96 STRUCTURE OF CLOUDS, &c. CHAP. 2. 18. 



posed that they may be composed of hollow 

 vesicles;* and, in this case, if the component 

 vesicles should contain an aeriform fluid, lighter 

 than common air, they would become buoyant, 

 and float in the atmosphere. It is not probable 

 that they contain hydrogen gas. For, if they 

 do, what can be the structure and component 

 parts of the vesicular bag itself? It cannot be 

 water. For, if electricity should preserve it 

 in the vesicular form, it could not prevent the 

 escape of its hydrogen from within. Could the 

 bag itself be water, it would never be oxygen 

 and hydrogen in a state of combination. The 

 oxygen would not affect or act upon the 

 hydrogen in the bag, because it had already 

 combined with its due proportion of hydrogen, 

 and become water. Nothing, then, but an 

 accession of more oxygen could convert the 

 contents of these vesicles into water. AVhen 

 the electricity is equalized, the water which 

 composes the bag or vesicle comes down in 

 Rain ; and the hydrogen, mixing with common 

 air, may be exploded by the electric spark, or, 

 meeting with oxygen, may explode sponta- 



* Idees sur la Meteorologie, par I. A. de Luc, vol. ii. p. 

 607. part iii. sect. vi. p. 100. 



