3 Its size is not reported. Tannery, Pour l'histoire de la science hellène 2 (Paris, 1930), pp. 84f., suggested that it was 9 times as great as the earth, thus forming an arithmetical sequence: 9–18–27. The order is given in A11, A18.
4 Kahn C. H., Anaximander and the Origins of Greek Cosmology (New York, 1960), p. 89.
6 Tannery's conjecture has become a classic: flame is seen only at the orifice, because inside the tube it does not emit light. The ρ does not have to obstruct the light of the fire inside, so it will be perfectly translucent for the light of the bodies behind it.
7 As I will argue in Section II we can give two different kinds of argument for this arrangement: one is to name the physical force(s) responsible for the constitution of the heavens, in particular the natural upward tendency of fire; while the other, which I am pursuing here, is to spell out the intrinsic motivation for a certain theory of the heavens to order the heavenly bodies in a particular way.
8 My proposal is the same as Kurt von Fritz's in a letter to Kahn, see Kahn, op. cit., p. 90 n. 3.
9 It is not clear from the testimonia whether the bands of the sun and the moon completely overlap, or only intersect at some point. In the standard illustrations the latter is presupposed (see e.g. Heath's T. L. Fig. 3 in his Aristarchus of Santos [Oxford, 1913], p. 35), and not argued for.
10 Cf. Diels H., ‘Über Anaximanders Kosmos’, AGP 10 (1897), 229 = Kleine Schriften, p. 14: ‘Es kann wohl keinem Zweifel unterliegen, dass die Helligkeit des Feuers zu dieser Rangordnung die Veranlassung gab.’ Diels, however, does not relate this ordering to the problem of visibility, and gives only a cosmogonical reason for it.
11 Kahn, op. cit., p. 90.
12 Some would accept on the testimony of 28A37 DK. that Parmenides postulated a fiery core of the earth.
13 Anaximander is apparently unique in this opinion. And it is no use to refer to the fact that he was the very first of the tradition, a trend-setter rather than follower of established orthodoxies. Popular thought and theology alike regarded celestial bodies as individual and permanent bodies (as e.g. the chariot of Ἥλιος etc.), and not as epiphenomena of some substructure, so Anaximander's departure from the tradition has to be explained.
14 The argument as recorded by Aristotle employs the Principle of Sufficient Reason. An extensive analysis in Barnes J., The Presocratic Philosophers (London, 1979), pp. 23–6.
15 W. K. C. Guthrie tried to assimilate the shape of the earth to that of the heavenly bodies by claiming that it was in fact a ring. See his A History of Greek Philosophy, i (Cambridge, 1962), p. 99, with notes 2 and 3.
16 Balance, as Classen H. J. rightly insists, is the keynote in Anaximander's fragment and parallels the role of order in his cosmology (‘Anaximander and Anaximenes: The Earliest Greek Theories of Change?’, Phronesis 12 (1977), 98). If my contention is right, balance will apply to the structure of the cosmos as well.