Joseph Grinnell (1877–1939) was the founding director of the Museum of Vertebrate Zoology, Berkeley (MVZ). He conducted extensive and intensive surveys of vertebrate species distribution throughout California, USA. The problem we track throughout this essay is that Grinnell’s carefully laid plans for his museum at the beginning of the twentieth century embodied a notion of locality and associated technology of recording that turned out to be hard to unify with a very different notion that emerged in mid-twentieth century ecology and is embedded in late twentieth century computing practices involving the collection, storage and retrieval of species locality data. Different concepts of space favor different protocols and ways of describing a species’ locality, and thus slowly entrench only certain courses of action in practice, which over time hinders the integration of data.