The history of the history of science reveals changing styles; for the late nineteenth century Galileo was the model of the empirical and positivistic scientist, formulating general laws à la Mach, as summaries of experimental data. For more recent writers such as Koyre and Burtt, Galileo was a Platonist whose revolutionary work sprang almost full-grown from his head and who did not do the experiments described in his dialogues — fortunately, for they would not have worked if he had done them. I do not propose to take up the issue of the Platonism or non-Platonism of Galileos mathematical science; the issue here is that these writers argue that Galileo did thought-experiments in the course of developing his theories, and referred to experience primarily as a final check in order to be sure that he hadnt gone wildly astray.¹ This rationalist picture of Galileo has been thoroughly criticized recently by Drake, Settle, and others,² who have shown that Galileo's notebooks give ample evidence that he did do experiments at the time when he was probably developing the analysis of freely falling bodies, and that these were reasonably accurate. Furthermore, even some of the experiments described in his dialogues give good results when done today. I shall argue for another, related point, that in any event, Galileo's own writings, particularly the Dialogue Concerning the Two Chief World Systems and the Two New Sciences, give ample evidence that he considered experience essential for the foundation of his science and not only as a check on the results.