But there is yet another striking thought connected with this new explanation of the phases of the moon. The explanation implies not merely the reflection of light by a dark body, but by a dark body of a particular form. Granted that reflections are in question, no body but a spherical one could give an appearance which the moon presents. The moon, then, is not merely a mass of earth, it is a spherical mass of earth. Here there were no flaws in the reasoning of Anaxagoras. By scientific induction he passed from observation to explanation. A new and most important element was added to the science of astronomy. Looking back from the latter-day stand-point, it would seem as if the mind of the philosopher must have taken one other step: the mind that had conceived sun, moon, stars, and earth to be of one substance might naturally, we should think, have reached out to the further induction that, since the moon is a sphere, the other cosmic bodies, including the earth, must be spheres also. But generalizer as he was, Anaxagoras was too rigidly scientific a thinker to make this assumption. The data at his command did not, as he analyzed them, seem to point to this conclusion. We have seen that Pythagoras probably, and Parmenides surely, out there in Italy had conceived the idea of the earth's rotundity, but the Pythagorean doctrines were not rapidly taken up in the mother-country, and Parmenides, it must be recalled, was a strict contemporary of Anaxagoras himself. It is no reproach, therefore, to the Clazomenaean philosopher that he should have held to the old idea that the earth is flat, or at most a convex disk--the latter being the Babylonian conception which probably dominated that Milesian school to which Anaxagoras harked back.