The Met has seventy-three wooden spindles from Peru, all given by the same man in 1933, and every one of them still has yarn on it.

Each has its own record, and the records are nearly identical. Here is one: “Wood, camelid hair.” Ten inches long, a quarter of an inch thick. Dated “10th–16th century (?)”, which is a span of about seven hundred years with a question mark after it. No place more exact than Peru. The others differ only in length, from about fifteen centimetres to twenty-six.

The photographs are where they stop being identical. There aren’t seventy-three; there are four, and each record points to one of them, so the spindle you looked up is somewhere in a group of fifteen to twenty-five, laid out like pencils. In one photograph a handwritten card sits at the bottom: “115-130”, the run of catalogue numbers in the shot. In every photograph the yarn is wound around the middle of each shaft, and the amount is different on every one. Some are thick with it. Some have a thin wrap you could unwind in a minute. Some yarn is dark and some is pale, and a few loose ends curl off the side.

A finished cloth hides how long it took. A spindle with yarn on it does the opposite. It holds exactly as much as was spun before someone stopped. That might have been the end of a day, or the moment the work was put away for good. I don’t know which, for any of these. The record doesn’t say where they were found. Elsewhere, on the dry Peruvian coast, spindles turn up with yarn still on them in workbaskets buried with their owners, but I can’t tell you that’s what happened to these seventy-three.

What I can say is that the museum can’t date any of them to within seven centuries, and yet each one shows, to the turn, how far its spinner got. The records make them all but identical. The photographs show seventy-three different stopping points.