26 SCIENCE IN GRECO-ROMAN ANTIQUITY 



science. 1 It is, moreover, in disagreement with the 

 perception of the motion, which, according to 

 Anaximander, brings the heavens back into the same 

 position every 24 hours. Therefore it is in a qual- 

 itative sense that primitive matter is, not infinite, 

 but indeterminate, that is to say, susceptible of 

 taking manifold and varied properties. The same 

 divergence of opinions exists as to the ideas of 

 Anaximander on the progressive constitution of the 

 universe. 



According to J. Burnet the ojzsioov is submitted to 

 shocks which shake it up and down, and which, in 

 certain regions, determine the opposition of heat and 

 cold. 2 The heat then appears as a sphere of flames, 

 which surrounds the cold, represented by a world 

 whose entire surface is covered with water. Under the 

 influence of the heat, part of the water evaporates and 

 changes into moist air. Owing to its force of expan- 

 sion, the air then penetrates the flaming sphere and 

 divides it into rings, in which the fire is imprisoned and 

 becomes invisible. It can, however, escape if an open- 

 ing has been left by chance on the ring, when, bursting 

 forth, it takes again its luminous consistency and forms 

 one of the heavenly bodies we see. This being so, the 

 eclipses of the sun or moon, and the waxing and waning 

 of the moon are easily explained. These phenomena 

 occur every time the openings of the solar or lunar rings 

 become completely or partially closed up. This explana- 

 tion appears, at first sight, surprisingly ingenious. 

 Anaximander, however, may have derived the inspira- 

 tion of his theory of rings from the appearance of the 

 Milky Way, and, on the other hand, he has but extended 

 to the heavenly bodies the explanation which he gave of 

 lightning and thunder, namely, a fire escaping through 

 the air contained in the clouds. " Anaximander held 



1 25 Tannery, Science hellene, p. 94. 



2 8 Burnet, Aurore, p. 62. 



