The author of the _Quæstio_ denies, however, that this was accomplished by the removal of part of the sphere of water from the underlying earth. The ocean cannot have been drawn away and heaped up, he says, for water would always flow down again, being fluid and naturally seeking its own level; nor can the sphere of water be eccentric to that of Earth so that the latter emerges from it in one part. The reason for this is interesting since it implies a curious and apparently novel theory of gravity. Aristotle’s theory was that all heavy things tend by a natural law towards the centre of the universe, and therefore necessarily range themselves in spheres round it; the writer might therefore have made short work of Water’s eccentric sphere by stating briefly that it could not exist; instead of this, however, he suggests a new law of gravity, apparently invented on the spur of the moment, according to which everything which has weight tends towards the centre of its own circumference. If then earth and water had two separate centres, they would fall naturally in different directions, or as he puts it “to different _downs_.” Yet we know that the law of gravity is the same for both, and therefore their spheres must be concentric. The centre of the universe must be also the centre of the sea.[626] It had apparently been urged (perhaps in the discussion at Mantua) that since water follows the moon in the tides, its sphere may also imitate hers in being eccentric, but although Dante takes for granted that a connection between moon and tides is proved (and indeed it had been noted by the Greeks),[627] he replies that imitation in one particular need not imply imitation in all. Again, if the sphere of earth emerged from the sphere of water in the way proposed, the outline of the emergent earth must be circular, instead of which it is that of a half-moon, or nearly so, its width from east to west being much greater than its length from north to south. ------------------------------------------------------------------------ [626] _Qu._ xv. 10, 11. [627] See p. 134. ------------------------------------------------------------------------ Dante has been blamed for assigning this shape to the emergent earth, but the comparison of Earth’s habitable quadrant with the illuminated quadrant of the moon is perfectly just, as we saw with Ristoro,[628] from whom Dante is perhaps quoting here. The qualifying “vel quasi”[629] is added because the truly habitable regions did not extend over the whole quadrant, but stopped short at the arctic circle. ------------------------------------------------------------------------ [628] See p. 247. [629] “Or nearly.” _Qu._ xix. 61-63. ------------------------------------------------------------------------