DECOMPOSITION. 161 



and the second a great deal lighter. Each, in its 

 separate state, may have a great variety of tem^ 

 peratures, and have its volume augmented by heat 

 or the removal of pressure, or diminished by pres- 

 sure or by cold ; and though that has not yet been 

 satisfactorily done by human experiment, there is 

 not the least doubt, that by sufficient cooling, both 

 might be condensed into liquids, and crystallized into 

 solids. We do not know that these elements of 

 water are absolutely simple ; but we call them so, 

 just because we are not able to resolve any of them 

 into two substances bearing different pioperties; 

 and the ancients thought water simple, and called it 

 an "element," for the same reason. But we can 

 work any of those (to us) simple substances through 

 a very great range of temperature, and still get them 

 back again in the very state with which we set out. 

 But bring them together in the proportions in which- 

 they form water, and apply a lighted match, and the 

 combustion is terrible, probably the most brilliant 

 display of the action of heat with which we are 

 acquainted, and perfectly irresistible in its effects. 

 When those elements are in sufficient quantity, and 

 free to mix with their natural rapidity, as much heat 

 would come out of the materials of a pitcher of 

 water, when passing from the state of separate airs 

 or gases to that of the compound liquid, as would 

 suffice to kindle the globe, or loosen from their 

 cohesion the particles of any substance in nature, 

 whether compound or simple. 



The progress of decomposition is always the 

 same as that which is produced by the action of 

 heat ; the solid is first changed to a liquid, and then 

 the liquid into an air or vapour ; but there are many 

 cases in which the process is altogether invisible ; 

 and there are others in which the two parts of it 

 follow each other so closely that we cannot distin- 

 guish them. There is no doubt, however, of the 

 perfect uniformity of the process ; and that whenever 



