PHYSICAL SCIENCE 



CHAP. PAGE 



XIV. There is nothing inconsistent with this in the explanation given 

 (I. xv.) of meteors. There may be an analogy with what is 

 observed in cases of fire when isolated groups of houses take fire 

 through gradual accumulation of heat . . . . . 65 



XV. Some (Stoics) think that spontaneous combustion takes place in 



the air 66 



XVI. Difference between a flash of lightning and a bolt ... 66 



XVII. Some explain the noise of thunder as due to hot meeting cold, 



as in the case of hot iron plunged into water .... 67 



XVIII. Anaximander attributes the effects to air and explains all the 

 phenomena by reference to it . . . . . . . 67 



XIX. Anaxagoras says it is the ether that acts on the lower atmo- 

 sphere to produce them : it sends out fire . . . . 68 



XX. Diogenes of Apollonia thinks that fire and air interact, pro- 

 ducing one another, as may be observed in the various phenomena 68 



XXL Authorities discarded : independent explanation. A flash and 



a bolt are fire in some form : they differ only in degree . . 69 



XXII. Analogy of fire on earth : it must apply above. Lightning due 

 either to impact or friction. Hurricanes are a sufficient cause of 



the former .......... 70 



XXIII. Clouds and air may through friction also be a cause. The fires 



so produced are insubstantial and evanescent . . . . 71 



XXIV. Fire by reason of its lightness levitates, just as water gravi- 

 tates. But in the case of a bolt it is forced down, contrary to its 

 nature, like a " weeping " tree . . . . . . 71 



XXV. But it is said that wet clouds produce fire. How ? . . 72 



XXVI. There is no inconsistency in the combination in the same 

 cloud of potential fire and water. A log burns at one end, 

 exudes moisture at the other. An island on each of two occa- 

 sions was thrown up by fire in the Aegean Sea, fire overcoming 

 water. And clouds are, as a matter of fact, required for light- 

 ning : exceptions to this are only apparent . . . . 72 



XXVII. Different kinds of thunder. The growling and the crashing, 



with their causes ......... 75 



XXVIII. In order to the sound of thunder, clouds of a particular 

 shape must meet in a particular way. A bladder does not burst 

 with a report if cut. A broad simultaneous blow over the whole 

 cloud is necessary to an explosion . . . . . . 76 



XXIX. The proper shape and the rupture of the cloud are necessary. 

 Compare drums, etc. . . . . . . . . 77 



XXX. According to some, clouds are not necessary to thunder : 

 witness eruptions of Etna and the overthrow of Cambyses' army, 

 where particles of sand were the medium of the thunder and 

 lightning. But in this case, too, a cloud was formed, it may be, 

 a denser, and more solid than one composed of mere air, before 



the sound was emitted . . . . . . . . 77 



XXXI. Strange effects of lightning 78 



XXXII. Portents and events, their undoubted and widespread con- 

 nection .......... 79 



XXXIII. Thunderbolts. Threefold division of the art of dealing 



with them . . . . . . . . . 8 1 



