LIGHTNING AND THUNDER. \^O^ Zl3 



L 



strata of air, and sustaining the weight of the ~ 

 air and vapor above, must have greater density than 

 the upper surface. Hence we may reasonably infer, 

 that this surface, and that of the earth below it, fulfill 

 the conditions necessary for a series of echoes. 



The hypothesis of simultaneous discharges, at differ- 

 ent distances, may also be true in certain instances : as 

 it is quite possible that such simultaneous discharges 

 frequently occur. But the succession of reports, often 

 following each other with marked regularity, and steadi- 

 ly decreasing in volume and intensity, is not fully ex- 

 plained by this hypothesis, while it is entirely in ac- 

 cordance with the character of a series of echoes. 



The re-adjustment of electric energy between differ- 

 ent parts of a large cloud, which must follow the pri- 

 mary discharge, gives rise to numerous minor discharges; 

 whose sound, mingling with that from the larger air 

 waves, causes the rumble ; analogous to the crackling 

 sound from similar minor discharges in the machine. A 

 premonitory rumble, from a similar cause, often precedes 

 the heavier discharge ; just as the crackling precedes 

 the discharge of the machine. 



If the cloud were a perfectly homogeneous conductor, 

 like a metal cylinder, this could not occur. But as it 

 is a mass of vapor, composed of drops insulated from 

 each other by air spaces, each particular drop having its 

 own electric charge ; and different parts of the cloud 

 having different densities, and hence differing in con- 

 ductivity and resistance; and condensation, with increase 

 of potential, following the discharge, as already shown, 

 such minor discharges, with the accompanying roar and 

 rumble, are inevitable. Also the development of the 

 residual, after the primary discharge, which, in a large 



