I wanted to make evolutionary biology experimental, and no one had thought of making biogeography experimental. How could you make biogeography experimental? And it dawned on me — because I was doing all this field work, more from the experience of natural history — that we weren’t going to be able to experiment with New Guinea or Fiji, or even a small island in the West Indies. Because what I had in mind was to eliminate all the species in a place where they could be eliminated without any real damage to the total fauna, and then study the return of those, and see how that accorded with the basic patterns predicted by the theory of island biogeography. And it dawned on me that whereas you have to have an island the size of Cuba, say, for a real population of woodpeckers or small mammals, that a very small island, like a mangrove island in the Florida Keys, would be an island for tiny insects where thousands of a species could live.