Bailly also found that the ancient fables, considered physically, appeared to belong to the northern regions of the earth. In 1779, Bailly published a second collection, forming a sequel to the former, and entitled _Letters on the Atlantis of Plato, and on the Ancient History of Asia_. Voltaire died before these new letters could be communicated to him. Bailly did not think that this circumstance ought to make him change the form of the discussion already employed in the former series; it is still Voltaire whom he addresses. The philosopher of Ferney thought it strange that there should be no knowledge of this ancient people, who, according to Bailly, had instructed the Indians. To answer this difficulty, the celebrated astronomer undertakes to prove that some nations have disappeared, without their existence being known to us by any thing beyond tradition. He cites five of these, and in the first rank the Atlantidae. Aristotle said that he thought Atlantis was a fiction of Plato's: "He who created it also destroyed it, like the walls that Homer built on the shores of Troy, and then made them disappear." Bailly does not join in this skepticism. According to him, Plato spoke seriously to the Athenians of a learned, polished people, but destroyed and forgotten. Only, he totally repudiates the idea of the Canaries being the remains of the ancient country of the Atlantidae, and now engulfed. Bailly rather places that nation at Spitzbergen, Greenland, or Nova Zembla, whose climate may have changed. We should also have to seek for the Garden of the Hesperides near the Pole; in short, the fable of the Phoenix may have arisen in the Gulf of the Obi, in a region where we must suppose the sun to have been annually absent during sixty-five days.