collection continued to grow. In 1982, Margaret decided to move on to "phase 2", with much more extensive removal of blown sand from above the artefact-rich layer. She suspected that the site might not be entirely redeposited, but that remnants of walls were present. This led to the calling-in of various expert second opinions. Always the answer was the same—there is no cash for a professional team to take over; try to gather all the finds you can before a storm destroys the whole deposit; get in touch if you are convinced you have an in situ wall.

From August to October, Margaret toiled on the beach, sometimes with helpers, more often accompanied only by gulls, sheep, rabbits, circling eagles and the occasional curious seal! By October the task of shovelling blown sand, excavating and sieving the midden layer, recording the position of all angular slabs, and initial sorting of finds was complete. Much more of the site must lie behind the sea-wall, we thought, and there it would remain undisturbed. The storms could do their worst without damaging any prehistori c remains.. .

But we had reckoned without the excessive storms of Autumn 1982! The Dalmore Burn changed its course, running the full length of the beach parallel to the shore, scouring sand from the sea-wall. The wall bulged here, split there, then collapsed along much of its length. The machair and even the cemetery extension were under threat of possible further erosion.

On November 19th, contractors acting on behalf of Comhairle nan Eilean (The Western Isles Islands Council) started work with a large mechanical excavator. The repair of the sea-wall involved digging large holes behind it to insert straining wires and anchor plates (acting