204 PHYSICAL SCIENCE BK. v 



combined their streams, they surpass in size regular, 

 constant rivers. The same thing may probably 

 happen in squalls ; they are short-lived whenever 

 they are alone. But when they have joined forces, 

 and the air expelled from several parts of the sky 

 at once has all combined in one, both force and 

 duration are added to them. 



XIII 



1 So, then, wind results from the breaking up of a 

 cloud, which breach is effected in several different 

 ways. The accumulation of air is burst sometimes 

 by the internal struggle, as it seeks to gain an exit ; 

 sometimes by the heat produced either simply by 

 the sun or else by the mutual ramming and friction 

 of the roaming bodies. 



At this point, if you have no objection, one may 

 raise the question why a whirlwind occurs. In 

 rivers, when their course has been without any 

 obstacle for a long distance, the channel is a straight, 



2 uniform one. But when they meet some boulder 

 that juts from the bank, the stream is driven back 

 and whirls the waters in a circle without a way of 

 escape, so that in their revolution they are con 

 stantly sucked in toward the centre to form a 

 whirlpool. In like manner the wind pours out in 

 full force as long as no obstacle stands in the way. 

 But when it is reflected from some jutting pro 

 jection, or is massed in a quarter which com 

 bines to form a thin downward channel, then it 

 revolves upon its own axis, and produces an 

 eddy similar to that in which, as we have just 



3 said, the water revolves. This revolving wind, 



