258 On the Fall of Stones from the Clouds. 



ter like pitch, observed around certain aerolites, must be consi- 

 dered as parts which have not been inflamed. If the substance^ 

 of which the balls of fire are composed, which precede the fall 

 of aerolites, were not very combustible, these balls would not be 

 of great extent, nor would the ignition be of long duration. The 

 meteorolites themselves do not contain any very inflammable 

 eubstancep, since they have been burnt before their fall. We also 

 add, that the phosphate of iron sometimes covers aerolites, as in 

 the case of that which fell in Russia in 1807. It must also be 

 observed, that nothing is wanting in the slimy ores of iron, but 

 nickel, to resemble in point of composition atmospheric stones, 

 as M. Vauqnelin has remarked*. As these ores are almost al- 

 ways formed in the midst of marshes, we may in some measure 

 thereby see the slow formation of aerolites. In fact, what we 

 have already said, shows wherefore in several cases cloudy wea- 

 ther precedes the fall of the stones ; for the cloud contains all 

 the substances of which they are formed, and in this sense th» 

 phsenomenon may to a certain extent be compared with the so 

 iution of salts. 



It seems, therefore, that in considering the force which sus» 

 tains the balls of fire as resulting from the inflammation or for* 

 mation of vapours, we do not admit a rash hypothesis. In facty 

 in all cases where these balls burn with little violence, the mat* 

 ter of which they are composed soon falls, while it reascend$ 

 the instant the inflammation becomes brisker. This is evident^ 

 from the effects of a fusee, in which the power of ignition raises 

 and even supports a great weight. This power also exercises 

 its action on aerolites ; and as it is opposed to gravity, it obliges 

 the stone to follow a mean direction between the two impulses 

 ■which are impressed upon it. Observations also show that 

 aerolites fall in proportion as the fire dies away, and that whenj 

 as at Connecticut, the burning increases and explosions take 

 place, the stone ascends and rises. 



From these facts we may conceive why all these phaeno* 

 mena are frequent in the hottest months and so rare in winter^ 

 and wherefore they appear in the evening frequently accom-* 

 panied by storms. The causes of these phaenomena, in appear- 

 ance so remote, have nevertheless some approximation : thus 

 rain is. nothing but the precipitation of the water which is con- 

 tinually rising into the air ; and aerolites probably only depend 

 on the precipitation of an infinity of substances which are inces- 

 santly evaporating, and the reaction of which upon each other 

 may form new combinations. This hypothesis will not appear 

 gratuitous, if we pay attention to the immense quantity of com*^ 



^ Annalts.iiu Musfum i'liist. i^at. tomeviii. p. 4^9. 



pound 



