EXTRACTS FROM M. PELTIER. 355 



of the same nature of the vesicules, the whole of the va- 

 pors is maintained, in one body ; if this force is insufficient, 

 the interior repulsions obtain the preponderance, and the 

 spout is divided. We shall frequently find these re-unions, 

 and these divisions, when we treat of clouds in a special 

 manner. 



Thus when each of the vesicules of vapor, preserves a 

 very great tension, and their insulation the one from the 

 other prevents that of the cloud from being considerable, 

 the whole of the cloud is attracted and not merely the elec- 

 tric charge of the surface ; the specific gravity of the cloud 

 is overcome by this attraction, the cloud lowers and ap- 

 proaches the earth. While this movement is effected, the 

 attraction increases in a proportion greater than the prox- 

 imity, as it increases inversely as the square of the distance ; 

 the cloud continues then to descend, and it would descend 

 with great velocity, if each vesicular particle could receive 

 rapidly, the electric quantities necessary to replace those 

 which escape by accelerated radiation, and which become 

 so much more indispensable as the density of the air aug- 

 ments. 



The feeble conductibility of the clouds does not, without 

 slowness, difficulty, and inequality, permit the electricity 

 of the superior clouds to arrive into the parts nearer the 

 earth. The electric diffusion, cannot be returned, but very 

 unequally, in bodies so varied, in their forms, and in the 

 relations (rapports) of their constituent parts. 



The disposition and constitution of the clouds, do not 

 permit but very rarely the descent of a portion of cloud 

 down to the ground, principally in our country, for as soon 

 as the density of the cloud permits a tolerable conduction, 

 the electric tension transfers itself to the exterior, the dis- 

 charges take place in mass, and the cloud rises. It is then 

 a storm and not a spout. 



We see then that a union of circumstances very favora- 



