Bob told me that it takes about seventeen minutes to read one zircon and it is necessary to read dozens from each rock to make the data reliable. In practice, the process seemed to involve about the same level of scattered activity, and about as much stimulation, as a trip to a laundromat. Bob seemed very happy, however; but then people from New Zealand very generally do.

The Earth Sciences compound was an odd combination of things—part offices, part labs, part machine shed. "We used to build everything here," Bennett said. "We even had our own glassblower, but he's retired. But we still have two full-time rock crushers." She caught my look of mild surprise. "We get through a lot of rocks. And they have to be very carefully prepared. You have to make sure there is no contamination from previous samples—no dust or anything. It's quite a meticulous process." She showed me the rock-crushing machines, which were indeed pristine, though the rock crushers had apparently gone for coffee. Beside the machines were large boxes containing rocks of all shapes and sizes. They do indeed get through a lot of rocks at the ANU.