ATMOSPHERIC ELECTRICITY. 367 



to have in producing these phenomena, M. Arago* remarks that there 

 is little hope of arriving at anything decisive on this subject. lie 

 cites the echoes which have continued during half a minute from the 

 sound of the discharge of a pistol, and from the observations he has 

 been able to collect, the longest peals of thunder are not more than 

 thirty-six, forty-one, and forty-five seconds. The objection has been 

 made that sailors hear the rolling of thunder in the open sea Avhere 

 then there is no terrestrial object capable of reflecting the sound ; but 

 observation proves that clouds possess, like other bodies, the property 

 of producing the reflection of sound. f Nevertheless a remark seems to 

 prove, according to M. Arago, that the rolling of thunder does not 

 always result from simple reflected sound when the flashes of lightning 

 succeed each other in a sky uniformly cloudy. Some are attended by 

 long reverberations while others are followed only by short and sharp 

 peals. It is evident that these remarkable differences cannot be ex- 

 plained by considering the rolling of the thunder as a simple phe- 

 nomenon of echo. 



Recourse has been had to another consideration derived from the 

 inec|uality of time which the noise, produced by the displacement ot 

 the air in the different points along the course of a flash of lightning, 

 takes to reach the ear by which it is perceived. Dr. Eobert IlookeJ 

 appears to be the first who observed this circumstance ; according to 

 him the different portions of the long line of the flash being in 

 general at different distances, the sounds generated, whether succes- 

 sively or simultaneously, will require different or unequal times to 

 reach the ear of the observer. Hence the length of the flash of lightning 

 v.'ould determine the duration of a peal of thunder, and to an observer 

 placed under the line of a flash of lightning, near its middle, the 

 same peal would produce reverberations only half as long as to one 

 placed at the extremity of the flash. This explanation, based on the 

 time necessary for the propagation of sound, has been generally adopted 

 by philosophers. This also accounts in the most rational manner for 

 the reverberation of thunder. As for the changes of intensity so fre- 

 quently mentioned, Helvig§ regards the zigzags of the flash of light- 

 ning as playing a very important part in this phenomenon. He founds 

 his opinion on an observation in which he saw a flash of lightning 

 Avhich, in descending towards the earth made four bends, and pro- 

 duced as many distinct and well defined explosions, but not all of 

 the same force. According to him a change in the direction of a 

 flash of lightning must occasion a change in the sound of the thunder. 

 And in fact, as M. Arago|| also observes, v/hen a flash of lightning 

 that is proceeding directly from an observer doubles upon itself for 

 a moment, it is evidently a necessary result that there will be an 

 increase of the noise. Farther, this increase v/ill in its turn be fol- 

 lowed by a sudden decrease, if by a second bend the flash is brought 

 again to move a little nearer to the visual line, and so on. The changes 



^ Annuaire pour 1S38, page 451. 



t Anaalcs da Cliiiuie et de Ph)-sique, tomeXS, pajcc 210. 1822. 



:]: The Pobthunious Works of Kobcrt Hooke, page 42i, Lon.lon, 1705. 



y Anualts de Gilbert, tome LI, page 13'J. 



II Annuaire pour 1838, page 45C. 



