The survey took place over two weekends in June, conducted by a quiet, methodical team of four with ground-penetrating radar equipment and a manner that suggested they were trying hard not to get anyone’s hopes up. Gerald served coffee and stayed out of their way, watching from the kitchen window as they moved the equipment in slow overlapping passes across his courgettes. On the second afternoon, they found something. A roughly circular area of soil disturbance approximately two metres across, at a depth of between forty and eighty centimetres, positioned toward the back of the garden. The kind of pattern that could indicate natural geological variation. Or, the team leader said carefully, the kind of pattern consistent with a deliberate burial or deposit of some age.
Gerald looked at the patch of ground, which looked identical to every other patch of ground. “What do we do now?” he asked.
“Now,” the team leader said, “it becomes an excavation question, not a survey question. That’s a different process, different permissions, different team.”
The permissions took until October—almost exactly a year after the spade had struck the pendant. The excavation team had six people with hand tools and a methodology that made Gerald’s engineering precision look reckless. They removed soil in horizontal layers of two centimetres, sieving every load. For three days, nothing. Then, on the fourth morning, Gerald arrived in the garden to find the lead archaeologist crouched at the base of the pit with her brush moving very carefully indeed.